fix: grind sort internalization

This PR ensures sorts are internalized by `grind`.

.claude/CLAUDE.md

@@ -1,89 +0,0 @@
-To build Lean you should use `make -j -C build/release`.
-
-## Running Tests
-
-See `doc/dev/testing.md` for full documentation. Quick reference:
-
-```bash
-# Full test suite (use after builds to verify correctness)
-make -j -C build/release test ARGS="-j$(nproc)"
-
-# Specific test by name (supports regex via ctest -R)
-make -j -C build/release test ARGS='-R grind_ematch --output-on-failure'
-
-# Rerun only previously failed tests
-make -j -C build/release test ARGS='--rerun-failed --output-on-failure'
-
-# Single test from tests/lean/run/ (quick check during development)
-cd tests/lean/run && ./test_single.sh example_test.lean
-
-# ctest directly (from stage1 build dir)
-cd build/release/stage1 && ctest -j$(nproc) --output-on-failure --timeout 300
-```
-
-The full test suite includes `tests/lean/`, `tests/lean/run/`, `tests/lean/interactive/`,
-`tests/compiler/`, `tests/pkg/`, Lake tests, and more. Using `make test` or `ctest` runs
-all of them; `test_single.sh` in `tests/lean/run/` only covers that one directory.
-
-## New features
-
-When asked to implement new features:
-* begin by reviewing existing relevant code and tests
-* write comprehensive tests first (expecting that these will initially fail)
-* and then iterate on the implementation until the tests pass.
-
-All new tests should go in `tests/lean/run/`. These tests don't have expected output; we just check there are no errors. You should use `#guard_msgs` to check for specific messages.
-
-## Success Criteria
-
-*Never* report success on a task unless you have verified both a clean build without errors, and that the relevant tests pass.
-
-## Build System Safety
-
-**NEVER manually delete build directories** (build/, stage0/, stage1/, etc.) even when builds fail.
-- ONLY use the project's documented build command: `make -j -C build/release`
-- If a build is broken, ask the user before attempting any manual cleanup
-
-## LSP and IDE Diagnostics
-
-After rebuilding, LSP diagnostics may be stale until the user interacts with files. Trust command-line test results over IDE diagnostics.
-
-## Update prompting when the user is frustrated
-
-If the user expresses frustration with you, stop and ask them to help update this `.claude/CLAUDE.md` file with missing guidance.
-
-## Creating pull requests
-
-Follow the commit convention in `doc/dev/commit_convention.md`.
-
-**Title format:** `<type>: <subject>` where type is one of: `feat`, `fix`, `doc`, `style`, `refactor`, `test`, `chore`, `perf`.
-Subject should use imperative present tense ("add" not "added"), no capitalization, no trailing period.
-
-**Body format:** The first paragraph must start with "This PR". This paragraph is automatically incorporated into release notes. Use imperative present tense. Include motivation and contrast with previous behavior when relevant.
-
-Example:
-```
-feat: add optional binder limit to `mkPatternFromTheorem`
-
-This PR adds a `num?` parameter to `mkPatternFromTheorem` to control how many
-leading quantifiers are stripped when creating a pattern.
-```
-
-**Changelog labels:** Add one `changelog-*` label to categorize the PR for release notes:
-- `changelog-language` - Language features and metaprograms
-- `changelog-tactics` - User facing tactics
-- `changelog-server` - Language server, widgets, and IDE extensions
-- `changelog-pp` - Pretty printing
-- `changelog-library` - Library
-- `changelog-compiler` - Compiler, runtime, and FFI
-- `changelog-lake` - Lake
-- `changelog-doc` - Documentation
-- `changelog-ffi` - FFI changes
-- `changelog-other` - Other changes
-- `changelog-no` - Do not include this PR in the release changelog
-
-If you're unsure which label applies, it's fine to omit the label and let reviewers add it.
-
-## CI Log Retrieval
-
-When CI jobs fail, investigate immediately - don't wait for other jobs to complete. Individual job logs are often available even while other jobs are still running. Try `gh run view <run-id> --log` or `gh run view <run-id> --log-failed`, or use `gh run view <run-id> --job=<job-id>` to target the specific failed job. Sleeping is fine when asked to monitor CI and no failures exist yet, but once any job fails, investigate that failure immediately.

.claude/commands/release.md

@@ -1,121 +0,0 @@
-# Release Management Command
-
-Execute the release process for a given version by running the release checklist and following its instructions.
-
-## Before Starting
-
-**IMPORTANT**: Before beginning the release process, read the in-file documentation:
-- Read `script/release_checklist.py` for what the checklist script does
-- Read `script/release_steps.py` for what the release steps script does
-
-These comments explain the scripts' behavior, which repositories get special handling, and how errors are handled.
-
-## Arguments
-- `version`: The version to release (e.g., v4.24.0)
-
-## Release Notes (Required for -rc1 releases)
-
-For first release candidates (`-rc1`), you must create release notes BEFORE the reference-manual toolchain bump PR can be merged.
-
-**Steps to create release notes:**
-
-1. Generate the release notes:
-   ```bash
-   cd /path/to/lean4
-   python3 script/release_notes.py --since <previous_version> > /tmp/release-notes-<version>.md
-   ```
-   Replace `<previous_version>` with the last stable release (e.g., `v4.27.0` when releasing `v4.28.0-rc1`).
-
-2. Review `/tmp/release-notes-<version>.md` for common issues:
-   - **Unterminated code blocks**: Look for code fences that aren't closed. Fetch original PR with `gh pr view <number>` to repair.
-   - **Truncated descriptions**: Some may end mid-sentence. Complete them from the original PR.
-   - **Markdown issues**: Other syntax problems that could cause parsing errors.
-
-3. Create the release notes file in the reference-manual repository:
-   - File path: `Manual/Releases/v<version>.lean` (e.g., `v4_28_0.lean`)
-   - Use Verso format with proper imports and `#doc (Manual)` block
-   - **Use `#` for headers, not `##`** (Verso uses level 1 for subsections)
-   - **Use plain ` ``` ` not ` ```lean `** (the latter executes code)
-   - **Wrap underscore identifiers in backticks**: `` `bv_decide` `` not `bv_decide`
-
-4. Update `Manual/Releases.lean`:
-   - Add import: `import Manual.Releases.«v4_28_0»`
-   - Add include: `{include 0 Manual.Releases.«v4_28_0»}`
-
-5. Build to verify: `lake build Manual.Releases.v4_28_0`
-
-6. Create a **separate PR** for release notes (not bundled with toolchain bump):
-   ```bash
-   git checkout -b v<version>-release-notes
-   gh pr create --title "doc: add v<version> release notes"
-   ```
-
-For subsequent RCs (`-rc2`, etc.) and stable releases, just update the version number in the existing release notes file title.
-
-See `doc/dev/release_checklist.md` section "Writing the release notes" for full details.
-
-## Process
-
-1. Run `script/release_checklist.py {version}` to check the current status
-2. **CRITICAL: If preliminary lean4 checks fail, STOP immediately and alert the user**
-   - Check for: release branch exists, CMake version correct, tag exists, release page exists, release notes file exists
-   - **IMPORTANT**: The release page is created AUTOMATICALLY by CI after pushing the tag - DO NOT create it manually
-   - **IMPORTANT**: For -rc1 releases, release notes must be created before proceeding
-   - Do NOT create any PRs or proceed with repository updates if these checks fail
-3. Create a todo list tracking all repositories that need updates
-4. **CRITICAL RULE: You can ONLY run `release_steps.py` for a repository if `release_checklist.py` explicitly says to do so**
-   - The checklist output will say "Run `script/release_steps.py {version} {repo_name}` to create it"
-   - If a repository shows "🟡 Dependencies not ready", you CANNOT create a PR for it yet
-   - You MUST rerun `release_checklist.py` before attempting to create PRs for any new repositories
-5. For each repository that the checklist says needs updating:
-   - Run `script/release_steps.py {version} {repo_name}` to create the PR
-   - Mark it complete when the PR is created
-6. After creating PRs, notify the user which PRs need review and merging
-7. **MANDATORY: Rerun `release_checklist.py` to check current status**
-   - Do this after creating each batch of PRs
-   - Do this after the user reports PRs have been merged
-   - NEVER assume a repository is ready without checking the checklist output
-8. As PRs are merged and tagged, dependent repositories will become ready
-9. Continue the cycle: run checklist → create PRs for ready repos → wait for merges → repeat
-10. Continue until all repositories are updated and the release is complete
-
-## Important Notes
-
-- **NEVER merge PRs autonomously** - always wait for the user to merge PRs themselves
-- The `release_steps.py` script is idempotent - it's safe to rerun
-- The `release_checklist.py` script is idempotent - it's safe to rerun
-- Some repositories depend on others (e.g., mathlib4 depends on batteries, aesop, etc.)
-- Wait for user to merge PRs before dependent repos can be updated
-- Alert user if anything unusual or scary happens
-- Use appropriate timeouts for long-running builds (verso can take 10+ minutes)
-- ProofWidgets4 uses semantic versioning (v0.0.X) - it's okay to create and push the next sequential tag yourself when needed for a release
-
-## PR Status Reporting
-
-Every time you run `release_checklist.py`, you MUST:
-1. Parse the output to identify ALL open PRs mentioned (lines with "✅ PR with title ... exists")
-2. Provide a summary to the user listing ALL open PRs that need review
-3. Group them by status:
-   - PRs for repositories that are blocked by dependencies (show these but note they're blocked)
-   - PRs for repositories that are ready to merge (highlight these)
-4. Format the summary clearly with PR numbers and URLs
-
-This summary should be provided EVERY time you run the checklist, not just after creating new PRs.
-The user needs to see the complete picture of what's waiting for review.
-
-## Nightly Infrastructure
-
-The nightly build system uses branches and tags across two repositories:
-
-- `leanprover/lean4` has **branches** `nightly` and `nightly-with-mathlib` tracking the latest nightly builds
-- `leanprover/lean4-nightly` has **dated tags** like `nightly-2026-01-23`
-
-When a nightly succeeds with mathlib, all three should point to the same commit. Don't confuse these: branches are in the main lean4 repo, dated tags are in lean4-nightly.
-
-## Error Handling
-
-**CRITICAL**: If something goes wrong or a command fails:
-- **DO NOT** try to manually reproduce the failing steps yourself
-- **DO NOT** try to fix things by running git commands or other manual operations
-- Both scripts are idempotent and designed to handle partial completion gracefully
-- If a script continues to fail after retrying, report the error to the user and wait for instructions

.gitattributes

@@ -4,7 +4,6 @@ RELEASES.md merge=union
 stage0/** binary linguist-generated
 # The following file is often manually edited, so do show it in diffs
 stage0/src/stdlib_flags.h -binary -linguist-generated
-doc/std/grove/GroveStdlib/Generated/** linguist-generated
 # These files should not have line endings translated on Windows, because
 # it throws off parser tests. Later lines override earlier ones, so the
 # runner code is still treated as ordinary text.

.github/ISSUE_TEMPLATE/bug_report.md

@@ -9,7 +9,7 @@ assignees: ''
 
 ### Prerequisites
 
-<!-- Please put an X between the brackets as you perform the following steps: -->
+Please put an X between the brackets as you perform the following steps:
 
 * [ ] Check that your issue is not already filed:
       https://github.com/leanprover/lean4/issues

.github/workflows/actionlint.yml

@@ -15,7 +15,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
     - name: Checkout
-      uses: actions/checkout@v6
+      uses: actions/checkout@v5
     - name: actionlint
       uses: raven-actions/actionlint@v2
       with:

.github/workflows/awaiting-manual.yml

@@ -12,7 +12,7 @@ jobs:
       - name: Check awaiting-manual label
         id: check-awaiting-manual-label
         if: github.event_name == 'pull_request'
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             const { labels, number: prNumber } = context.payload.pull_request;

.github/workflows/awaiting-mathlib.yml

@@ -12,7 +12,7 @@ jobs:
       - name: Check awaiting-mathlib label
         id: check-awaiting-mathlib-label
         if: github.event_name == 'pull_request'
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             const { labels, number: prNumber } = context.payload.pull_request;

.github/workflows/build-template.yml

@@ -3,6 +3,9 @@ name: build-template
 on:
   workflow_call:
     inputs:
+      check-level:
+        type: string
+        required: true
       config:
         type: string
         required: true
@@ -67,13 +70,13 @@ jobs:
         if: runner.os == 'macOS'
       - name: Checkout
         if: (!endsWith(matrix.os, '-with-cache'))
-        uses: actions/checkout@v6
+        uses: actions/checkout@v5
         with:
           # the default is to use a virtual merge commit between the PR and master: just use the PR
           ref: ${{ github.event.pull_request.head.sha }}
       - name: Namespace Checkout
         if: endsWith(matrix.os, '-with-cache')
-        uses: namespacelabs/nscloud-checkout-action@v8
+        uses: namespacelabs/nscloud-checkout-action@v7
         with:
           ref: ${{ github.event.pull_request.head.sha }}
       - name: Open Nix shell once
@@ -102,7 +105,7 @@ jobs:
         if: matrix.cmultilib
       - name: Restore Cache
         id: restore-cache
-        uses: actions/cache/restore@v5
+        uses: actions/cache/restore@v4
         with:
           # NOTE: must be in sync with `save` below and with `restore-cache` in `update-stage0.yml`
           path: |
@@ -113,10 +116,10 @@ jobs:
             build/stage1/**/*.ir
             build/stage1/**/*.c
             build/stage1/**/*.c.o*' || '' }}
-          key: ${{ matrix.name }}-build-v4-${{ github.sha }}
+          key: ${{ matrix.name }}-build-v3-${{ github.sha }}
           # fall back to (latest) previous cache
           restore-keys: |
-            ${{ matrix.name }}-build-v4
+            ${{ matrix.name }}-build-v3
       # open nix-shell once for initial setup
       - name: Setup
         run: |
@@ -175,7 +178,7 @@ jobs:
         # Caching on cancellation created some mysterious issues perhaps related to improper build
         # shutdown
         if: steps.restore-cache.outputs.cache-hit != 'true' && !cancelled()
-        uses: actions/cache/save@v5
+        uses: actions/cache/save@v4
         with:
           # NOTE: must be in sync with `restore` above
           path: |
@@ -213,21 +216,21 @@ jobs:
           else
             ${{ matrix.tar || 'tar' }} cf - $dir | zstd -T0 --no-progress -o pack/$dir.tar.zst
           fi
-      - uses: actions/upload-artifact@v5
+      - uses: actions/upload-artifact@v4
         if: matrix.release
         with:
           name: build-${{ matrix.name }}
           path: pack/*
       - name: Lean stats
         run: |
-          build/$TARGET_STAGE/bin/lean --stats src/Lean.lean
+          build/$TARGET_STAGE/bin/lean --stats src/Lean.lean -Dexperimental.module=true
         if: ${{ !matrix.cross }}
       - name: Test
         id: test
         run: |
           ulimit -c unlimited  # coredumps
           time ctest --preset ${{ matrix.CMAKE_PRESET || 'release' }} --test-dir build/$TARGET_STAGE -j$NPROC --output-junit test-results.xml ${{ matrix.CTEST_OPTIONS }}
-        if: matrix.test
+        if: (matrix.wasm || !matrix.cross) && (inputs.check-level >= 1 || matrix.test)
       - name: Test Summary
         uses: test-summary/action@v2
         with:

.github/workflows/check-prelude.yml

@@ -7,7 +7,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout
-        uses: actions/checkout@v6
+        uses: actions/checkout@v5
         with:
           # the default is to use a virtual merge commit between the PR and master: just use the PR
           ref: ${{ github.event.pull_request.head.sha }}

.github/workflows/check-stage0.yml

@@ -8,11 +8,11 @@ jobs:
   check-stage0-on-queue:
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/checkout@v6
+    - uses: actions/checkout@v5
       with:
         ref: ${{ github.event.pull_request.head.sha }}
+        filter: blob:none
         fetch-depth: 0
-        filter: tree:0
 
     - name: Find base commit
       if: github.event_name == 'pull_request'
@@ -31,7 +31,7 @@ jobs:
 
     - if: github.event_name == 'pull_request'
       name: Set label
-      uses: actions/github-script@v8
+      uses: actions/github-script@v7
       with:
         script: |
           const { owner, repo, number: issue_number  } = context.issue;

.github/workflows/check-stdlib-flags.yml

@@ -1,57 +0,0 @@
-name: Check stdlib_flags.h modifications
-
-on:
-  pull_request:
-    types: [opened, synchronize, reopened, labeled, unlabeled]
-
-jobs:
-  check-stdlib-flags:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Check if stdlib_flags.h was modified
-        uses: actions/github-script@v8
-        with:
-          script: |
-            // Get the list of files changed in this PR
-            const files = await github.paginate(
-              github.rest.pulls.listFiles,
-              {
-                owner: context.repo.owner,
-                repo: context.repo.repo,
-                pull_number: context.payload.pull_request.number,
-              }
-            );
-
-            // Check if stdlib_flags.h was modified
-            const stdlibFlagsModified = files.some(file =>
-              file.filename === 'src/stdlib_flags.h'
-            );
-
-            if (stdlibFlagsModified) {
-              console.log('src/stdlib_flags.h was modified in this PR');
-
-              // Check if the unlock label is present
-
-              const { data: pr } = await github.rest.pulls.get({
-                owner: context.repo.owner,
-                repo: context.repo.repo,
-                pull_number: context.issue.number,
-              });
-
-              const hasUnlockLabel = pr.labels.some(label =>
-                label.name === 'unlock-upstream-stdlib-flags'
-              );
-
-              if (!hasUnlockLabel) {
-                core.setFailed(
-                  'src/stdlib_flags.h was modified. This is likely a mistake. If you would like to change ' +
-                  'bootstrapping settings or request a stage0 update, you should modify stage0/src/stdlib_flags.h. ' +
-                  'If you really want to change src/stdlib_flags.h (which should be extremely rare), set the ' +
-                  'unlock-upstream-stdlib-flags label.'
-                );
-              } else {
-                console.log('Found unlock-upstream-stdlib-flags');
-              }
-            } else {
-              console.log('src/stdlib_flags.h was not modified');
-            }

.github/workflows/ci.yml

@@ -31,6 +31,10 @@ jobs:
   configure:
     runs-on: ubuntu-latest
     outputs:
+      # 0: PRs without special label
+      # 1: PRs with `merge-ci` label, merge queue checks, master commits
+      # 2: PRs with `release-ci` label, releases (incl. nightlies)
+      check-level: ${{ steps.set-level.outputs.check-level }}
       # The build matrix, dynamically generated here
       matrix: ${{ steps.set-matrix.outputs.matrix }}
       # secondary build jobs that should not block the CI success/merge queue
@@ -50,9 +54,9 @@ jobs:
 
     steps:
       - name: Checkout
-        uses: actions/checkout@v6
+        uses: actions/checkout@v5
         # don't schedule nightlies on forks
-        if: github.event_name == 'schedule' && github.repository == 'leanprover/lean4' || inputs.action == 'release nightly' || (startsWith(github.ref, 'refs/tags/') && github.repository == 'leanprover/lean4')
+        if: github.event_name == 'schedule' && github.repository == 'leanprover/lean4' || inputs.action == 'release nightly'
       - name: Set Nightly
         if: github.event_name == 'schedule' && github.repository == 'leanprover/lean4' || inputs.action == 'release nightly'
         id: set-nightly
@@ -60,23 +64,10 @@ jobs:
           if [[ -n '${{ secrets.PUSH_NIGHTLY_TOKEN }}' ]]; then
             git remote add nightly https://foo:'${{ secrets.PUSH_NIGHTLY_TOKEN }}'@github.com/${{ github.repository_owner }}/lean4-nightly.git
             git fetch nightly --tags
-            if [[ '${{ github.event_name }}' == 'workflow_dispatch' ]]; then
-              # Manual re-release: create a revision of the most recent nightly
-              BASE_NIGHTLY=$(git tag -l 'nightly-*' | sort -rV | head -1)
-              # Strip any existing -revK suffix to get the base date tag
-              BASE_NIGHTLY="${BASE_NIGHTLY%%-rev*}"
-              REV=1
-              while git rev-parse "refs/tags/${BASE_NIGHTLY}-rev${REV}" >/dev/null 2>&1; do
-                REV=$((REV + 1))
-              done
-              LEAN_VERSION_STRING="${BASE_NIGHTLY}-rev${REV}"
+            LEAN_VERSION_STRING="nightly-$(date -u +%F)"
+            # do nothing if commit already has a different tag
+            if [[ "$(git name-rev --name-only --tags --no-undefined HEAD 2> /dev/null || echo "$LEAN_VERSION_STRING")" == "$LEAN_VERSION_STRING" ]]; then
               echo "nightly=$LEAN_VERSION_STRING" >> "$GITHUB_OUTPUT"
-            else
-              # Scheduled: do nothing if commit already has a different tag
-              LEAN_VERSION_STRING="nightly-$(date -u +%F)"
-              if [[ "$(git name-rev --name-only --tags --no-undefined HEAD 2> /dev/null || echo "$LEAN_VERSION_STRING")" == "$LEAN_VERSION_STRING" ]]; then
-                echo "nightly=$LEAN_VERSION_STRING" >> "$GITHUB_OUTPUT"
-              fi
             fi
           fi
 
@@ -119,54 +110,6 @@ jobs:
           TAG_NAME="${GITHUB_REF##*/}"
           echo "RELEASE_TAG=$TAG_NAME" >> "$GITHUB_OUTPUT"
 
-      - name: Validate CMakeLists.txt version matches tag
-        if: steps.set-release.outputs.RELEASE_TAG != ''
-        run: |
-          echo "Validating CMakeLists.txt version matches tag ${{ steps.set-release.outputs.RELEASE_TAG }}"
-
-          # Extract version values from CMakeLists.txt
-          CMAKE_MAJOR=$(grep -E "^set\(LEAN_VERSION_MAJOR " src/CMakeLists.txt | grep -oE '[0-9]+')
-          CMAKE_MINOR=$(grep -E "^set\(LEAN_VERSION_MINOR " src/CMakeLists.txt | grep -oE '[0-9]+')
-          CMAKE_PATCH=$(grep -E "^set\(LEAN_VERSION_PATCH " src/CMakeLists.txt | grep -oE '[0-9]+')
-          CMAKE_IS_RELEASE=$(grep -m 1 -E "^set\(LEAN_VERSION_IS_RELEASE " src/CMakeLists.txt | sed -nE 's/^set\(LEAN_VERSION_IS_RELEASE ([0-9]+)\).*/\1/p')
-
-          # Expected values from tag parsing
-          TAG_MAJOR="${{ steps.set-release.outputs.LEAN_VERSION_MAJOR }}"
-          TAG_MINOR="${{ steps.set-release.outputs.LEAN_VERSION_MINOR }}"
-          TAG_PATCH="${{ steps.set-release.outputs.LEAN_VERSION_PATCH }}"
-
-          ERRORS=""
-
-          if [[ "$CMAKE_MAJOR" != "$TAG_MAJOR" ]]; then
-            ERRORS+="LEAN_VERSION_MAJOR: expected $TAG_MAJOR, found $CMAKE_MAJOR\n"
-          fi
-          if [[ "$CMAKE_MINOR" != "$TAG_MINOR" ]]; then
-            ERRORS+="LEAN_VERSION_MINOR: expected $TAG_MINOR, found $CMAKE_MINOR\n"
-          fi
-          if [[ "$CMAKE_PATCH" != "$TAG_PATCH" ]]; then
-            ERRORS+="LEAN_VERSION_PATCH: expected $TAG_PATCH, found $CMAKE_PATCH\n"
-          fi
-          if [[ "$CMAKE_IS_RELEASE" != "1" ]]; then
-            ERRORS+="LEAN_VERSION_IS_RELEASE: expected 1, found $CMAKE_IS_RELEASE\n"
-          fi
-
-          if [[ -n "$ERRORS" ]]; then
-            echo "::error::Version mismatch between tag and src/CMakeLists.txt"
-            echo ""
-            echo "Tag ${{ steps.set-release.outputs.RELEASE_TAG }} expects version $TAG_MAJOR.$TAG_MINOR.$TAG_PATCH"
-            echo "But src/CMakeLists.txt has mismatched values:"
-            echo -e "$ERRORS"
-            echo ""
-            echo "Fix src/CMakeLists.txt, delete the tag, and re-tag."
-            exit 1
-          fi
-
-          echo "Version validation passed: $TAG_MAJOR.$TAG_MINOR.$TAG_PATCH"
-
-      # 0: PRs without special label
-      # 1: PRs with `merge-ci` label, merge queue checks, master commits
-      # 2: nightlies
-      # 3: PRs with `release-ci` label, full releases
       - name: Set check level
         id: set-level
         # We do not use github.event.pull_request.labels.*.name here because
@@ -174,39 +117,31 @@ jobs:
         # rerun the workflow run after setting the `release-ci`/`merge-ci` labels.
         run: |
           check_level=0
-          fast=false
 
-          if [[ -n "${{ steps.set-release.outputs.RELEASE_TAG }}" || -n "${{ steps.set-release-custom.outputs.RELEASE_TAG }}" ]]; then
-            check_level=3
-          elif [[ -n "${{ steps.set-nightly.outputs.nightly }}" ]]; then
+          if [[ -n "${{ steps.set-nightly.outputs.nightly }}" || -n "${{ steps.set-release.outputs.RELEASE_TAG }}" || -n "${{ steps.set-release-custom.outputs.RELEASE_TAG }}" ]]; then
             check_level=2
           elif [[ "${{ github.event_name }}" != "pull_request" ]]; then
             check_level=1
           else
             labels="$(gh api repos/${{ github.repository_owner }}/${{ github.event.repository.name }}/pulls/${{ github.event.pull_request.number }} --jq '.labels')"
             if echo "$labels" | grep -q "release-ci"; then
-              check_level=3
+              check_level=2
             elif echo "$labels" | grep -q "merge-ci"; then
               check_level=1
             fi
-            if echo "$labels" | grep -q "fast-ci"; then
-              fast=true
-            fi
           fi
 
           echo "check-level=$check_level" >> "$GITHUB_OUTPUT"
-          echo "fast=$fast" >> "$GITHUB_OUTPUT"
         env:
           GH_TOKEN: ${{ github.token }}
 
       - name: Configure build matrix
         id: set-matrix
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             const level = ${{ steps.set-level.outputs.check-level }};
-            const fast = ${{ steps.set-level.outputs.fast }};
-            console.log(`level: ${level}, fast: ${fast}`);
+            console.log(`level: ${level}`);
             // use large runners where available (original repo)
             let large = ${{ github.repository == 'leanprover/lean4' }};
             const isPr = "${{ github.event_name }}" == "pull_request";
@@ -217,8 +152,7 @@ jobs:
                 "name": "Linux LLVM",
                 "os": "ubuntu-latest",
                 "release": false,
-                "enabled": level >= 2,
-                "test": true,
+                "check-level": 2,
                 "shell": "nix develop .#oldGlibc -c bash -euxo pipefail {0}",
                 "llvm-url": "https://github.com/leanprover/lean-llvm/releases/download/19.1.2/lean-llvm-x86_64-linux-gnu.tar.zst",
                 "prepare-llvm": "../script/prepare-llvm-linux.sh lean-llvm*",
@@ -231,19 +165,17 @@ jobs:
               {
                 // portable release build: use channel with older glibc (2.26)
                 "name": "Linux release",
-                // usually not a bottleneck so make exclusive to `fast-ci`
-                "os": large && fast ? "nscloud-ubuntu-22.04-amd64-8x16-with-cache" : "ubuntu-latest",
+                "os": "ubuntu-latest",
                 "release": true,
                 // Special handling for release jobs. We want:
-                // 1. To run it in PRs so developers get PR toolchains (so secondary without tests is sufficient)
+                // 1. To run it in PRs so developers get PR toolchains (so secondary is sufficient)
                 // 2. To skip it in merge queues as it takes longer than the
                 //    Linux lake build and adds little value in the merge queue
                 // 3. To run it in release (obviously)
                 // 4. To run it for pushes to master so that pushes to master have a Linux toolchain
                 //    available as an artifact for Grove to use.
-                "enabled": isPr || level != 1 || isPushToMaster,
-                "test": level >= 1,
-                "secondary": level == 0,
+                "check-level": (isPr || isPushToMaster) ? 0 : 2,
+                "secondary": isPr,
                 "shell": "nix develop .#oldGlibc -c bash -euxo pipefail {0}",
                 "llvm-url": "https://github.com/leanprover/lean-llvm/releases/download/19.1.2/lean-llvm-x86_64-linux-gnu.tar.zst",
                 "prepare-llvm": "../script/prepare-llvm-linux.sh lean-llvm*",
@@ -254,50 +186,36 @@ jobs:
               {
                 "name": "Linux Lake",
                 "os": large ? "nscloud-ubuntu-22.04-amd64-8x16-with-cache" : "ubuntu-latest",
-                "enabled": true,
+                "check-level": 0,
+                "test": true,
                 "check-rebootstrap": level >= 1,
                 "check-stage3": level >= 2,
-                "test": true,
                 // NOTE: `test-speedcenter` currently seems to be broken on `ubuntu-latest`
                 "test-speedcenter": large && level >= 2,
-                // We are not warning-free yet on all platforms, start here
-                "CMAKE_OPTIONS": "-DLEAN_EXTRA_CXX_FLAGS=-Werror",
+                // made explicit until it can be assumed to have propagated to PRs
+                "CMAKE_OPTIONS": "-DUSE_LAKE=ON",
               },
               {
                 "name": "Linux Reldebug",
                 "os": "ubuntu-latest",
-                "enabled": level >= 2,
-                "test": true,
+                "check-level": 2,
                 "CMAKE_PRESET": "reldebug",
               },
-              {
+              // TODO: suddenly started failing in CI
+              /*{
                 "name": "Linux fsanitize",
-                // Always run on large if available, more reliable regarding timeouts
-                "os": large ? "nscloud-ubuntu-22.04-amd64-16x32-with-cache" : "ubuntu-latest",
-                "enabled": level >= 2,
-                // do not fail nightlies on this for now
-                "secondary": level <= 2,
-                "test": true,
+                "os": "ubuntu-latest",
+                "check-level": 2,
                 // turn off custom allocator & symbolic functions to make LSAN do its magic
                 "CMAKE_PRESET": "sanitize",
-                // * `StackOverflow*` correctly triggers ubsan.
-                // * `reverse-ffi` fails to link in sanitizers.
-                // * `interactive` and `async_select_channel` fail nondeterministically, would need
-                //   to be investigated..
-                // * 9366 is too close to timeout.
-                // * `bv_` sometimes times out calling into cadical even though we should be using
-                //   the standard compile flags for it.
-                // * `grind_guide` always times out.
-                // * `pkg/|lake/` tests sometimes time out (likely even hang), related to Lake CI
-                //   failures?
-                "CTEST_OPTIONS": "-E 'StackOverflow|reverse-ffi|interactive|async_select_channel|9366|run/bv_|grind_guide|grind_bitvec2|grind_constProp|grind_indexmap|grind_list|grind_lint|grind_array_attach|grind_ite_trace|pkg/|lake/'"
-              },
+                // exclude seriously slow/problematic tests (laketests crash)
+                "CTEST_OPTIONS": "-E 'interactivetest|leanpkgtest|laketest|benchtest'"
+              },*/
               {
                 "name": "macOS",
-                "os": "macos-15-intel",
+                "os": "macos-13",
                 "release": true,
-                "test": false,  // Tier 2 platform
-                "enabled": level >= 2,
+                "check-level": 2,
                 "shell": "bash -euxo pipefail {0}",
                 "llvm-url": "https://github.com/leanprover/lean-llvm/releases/download/19.1.2/lean-llvm-x86_64-apple-darwin.tar.zst",
                 "prepare-llvm": "../script/prepare-llvm-macos.sh lean-llvm*",
@@ -308,7 +226,7 @@ jobs:
               {
                 "name": "macOS aarch64",
                 // standard GH runner only comes with 7GB so use large runner if possible when running tests
-                "os": large && (fast || level >= 1) ? "nscloud-macos-sequoia-arm64-6x14" : "macos-15",
+                "os": large && !isPr ? "nscloud-macos-sonoma-arm64-6x14" : "macos-14",
                 "CMAKE_OPTIONS": "-DLEAN_INSTALL_SUFFIX=-darwin_aarch64",
                 "release": true,
                 "shell": "bash -euxo pipefail {0}",
@@ -317,16 +235,14 @@ jobs:
                 "binary-check": "otool -L",
                 "tar": "gtar", // https://github.com/actions/runner-images/issues/2619
                 // See "Linux release" for release job levels; Grove is not a concern here
-                "enabled": isPr || level != 1,
-                "test": level >= 1,
-                "secondary": level == 0,
+                "check-level": isPr ? 0 : 2,
+                "secondary": isPr,
               },
               {
                 "name": "Windows",
-                "os": large && (fast || level >= 2) ? "namespace-profile-windows-amd64-4x16" : "windows-2022",
+                "os": large && level == 2 ? "namespace-profile-windows-amd64-4x16" : "windows-2022",
                 "release": true,
-                "enabled": level >= 2,
-                "test": true,
+                "check-level": 2,
                 "shell": "msys2 {0}",
                 "CMAKE_OPTIONS": "-G \"Unix Makefiles\"",
                 "llvm-url": "https://github.com/leanprover/lean-llvm/releases/download/19.1.2/lean-llvm-x86_64-w64-windows-gnu.tar.zst",
@@ -338,8 +254,7 @@ jobs:
                 "os": "nscloud-ubuntu-22.04-arm64-4x16",
                 "CMAKE_OPTIONS": "-DLEAN_INSTALL_SUFFIX=-linux_aarch64",
                 "release": true,
-                "enabled": level >= 2,
-                "test": true,
+                "check-level": 2,
                 "shell": "nix develop .#oldGlibcAArch -c bash -euxo pipefail {0}",
                 "llvm-url": "https://github.com/leanprover/lean-llvm/releases/download/19.1.2/lean-llvm-aarch64-linux-gnu.tar.zst",
                 "prepare-llvm": "../script/prepare-llvm-linux.sh lean-llvm*",
@@ -352,7 +267,7 @@ jobs:
               //  "CMAKE_OPTIONS": "-DSTAGE0_USE_GMP=OFF -DSTAGE0_LEAN_EXTRA_CXX_FLAGS='-m32' -DSTAGE0_LEANC_OPTS='-m32' -DSTAGE0_MMAP=OFF -DUSE_GMP=OFF -DLEAN_EXTRA_CXX_FLAGS='-m32' -DLEANC_OPTS='-m32' -DMMAP=OFF -DLEAN_INSTALL_SUFFIX=-linux_x86 -DCMAKE_LIBRARY_PATH=/usr/lib/i386-linux-gnu/ -DSTAGE0_CMAKE_LIBRARY_PATH=/usr/lib/i386-linux-gnu/ -DPKG_CONFIG_EXECUTABLE=/usr/bin/i386-linux-gnu-pkg-config",
               //  "cmultilib": true,
               //  "release": true,
-              //  "enabled": level >= 2,
+              //  "check-level": 2,
               //  "cross": true,
               //  "shell": "bash -euxo pipefail {0}"
               //}
@@ -364,7 +279,7 @@ jobs:
               //   "wasm": true,
               //   "cmultilib": true,
               //   "release": true,
-              //   "enabled": level >= 2,
+              //   "check-level": 2,
               //   "cross": true,
               //   "shell": "bash -euxo pipefail {0}",
               //   // Just a few selected tests because wasm is slow
@@ -378,7 +293,7 @@ jobs:
               }
             }
             console.log(`matrix:\n${JSON.stringify(matrix, null, 2)}`);
-            matrix = matrix.filter((job) => job["enabled"]);
+            matrix = matrix.filter((job) => level >= job["check-level"]);
             core.setOutput('matrix', matrix.filter((job) => !job["secondary"]));
             core.setOutput('matrix-secondary', matrix.filter((job) => job["secondary"]));
 
@@ -388,6 +303,7 @@ jobs:
     uses: ./.github/workflows/build-template.yml
     with:
       config: ${{needs.configure.outputs.matrix}}
+      check-level: ${{ needs.configure.outputs.check-level }}
       nightly: ${{ needs.configure.outputs.nightly }}
       LEAN_VERSION_MAJOR: ${{ needs.configure.outputs.LEAN_VERSION_MAJOR }}
       LEAN_VERSION_MINOR: ${{ needs.configure.outputs.LEAN_VERSION_MINOR }}
@@ -403,6 +319,7 @@ jobs:
     uses: ./.github/workflows/build-template.yml
     with:
       config: ${{needs.configure.outputs.matrix-secondary}}
+      check-level: ${{ needs.configure.outputs.check-level }}
       nightly: ${{ needs.configure.outputs.nightly }}
       LEAN_VERSION_MAJOR: ${{ needs.configure.outputs.LEAN_VERSION_MAJOR }}
       LEAN_VERSION_MINOR: ${{ needs.configure.outputs.LEAN_VERSION_MINOR }}
@@ -433,7 +350,7 @@ jobs:
         content: |
           A build of `${{ github.ref_name }}`, triggered by event `${{ github.event_name }}`, [failed](https://github.com/${{ github.repository }}/actions/runs/${{ github.run_id }}).
     - if: contains(needs.*.result, 'failure')
-      uses: actions/github-script@v8
+      uses: actions/github-script@v7
       with:
         script: |
             core.setFailed('Some jobs failed')
@@ -446,11 +363,11 @@ jobs:
     runs-on: ubuntu-latest
     needs: build
     steps:
-      - uses: actions/download-artifact@v7
+      - uses: actions/download-artifact@v5
         with:
           path: artifacts
       - name: Release
-        uses: softprops/action-gh-release@a06a81a03ee405af7f2048a818ed3f03bbf83c7b
+        uses: softprops/action-gh-release@72f2c25fcb47643c292f7107632f7a47c1df5cd8
         with:
           files: artifacts/*/*
           fail_on_unmatched_files: true
@@ -471,14 +388,12 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout
-        uses: actions/checkout@v6
+        uses: actions/checkout@v5
         with:
           # needed for tagging
           fetch-depth: 0
-          # Doesn't seem to be working when additionally fetching from lean4-nightly
-          #filter: tree:0
           token: ${{ secrets.PUSH_NIGHTLY_TOKEN }}
-      - uses: actions/download-artifact@v7
+      - uses: actions/download-artifact@v5
         with:
           path: artifacts
       - name: Prepare Nightly Release
@@ -488,7 +403,7 @@ jobs:
           git tag "${{ needs.configure.outputs.nightly }}"
           git push nightly "${{ needs.configure.outputs.nightly }}"
           git push -f origin refs/tags/${{ needs.configure.outputs.nightly }}:refs/heads/nightly
-          last_tag="$(git log HEAD^ --simplify-by-decoration --pretty="format:%d" | grep -o "nightly-[^ ,)]*" | head -n 1)"
+          last_tag="$(git log HEAD^ --simplify-by-decoration --pretty="format:%d" | grep -o "nightly-[-0-9]*" | head -n 1)"
           echo -e "*Changes since ${last_tag}:*\n\n" > diff.md
           git show "$last_tag":RELEASES.md > old.md
           #./script/diff_changelogs.py old.md doc/changes.md >> diff.md
@@ -496,7 +411,7 @@ jobs:
           echo -e "\n*Full commit log*\n" >> diff.md
           git log --oneline "$last_tag"..HEAD | sed 's/^/* /' >> diff.md
       - name: Release Nightly
-        uses: softprops/action-gh-release@a06a81a03ee405af7f2048a818ed3f03bbf83c7b
+        uses: softprops/action-gh-release@72f2c25fcb47643c292f7107632f7a47c1df5cd8
         with:
           body_path: diff.md
           prerelease: true
@@ -511,18 +426,8 @@ jobs:
           gh workflow -R leanprover/release-index run update-index.yml
         env:
           GITHUB_TOKEN: ${{ secrets.RELEASE_INDEX_TOKEN }}
-      - name: Generate mathlib nightly-testing app token
-        id: mathlib-app-token
-        uses: actions/create-github-app-token@29824e69f54612133e76f7eaac726eef6c875baf # v2.2.1
-        continue-on-error: true
-        with:
-          app-id: ${{ secrets.MATHLIB_NIGHTLY_TESTING_APP_ID }}
-          private-key: ${{ secrets.MATHLIB_NIGHTLY_TESTING_PRIVATE_KEY }}
-          owner: leanprover-community
-          repositories: mathlib4-nightly-testing
       - name: Update toolchain on mathlib4's nightly-testing branch
-        if: steps.mathlib-app-token.outcome == 'success'
         run: |
-          gh workflow -R leanprover-community/mathlib4-nightly-testing run nightly_bump_and_merge.yml
+          gh workflow -R leanprover-community/mathlib4-nightly-testing run nightly_bump_toolchain.yml
         env:
-          GITHUB_TOKEN: ${{ steps.mathlib-app-token.outputs.token }}
+          GITHUB_TOKEN: ${{ secrets.MATHLIB4_BOT }}

.github/workflows/copyright-header.yml

@@ -6,7 +6,7 @@ jobs:
   check-lean-files:
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/checkout@v6
+    - uses: actions/checkout@v5
 
     - name: Verify .lean files start with a copyright header.
       run: |

.github/workflows/grove.yml

@@ -51,7 +51,7 @@ jobs:
       - name: Fetch upstream invalidated facts
         if: ${{ steps.should-run.outputs.should-run == 'true' && steps.workflow-info.outputs.pullRequestNumber != '' }}
         id: fetch-upstream
-        uses: TwoFx/grove-action/fetch-upstream@v0.5
+        uses: TwoFx/grove-action/fetch-upstream@v0.4
         with:
           artifact-name: grove-invalidated-facts
           base-ref: master
@@ -65,7 +65,6 @@ jobs:
           workflow: ci.yml
           path: artifacts
           name: "build-Linux release"
-          allow_forks: true
           name_is_regexp: true
 
       - name: Unpack toolchain
@@ -96,7 +95,7 @@ jobs:
       - name: Build
         if: ${{ steps.should-run.outputs.should-run == 'true' }}
         id: build
-        uses: TwoFx/grove-action/build@v0.5
+        uses: TwoFx/grove-action/build@v0.4
         with:
           project-path: doc/std/grove
           script-name: grove-stdlib
@@ -111,7 +110,7 @@ jobs:
       # material.
       - id: deploy-alias
         if: ${{ steps.should-run.outputs.should-run == 'true' }}
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         name: Compute Alias
         with:
           result-encoding: string

.github/workflows/labels-from-comments.yml

@@ -17,7 +17,7 @@ jobs:
 
     steps:
     - name: Add label based on comment
-      uses: actions/github-script@v8
+      uses: actions/github-script@v7
       with:
         github-token: ${{ secrets.GITHUB_TOKEN }}
         script: |

.github/workflows/pr-body.yml

@@ -11,7 +11,7 @@ jobs:
     steps:
       - name: Check PR body
         if: github.event_name == 'pull_request'
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             const { title, body, labels, draft } = context.payload.pull_request;

.github/workflows/pr-release.yml

@@ -20,9 +20,7 @@ on:
 jobs:
   on-success:
     runs-on: ubuntu-latest
-    # Run even if CI fails, as long as build artifacts are available
-    # The "Verify release artifacts exist" step will fail if necessary artifacts are missing
-    if: github.event.workflow_run.event == 'pull_request' && github.repository == 'leanprover/lean4'
+    if: github.event.workflow_run.conclusion == 'success' && github.event.workflow_run.event == 'pull_request' && github.repository == 'leanprover/lean4'
     steps:
       - name: Retrieve information about the original workflow
         uses: potiuk/get-workflow-origin@v1_1 # https://github.com/marketplace/actions/get-workflow-origin
@@ -43,19 +41,6 @@ jobs:
           name: build-.*
           name_is_regexp: true
 
-      # Verify artifacts were downloaded before any side effects (tag creation, release deletion).
-      - name: Verify release artifacts exist
-        if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
-        run: |
-          shopt -s nullglob
-          files=(artifacts/*/*)
-          if [ ${#files[@]} -eq 0 ]; then
-            echo "::error::No artifacts found matching artifacts/*/*"
-            exit 1
-          fi
-          echo "Found ${#files[@]} artifacts to upload:"
-          printf '%s\n' "${files[@]}"
-
       - name: Push tag
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
         run: |
@@ -63,62 +48,60 @@ jobs:
           git -C lean4.git remote add origin https://github.com/${{ github.repository_owner }}/lean4.git
           git -C lean4.git fetch -n origin master
           git -C lean4.git fetch -n origin "${{ steps.workflow-info.outputs.sourceHeadSha }}"
-
+          
           # Create both the original tag and the SHA-suffixed tag
           SHORT_SHA="${{ steps.workflow-info.outputs.sourceHeadSha }}"
           SHORT_SHA="${SHORT_SHA:0:7}"
-
+          
           # Export the short SHA for use in subsequent steps
           echo "SHORT_SHA=${SHORT_SHA}" >> "$GITHUB_ENV"
 
           git -C lean4.git tag -f pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }} "${{ steps.workflow-info.outputs.sourceHeadSha }}"
           git -C lean4.git tag -f pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-"${SHORT_SHA}" "${{ steps.workflow-info.outputs.sourceHeadSha }}"
-
+          
           git -C lean4.git remote add pr-releases https://foo:'${{ secrets.PR_RELEASES_TOKEN }}'@github.com/${{ github.repository_owner }}/lean4-pr-releases.git
           git -C lean4.git push -f pr-releases pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}
           git -C lean4.git push -f pr-releases pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-"${SHORT_SHA}"
-      - name: Delete existing releases if present
+      - name: Delete existing release if present
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
         run: |
-          # Delete any existing releases for this PR.
-          # The short format release is always recreated with the latest commit.
-          # The SHA-suffixed release should be unique per commit, but delete just in case.
+          # Try to delete any existing release for the current PR (just the version without the SHA suffix).
           gh release delete --repo ${{ github.repository_owner }}/lean4-pr-releases pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }} -y || true
-          gh release delete --repo ${{ github.repository_owner }}/lean4-pr-releases pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-${{ env.SHORT_SHA }} -y || true
         env:
           GH_TOKEN: ${{ secrets.PR_RELEASES_TOKEN }}
-      # We use `gh release create` instead of `softprops/action-gh-release` because
-      # the latter enumerates all releases to check for existing ones, which fails
-      # when the repository has more than 10000 releases (GitHub API pagination limit).
-      # Upstream fix: https://github.com/softprops/action-gh-release/pull/725
       - name: Release (short format)
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
-        run: |
-          # There are coredump files in deeper subdirectories; artifacts/*/* gets the release archives.
-          gh release create \
-            --repo ${{ github.repository_owner }}/lean4-pr-releases \
-            --title "Release for PR ${{ steps.workflow-info.outputs.pullRequestNumber }}" \
-            --notes "" \
-            pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }} \
-            artifacts/*/*
+        uses: softprops/action-gh-release@72f2c25fcb47643c292f7107632f7a47c1df5cd8
+        with:
+          name: Release for PR ${{ steps.workflow-info.outputs.pullRequestNumber }}
+          # There are coredumps files here as well, but all in deeper subdirectories.
+          files: artifacts/*/*
+          fail_on_unmatched_files: true
+          draft: false
+          tag_name: pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}
+          repository: ${{ github.repository_owner }}/lean4-pr-releases
         env:
-          GH_TOKEN: ${{ secrets.PR_RELEASES_TOKEN }}
+          # The token used here must have `workflow` privileges.
+          GITHUB_TOKEN: ${{ secrets.PR_RELEASES_TOKEN }}
 
       - name: Release (SHA-suffixed format)
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
-        run: |
-          gh release create \
-            --repo ${{ github.repository_owner }}/lean4-pr-releases \
-            --title "Release for PR ${{ steps.workflow-info.outputs.pullRequestNumber }} (${{ steps.workflow-info.outputs.sourceHeadSha }})" \
-            --notes "" \
-            pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-${{ env.SHORT_SHA }} \
-            artifacts/*/*
+        uses: softprops/action-gh-release@72f2c25fcb47643c292f7107632f7a47c1df5cd8
+        with:
+          name: Release for PR ${{ steps.workflow-info.outputs.pullRequestNumber }} (${{ steps.workflow-info.outputs.sourceHeadSha }})
+          # There are coredumps files here as well, but all in deeper subdirectories.
+          files: artifacts/*/*
+          fail_on_unmatched_files: true
+          draft: false
+          tag_name: pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-${{ env.SHORT_SHA }}
+          repository: ${{ github.repository_owner }}/lean4-pr-releases
         env:
-          GH_TOKEN: ${{ secrets.PR_RELEASES_TOKEN }}
+          # The token used here must have `workflow` privileges.
+          GITHUB_TOKEN: ${{ secrets.PR_RELEASES_TOKEN }}
 
       - name: Report release status (short format)
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             await github.rest.repos.createCommitStatus({
@@ -132,7 +115,7 @@ jobs:
 
       - name: Report release status (SHA-suffixed format)
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             await github.rest.repos.createCommitStatus({
@@ -144,9 +127,9 @@ jobs:
               description: "${{ github.repository_owner }}/lean4-pr-releases:pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-${{ env.SHORT_SHA }}",
             });
 
-      - name: Add toolchain-available label
+      - name: Add label
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             await github.rest.issues.addLabels({
@@ -170,18 +153,6 @@ jobs:
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
         uses: dcarbone/install-jq-action@v3.2.0
 
-      # Generate a token for posting comments to Lean PRs about mathlib compatibility.
-      # This app is in the leanprover org and installed on leanprover/lean4.
-      - name: Generate GitHub App token for Lean PR comments
-        if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
-        id: mathlib-comment-token
-        uses: actions/create-github-app-token@3ff1caaa28b64c9cc276ce0a02e2ff584f3900c5 # v2.0.2
-        with:
-          app-id: ${{ secrets.MATHLIB_LEAN_PR_TESTING_APP_ID }}
-          private-key: ${{ secrets.MATHLIB_LEAN_PR_TESTING_PRIVATE_KEY }}
-          owner: leanprover
-          repositories: lean4
-
       # Check that the most recently nightly coincides with 'git merge-base HEAD master'
       - name: Check merge-base and nightly-testing-YYYY-MM-DD for Mathlib/Batteries
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' }}
@@ -195,14 +166,22 @@ jobs:
           if [ "$NIGHTLY_SHA" = "$MERGE_BASE_SHA" ]; then
             echo "The merge base of this PR coincides with the nightly release"
 
+            BATTERIES_REMOTE_TAGS="$(git ls-remote https://github.com/leanprover-community/batteries.git nightly-testing-"$MOST_RECENT_NIGHTLY")"
             MATHLIB_REMOTE_TAGS="$(git ls-remote https://github.com/leanprover-community/mathlib4-nightly-testing.git nightly-testing-"$MOST_RECENT_NIGHTLY")"
 
-            if [[ -n "$MATHLIB_REMOTE_TAGS" ]]; then
-              echo "... and Mathlib has a 'nightly-testing-$MOST_RECENT_NIGHTLY' tag."
+            if [[ -n "$BATTERIES_REMOTE_TAGS" ]]; then
+              echo "... and Batteries has a 'nightly-testing-$MOST_RECENT_NIGHTLY' tag."
               MESSAGE=""
+
+              if [[ -n "$MATHLIB_REMOTE_TAGS" ]]; then
+                echo "... and Mathlib has a 'nightly-testing-$MOST_RECENT_NIGHTLY' tag."
+              else
+                echo "... but Mathlib does not yet have a 'nightly-testing-$MOST_RECENT_NIGHTLY' tag."
+                MESSAGE="- ❗ Mathlib CI can not be attempted yet, as the \`nightly-testing-$MOST_RECENT_NIGHTLY\` tag does not exist there yet. We will retry when you push more commits. If you rebase your branch onto \`nightly-with-mathlib\`, Mathlib CI should run now."
+              fi
             else
-              echo "... but Mathlib does not yet have a 'nightly-testing-$MOST_RECENT_NIGHTLY' tag."
-              MESSAGE="- ❗ Mathlib CI can not be attempted yet, as the \`nightly-testing-$MOST_RECENT_NIGHTLY\` tag does not exist there yet. We will retry when you push more commits. If you rebase your branch onto \`nightly-with-mathlib\`, Mathlib CI should run now."
+              echo "... but Batteries does not yet have a 'nightly-testing-$MOST_RECENT_NIGHTLY' tag."
+              MESSAGE="- ❗ Batteries CI can not be attempted yet, as the \`nightly-testing-$MOST_RECENT_NIGHTLY\` tag does not exist there yet. We will retry when you push more commits. If you rebase your branch onto \`nightly-with-mathlib\`, Batteries CI should run now."
             fi
           else
             echo "The most recently nightly tag on this branch has SHA: $NIGHTLY_SHA"
@@ -216,13 +195,12 @@ jobs:
 
           if [[ -n "$MESSAGE" ]]; then
             # Check if force-mathlib-ci label is present
-            # Use GITHUB_TOKEN for read-only label fetch (MATHLIB4_COMMENT_BOT is only for posting comments)
             LABELS="$(curl --retry 3 --location --silent \
-                          -H "Authorization: token ${{ secrets.GITHUB_TOKEN }}" \
+                          -H "Authorization: token ${{ secrets.MATHLIB4_COMMENT_BOT }}" \
                           -H "Accept: application/vnd.github.v3+json" \
                           "https://api.github.com/repos/leanprover/lean4/issues/${{ steps.workflow-info.outputs.pullRequestNumber }}/labels" \
                           | jq -r '.[].name')"
-
+            
             if echo "$LABELS" | grep -q "^force-mathlib-ci$"; then
               echo "force-mathlib-ci label detected, forcing CI despite issues"
               MESSAGE="Forcing Mathlib CI because the \`force-mathlib-ci\` label is present, despite problem: $MESSAGE"
@@ -239,10 +217,10 @@ jobs:
 
             # Use GitHub API to check if a comment already exists
             existing_comment="$(curl --retry 3 --location --silent \
-                                    -H "Authorization: token ${{ steps.mathlib-comment-token.outputs.token }}" \
+                                    -H "Authorization: token ${{ secrets.MATHLIB4_COMMENT_BOT }}" \
                                     -H "Accept: application/vnd.github.v3+json" \
                                     "https://api.github.com/repos/leanprover/lean4/issues/${{ steps.workflow-info.outputs.pullRequestNumber }}/comments" \
-                                    | jq 'first(.[] | select(.body | test("^- . Mathlib") or startswith("Mathlib CI status")) | select(.user.login == "mathlib-lean-pr-testing[bot]"))')"
+                                    | jq 'first(.[] | select(.body | test("^- . Mathlib") or startswith("Mathlib CI status")) | select(.user.login == "leanprover-community-bot"))')"
             existing_comment_id="$(echo "$existing_comment" | jq -r .id)"
             existing_comment_body="$(echo "$existing_comment" | jq -r .body)"
 
@@ -252,14 +230,14 @@ jobs:
               echo "Posting message to the comments: $MESSAGE"
 
               # Append new result to the existing comment or post a new comment
-              # Use the mathlib-lean-pr-testing app token so Mathlib CI can subsequently edit the comment.
+              # It's essential we use the MATHLIB4_COMMENT_BOT token here, so that Mathlib CI can subsequently edit the comment.
               if [ -z "$existing_comment_id" ]; then
                 INTRO="Mathlib CI status ([docs](https://leanprover-community.github.io/contribute/tags_and_branches.html)):"
                 # Post new comment with a bullet point
                 echo "Posting as new comment at leanprover/lean4/issues/${{ steps.workflow-info.outputs.pullRequestNumber }}/comments"
                 curl -L -s \
                   -X POST \
-                  -H "Authorization: token ${{ steps.mathlib-comment-token.outputs.token }}" \
+                  -H "Authorization: token ${{ secrets.MATHLIB4_COMMENT_BOT }}" \
                   -H "Accept: application/vnd.github.v3+json" \
                   -d "$(jq --null-input --arg intro "$INTRO" --arg val "$MESSAGE" '{"body":($intro + "\n" + $val)}')" \
                   "https://api.github.com/repos/leanprover/lean4/issues/${{ steps.workflow-info.outputs.pullRequestNumber }}/comments"
@@ -268,7 +246,7 @@ jobs:
                 echo "Appending to existing comment at leanprover/lean4/issues/${{ steps.workflow-info.outputs.pullRequestNumber }}/comments"
                 curl -L -s \
                   -X PATCH \
-                  -H "Authorization: token ${{ steps.mathlib-comment-token.outputs.token }}" \
+                  -H "Authorization: token ${{ secrets.MATHLIB4_COMMENT_BOT }}" \
                   -H "Accept: application/vnd.github.v3+json" \
                   -d "$(jq --null-input --arg existing "$existing_comment_body" --arg message "$MESSAGE" '{"body":($existing + "\n" + $message)}')" \
                   "https://api.github.com/repos/leanprover/lean4/issues/comments/$existing_comment_id"
@@ -323,7 +301,7 @@ jobs:
                           -H "Accept: application/vnd.github.v3+json" \
                           "https://api.github.com/repos/leanprover/lean4/issues/${{ steps.workflow-info.outputs.pullRequestNumber }}/labels" \
                           | jq -r '.[].name')"
-
+            
             if echo "$LABELS" | grep -q "^force-manual-ci$"; then
               echo "force-manual-ci label detected, forcing CI despite issues"
               MESSAGE="Forcing reference manual CI because the \`force-manual-ci\` label is present, despite problem: $MESSAGE"
@@ -390,7 +368,7 @@ jobs:
 
       - name: Report mathlib base
         if: ${{ steps.workflow-info.outputs.pullRequestNumber != '' && steps.ready.outputs.mathlib_ready == 'true' }}
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             const description =
@@ -409,18 +387,6 @@ jobs:
       # We next automatically create a Batteries branch using this toolchain.
       # Batteries doesn't itself have a mechanism to report results of CI from this branch back to Lean
       # Instead this is taken care of by Mathlib CI, which will fail if Batteries fails.
-
-      # Generate a token from the mathlib-nightly-testing GitHub App for cross-org access
-      - name: Generate GitHub App token for leanprover-community repos
-        if: steps.workflow-info.outputs.pullRequestNumber != '' && steps.ready.outputs.mathlib_ready == 'true'
-        id: mathlib-app-token
-        uses: actions/create-github-app-token@3ff1caaa28b64c9cc276ce0a02e2ff584f3900c5 # v2.0.2
-        with:
-          app-id: ${{ secrets.MATHLIB_NIGHTLY_TESTING_APP_ID }}
-          private-key: ${{ secrets.MATHLIB_NIGHTLY_TESTING_PRIVATE_KEY }}
-          owner: leanprover-community
-          repositories: batteries,mathlib4-nightly-testing
-
       - name: Cleanup workspace
         if: steps.workflow-info.outputs.pullRequestNumber != '' && steps.ready.outputs.mathlib_ready == 'true'
         run: |
@@ -429,13 +395,12 @@ jobs:
       # Checkout the Batteries repository with all branches
       - name: Checkout Batteries repository
         if: steps.workflow-info.outputs.pullRequestNumber != '' && steps.ready.outputs.mathlib_ready == 'true'
-        uses: actions/checkout@v6
+        uses: actions/checkout@v5
         with:
           repository: leanprover-community/batteries
-          token: ${{ steps.mathlib-app-token.outputs.token }}
+          token: ${{ secrets.MATHLIB4_BOT }}
           ref: nightly-testing
           fetch-depth: 0 # This ensures we check out all tags and branches.
-          filter: tree:0
 
       - name: Check if tag exists
         if: steps.workflow-info.outputs.pullRequestNumber != '' && steps.ready.outputs.mathlib_ready == 'true'
@@ -460,7 +425,7 @@ jobs:
             git switch -c lean-pr-testing-${{ steps.workflow-info.outputs.pullRequestNumber }} "$BASE"
             echo "leanprover/lean4-pr-releases:pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-${{ env.SHORT_SHA }}" > lean-toolchain
             git add lean-toolchain
-            git commit --allow-empty -m "Update lean-toolchain for testing https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
+            git commit -m "Update lean-toolchain for testing https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
           else
             echo "Branch already exists, updating lean-toolchain."
             git switch lean-pr-testing-${{ steps.workflow-info.outputs.pullRequestNumber }}
@@ -469,7 +434,7 @@ jobs:
             git merge "$BASE" --strategy-option ours --no-commit --allow-unrelated-histories
             echo "leanprover/lean4-pr-releases:pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-${{ env.SHORT_SHA }}" > lean-toolchain
             git add lean-toolchain
-            git commit --allow-empty -m "Update lean-toolchain for https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
+            git commit -m "Update lean-toolchain for https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
           fi
 
       - name: Push changes
@@ -489,13 +454,12 @@ jobs:
       # Checkout the mathlib4 repository with all branches
       - name: Checkout mathlib4 repository
         if: steps.workflow-info.outputs.pullRequestNumber != '' && steps.ready.outputs.mathlib_ready == 'true'
-        uses: actions/checkout@v6
+        uses: actions/checkout@v5
         with:
           repository: leanprover-community/mathlib4-nightly-testing
-          token: ${{ steps.mathlib-app-token.outputs.token }}
+          token: ${{ secrets.MATHLIB4_BOT }}
           ref: nightly-testing
           fetch-depth: 0 # This ensures we check out all tags and branches.
-          filter: tree:0
 
       - name: install elan
         run: |
@@ -530,7 +494,7 @@ jobs:
             sed -i 's,require "leanprover-community" / "batteries" @ git ".\+",require "leanprover-community" / "batteries" @ git "lean-pr-testing-${{ steps.workflow-info.outputs.pullRequestNumber }}",' lakefile.lean
             lake update batteries
             git add lakefile.lean lake-manifest.json
-            git commit --allow-empty -m "Update lean-toolchain for testing https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
+            git commit -m "Update lean-toolchain for testing https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
           else
             echo "Branch already exists, updating lean-toolchain and bumping Batteries."
             git switch lean-pr-testing-${{ steps.workflow-info.outputs.pullRequestNumber }}
@@ -541,7 +505,7 @@ jobs:
             git add lean-toolchain
             lake update batteries
             git add lake-manifest.json
-            git commit --allow-empty -m "Update lean-toolchain for https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
+            git commit -m "Update lean-toolchain for https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
           fi
 
       - name: Push changes
@@ -549,18 +513,6 @@ jobs:
         run: |
           git push origin lean-pr-testing-${{ steps.workflow-info.outputs.pullRequestNumber }}
 
-      - name: Add mathlib4-nightly-available label
-        if: steps.workflow-info.outputs.pullRequestNumber != '' && steps.ready.outputs.mathlib_ready == 'true'
-        uses: actions/github-script@v8
-        with:
-          script: |
-            await github.rest.issues.addLabels({
-              issue_number: ${{ steps.workflow-info.outputs.pullRequestNumber }},
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              labels: ['mathlib4-nightly-available']
-            })
-
       # We next automatically create a reference manual branch using this toolchain.
       # Reference manual CI will be responsible for reporting back success or failure
       # to the PR comments asynchronously (and thus transitively SubVerso/Verso).
@@ -572,13 +524,12 @@ jobs:
       # Checkout the reference manual repository with all branches
       - name: Checkout mathlib4 repository
         if: steps.workflow-info.outputs.pullRequestNumber != '' && steps.reference-manual-ready.outputs.manual_ready == 'true'
-        uses: actions/checkout@v6
+        uses: actions/checkout@v5
         with:
           repository: leanprover/reference-manual
           token: ${{ secrets.MANUAL_PR_BOT }}
           ref: nightly-testing
           fetch-depth: 0 # This ensures we check out all tags and branches.
-          filter: tree:0
 
       - name: Check if tag in reference manual exists
         if: steps.workflow-info.outputs.pullRequestNumber != '' && steps.reference-manual-ready.outputs.manual_ready == 'true'
@@ -604,7 +555,7 @@ jobs:
             echo "leanprover/lean4-pr-releases:pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-${{ env.SHORT_SHA }}" > lean-toolchain
             git add lean-toolchain
             git add lakefile.lean lake-manifest.json
-            git commit --allow-empty -m "Update lean-toolchain for testing https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
+            git commit -m "Update lean-toolchain for testing https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
           else
             echo "Branch already exists, updating lean-toolchain."
             git switch lean-pr-testing-${{ steps.workflow-info.outputs.pullRequestNumber }}
@@ -614,7 +565,7 @@ jobs:
             echo "leanprover/lean4-pr-releases:pr-release-${{ steps.workflow-info.outputs.pullRequestNumber }}-${{ env.SHORT_SHA }}" > lean-toolchain
             git add lean-toolchain
             git add lake-manifest.json
-            git commit --allow-empty -m "Update lean-toolchain for https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
+            git commit -m "Update lean-toolchain for https://github.com/leanprover/lean4/pull/${{ steps.workflow-info.outputs.pullRequestNumber }}"
           fi
 
       - name: Push changes

.github/workflows/pr-title.yml

@@ -10,11 +10,11 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Check PR title
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             const msg = context.payload.pull_request? context.payload.pull_request.title : context.payload.merge_group.head_commit.message;
             console.log(`Message: ${msg}`)
-            if (!/^(feat|fix|doc|style|refactor|test|chore|perf): (?![A-Z][a-z]).*[^.]($|\n\n)/.test(msg)) {
+            if (!/^(feat|fix|doc|style|refactor|test|chore|perf): .*[^.]($|\n\n)/.test(msg)) {
               core.setFailed('PR title does not follow the Commit Convention (https://leanprover.github.io/lean4/doc/dev/commit_convention.html).');
             }

.github/workflows/stale.yml

@@ -11,7 +11,7 @@ jobs:
   stale:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/stale@v10
+      - uses: actions/stale@v9
         with:
           days-before-stale: -1
           days-before-pr-stale: 30

.github/workflows/update-stage0.yml

@@ -27,7 +27,7 @@ jobs:
     # This action should push to an otherwise protected branch, so it
     # uses a deploy key with write permissions, as suggested at
     # https://stackoverflow.com/a/76135647/946226
-    - uses: actions/checkout@v6
+    - uses: actions/checkout@v5
       with:
         ssh-key: ${{secrets.STAGE0_SSH_KEY}}
     - run: echo "should_update_stage0=yes" >> "$GITHUB_ENV"
@@ -58,7 +58,7 @@ jobs:
       shell: 'nix develop -c bash -euxo pipefail {0}'
     - name: Restore Cache
       if: env.should_update_stage0 == 'yes'
-      uses: actions/cache/restore@v5
+      uses: actions/cache/restore@v4
       with:
         # NOTE: must be in sync with `restore-cache` in `build-template.yml`
         path: |
@@ -69,10 +69,10 @@ jobs:
           build/stage1/**/*.ir
           build/stage1/**/*.c
           build/stage1/**/*.c.o*
-        key: Linux Lake-build-v4-${{ github.sha }}
+        key: Linux Lake-build-v3-${{ github.sha }}
         # fall back to (latest) previous cache
         restore-keys: |
-          Linux Lake-build-v4
+          Linux Lake-build-v3
     - if: env.should_update_stage0 == 'yes'
       # sync options with `Linux Lake` to ensure cache reuse
       run: |

.gitignore

@@ -20,6 +20,7 @@ tasks.json
 settings.json
 .gdb_history
 .vscode/*
+!.vscode/settings.json
 script/__pycache__
 *.produced.out
 CMakeSettings.json

CMakeLists.txt

@@ -10,22 +10,22 @@ option(USE_MIMALLOC "use mimalloc" ON)
 get_cmake_property(vars CACHE_VARIABLES)
 foreach(var ${vars})
   get_property(currentHelpString CACHE "${var}" PROPERTY HELPSTRING)
-  if(var MATCHES "STAGE0_(.*)")
+  if("${var}" MATCHES "STAGE0_(.*)")
     list(APPEND STAGE0_ARGS "-D${CMAKE_MATCH_1}=${${var}}")
-  elseif(var MATCHES "STAGE1_(.*)")
+  elseif("${var}" MATCHES "STAGE1_(.*)")
     list(APPEND STAGE1_ARGS "-D${CMAKE_MATCH_1}=${${var}}")
-  elseif(currentHelpString MATCHES "No help, variable specified on the command line." OR currentHelpString STREQUAL "")
+  elseif("${currentHelpString}" MATCHES "No help, variable specified on the command line." OR "${currentHelpString}" STREQUAL "")
     list(APPEND CL_ARGS "-D${var}=${${var}}")
-    if(var MATCHES "USE_GMP|CHECK_OLEAN_VERSION|LEAN_VERSION_.*|LEAN_SPECIAL_VERSION_DESC")
+    if("${var}" MATCHES "USE_GMP|CHECK_OLEAN_VERSION|LEAN_VERSION_.*|LEAN_SPECIAL_VERSION_DESC")
       # must forward options that generate incompatible .olean format
       list(APPEND STAGE0_ARGS "-D${var}=${${var}}")
-    elseif(var MATCHES "LLVM*|PKG_CONFIG|USE_LAKE|USE_MIMALLOC")
+    elseif("${var}" MATCHES "LLVM*|PKG_CONFIG|USE_LAKE|USE_MIMALLOC")
       list(APPEND STAGE0_ARGS "-D${var}=${${var}}")
     endif()
-  elseif(var MATCHES "USE_MIMALLOC")
+  elseif("${var}" MATCHES "USE_MIMALLOC")
     list(APPEND CL_ARGS "-D${var}=${${var}}")
     list(APPEND STAGE0_ARGS "-D${var}=${${var}}")
-  elseif((var MATCHES "CMAKE_.*") AND NOT (var MATCHES "CMAKE_BUILD_TYPE") AND NOT (var MATCHES "CMAKE_HOME_DIRECTORY"))
+  elseif(("${var}" MATCHES "CMAKE_.*") AND NOT ("${var}" MATCHES "CMAKE_BUILD_TYPE") AND NOT ("${var}" MATCHES "CMAKE_HOME_DIRECTORY"))
     list(APPEND PLATFORM_ARGS "-D${var}=${${var}}")
   endif()
 endforeach()
@@ -34,19 +34,17 @@ include(ExternalProject)
 project(LEAN CXX C)
 
 if(NOT (DEFINED STAGE0_CMAKE_EXECUTABLE_SUFFIX))
-  set(STAGE0_CMAKE_EXECUTABLE_SUFFIX "${CMAKE_EXECUTABLE_SUFFIX}")
+    set(STAGE0_CMAKE_EXECUTABLE_SUFFIX "${CMAKE_EXECUTABLE_SUFFIX}")
 endif()
 
 # Don't do anything with cadical on wasm
-if(NOT CMAKE_SYSTEM_NAME MATCHES "Emscripten")
+if (NOT ${CMAKE_SYSTEM_NAME} MATCHES "Emscripten")
   find_program(CADICAL cadical)
   if(NOT CADICAL)
     set(CADICAL_CXX c++)
-    if(CADICAL_USE_CUSTOM_CXX)
+    if (CADICAL_USE_CUSTOM_CXX)
       set(CADICAL_CXX ${CMAKE_CXX_COMPILER})
-      # Use same platform flags as for Lean executables, in particular from `prepare-llvm-linux.sh`,
-      # but not Lean-specific `LEAN_EXTRA_CXX_FLAGS` such as fsanitize.
-      set(CADICAL_CXXFLAGS "${CMAKE_CXX_FLAGS}")
+      set(CADICAL_CXXFLAGS "${LEAN_EXTRA_CXX_FLAGS}")
       set(CADICAL_LDFLAGS "-Wl,-rpath=\\$$ORIGIN/../lib")
     endif()
     find_program(CCACHE ccache)
@@ -54,51 +52,42 @@ if(NOT CMAKE_SYSTEM_NAME MATCHES "Emscripten")
       set(CADICAL_CXX "${CCACHE} ${CADICAL_CXX}")
     endif()
     # missing stdio locking API on Windows
-    if(CMAKE_SYSTEM_NAME MATCHES "Windows")
+    if(${CMAKE_SYSTEM_NAME} MATCHES "Windows")
       string(APPEND CADICAL_CXXFLAGS " -DNUNLOCKED")
     endif()
     string(APPEND CADICAL_CXXFLAGS " -DNCLOSEFROM")
-    ExternalProject_Add(
-      cadical
+    ExternalProject_add(cadical
       PREFIX cadical
       GIT_REPOSITORY https://github.com/arminbiere/cadical
       GIT_TAG rel-2.1.2
       CONFIGURE_COMMAND ""
-      BUILD_COMMAND
-        $(MAKE) -f ${CMAKE_SOURCE_DIR}/src/cadical.mk CMAKE_EXECUTABLE_SUFFIX=${CMAKE_EXECUTABLE_SUFFIX}
-        CXX=${CADICAL_CXX} CXXFLAGS=${CADICAL_CXXFLAGS} LDFLAGS=${CADICAL_LDFLAGS}
+      BUILD_COMMAND $(MAKE) -f ${CMAKE_SOURCE_DIR}/src/cadical.mk
+        CMAKE_EXECUTABLE_SUFFIX=${CMAKE_EXECUTABLE_SUFFIX}
+        CXX=${CADICAL_CXX}
+        CXXFLAGS=${CADICAL_CXXFLAGS}
+        LDFLAGS=${CADICAL_LDFLAGS}
       BUILD_IN_SOURCE ON
-      INSTALL_COMMAND ""
-    )
-    set(
-      CADICAL
-      ${CMAKE_BINARY_DIR}/cadical/cadical${CMAKE_EXECUTABLE_SUFFIX}
-      CACHE FILEPATH
-      "path to cadical binary"
-      FORCE
-    )
+      INSTALL_COMMAND "")
+    set(CADICAL ${CMAKE_BINARY_DIR}/cadical/cadical${CMAKE_EXECUTABLE_SUFFIX} CACHE FILEPATH "path to cadical binary" FORCE)
     list(APPEND EXTRA_DEPENDS cadical)
   endif()
   list(APPEND CL_ARGS -DCADICAL=${CADICAL})
 endif()
 
-if(USE_MIMALLOC)
-  ExternalProject_Add(
-    mimalloc
+if (USE_MIMALLOC)
+  ExternalProject_add(mimalloc
     PREFIX mimalloc
     GIT_REPOSITORY https://github.com/microsoft/mimalloc
     GIT_TAG v2.2.3
     # just download, we compile it as part of each stage as it is small
     CONFIGURE_COMMAND ""
     BUILD_COMMAND ""
-    INSTALL_COMMAND ""
-  )
+    INSTALL_COMMAND "")
   list(APPEND EXTRA_DEPENDS mimalloc)
 endif()
 
-if(NOT STAGE1_PREV_STAGE)
-  ExternalProject_Add(
-    stage0
+if (NOT STAGE1_PREV_STAGE)
+  ExternalProject_add(stage0
     SOURCE_DIR "${LEAN_SOURCE_DIR}/stage0"
     SOURCE_SUBDIR src
     BINARY_DIR stage0
@@ -106,49 +95,38 @@ if(NOT STAGE1_PREV_STAGE)
     # (however, CI will override this as we need to embed the githash into the stage 1 library built
     # by stage 0)
     CMAKE_ARGS -DSTAGE=0 -DUSE_GITHASH=OFF ${PLATFORM_ARGS} ${STAGE0_ARGS}
-    BUILD_ALWAYS
-      ON # cmake doesn't auto-detect changes without a download method
-    INSTALL_COMMAND
-      "" # skip install
+    BUILD_ALWAYS ON  # cmake doesn't auto-detect changes without a download method
+    INSTALL_COMMAND ""  # skip install
     DEPENDS ${EXTRA_DEPENDS}
   )
   list(APPEND EXTRA_DEPENDS stage0)
 endif()
-ExternalProject_Add(
-  stage1
+ExternalProject_add(stage1
   SOURCE_DIR "${LEAN_SOURCE_DIR}"
   SOURCE_SUBDIR src
   BINARY_DIR stage1
-  CMAKE_ARGS
-    -DSTAGE=1 -DPREV_STAGE=${CMAKE_BINARY_DIR}/stage0
-    -DPREV_STAGE_CMAKE_EXECUTABLE_SUFFIX=${STAGE0_CMAKE_EXECUTABLE_SUFFIX} ${CL_ARGS} ${STAGE1_ARGS}
+  CMAKE_ARGS -DSTAGE=1 -DPREV_STAGE=${CMAKE_BINARY_DIR}/stage0 -DPREV_STAGE_CMAKE_EXECUTABLE_SUFFIX=${STAGE0_CMAKE_EXECUTABLE_SUFFIX} ${CL_ARGS} ${STAGE1_ARGS}
   BUILD_ALWAYS ON
   INSTALL_COMMAND ""
   DEPENDS ${EXTRA_DEPENDS}
   STEP_TARGETS configure
 )
-ExternalProject_Add(
-  stage2
+ExternalProject_add(stage2
   SOURCE_DIR "${LEAN_SOURCE_DIR}"
   SOURCE_SUBDIR src
   BINARY_DIR stage2
-  CMAKE_ARGS
-    -DSTAGE=2 -DPREV_STAGE=${CMAKE_BINARY_DIR}/stage1 -DPREV_STAGE_CMAKE_EXECUTABLE_SUFFIX=${CMAKE_EXECUTABLE_SUFFIX}
-    ${CL_ARGS}
+  CMAKE_ARGS -DSTAGE=2 -DPREV_STAGE=${CMAKE_BINARY_DIR}/stage1 -DPREV_STAGE_CMAKE_EXECUTABLE_SUFFIX=${CMAKE_EXECUTABLE_SUFFIX} ${CL_ARGS}
   BUILD_ALWAYS ON
   INSTALL_COMMAND ""
   DEPENDS stage1
   EXCLUDE_FROM_ALL ON
   STEP_TARGETS configure
 )
-ExternalProject_Add(
-  stage3
+ExternalProject_add(stage3
   SOURCE_DIR "${LEAN_SOURCE_DIR}"
   SOURCE_SUBDIR src
   BINARY_DIR stage3
-  CMAKE_ARGS
-    -DSTAGE=3 -DPREV_STAGE=${CMAKE_BINARY_DIR}/stage2 -DPREV_STAGE_CMAKE_EXECUTABLE_SUFFIX=${CMAKE_EXECUTABLE_SUFFIX}
-    ${CL_ARGS}
+  CMAKE_ARGS -DSTAGE=3 -DPREV_STAGE=${CMAKE_BINARY_DIR}/stage2 -DPREV_STAGE_CMAKE_EXECUTABLE_SUFFIX=${CMAKE_EXECUTABLE_SUFFIX} ${CL_ARGS}
   BUILD_ALWAYS ON
   INSTALL_COMMAND ""
   DEPENDS stage2
@@ -157,14 +135,24 @@ ExternalProject_Add(
 
 # targets forwarded to appropriate stages
 
-add_custom_target(update-stage0 COMMAND $(MAKE) -C stage1 update-stage0 DEPENDS stage1)
+add_custom_target(update-stage0
+  COMMAND $(MAKE) -C stage1 update-stage0
+  DEPENDS stage1)
 
-add_custom_target(update-stage0-commit COMMAND $(MAKE) -C stage1 update-stage0-commit DEPENDS stage1)
+add_custom_target(update-stage0-commit
+  COMMAND $(MAKE) -C stage1 update-stage0-commit
+  DEPENDS stage1)
 
-add_custom_target(test COMMAND $(MAKE) -C stage1 test DEPENDS stage1)
+add_custom_target(test
+  COMMAND $(MAKE) -C stage1 test
+  DEPENDS stage1)
 
-add_custom_target(clean-stdlib COMMAND $(MAKE) -C stage1 clean-stdlib DEPENDS stage1)
+add_custom_target(clean-stdlib
+  COMMAND $(MAKE) -C stage1 clean-stdlib
+  DEPENDS stage1)
 
 install(CODE "execute_process(COMMAND make -C stage1 install)")
 
-add_custom_target(check-stage3 COMMAND diff "stage2/bin/lean" "stage3/bin/lean" DEPENDS stage3)
+add_custom_target(check-stage3
+  COMMAND diff "stage2/bin/lean" "stage3/bin/lean"
+  DEPENDS stage3)

CMakePresets.json

@@ -41,7 +41,7 @@
         "SMALL_ALLOCATOR": "OFF",
         "USE_MIMALLOC": "OFF",
         "BSYMBOLIC": "OFF",
-        "LEAN_TEST_VARS": "MAIN_STACK_SIZE=16000 LSAN_OPTIONS=max_leaks=10"
+        "LEAN_TEST_VARS": "MAIN_STACK_SIZE=16000"
       },
       "generator": "Unix Makefiles",
       "binaryDir": "${sourceDir}/build/sanitize"

CODEOWNERS

@@ -7,9 +7,9 @@
 /.github/ @kim-em
 /RELEASES.md @kim-em
 /src/kernel/ @leodemoura
-/src/library/compiler/ @hargoniX
+/src/library/compiler/ @zwarich
 /src/lake/ @tydeu
-/src/Lean/Compiler/ @leodemoura @hargoniX
+/src/Lean/Compiler/ @leodemoura @zwarich
 /src/Lean/Data/Lsp/ @mhuisi
 /src/Lean/Elab/Deriving/ @kim-em
 /src/Lean/Elab/Tactic/ @kim-em

doc/dev/bootstrap.md

@@ -6,7 +6,7 @@ building Lean itself - which is needed to again build those parts. This cycle is
 broken by using pre-built C files checked into the repository (which ultimately
 go back to a point where the Lean compiler was not written in Lean) in place of
 these Lean inputs and then compiling everything in multiple stages up to a fixed
-point. The build directory is organized into these stages:
+point. The build directory is organized in these stages:
 
 ```bash
 stage0/
@@ -72,14 +72,11 @@ update the archived C source code of the stage 0 compiler in `stage0/src`.
 
 The github repository will automatically update stage0 on `master` once
 `src/stdlib_flags.h` and `stage0/src/stdlib_flags.h` are out of sync.
-To trigger this, modify `stage0/src/stdlib_flags.h` (e.g., by adding or changing
-a comment). When `update-stage0` runs, it will overwrite `stage0/src/stdlib_flags.h`
-with the contents of `src/stdlib_flags.h`, bringing them back in sync.
 
 NOTE: A full rebuild of stage 1 will only be triggered when the *committed* contents of `stage0/` are changed.
 Thus if you change files in it manually instead of through `update-stage0-commit` (see below) or fetching updates from git, you either need to commit those changes first or run `make -C build/release clean-stdlib`.
 The same is true for further stages except that a rebuild of them is retriggered on any committed change, not just to a specific directory.
-Thus when debugging e.g. stage 2 failures, you can resume the build from these failures on but you may want to explicitly call `clean-stdlib` to either observe changes from `.olean` files of modules that built successfully or to check that you did not break modules that built successfully at some prior point.
+Thus when debugging e.g. stage 2 failures, you can resume the build from these failures on but may want to explicitly call `clean-stdlib` to either observe changes from `.olean` files of modules that built successfully or to check that you did not break modules that built successfully at some prior point.
 
 If you have write access to the lean4 repository, you can also manually
 trigger that process, for example to be able to use new features in the compiler itself.
@@ -101,7 +98,7 @@ The script `script/rebase-stage0.sh` can be used for that.
 
 The CI should prevent PRs with changes to stage0 (besides `stdlib_flags.h`)
 from entering `master` through the (squashing!) merge queue, and label such PRs
-with the `changes-stage0` label. Such PRs should have a cleaned-up history,
+with the `changes-stage0` label. Such PRs should have a cleaned up history,
 with separate stage0 update commits; then coordinate with the admins to merge
 your PR using rebase merge, bypassing the merge queue.
 

doc/dev/ffi.md

@@ -1,9 +1,190 @@
 # Foreign Function Interface
 
-The Lean FFI documentation is now part of the [Lean language reference](https://lean-lang.org/doc/reference/latest/).
+NOTE: The current interface was designed for internal use in Lean and should be considered **unstable**.
+It will be refined and extended in the future.
 
- * [General FFI](https://lean-lang.org/doc/reference/latest/find/?domain=Verso.Genre.Manual.section&name=ffi)
- * [Representation of inductive types](https://lean-lang.org/doc/reference/latest/find/?domain=Verso.Genre.Manual.section&name=inductive-types-ffi)
- * [String](https://lean-lang.org/doc/reference/latest/find/?domain=Verso.Genre.Manual.section&name=string-ffi)
- * [Array](https://lean-lang.org/doc/reference/latest/find/?domain=Verso.Genre.Manual.section&name=array-ffi)
+As Lean is written partially in Lean itself and partially in C++, it offers efficient interoperability between the two languages (or rather, between Lean and any language supporting C interfaces).
+This support is however currently limited to transferring Lean data types; in particular, it is not possible yet to pass or return compound data structures such as C `struct`s by value from or to Lean.
 
+There are two primary attributes for interoperating with other languages:
+* `@[extern "sym"] constant leanSym : ...` binds a Lean declaration to the external symbol `sym`.
+  It can also be used with `def` to provide an internal definition, but ensuring consistency of both definitions is up to the user.
+* `@[export sym] def leanSym : ...` exports `leanSym` under the unmangled symbol name `sym`.
+
+For simple examples of how to call foreign code from Lean and vice versa, see <https://github.com/leanprover/lean4/blob/master/src/lake/examples/ffi> and <https://github.com/leanprover/lean4/blob/master/src/lake/examples/reverse-ffi>, respectively.
+
+## The Lean ABI
+
+The Lean Application Binary Interface (ABI) describes how the signature of a Lean declaration is encoded as a native calling convention.
+It is based on the standard C ABI and calling convention of the target platform.
+For a Lean declaration marked with either `@[extern "sym"]` or `@[export sym]` for some symbol name `sym`, let `α₁ → ... → αₙ → β` be the normalized declaration's type.
+If `n` is 0, the corresponding C declaration is
+```c
+extern s sym;
+```
+where `s` is the C translation of `β` as specified in the next section.
+In the case of an `@[extern]` definition, the symbol's value is guaranteed to be initialized only after calling the Lean module's initializer or that of an importing module; see [Initialization](#initialization).
+
+If `n` is greater than 0, the corresponding C declaration is
+```c
+s sym(t₁, ..., tₘ);
+```
+where the parameter types `tᵢ` are the C translation of the `αᵢ` as in the next section.
+In the case of `@[extern]` all *irrelevant* types are removed first; see next section.
+
+### Translating Types from Lean to C
+
+* The integer types `UInt8`, ..., `UInt64`, `USize` are represented by the C types `uint8_t`, ..., `uint64_t`, `size_t`, respectively
+* `Char` is represented by `uint32_t`
+* `Float` is represented by `double`
+* An *enum* inductive type of at least 2 and at most 2^32 constructors, each of which with no parameters, is represented by the first type of `uint8_t`, `uint16_t`, `uint32_t` that is sufficient to represent all constructor indices.
+
+  For example, the type `Bool` is represented as `uint8_t` with values `0` for `false` and `1` for `true`.
+* `Decidable α` is represented the same way as `Bool`
+* An inductive type with a *trivial structure*, that is,
+  * it is none of the types described above
+  * it is not marked `unsafe`
+  * it has a single constructor with a single parameter of *relevant* type
+
+  is represented by the representation of that parameter's type.
+
+  For example, `{ x : α // p }`, the `Subtype` structure of a value of type `α` and an irrelevant proof, is represented by the representation of `α`.
+  Similarly, the signed integer types `Int8`, ..., `Int64`, `ISize` are also represented by the unsigned C types `uint8_t`, ..., `uint64_t`, `size_t`, respectively, because they have a trivial structure.
+* `Nat` and `Int` are represented by `lean_object *`.
+  Their runtime values is either a pointer to an opaque bignum object or, if the lowest bit of the "pointer" is 1 (`lean_is_scalar`), an encoded unboxed natural number or integer (`lean_box`/`lean_unbox`).
+* A universe `Sort u`, type constructor `... → Sort u`, or proposition `p : Prop` is *irrelevant* and is either statically erased (see above) or represented as a `lean_object *` with the runtime value `lean_box(0)`
+* Any other type is represented by `lean_object *`.
+  Its runtime value is a pointer to an object of a subtype of `lean_object` (see the "Inductive types" section below) or the unboxed value `lean_box(cidx)` for the `cidx`th constructor of an inductive type if this constructor does not have any relevant parameters.
+
+  Example: the runtime value of `u : Unit` is always `lean_box(0)`.
+
+#### Inductive types
+
+For inductive types which are in the fallback `lean_object *` case above and not trivial constructors, the type is stored as a `lean_ctor_object`, and `lean_is_ctor` will return true. A `lean_ctor_object` stores the constructor index in the header, and the fields are stored in the `m_objs` portion of the object.
+
+The memory order of the fields is derived from the types and order of the fields in the declaration. They are ordered as follows:
+
+* Non-scalar fields stored as `lean_object *`
+* Fields of type `USize`
+* Other scalar fields, in decreasing order by size
+
+Within each group the fields are ordered in declaration order. Trivial wrapper types count as their underlying wrapped type for this purpose.
+
+* To access fields of the first kind, use `lean_ctor_get(val, i)` to get the `i`th non-scalar field.
+* To access `USize` fields, use `lean_ctor_get_usize(val, n+i)` to get the `i`th usize field and `n` is the total number of fields of the first kind.
+* To access other scalar fields, use `lean_ctor_get_uintN(val, off)` or `lean_ctor_get_usize(val, off)` as appropriate. Here `off` is the byte offset of the field in the structure, starting at `n*sizeof(void*)` where `n` is the number of fields of the first two kinds.
+
+For example, a structure such as
+```lean
+structure S where
+  ptr_1 : Array Nat
+  usize_1 : USize
+  sc64_1 : UInt64
+  sc64_2 : { x : UInt64 // x > 0 } -- wrappers of scalars count as scalars
+  sc64_3 : Float -- `Float` is 64 bit
+  sc8_1 : Bool
+  sc16_1 : UInt16
+  sc8_2 : UInt8
+  sc64_4 : UInt64
+  usize_2 : USize
+  sc32_1 : Char -- trivial wrapper around `UInt32`
+  sc32_2 : UInt32
+  sc16_2 : UInt16
+```
+would get re-sorted into the following memory order:
+
+* `S.ptr_1` - `lean_ctor_get(val, 0)`
+* `S.usize_1` - `lean_ctor_get_usize(val, 1)`
+* `S.usize_2` - `lean_ctor_get_usize(val, 2)`
+* `S.sc64_1` - `lean_ctor_get_uint64(val, sizeof(void*)*3)`
+* `S.sc64_2` - `lean_ctor_get_uint64(val, sizeof(void*)*3 + 8)`
+* `S.sc64_3` - `lean_ctor_get_float(val, sizeof(void*)*3 + 16)`
+* `S.sc64_4` - `lean_ctor_get_uint64(val, sizeof(void*)*3 + 24)`
+* `S.sc32_1` - `lean_ctor_get_uint32(val, sizeof(void*)*3 + 32)`
+* `S.sc32_2` - `lean_ctor_get_uint32(val, sizeof(void*)*3 + 36)`
+* `S.sc16_1` - `lean_ctor_get_uint16(val, sizeof(void*)*3 + 40)`
+* `S.sc16_2` - `lean_ctor_get_uint16(val, sizeof(void*)*3 + 42)`
+* `S.sc8_1` - `lean_ctor_get_uint8(val, sizeof(void*)*3 + 44)`
+* `S.sc8_2` - `lean_ctor_get_uint8(val, sizeof(void*)*3 + 45)`
+
+### Borrowing
+
+By default, all `lean_object *` parameters of an `@[extern]` function are considered *owned*, i.e. the external code is passed a "virtual RC token" and is responsible for passing this token along to another consuming function (exactly once) or freeing it via `lean_dec`.
+To reduce reference counting overhead, parameters can be marked as *borrowed* by prefixing their type with `@&`.
+Borrowed objects must only be passed to other non-consuming functions (arbitrarily often) or converted to owned values using `lean_inc`.
+In `lean.h`, the `lean_object *` aliases `lean_obj_arg` and `b_lean_obj_arg` are used to mark this difference on the C side.
+
+Return values and `@[export]` parameters are always owned at the moment.
+
+## Initialization
+
+When including Lean code as part of a larger program, modules must be *initialized* before accessing any of their declarations.
+Module initialization entails
+* initialization of all "constants" (nullary functions), including closed terms lifted out of other functions
+* execution of all `[init]` functions
+* execution of all `[builtin_init]` functions, if the `builtin` parameter of the module initializer has been set
+
+The module initializer is automatically run with the `builtin` flag for executables compiled from Lean code and for "plugins" loaded with `lean --plugin`.
+For all other modules imported by `lean`, the initializer is run without `builtin`.
+Thus `[init]` functions are run iff their module is imported, regardless of whether they have native code available or not, while `[builtin_init]` functions are only run for native executable or plugins, regardless of whether their module is imported or not.
+`lean` uses built-in initializers for e.g. registering basic parsers that should be available even without importing their module (which is necessary for bootstrapping).
+
+The initializer for module `A.B` is called `initialize_A_B` and will automatically initialize any imported modules.
+Module initializers are idempotent (when run with the same `builtin` flag), but not thread-safe.
+
+**Important for process-related functionality**: If your application needs to use process-related functions from libuv, such as `Std.Internal.IO.Process.getProcessTitle` and `Std.Internal.IO.Process.setProcessTitle`, you must call `lean_setup_args(argc, argv)` (which returns a potentially modified `argv` that must be used in place of the original) **before** calling `lean_initialize()` or `lean_initialize_runtime_module()`. This sets up process handling capabilities correctly, which is essential for certain system-level operations that Lean's runtime may depend on.
+
+Together with initialization of the Lean runtime, you should execute code like the following exactly once before accessing any Lean declarations:
+
+```c
+void lean_initialize_runtime_module();
+void lean_initialize();
+char ** lean_setup_args(int argc, char ** argv);
+
+lean_object * initialize_A_B(uint8_t builtin, lean_object *);
+lean_object * initialize_C(uint8_t builtin, lean_object *);
+...
+
+argv = lean_setup_args(argc, argv); // if using process-related functionality
+lean_initialize_runtime_module();
+//lean_initialize();  // necessary (and replaces `lean_initialize_runtime_module`) if you (indirectly) access the `Lean` package
+
+lean_object * res;
+// use same default as for Lean executables
+uint8_t builtin = 1;
+res = initialize_A_B(builtin, lean_io_mk_world());
+if (lean_io_result_is_ok(res)) {
+    lean_dec_ref(res);
+} else {
+    lean_io_result_show_error(res);
+    lean_dec(res);
+    return ...;  // do not access Lean declarations if initialization failed
+}
+res = initialize_C(builtin, lean_io_mk_world());
+if (lean_io_result_is_ok(res)) {
+...
+
+//lean_init_task_manager();  // necessary if you (indirectly) use `Task`
+lean_io_mark_end_initialization();
+```
+
+In addition, any other thread not spawned by the Lean runtime itself must be initialized for Lean use by calling
+```c
+void lean_initialize_thread();
+```
+and should be finalized in order to free all thread-local resources by calling
+```c
+void lean_finalize_thread();
+```
+
+## `@[extern]` in the Interpreter
+
+The interpreter can run Lean declarations for which symbols are available in loaded shared libraries, which includes `@[extern]` declarations.
+Thus to e.g. run `#eval` on such a declaration, you need to
+1. compile (at least) the module containing the declaration and its dependencies into a shared library, and then
+1. pass this library to `lean --load-dynlib=` to run code `import`ing this module.
+
+Note that it is not sufficient to load the foreign library containing the external symbol because the interpreter depends on code that is emitted for each `@[extern]` declaration.
+Thus it is not possible to interpret an `@[extern]` declaration in the same file.
+
+See [`tests/compiler/foreign`](https://github.com/leanprover/lean4/tree/master/tests/compiler/foreign/) for an example.

doc/dev/release_checklist.md

@@ -30,7 +30,7 @@ We'll use `v4.6.0` as the intended release version as a running example.
     run `script/release_notes.py --since v4.5.0` on the `releases/v4.6.0` branch,
     and see the section "Writing the release notes" below for more information.
   - Release notes live in https://github.com/leanprover/reference-manual, in e.g. `Manual/Releases/v4.6.0.lean`.
-    It's best if you update these at the same time as you update the `lean-toolchain` for the `reference-manual` repository, see below.
+    It's best if you update these at the same time as a you update the `lean-toolchain` for the `reference-manual` repository, see below.
 - Go to https://github.com/leanprover/lean4/releases and verify that the `v4.6.0` release appears.
   - Verify on Github that "Set as the latest release" is checked.
 - Next, we will move a curated list of downstream repos to the latest stable release.
@@ -54,7 +54,7 @@ We'll use `v4.6.0` as the intended release version as a running example.
     - `verso`:
       - The `subverso` dependency is unusual in that it needs to be compatible with _every_ Lean release simultaneously.
         Usually you don't need to do anything.
-        If you think something is wrong here, please contact David Thrane Christiansen (@david-christiansen)
+        If you think something is wrong here please contact David Thrane Christiansen (@david-christiansen)
       - Warnings during `lake update` and `lake build` are expected.
     - `reference-manual`: the release notes generated by `script/release_notes.py` as described above must be included in
       `Manual/Releases/v4.6.0.lean`, and `import` and `include` statements adding in `Manual/Releases.lean`. 
@@ -69,10 +69,6 @@ We'll use `v4.6.0` as the intended release version as a running example.
     - `repl`:
       There are two copies of `lean-toolchain`/`lakefile.lean`:
       in the root, and in `test/Mathlib/`. Edit both, and run `lake update` in both directories.
-    - `lean-fro.org`:
-      After updating the toolchains and running `lake update`, you must run `scripts/update.sh` to regenerate
-      the site content. This script updates generated files that depend on the Lean version.
-      The `release_steps.py` script handles this automatically.
 - An awkward situation that sometimes occurs (e.g. with Verso) is that the `master`/`main` branch has already been moved
   to a nightly toolchain that comes *after* the stable toolchain we are
   targeting. In this case it is necessary to create a branch `releases/v4.6.0` from the last commit which was on
@@ -218,21 +214,6 @@ Please read https://leanprover-community.github.io/contribute/tags_and_branches.
 
 # Writing the release notes
 
-Release notes content is only written for the first release candidate (`-rc1`). For subsequent RCs and stable releases,
-just update the title in the existing release notes file (see "Release notes title format" below).
-
-## Release notes title format
-
-The title in the `#doc (Manual)` line must follow these formats:
-
-- **For -rc1**: `"Lean 4.7.0-rc1 (2024-03-15)"` — Include the RC suffix and the release date
-- **For -rc2, -rc3, etc.**: `"Lean 4.7.0-rc2 (2024-03-20)"` — Update the RC number and date
-- **For stable release**: `"Lean 4.7.0 (2024-04-01)"` — Remove the RC suffix but keep the date
-
-The date should be the actual date when the tag was pushed (or when CI completed and created the release page).
-
-## Generating the release notes
-
 Release notes are automatically generated from the commit history, using `script/release_notes.py`.
 
 Run this as `script/release_notes.py --since v4.6.0`, where `v4.6.0` is the *previous* release version.
@@ -247,113 +228,4 @@ Some judgement is required here: ignore commits which look minor,
 but manually add items to the release notes for significant PRs that were rebase-merged.
 
 There can also be pre-written entries in `./releases_drafts`, which should be all incorporated in the release notes and then deleted from the branch.
-
-## Reviewing and fixing the generated markdown
-
-Before adding the release notes to the reference manual, carefully review the generated markdown for these common issues:
-
-1. **Unterminated code blocks**: PR descriptions sometimes have unclosed code fences. Look for code blocks
-   that don't have a closing ` ``` `. If found, fetch the original PR description with `gh pr view <number>`
-   and repair the code block with the complete content.
-
-2. **Truncated descriptions**: Some PR descriptions may end abruptly mid-sentence. Review these and complete
-   the descriptions based on the original PR.
-
-3. **Markdown syntax issues**: Check for other markdown problems that could cause parsing errors.
-
-## Creating the release notes file
-
-The release notes go in `Manual/Releases/v4_7_0.lean` in the reference-manual repository.
-
-The file structure must follow the Verso format:
-
-```lean
-/-
-Copyright (c) 2025 Lean FRO LLC. All rights reserved.
-Released under Apache 2.0 license as described in the file LICENSE.
-Author: <Your Name>
--/
-
-import VersoManual
-import Manual.Meta
-import Manual.Meta.Markdown
-
-open Manual
-open Verso.Genre
-open Verso.Genre.Manual
-open Verso.Genre.Manual.InlineLean
-
-#doc (Manual) "Lean 4.7.0-rc1 (2024-03-15)" =>
-%%%
-tag := "release-v4.7.0"
-file := "v4.7.0"
-%%%
-
-<release notes content here>
-```
-
-**Important formatting rules for Verso:**
-- Use `#` for section headers inside the document, not `##` (Verso uses header level 1 for subsections)
-- Use plain ` ``` ` for code blocks, not ` ```lean ` (the latter will cause Lean to execute the code)
-- Identifiers with underscores like `bv_decide` should be wrapped in backticks: `` `bv_decide` ``
-  (otherwise the underscore may be interpreted as markdown emphasis)
-
-## Updating Manual/Releases.lean
-
-After creating the release notes file, update `Manual/Releases.lean` to include it:
-
-1. Add the import near the top with other version imports:
-   ```lean
-   import Manual.Releases.«v4_7_0»
-   ```
-
-2. Add the include statement after the other includes:
-   ```lean
-   {include 0 Manual.Releases.«v4_7_0»}
-   ```
-
-## Building and verifying
-
-Build the release notes to check for errors:
-```bash
-lake build Manual.Releases.v4_7_0
-```
-
-Common errors and fixes:
-- "Wrong header nesting - got ## but expected at most #": Change `##` to `#`
-- "Tactic 'X' failed" or similar: Code is being executed; change ` ```lean ` to ` ``` `
-- "'_'" errors: Underscore in identifier being parsed as emphasis; wrap in backticks
-
-## Creating the PR
-
-**Important: Timing with the reference-manual tag**
-
-The reference-manual repository deploys documentation when a version tag is pushed. If you merge
-release notes AFTER the tag is created, the deployed documentation won't include them.
-
-You have two options:
-
-1. **Preferred**: Include the release notes in the same PR as the toolchain bump (or merge the
-   release notes PR before creating the tag). This ensures the tag includes the release notes.
-
-2. **If release notes are merged after the tag**: You must regenerate the tag to trigger a new deployment:
-   ```bash
-   cd /path/to/reference-manual
-   git fetch origin
-   git tag -d v4.7.0-rc1  # Delete local tag
-   git tag v4.7.0-rc1 origin/main  # Create tag at current main (which has release notes)
-   git push origin :refs/tags/v4.7.0-rc1  # Delete remote tag
-   git push origin v4.7.0-rc1  # Push new tag (triggers Deploy workflow)
-   ```
-
-If creating a separate PR for release notes:
-```bash
-git checkout -b v4.7.0-release-notes
-git add Manual/Releases/v4_7_0.lean Manual/Releases.lean
-git commit -m "doc: add v4.7.0 release notes"
-git push -u origin v4.7.0-release-notes
-gh pr create --title "doc: add v4.7.0 release notes" --body "This PR adds the release notes for Lean v4.7.0."
-```
-
-See `./releases_drafts/README.md` for more information about pre-written release note entries.
 See `./releases_drafts/README.md` for more information.

doc/dev/testing.md

@@ -51,10 +51,6 @@ All these tests are included by [src/shell/CMakeLists.txt](https://github.com/le
   codes and do not check the expected output even though output is
   produced, it is ignored.
 
-  **Note:** Tests in this directory run with `-Dlinter.all=false` to reduce noise.
-  If your test needs to verify linter behavior (e.g., deprecation warnings),
-  explicitly enable the relevant linter with `set_option linter.<name> true`.
-
 - [`tests/lean/interactive`](https://github.com/leanprover/lean4/tree/master/tests/lean/interactive/): are designed to test server requests at a
   given position in the input file. Each .lean file contains comments
   that indicate how to simulate a client request at that position.
@@ -63,7 +59,7 @@ All these tests are included by [src/shell/CMakeLists.txt](https://github.com/le
     open Foo in
     theorem tst2 (h : a ≤ b) : a + 2 ≤ b + 2 :=
     Bla.
-      --^ completion
+      --^ textDocument/completion
     ```
     In this example, the test driver [`test_single.sh`](https://github.com/leanprover/lean4/tree/master/tests/lean/interactive/test_single.sh) will simulate an
     auto-completion request at `Bla.`. The expected output is stored in

doc/examples/IJCAR2026/README.md

@@ -1,6 +0,0 @@
-# IJCAR 2026: `grind`, An SMT-Inspired Tactic for Lean 4
-
-Ancillary materials for the paper.
-
-- `examples.lean`: interactive examples from the paper
-- `analyze_grind_loc.py`: script used for the evaluation section, analyzing `grind` adoption and lines-of-code changes in Mathlib

doc/examples/IJCAR2026/analyze_grind_loc.py

@@ -1,401 +0,0 @@
-#!/usr/bin/env python3
-"""
-Analyze grind adoption LoC changes in mathlib.
-
-For each theorem/lemma in master that uses grind, find the most recent
-commit where it didn't use grind, and measure the LoC change.
-
-This script was used in preparing the "Evaluation" section of the grind paper.
-"""
-
-import subprocess
-import re
-import csv
-import sys
-from pathlib import Path
-from dataclasses import dataclass
-from concurrent.futures import ThreadPoolExecutor, as_completed
-from typing import Iterator
-from functools import lru_cache
-
-
-@dataclass
-class GrindUsage:
-    file: str
-    line_no: int
-    decl_name: str
-    decl_type: str  # theorem, lemma, def, example, etc.
-
-
-@dataclass
-class LocChange:
-    file: str
-    decl_name: str
-    decl_type: str
-    old_loc: int
-    new_loc: int
-    loc_saved: int
-    commit_sha: str
-    commit_date: str
-
-
-def run_git(args: list[str], repo: str = ".") -> str:
-    """Run a git command and return stdout."""
-    result = subprocess.run(
-        ["git", "-C", repo] + args,
-        capture_output=True, text=True, check=True
-    )
-    return result.stdout
-
-
-def run_git_safe(args: list[str], repo: str = ".") -> str | None:
-    """Run a git command, return None on failure."""
-    result = subprocess.run(
-        ["git", "-C", repo] + args,
-        capture_output=True, text=True
-    )
-    if result.returncode != 0:
-        return None
-    return result.stdout
-
-
-@lru_cache(maxsize=4096)
-def get_file_at_commit(repo: str, commit: str, file_path: str) -> str | None:
-    """Get file contents at a specific commit (cached)."""
-    return run_git_safe(["show", f"{commit}:{file_path}"], repo)
-
-
-def find_grind_usages(repo: str = ".") -> tuple[list[GrindUsage], int, int]:
-    """Find all declarations using grind in current master.
-    Returns (usages, total_grind_calls, grind_in_decls) where:
-    - total_grind_calls is the count of grind tactic calls (after filtering comments/attrs)
-    - grind_in_decls is the count of those that are inside named declarations
-    """
-    # Use git grep to find lines containing 'grind' (excludes lake packages)
-    result = run_git(["grep", "-n", "grind", "master", "--", "Mathlib/"], repo)
-
-    usages = []
-    seen = set()  # (file, decl_name) to dedupe
-    total_grind_calls = 0
-    grind_in_decls = 0
-
-    for line in result.strip().split('\n'):
-        if not line:
-            continue
-        # Format: master:path/to/file.lean:123:line content
-        match = re.match(r'^master:(.+\.lean):(\d+):(.*)$', line)
-        if not match:
-            continue
-
-        file_path, line_no_str, content = match.groups()
-        line_no = int(line_no_str)
-
-        # Skip comments and attributes (not tactic calls)
-        content_stripped = content.strip()
-        if content_stripped.startswith('--') or content_stripped.startswith('/-'):
-            continue
-        if content_stripped.startswith('attribute'):
-            continue
-        if '@[' in content and 'grind' in content:
-            # Could be an attribute like @[grind =], skip
-            if 'by' not in content and ':=' not in content:
-                continue
-
-        total_grind_calls += 1
-
-        # Find the declaration this grind belongs to
-        decl_name, decl_type = find_decl_at_line(repo, file_path, line_no)
-        if decl_name is None:
-            continue
-
-        grind_in_decls += 1
-
-        key = (file_path, decl_name)
-        if key in seen:
-            continue
-        seen.add(key)
-
-        usages.append(GrindUsage(
-            file=file_path,
-            line_no=line_no,
-            decl_name=decl_name,
-            decl_type=decl_type
-        ))
-
-    return usages, total_grind_calls, grind_in_decls
-
-
-def find_decl_at_line(repo: str, file_path: str, grind_line: int) -> tuple[str | None, str | None]:
-    """
-    Find the declaration name and type that contains the grind at the given line.
-    Search backwards from grind_line to find the most recent declaration.
-    """
-    # Get file content at master
-    content = get_file_at_commit(repo, "master", file_path)
-    if content is None:
-        return None, None
-
-    lines = content.split('\n')
-
-    # Search backwards from grind_line for a declaration
-    # Match declarations with optional leading modifiers and attributes
-    decl_pattern = re.compile(r'^(?:@\[.*?\]\s*)*(?:private\s+|protected\s+|noncomputable\s+|scoped\s+)*(theorem|lemma|def|example|instance|abbrev|structure|class)\s+(\w+)')
-
-    for i in range(grind_line - 1, -1, -1):
-        if i >= len(lines):
-            continue
-        line = lines[i]
-        match = decl_pattern.match(line)
-        if match:
-            return match.group(2), match.group(1)
-
-    return None, None
-
-
-def find_grind_introduction_commit(repo: str, file_path: str, decl_name: str) -> str | None:
-    """
-    Find the commit that introduced grind to this declaration.
-    Returns None if the declaration was born with grind.
-    """
-    # First, find the line range of the declaration in master
-    content = get_file_at_commit(repo, "master", file_path)
-    if content is None:
-        return None
-
-    lines = content.split('\n')
-    decl_start = None
-    decl_end = None
-
-    # Find declaration start
-    decl_pattern = re.compile(rf'^(?:@\[.*?\]\s*)*(?:private\s+|protected\s+|noncomputable\s+|scoped\s+)*(theorem|lemma|def|example|instance|abbrev|structure|class)\s+{re.escape(decl_name)}\b')
-    for i, line in enumerate(lines):
-        if decl_pattern.match(line):
-            decl_start = i
-            break
-
-    if decl_start is None:
-        return None
-
-    # Find declaration end (next top-level declaration or EOF)
-    end_patterns = re.compile(r'^(?:private\s+|protected\s+|noncomputable\s+|scoped\s+)*(theorem|lemma|def|example|instance|abbrev|structure|class|namespace|section|end\s|@\[|#|/-)')
-    for i in range(decl_start + 1, len(lines)):
-        line = lines[i]
-        if line and not line[0].isspace() and end_patterns.match(line):
-            decl_end = i
-            break
-    if decl_end is None:
-        decl_end = len(lines)
-
-    # Find grind line within declaration
-    grind_line = None
-    for i in range(decl_start, decl_end):
-        if 'grind' in lines[i]:
-            grind_line = i + 1  # 1-indexed
-            break
-
-    if grind_line is None:
-        return None
-
-    # Use git blame to find when that grind line was added
-    blame_result = run_git_safe(["blame", "-L", f"{grind_line},{grind_line}", "--porcelain", "master", "--", file_path], repo)
-    if blame_result is None:
-        return None
-
-    # First line of porcelain output is the commit SHA
-    first_line = blame_result.split('\n')[0]
-    commit_sha = first_line.split()[0]
-
-    # Check if this declaration existed before this commit (without grind)
-    parent_sha = run_git_safe(["rev-parse", f"{commit_sha}^"], repo)
-    if parent_sha is None:
-        return None  # Initial commit, born with grind
-    parent_sha = parent_sha.strip()
-
-    # Check if declaration existed in parent
-    parent_content = get_file_at_commit(repo, parent_sha, file_path)
-    if parent_content is None:
-        # File didn't exist in parent - might be new file or renamed
-        return None
-
-    # Check if declaration existed and didn't have grind
-    if decl_name not in parent_content:
-        return None  # Declaration didn't exist - born with grind
-
-    # Check if it already had grind in parent
-    parent_lines = parent_content.split('\n')
-    in_decl = False
-    for line in parent_lines:
-        if decl_pattern.match(line):
-            in_decl = True
-        elif in_decl:
-            if line and not line[0].isspace() and end_patterns.match(line):
-                break
-            if 'grind' in line:
-                # Already had grind in parent — not the introduction commit
-                return None
-
-    return commit_sha
-
-
-def extract_proof_loc(repo: str, file_path: str, decl_name: str, commit: str) -> int | None:
-    """
-    Extract the number of lines in a declaration's proof at a given commit.
-    Returns None if the declaration doesn't exist at that commit.
-    """
-    content = get_file_at_commit(repo, commit, file_path)
-    if content is None:
-        return None
-
-    lines = content.split('\n')
-
-    # Find declaration start
-    decl_pattern = re.compile(rf'^(?:@\[.*?\]\s*)*(?:private\s+|protected\s+|noncomputable\s+|scoped\s+)*(theorem|lemma|def|example|instance|abbrev|structure|class)\s+{re.escape(decl_name)}\b')
-    decl_start = None
-    for i, line in enumerate(lines):
-        if decl_pattern.match(line):
-            decl_start = i
-            break
-
-    if decl_start is None:
-        return None
-
-    # Find declaration end
-    end_patterns = re.compile(r'^(?:private\s+|protected\s+|noncomputable\s+|scoped\s+)*(theorem|lemma|def|example|instance|abbrev|structure|class|namespace|section|end\s|@\[|#|/-)')
-    decl_end = None
-    for i in range(decl_start + 1, len(lines)):
-        line = lines[i]
-        if line and not line[0].isspace() and end_patterns.match(line):
-            decl_end = i
-            break
-    if decl_end is None:
-        decl_end = len(lines)
-
-    # Count non-empty lines in declaration
-    loc = sum(1 for i in range(decl_start, decl_end) if lines[i].strip())
-    return loc
-
-
-def get_commit_date(repo: str, sha: str) -> str:
-    """Get the date of a commit."""
-    result = run_git(["log", "-1", "--format=%ci", sha], repo)
-    return result.strip().split()[0]  # Just the date part
-
-
-def analyze_usage_detailed(repo: str, usage: GrindUsage) -> tuple[LocChange | None, str]:
-    """Analyze a single grind usage, returning (result, skip_reason)."""
-    commit = find_grind_introduction_commit(repo, usage.file, usage.decl_name)
-    if commit is None:
-        return None, "born_with_grind"
-
-    parent = run_git_safe(["rev-parse", f"{commit}^"], repo)
-    if parent is None:
-        return None, "no_parent"
-    parent = parent.strip()
-
-    old_loc = extract_proof_loc(repo, usage.file, usage.decl_name, parent)
-    new_loc = extract_proof_loc(repo, usage.file, usage.decl_name, "master")
-
-    if old_loc is None:
-        return None, "old_loc_failed"
-    if new_loc is None:
-        return None, "new_loc_failed"
-
-    commit_date = get_commit_date(repo, commit)
-
-    return LocChange(
-        file=usage.file,
-        decl_name=usage.decl_name,
-        decl_type=usage.decl_type,
-        old_loc=old_loc,
-        new_loc=new_loc,
-        loc_saved=old_loc - new_loc,
-        commit_sha=commit[:12],
-        commit_date=commit_date
-    ), "success"
-
-
-def main(repo: str = "."):
-    print("Finding grind usages in master...", file=sys.stderr)
-    usages, total_grind_calls, grind_in_decls = find_grind_usages(repo)
-    print(f"Found {len(usages)} declarations using grind ({grind_in_decls}/{total_grind_calls} grind calls)", file=sys.stderr)
-
-    print("Analyzing git history (this may take a while)...", file=sys.stderr)
-    results: list[LocChange] = []
-    skip_reasons: dict[str, int] = {}
-
-    with ThreadPoolExecutor(max_workers=64) as executor:
-        futures = {executor.submit(analyze_usage_detailed, repo, usage): usage for usage in usages}
-
-        for i, future in enumerate(as_completed(futures)):
-            if (i + 1) % 50 == 0:
-                print(f"  Progress: {i + 1}/{len(usages)}", file=sys.stderr, flush=True)
-
-            result, reason = future.result()
-            if result:
-                results.append(result)
-            else:
-                skip_reasons[reason] = skip_reasons.get(reason, 0) + 1
-
-    total_skipped = sum(skip_reasons.values())
-    print(f"\nAnalyzed {len(results)} declarations, skipped {total_skipped}:", file=sys.stderr)
-    for reason, count in sorted(skip_reasons.items(), key=lambda x: -x[1]):
-        print(f"  - {reason}: {count}", file=sys.stderr)
-
-    # Sort by LoC saved (descending)
-    results.sort(key=lambda r: r.loc_saved, reverse=True)
-
-    # Output CSV
-    writer = csv.writer(sys.stdout)
-    writer.writerow(["file", "declaration", "type", "old_loc", "new_loc", "loc_saved", "commit", "date"])
-    for r in results:
-        writer.writerow([r.file, r.decl_name, r.decl_type, r.old_loc, r.new_loc, r.loc_saved, r.commit_sha, r.commit_date])
-
-    # Summary stats to stderr
-    total_old = sum(r.old_loc for r in results) if results else 0
-    total_new = sum(r.new_loc for r in results) if results else 0
-    total_saved = sum(r.loc_saved for r in results) if results else 0
-    avg_saved = total_saved / len(results) if results else 0
-
-    print("\n" + "=" * 60, file=sys.stderr)
-    print("GRIND ADOPTION LOC ANALYSIS", file=sys.stderr)
-    print("=" * 60, file=sys.stderr)
-
-    print("\n## Declaration Counts\n", file=sys.stderr)
-    print(f"  Total grind tactic calls:         {total_grind_calls}", file=sys.stderr)
-    print(f"  In named declarations:            {grind_in_decls} ({total_grind_calls - grind_in_decls} in anonymous/other)", file=sys.stderr)
-    print(f"  Unique declarations:              {len(usages)}", file=sys.stderr)
-    print(f"  Converted to grind:               {len(results)}", file=sys.stderr)
-    print(f"  Born with grind:                  {skip_reasons.get('born_with_grind', 0)}", file=sys.stderr)
-    if skip_reasons.get('old_loc_failed', 0) > 0:
-        print(f"  Could not trace history:          {skip_reasons.get('old_loc_failed', 0)}", file=sys.stderr)
-
-    print("\n## Lines of Code Impact\n", file=sys.stderr)
-    print(f"  Total LoC before grind:           {total_old}", file=sys.stderr)
-    print(f"  Total LoC after grind:            {total_new}", file=sys.stderr)
-    print(f"  Total LoC saved:                  {total_saved}", file=sys.stderr)
-    print(f"  Average LoC saved per theorem:    {avg_saved:.1f}", file=sys.stderr)
-    big_savings = sum(1 for r in results if r.loc_saved >= 10)
-    print(f"  Declarations shrunk by 10+ lines: {big_savings}", file=sys.stderr)
-
-    if results:
-        print("\n## Top 10 Biggest LoC Savings\n", file=sys.stderr)
-        for r in results[:10]:
-            print(f"  {r.loc_saved:+4d} lines: {r.decl_name} ({r.file})", file=sys.stderr)
-
-        # Show any that got bigger (negative savings)
-        got_bigger = [r for r in results if r.loc_saved < 0]
-        if got_bigger:
-            print(f"\n## Declarations That Got Bigger ({len(got_bigger)} total)\n", file=sys.stderr)
-            print("  (showing 5 worst):", file=sys.stderr)
-            for r in got_bigger[-5:]:  # Show worst 5
-                print(f"  {r.loc_saved:+4d} lines: {r.decl_name} ({r.file})", file=sys.stderr)
-
-    print("\n" + "=" * 60, file=sys.stderr)
-
-
-if __name__ == "__main__":
-    import argparse
-    parser = argparse.ArgumentParser(description="Analyze grind LoC savings")
-    parser.add_argument("--repo", "-r", default=".", help="Repository path")
-    args = parser.parse_args()
-    main(args.repo)

doc/examples/IJCAR2026/examples.lean

@@ -1,127 +0,0 @@
-/- Examples from the paper "grind: An SMT-Inspired Tactic for Lean 4" -/
-open Lean Grind
-
-/- Congruence closure. -/
-
-example (f : Nat → Nat) (h : a = b) : f (f b) = f (f a) := by grind
-
-/-
-E-matching.
-
-Any `f` that is the left inverse of `g` would work on this example.
--/
-def f (x : Nat) := x - 1
-def g (x : Nat) := x + 1
-
-@[grind =] theorem fg : f (g x) = x := by simp [f, g]
-example : f a = b → a = g c → b = c := by grind
-
-/-
-Any `R` that is transitive and symmetric would work on this example.
--/
-def R : Nat → Nat → Prop := (· % 7 = · % 7)
-@[grind →] theorem Rtrans : R x y → R y z → R x z := by grind [R]
-@[grind →] theorem Rsymm  : R x y → R y x := by grind [R]
-example : R a b → R c b → R d c → R a d := by grind
-
-/- Big step operational semantics example. -/
-
-abbrev Variable := String
-def State := Variable → Nat
-inductive Stmt : Type where
-  | skip : Stmt
-  | assign : Variable → (State → Nat) → Stmt
-  | seq : Stmt → Stmt → Stmt
-  | ifThenElse : (State → Prop) → Stmt → Stmt → Stmt
-  | whileDo : (State → Prop) → Stmt → Stmt
-infix:60 ";; " => Stmt.seq
-export Stmt (skip assign seq ifThenElse whileDo)
-set_option quotPrecheck false in
-notation s:70 "[" x:70 "↦" n:70 "]" => (fun v ↦ if v = x then n else s v)
-inductive BigStep : Stmt → State → State → Prop where
-  | skip (s : State) : BigStep skip s s
-  | assign (x : Variable) (a : State → Nat) (s : State) : BigStep (assign x a) s (s[x ↦ a s])
-  | seq {S T : Stmt} {s t u : State} (hS : BigStep S s t) (hT : BigStep T t u) :
-    BigStep (S;; T) s u
-  | if_true {B : State → Prop} {s t : State} (hcond : B s) (S T : Stmt) (hbody : BigStep S s t) :
-    BigStep (ifThenElse B S T) s t
-  | if_false {B : State → Prop} {s t : State} (hcond : ¬ B s) (S T : Stmt) (hbody : BigStep T s t) :
-    BigStep (ifThenElse B S T) s t
-  | while_true {B S s t u} (hcond : B s) (hbody : BigStep S s t) (hrest : BigStep (whileDo B S) t u) :
-    BigStep (whileDo B S) s u
-  | while_false {B S s} (hcond : ¬ B s) : BigStep (whileDo B S) s s
-notation:55 "(" S:55 "," s:55 ")" " ==> " t:55 => BigStep S s t
-
-example {B S T s t} (hcond : B s) : (ifThenElse B S T, s) ==> t → (S, s) ==> t := by
-  grind [cases BigStep]
-
-theorem cases_if_of_true {B S T s t} (hcond : B s) : (ifThenElse B S T, s) ==> t → (S, s) ==> t := by
-  grind [cases BigStep]
-
-theorem cases_if_of_false {B S T s t} (hcond : ¬ B s) : (ifThenElse B S T, s) ==> t → (T, s) ==> t := by
-  grind [cases BigStep]
-
-example {B S T s t} : (ifThenElse B S T, s) ==> t ↔ (B s ∧ (S, s) ==> t) ∨ (¬ B s ∧ (T, s) ==> t) := by
-  grind [BigStep] -- shortcut for `cases BigStep` and `intro BigStep`
-
-attribute [grind] BigStep
-theorem if_iff {B S T s t} : (ifThenElse B S T, s) ==>
-    t ↔ (B s ∧ (S, s) ==> t) ∨ (¬ B s ∧ (T, s) ==> t) := by grind
-
-/- Dependent pattern matching. -/
-
-inductive Vec (α : Type u) : Nat → Type u
-  | nil : Vec α 0
-  | cons : α → Vec α n → Vec α (n+1)
-@[grind =] def Vec.head : Vec α (n+1) → α
-  | .cons a _ => a
-example (as bs : Vec Int (n+1)) : as.head = bs.head
-    → (match as, bs with
-       | .cons a _, .cons b _ => a + b) = 2 * as.head := by grind
-
-/- Theory solvers. -/
-
-example [CommRing α] (a b c : α) :
-    a + b + c = 3 →
-    a^2 + b^2 + c^2 = 5 →
-    a^3 + b^3 + c^3 = 7 →
-    a^4 + b^4 + c^4 = 9 := by grind
-
-example (x : BitVec 8) : (x - 16) * (x + 16) = x^2 := by grind
-
-example [CommSemiring α] [AddRightCancel α] (x y : α) :
-    x^2*y = 1 → x*y^2 = y → y*x = 1 := by grind
-
-example (a b : UInt32) : a ≤ 2 → b ≤ 3 → a + b ≤ 5 := by grind
-
-example [LE α] [Std.IsLinearPreorder α] (a b c d : α) :
-    a ≤ b → ¬ (c ≤ b) → ¬ (d ≤ c) → a ≤ d := by grind
-
-/- Theory combination. -/
-
-example [CommRing α] [NoNatZeroDivisors α]
-    (a b c : α) (f : α → Nat) :
-    a + b + c = 3 → a^2 + b^2 + c^2 = 5 → a^3 + b^3 + c^3 = 7 →
-    f (a^4 + b^4) + f (9 - c^4) ≠ 1 := by grind
-
-/- Interactive mode. -/
-
--- Remark: Mathlib contains the definition of `Real`, `sin`, and `cos`.
-axiom Real : Type
-instance : Lean.Grind.CommRing Real := sorry
-
-axiom cos : Real → Real
-axiom sin : Real → Real
-axiom trig_identity : ∀ x, (cos x)^2 + (sin x)^2 = 1
-
--- Manually specify the patterns for `trig_identity`
-grind_pattern trig_identity => cos x
-grind_pattern trig_identity => sin x
-
-example : (cos x + sin x)^2 = 2 * cos x * sin x + 1 := by
-  grind? -- Provides code action
-
-example : (cos x + sin x)^2 = 2 * cos x * sin x + 1 := by
-  grind =>
-    instantiate only [trig_identity]
-    ring

doc/examples/palindromes.lean

@@ -94,8 +94,10 @@ theorem List.palindrome_of_eq_reverse (h : as.reverse = as) : Palindrome as := b
   next   => exact Palindrome.nil
   next a => exact Palindrome.single a
   next a b as ih =>
-    obtain ⟨rfl, h, -⟩ := by simpa using h
-    exact Palindrome.sandwich b (ih h)
+    have : a = b := by simp_all
+    subst this
+    have : as.reverse = as := by simp_all
+    exact Palindrome.sandwich a (ih this)
 
 /-!
 We now define a function that returns `true` iff `as` is a palindrome.

doc/std/grove/GroveStdlib/Generated.lean

@@ -4,7 +4,6 @@ import GroveStdlib.Generated.«associative-creation-operations»
 import GroveStdlib.Generated.«associative-modification-operations»
 import GroveStdlib.Generated.«associative-create-then-query»
 import GroveStdlib.Generated.«associative-all-operations-covered»
-import GroveStdlib.Generated.«slice-producing»
 
 /-
 This file is autogenerated by grove. You can manually edit it, for example to resolve merge
@@ -21,4 +20,3 @@ def restoreState : RestoreStateM Unit := do
   «associative-modification-operations».restoreState
   «associative-create-then-query».restoreState
   «associative-all-operations-covered».restoreState
-  «slice-producing».restoreState

doc/std/grove/GroveStdlib/Generated/slice-producing.lean

@@ -1,459 +0,0 @@
-import Grove.Framework
-
-/-
-This file is autogenerated by grove. You can manually edit it, for example to resolve merge
-conflicts, but be careful.
--/
-
-open Grove.Framework Widget
-
-namespace GroveStdlib.Generated.«slice-producing»
-
-def «c8a13d6d-7ed6-4cd1-a386-23e2d55ce6f7» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "c8a13d6d-7ed6-4cd1-a386-23e2d55ce6f7"
-  rowId := "c8a13d6d-7ed6-4cd1-a386-23e2d55ce6f7"
-  rowState := #[⟨"String", "String.slice", Declaration.def {
-    name := `String.slice
-    renderedStatement := "String.slice (s : String) (startInclusive endExclusive : s.Pos)\n  (h : startInclusive ≤ endExclusive) : String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.slice", Declaration.def {
-    name := `String.Slice.slice
-    renderedStatement := "String.Slice.slice (s : String.Slice) (newStart newEnd : s.Pos) (h : newStart ≤ newEnd) :\n  String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"string-pos-forwards", "String.Pos.slice", Declaration.def {
-    name := `String.Pos.slice
-    renderedStatement := "String.Pos.slice {s : String} (pos p₀ p₁ : s.Pos) (h₁ : p₀ ≤ pos) (h₂ : pos ≤ p₁) :\n  (s.slice p₀ p₁ ⋯).Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-pos-backwards", "String.Pos.ofSlice", Declaration.def {
-    name := `String.Pos.ofSlice
-    renderedStatement := "String.Pos.ofSlice {s : String} {p₀ p₁ : s.Pos} {h : p₀ ≤ p₁} (pos : (s.slice p₀ p₁ h).Pos) : s.Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-slice-pos-forwards", "String.Slice.Pos.slice", Declaration.def {
-    name := `String.Slice.Pos.slice
-    renderedStatement := "String.Slice.Pos.slice {s : String.Slice} (pos p₀ p₁ : s.Pos) (h₁ : p₀ ≤ pos) (h₂ : pos ≤ p₁) :\n  (s.slice p₀ p₁ ⋯).Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-slice-pos-backwards", "String.Slice.Pos.ofSlice", Declaration.def {
-    name := `String.Slice.Pos.ofSlice
-    renderedStatement := "String.Slice.Pos.ofSlice {s : String.Slice} {p₀ p₁ : s.Pos} {h : p₀ ≤ p₁}\n  (pos : (s.slice p₀ p₁ h).Pos) : s.Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-pos-noproof", "String.Pos.sliceOrPanic", Declaration.def {
-    name := `String.Pos.sliceOrPanic
-    renderedStatement := "String.Pos.sliceOrPanic {s : String} (pos p₀ p₁ : s.Pos) {h : p₀ ≤ p₁} : (s.slice p₀ p₁ h).Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-slice-pos-noproof", "String.Slice.Pos.sliceOrPanic", Declaration.def {
-    name := `String.Slice.Pos.sliceOrPanic
-    renderedStatement := "String.Slice.Pos.sliceOrPanic {s : String.Slice} (pos p₀ p₁ : s.Pos) {h : p₀ ≤ p₁} :\n  (s.slice p₀ p₁ h).Pos"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .done
-    comment := ""
-  }
-def «21b4fdfd-f8b3-44f5-a59e-57f1dc1d6819» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "21b4fdfd-f8b3-44f5-a59e-57f1dc1d6819"
-  rowId := "21b4fdfd-f8b3-44f5-a59e-57f1dc1d6819"
-  rowState := #[⟨"String", "String.slice?", Declaration.def {
-    name := `String.slice?
-    renderedStatement := "String.slice? (s : String) (startInclusive endExclusive : s.Pos) : Option String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.slice?", Declaration.def {
-    name := `String.Slice.slice?
-    renderedStatement := "String.Slice.slice? (s : String.Slice) (newStart newEnd : s.Pos) : Option String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .postponed
-    comment := "Would be good to have better support"
-  }
-def «6f2b6ecb-2f0c-4e45-9da3-eb7f2e15eff0» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "6f2b6ecb-2f0c-4e45-9da3-eb7f2e15eff0"
-  rowId := "6f2b6ecb-2f0c-4e45-9da3-eb7f2e15eff0"
-  rowState := #[⟨"String", "String.slice!", Declaration.def {
-    name := `String.slice!
-    renderedStatement := "String.slice! (s : String) (p₁ p₂ : s.Pos) : String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.slice!", Declaration.def {
-    name := `String.Slice.slice!
-    renderedStatement := "String.Slice.slice! (s : String.Slice) (newStart newEnd : s.Pos) : String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"string-pos-forwards", "String.Pos.slice!", Declaration.def {
-    name := `String.Pos.slice!
-    renderedStatement := "String.Pos.slice! {s : String} (pos p₀ p₁ : s.Pos) : (s.slice! p₀ p₁).Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-pos-backwards", "String.Pos.ofSlice!", Declaration.def {
-    name := `String.Pos.ofSlice!
-    renderedStatement := "String.Pos.ofSlice! {s : String} {p₀ p₁ : s.Pos} (pos : (s.slice! p₀ p₁).Pos) : s.Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-slice-pos-forwards", "String.Slice.Pos.slice!", Declaration.def {
-    name := `String.Slice.Pos.slice!
-    renderedStatement := "String.Slice.Pos.slice! {s : String.Slice} (pos p₀ p₁ : s.Pos) : (s.slice! p₀ p₁).Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-slice-pos-backwards", "String.Slice.Pos.ofSlice!", Declaration.def {
-    name := `String.Slice.Pos.ofSlice!
-    renderedStatement := "String.Slice.Pos.ofSlice! {s : String.Slice} {p₀ p₁ : s.Pos} (pos : (s.slice! p₀ p₁).Pos) : s.Pos"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .done
-    comment := ""
-  }
-def «a3bdf66d-bc11-4019-aee9-2f1c1701de52» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "a3bdf66d-bc11-4019-aee9-2f1c1701de52"
-  rowId := "a3bdf66d-bc11-4019-aee9-2f1c1701de52"
-  rowState := #[⟨"String", "String.trimAsciiStart", Declaration.def {
-    name := `String.trimAsciiStart
-    renderedStatement := "String.trimAsciiStart (s : String) : String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.trimAsciiStart", Declaration.def {
-    name := `String.Slice.trimAsciiStart
-    renderedStatement := "String.Slice.trimAsciiStart (s : String.Slice) : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing `of` version at least"
-  }
-def «f12b2730-7a4d-465c-8a6d-9d051c300fd5» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "f12b2730-7a4d-465c-8a6d-9d051c300fd5"
-  rowId := "f12b2730-7a4d-465c-8a6d-9d051c300fd5"
-  rowState := #[⟨"String", "String.trimAsciiEnd", Declaration.def {
-    name := `String.trimAsciiEnd
-    renderedStatement := "String.trimAsciiEnd (s : String) : String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.trimAsciiEnd", Declaration.def {
-    name := `String.Slice.trimAsciiEnd
-    renderedStatement := "String.Slice.trimAsciiEnd (s : String.Slice) : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing `of` version at least"
-  }
-def «32307b55-d6d1-4756-a947-dbe4dfde573c» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "32307b55-d6d1-4756-a947-dbe4dfde573c"
-  rowId := "32307b55-d6d1-4756-a947-dbe4dfde573c"
-  rowState := #[⟨"String", "String.trimAscii", Declaration.def {
-    name := `String.trimAscii
-    renderedStatement := "String.trimAscii (s : String) : String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.trimAscii", Declaration.def {
-    name := `String.Slice.trimAscii
-    renderedStatement := "String.Slice.trimAscii (s : String.Slice) : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing `of` version at least\n"
-  }
-def «dce95a38-f55a-4d6a-ae79-078ffe4b5c15» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "dce95a38-f55a-4d6a-ae79-078ffe4b5c15"
-  rowId := "dce95a38-f55a-4d6a-ae79-078ffe4b5c15"
-  rowState := #[⟨"String", "String.toSlice", Declaration.def {
-    name := `String.toSlice
-    renderedStatement := "String.toSlice (s : String) : String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"string-pos-forwards", "String.Pos.toSlice", Declaration.def {
-    name := `String.Pos.toSlice
-    renderedStatement := "String.Pos.toSlice {s : String} (pos : s.Pos) : s.toSlice.Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-pos-backwards", "String.Pos.ofToSlice", Declaration.def {
-    name := `String.Pos.ofToSlice
-    renderedStatement := "String.Pos.ofToSlice {s : String} (pos : s.toSlice.Pos) : s.Pos"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .done
-    comment := ""
-  }
-def «005a3f30-5dab-493f-b168-32c36a2bdf7c» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "005a3f30-5dab-493f-b168-32c36a2bdf7c"
-  rowId := "005a3f30-5dab-493f-b168-32c36a2bdf7c"
-  rowState := #[⟨"String.Slice", "String.Slice.str", Declaration.def {
-    name := `String.Slice.str
-    renderedStatement := "String.Slice.str (self : String.Slice) : String"
-    isDeprecated := false
-  }
-⟩,⟨"string-slice-pos-forwards", "String.Slice.Pos.str", Declaration.def {
-    name := `String.Slice.Pos.str
-    renderedStatement := "String.Slice.Pos.str {s : String.Slice} (pos : s.Pos) : s.str.Pos"
-    isDeprecated := false
-  }
-⟩,⟨"string-slice-pos-backwards", "String.Slice.Pos.ofStr", Declaration.def {
-    name := `String.Slice.Pos.ofStr
-    renderedStatement := "String.Slice.Pos.ofStr {s : String.Slice} (pos : s.str.Pos) (h₁ : s.startInclusive ≤ pos)\n  (h₂ : pos ≤ s.endExclusive) : s.Pos"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing `no proof` version\n"
-  }
-def «5f1a154c-ae2f-43a1-9409-2ce95b163ef3» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "5f1a154c-ae2f-43a1-9409-2ce95b163ef3"
-  rowId := "5f1a154c-ae2f-43a1-9409-2ce95b163ef3"
-  rowState := #[⟨"String", "String.drop", Declaration.def {
-    name := `String.drop
-    renderedStatement := "String.drop (s : String) (n : Nat) : String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.drop", Declaration.def {
-    name := `String.Slice.drop
-    renderedStatement := "String.Slice.drop (s : String.Slice) (n : Nat) : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing position transformations"
-  }
-def «179518d1-ad07-4b2b-8ffe-3b7616e4c4ab» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "179518d1-ad07-4b2b-8ffe-3b7616e4c4ab"
-  rowId := "179518d1-ad07-4b2b-8ffe-3b7616e4c4ab"
-  rowState := #[⟨"String", "String.take", Declaration.def {
-    name := `String.take
-    renderedStatement := "String.take (s : String) (n : Nat) : String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.take", Declaration.def {
-    name := `String.Slice.take
-    renderedStatement := "String.Slice.take (s : String.Slice) (n : Nat) : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing position transformations"
-  }
-def «55c587fd-a7a8-4633-a4ae-e2c4e768ad28» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "55c587fd-a7a8-4633-a4ae-e2c4e768ad28"
-  rowId := "55c587fd-a7a8-4633-a4ae-e2c4e768ad28"
-  rowState := #[⟨"String", "String.dropWhile", Declaration.def {
-    name := `String.dropWhile
-    renderedStatement := "String.dropWhile {ρ : Type} (s : String) (pat : ρ) [String.Slice.Pattern.ForwardPattern pat] :\n  String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.dropWhile", Declaration.def {
-    name := `String.Slice.dropWhile
-    renderedStatement := "String.Slice.dropWhile {ρ : Type} (s : String.Slice) (pat : ρ)\n  [String.Slice.Pattern.ForwardPattern pat] : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing position transformations"
-  }
-def «d4444684-4279-4400-9be2-561a7cdb32c1» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "d4444684-4279-4400-9be2-561a7cdb32c1"
-  rowId := "d4444684-4279-4400-9be2-561a7cdb32c1"
-  rowState := #[⟨"String", "String.takeWhile", Declaration.def {
-    name := `String.takeWhile
-    renderedStatement := "String.takeWhile {ρ : Type} (s : String) (pat : ρ) [String.Slice.Pattern.ForwardPattern pat] :\n  String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.takeWhile", Declaration.def {
-    name := `String.Slice.takeWhile
-    renderedStatement := "String.Slice.takeWhile {ρ : Type} (s : String.Slice) (pat : ρ)\n  [String.Slice.Pattern.ForwardPattern pat] : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing position transformations"
-  }
-def «1c9e6689-65a0-4d4b-b001-256e83917d98» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "1c9e6689-65a0-4d4b-b001-256e83917d98"
-  rowId := "1c9e6689-65a0-4d4b-b001-256e83917d98"
-  rowState := #[⟨"String", "String.dropEndWhile", Declaration.def {
-    name := `String.dropEndWhile
-    renderedStatement := "String.dropEndWhile {ρ : Type} (s : String) (pat : ρ) [String.Slice.Pattern.BackwardPattern pat] :\n  String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.dropEndWhile", Declaration.def {
-    name := `String.Slice.dropEndWhile
-    renderedStatement := "String.Slice.dropEndWhile {ρ : Type} (s : String.Slice) (pat : ρ)\n  [String.Slice.Pattern.BackwardPattern pat] : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing position transformations"
-  }
-def «b836052b-3470-4a8e-8989-6951c898de37» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "b836052b-3470-4a8e-8989-6951c898de37"
-  rowId := "b836052b-3470-4a8e-8989-6951c898de37"
-  rowState := #[⟨"String", "String.takeEndWhile", Declaration.def {
-    name := `String.takeEndWhile
-    renderedStatement := "String.takeEndWhile {ρ : Type} (s : String) (pat : ρ) [String.Slice.Pattern.BackwardPattern pat] :\n  String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.takeEndWhile", Declaration.def {
-    name := `String.Slice.takeEndWhile
-    renderedStatement := "String.Slice.takeEndWhile {ρ : Type} (s : String.Slice) (pat : ρ)\n  [String.Slice.Pattern.BackwardPattern pat] : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing position transformations"
-  }
-def «5aa777d8-9642-43d8-9e20-30400fb8bb9d» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "5aa777d8-9642-43d8-9e20-30400fb8bb9d"
-  rowId := "5aa777d8-9642-43d8-9e20-30400fb8bb9d"
-  rowState := #[⟨"String", "String.dropPrefix", Declaration.def {
-    name := `String.dropPrefix
-    renderedStatement := "String.dropPrefix {ρ : Type} (s : String) (pat : ρ) [String.Slice.Pattern.ForwardPattern pat] :\n  String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.dropPrefix", Declaration.def {
-    name := `String.Slice.dropPrefix
-    renderedStatement := "String.Slice.dropPrefix {ρ : Type} (s : String.Slice) (pat : ρ)\n  [String.Slice.Pattern.ForwardPattern pat] : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing position transformations"
-  }
-def «80e3869d-fcfe-459d-8433-fe221f7b3c7a» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "80e3869d-fcfe-459d-8433-fe221f7b3c7a"
-  rowId := "80e3869d-fcfe-459d-8433-fe221f7b3c7a"
-  rowState := #[⟨"String", "String.dropSuffix", Declaration.def {
-    name := `String.dropSuffix
-    renderedStatement := "String.dropSuffix {ρ : Type} (s : String) (pat : ρ) [String.Slice.Pattern.BackwardPattern pat] :\n  String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.dropSuffix", Declaration.def {
-    name := `String.Slice.dropSuffix
-    renderedStatement := "String.Slice.dropSuffix {ρ : Type} (s : String.Slice) (pat : ρ)\n  [String.Slice.Pattern.BackwardPattern pat] : String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .bad
-    comment := "Missing position transformations"
-  }
-def «4feda3e0-903b-4d52-b34e-0af70f7866e0» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "4feda3e0-903b-4d52-b34e-0af70f7866e0"
-  rowId := "4feda3e0-903b-4d52-b34e-0af70f7866e0"
-  rowState := #[⟨"String", "String.dropPrefix?", Declaration.def {
-    name := `String.dropPrefix?
-    renderedStatement := "String.dropPrefix? {ρ : Type} (s : String) (pat : ρ) [String.Slice.Pattern.ForwardPattern pat] :\n  Option String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.dropPrefix?", Declaration.def {
-    name := `String.Slice.dropPrefix?
-    renderedStatement := "String.Slice.dropPrefix? {ρ : Type} (s : String.Slice) (pat : ρ)\n  [String.Slice.Pattern.ForwardPattern pat] : Option String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .postponed
-    comment := "Missing position transformations"
-  }
-def «45ca44c8-fbd5-4400-8297-a60778f302b0» : AssociationTable.Fact .declaration where
-  widgetId := "slice-producing"
-  factId := "45ca44c8-fbd5-4400-8297-a60778f302b0"
-  rowId := "45ca44c8-fbd5-4400-8297-a60778f302b0"
-  rowState := #[⟨"String", "String.dropSuffix?", Declaration.def {
-    name := `String.dropSuffix?
-    renderedStatement := "String.dropSuffix? {ρ : Type} (s : String) (pat : ρ) [String.Slice.Pattern.BackwardPattern pat] :\n  Option String.Slice"
-    isDeprecated := false
-  }
-⟩,⟨"String.Slice", "String.Slice.dropSuffix?", Declaration.def {
-    name := `String.Slice.dropSuffix?
-    renderedStatement := "String.Slice.dropSuffix? {ρ : Type} (s : String.Slice) (pat : ρ)\n  [String.Slice.Pattern.BackwardPattern pat] : Option String.Slice"
-    isDeprecated := false
-  }
-⟩,]
-  metadata := {
-    status := .postponed
-    comment := "Missing position transformations"
-  }
-
-def table : AssociationTable.Data .declaration where
-  widgetId := "slice-producing"
-  rows := #[
-    ⟨"c8a13d6d-7ed6-4cd1-a386-23e2d55ce6f7", "slice", #[⟨"String", "String.slice"⟩,⟨"String.Slice", "String.Slice.slice"⟩,⟨"string-pos-forwards", "String.Pos.slice"⟩,⟨"string-pos-backwards", "String.Pos.ofSlice"⟩,⟨"string-slice-pos-forwards", "String.Slice.Pos.slice"⟩,⟨"string-slice-pos-backwards", "String.Slice.Pos.ofSlice"⟩,⟨"string-pos-noproof", "String.Pos.sliceOrPanic"⟩,⟨"string-slice-pos-noproof", "String.Slice.Pos.sliceOrPanic"⟩,]⟩,
-    ⟨"21b4fdfd-f8b3-44f5-a59e-57f1dc1d6819", "slice?", #[⟨"String", "String.slice?"⟩,⟨"String.Slice", "String.Slice.slice?"⟩,]⟩,
-    ⟨"6f2b6ecb-2f0c-4e45-9da3-eb7f2e15eff0", "slice!", #[⟨"String", "String.slice!"⟩,⟨"String.Slice", "String.Slice.slice!"⟩,⟨"string-pos-forwards", "String.Pos.slice!"⟩,⟨"string-pos-backwards", "String.Pos.ofSlice!"⟩,⟨"string-slice-pos-forwards", "String.Slice.Pos.slice!"⟩,⟨"string-slice-pos-backwards", "String.Slice.Pos.ofSlice!"⟩,]⟩,
-    ⟨"a3bdf66d-bc11-4019-aee9-2f1c1701de52", "trimAsciiStart", #[⟨"String", "String.trimAsciiStart"⟩,⟨"String.Slice", "String.Slice.trimAsciiStart"⟩,]⟩,
-    ⟨"f12b2730-7a4d-465c-8a6d-9d051c300fd5", "trimAsciiEnd", #[⟨"String", "String.trimAsciiEnd"⟩,⟨"String.Slice", "String.Slice.trimAsciiEnd"⟩,]⟩,
-    ⟨"32307b55-d6d1-4756-a947-dbe4dfde573c", "trimAscii", #[⟨"String", "String.trimAscii"⟩,⟨"String.Slice", "String.Slice.trimAscii"⟩,]⟩,
-    ⟨"dce95a38-f55a-4d6a-ae79-078ffe4b5c15", "toSlice", #[⟨"String", "String.toSlice"⟩,⟨"string-pos-forwards", "String.Pos.toSlice"⟩,⟨"string-pos-backwards", "String.Pos.ofToSlice"⟩,]⟩,
-    ⟨"005a3f30-5dab-493f-b168-32c36a2bdf7c", "str", #[⟨"String.Slice", "String.Slice.str"⟩,⟨"string-slice-pos-forwards", "String.Slice.Pos.str"⟩,⟨"string-slice-pos-backwards", "String.Slice.Pos.ofStr"⟩,]⟩,
-    ⟨"5f1a154c-ae2f-43a1-9409-2ce95b163ef3", "drop", #[⟨"String", "String.drop"⟩,⟨"String.Slice", "String.Slice.drop"⟩,]⟩,
-    ⟨"179518d1-ad07-4b2b-8ffe-3b7616e4c4ab", "take", #[⟨"String", "String.take"⟩,⟨"String.Slice", "String.Slice.take"⟩,]⟩,
-    ⟨"55c587fd-a7a8-4633-a4ae-e2c4e768ad28", "dropWhile", #[⟨"String", "String.dropWhile"⟩,⟨"String.Slice", "String.Slice.dropWhile"⟩,]⟩,
-    ⟨"d4444684-4279-4400-9be2-561a7cdb32c1", "takeWhile", #[⟨"String", "String.takeWhile"⟩,⟨"String.Slice", "String.Slice.takeWhile"⟩,]⟩,
-    ⟨"1c9e6689-65a0-4d4b-b001-256e83917d98", "dropEndWhile", #[⟨"String", "String.dropEndWhile"⟩,⟨"String.Slice", "String.Slice.dropEndWhile"⟩,]⟩,
-    ⟨"b836052b-3470-4a8e-8989-6951c898de37", "takeEndWhile", #[⟨"String", "String.takeEndWhile"⟩,⟨"String.Slice", "String.Slice.takeEndWhile"⟩,]⟩,
-    ⟨"5aa777d8-9642-43d8-9e20-30400fb8bb9d", "dropPrefix", #[⟨"String", "String.dropPrefix"⟩,⟨"String.Slice", "String.Slice.dropPrefix"⟩,]⟩,
-    ⟨"80e3869d-fcfe-459d-8433-fe221f7b3c7a", "dropSuffix", #[⟨"String", "String.dropSuffix"⟩,⟨"String.Slice", "String.Slice.dropSuffix"⟩,]⟩,
-    ⟨"4feda3e0-903b-4d52-b34e-0af70f7866e0", "dropPrefix?", #[⟨"String", "String.dropPrefix?"⟩,⟨"String.Slice", "String.Slice.dropPrefix?"⟩,]⟩,
-    ⟨"45ca44c8-fbd5-4400-8297-a60778f302b0", "dropSuffix?", #[⟨"String", "String.dropSuffix?"⟩,⟨"String.Slice", "String.Slice.dropSuffix?"⟩,]⟩,
-  ]
-  facts := #[
-    «c8a13d6d-7ed6-4cd1-a386-23e2d55ce6f7»,
-    «21b4fdfd-f8b3-44f5-a59e-57f1dc1d6819»,
-    «6f2b6ecb-2f0c-4e45-9da3-eb7f2e15eff0»,
-    «a3bdf66d-bc11-4019-aee9-2f1c1701de52»,
-    «f12b2730-7a4d-465c-8a6d-9d051c300fd5»,
-    «32307b55-d6d1-4756-a947-dbe4dfde573c»,
-    «dce95a38-f55a-4d6a-ae79-078ffe4b5c15»,
-    «005a3f30-5dab-493f-b168-32c36a2bdf7c»,
-    «5f1a154c-ae2f-43a1-9409-2ce95b163ef3»,
-    «179518d1-ad07-4b2b-8ffe-3b7616e4c4ab»,
-    «55c587fd-a7a8-4633-a4ae-e2c4e768ad28»,
-    «d4444684-4279-4400-9be2-561a7cdb32c1»,
-    «1c9e6689-65a0-4d4b-b001-256e83917d98»,
-    «b836052b-3470-4a8e-8989-6951c898de37»,
-    «5aa777d8-9642-43d8-9e20-30400fb8bb9d»,
-    «80e3869d-fcfe-459d-8433-fe221f7b3c7a»,
-    «4feda3e0-903b-4d52-b34e-0af70f7866e0»,
-    «45ca44c8-fbd5-4400-8297-a60778f302b0»,
-  ]
-
-def restoreState : RestoreStateM Unit := do
-  addAssociationTable table

doc/std/grove/GroveStdlib/Std.lean

@@ -15,7 +15,7 @@ namespace GroveStdlib
 namespace Std
 
 def introduction : Node :=
-  .text ⟨"introduction", "Welcome to the interactive Lean standard library outline!"⟩
+  .text "Welcome to the interactive Lean standard library outline!"
 
 end Std
 

doc/std/grove/GroveStdlib/Std/CoreTypesAndOperations/StringsAndFormatting.lean

@@ -11,87 +11,9 @@ namespace GroveStdlib.Std.CoreTypesAndOperations
 
 namespace StringsAndFormatting
 
-open Lean Meta
-
-def introduction : Text where
-  id := "string-introduction"
-  content := Grove.Markdown.render [
-    .h1 "The Lean string library",
-    .text "The Lean standard library contains a fully-featured string library, centered around the types `String` and `String.Slice`.",
-    .text "`String` is defined as the subtype of `ByteArray` of valid UTF-8 strings. A `String.Slice` is a `String` together with a start and end position.",
-    .text "`String` is equivalent to `List Char`, but it has a more efficient runtime representation. While the logical model based on `ByteArray` is overwritten in the runtime, the runtime implementation is very similar to the logical model, with the main difference being that the length of a string in Unicode code points is cached in the runtime implementation.",
-    .text "We are considering removing this feature in the future (i.e., deprecating `String.length`), as the number of UTF-8 codepoints in a string is not particularly useful, and if needed it can be computed in linear time using `s.positions.count`."
-  ]
-
-def highLevelStringTypes : List Lean.Name :=
-  [`String, `String.Slice, `String.Pos, `String.Slice.Pos]
-
-def creatingStringsAndSlices : Text where
-  id := "transforming-strings-and-slices"
-  content := Grove.Markdown.render [
-    .h2 "Transforming strings and slices",
-    .text "The Lean standard library contains a number of functions that take one or more strings and slices and return a string or a slice.",
-    .text "If possible, these functions should avoid allocating a new string, and return a slice of their input(s) instead.",
-    .text "Usually, for every operation `f`, there will be functions `String.f` and `String.Slice.f`, where `String.f s` is defined as `String.Slice.f s.toSlice`.",
-    .text "In particular, functions that transform strings and slices should live in the `String` and `String.Slice` namespaces even if they involve a `String.Pos`/`String.Slice.Pos` (like `String.sliceTo`), for reasons that will become clear shortly.",
-
-    .h3 "Transforming positions",
-    .text "Since positions on strings and slices are dependent on the string or slice, whenever users transform a string/slice, they will be interested in interpreting positions on the original string/slice as positions on the result, or vice versa.",
-    .text "Consequently, every operation that transforms a string or slice should come with a corresponding set of transformations between positions, usually in both directions, possibly with one of the directions being conditional.",
-    .text "For example, given a string `s` and a position `p` on `s`, we have the slice `s.sliceFrom p`, which is the slice from `p` to the end of `s`. A position on `s.sliceFrom p` can always be interpreted as a position on `s`. This is the \"backwards\" transformation. Conversely, a position `q` on `s` can be interpreted as a position on `s.sliceFrom p` as long as `p ≤ q`. This is the conditional forwards direction.",
-    .text "The convention for naming these transformations is that the forwards transformation should have the same name as the transformation on strings/slices, but it should be located in the `String.Pos` or `String.Slice.Pos` namespace, depending on the type of the starting position (so that dot notation is possible for the forward direction). The backwards transformation should have the same name as the operation on strings/slices, but with an `of` prefix, and live in the same namespace as the forwards transformation (so in general dot notation will not be available).",
-    .text "So, in the `sliceFrom` example, the forward direction would be called `String.Pos.sliceFrom`, while the backwards direction should be called `String.Pos.ofSliceFrom` (not `String.Slice.Pos.ofSliceFrom`).",
-    .text "If one of the directions is conditional, it should have a corresponding panicking operation that does not require a proof; in our example this would be `String.Pos.sliceFrom!`.",
-    .text "Sometimes there is a name clash for the panicking operations if the operation on strings is already panicking. For example, there are both `String.slice` and `String.slice!`. If the original operation is already panicking, we only provide panicking transformation operations. But now `String.Pos.slice!` could refer both to the panicking forwards transformation associated with `String.slice`, and also to the (only) forwards transformation associated with `String.slice!`. In this situation, we use an `orPanic` suffix to disambiguate. So the panicking forwards operation associated with `String.slice` is called `String.Pos.sliceOrPanic`, and the forwards operation associated with `String.slice!` is called `String.Pos.slice!`."
-  ]
-
--- TODO: also include the `HAppend` instance(s)
-def sliceProducing : AssociationTable (β := Alias Lean.Name) .declaration
-    [`String, `String.Slice,
-     Alias.mk `String.Pos "string-pos-forwards" "String.Pos (forwards)",
-     Alias.mk `String.Pos "string-pos-backwards" "String.Pos (backwards)",
-     Alias.mk `String.Pos "string-pos-noproof" "String.Pos (no proof)",
-     Alias.mk `String.Slice.Pos "string-slice-pos-forwards" "String.Slice.Pos (forwards)",
-     Alias.mk `String.Slice.Pos "string-slice-pos-backwards" "String.Slice.Pos (backwards)",
-     Alias.mk `String.Slice.Pos "string-slice-pos-noproof" "String.Slice.Pos (no proof)"] where
-  id := "slice-producing"
-  title := "String functions returning strings or slices"
-  description := "Operations on strings and string slices that themselves return a new string slice."
-  dataSources n := DataSource.definitionsInNamespace n.inner
-
-def sliceProducingComplete : Assertion where
-  widgetId := "slice-producing-complete"
-  title := "Slice-producing table is complete"
-  description := "All functions in the `String.**` namespace that return a string or a slice are covered in the table"
-  check := do
-    let mut ans := #[]
-    let covered := Std.HashSet.ofArray (← valuesInAssociationTable sliceProducing)
-    let pred : DataSource.DeclarationPredicate :=
-      DataSource.DeclarationPredicate.all [.isDefinition, .not .isDeprecated,
-        .notInNamespace `String.Pos.Raw, .notInNamespace `String.Legacy,
-        .not .isInstance]
-    let env ← getEnv
-    for name in ← declarationsMatching `String pred do
-      let some c := env.find? name | continue
-      if c.type.getForallBody.getUsedConstants.any (fun n => n == ``String || n == ``String.Slice) then
-        let success : Bool := name.toString ∈ covered
-        ans := ans.push {
-          assertionId := name.toString
-          description := s!"`{name}` should appear in the table."
-          passed := success
-          message := s!"`{name}` was{if success then "" else " not"} found in the table."
-        }
-    return ans
-
 end StringsAndFormatting
 
-open StringsAndFormatting
-
 def stringsAndFormatting : Node :=
-  .section "strings-and-formatting" "Strings and formatting"
-    #[.text introduction,
-      .text creatingStringsAndSlices,
-      .associationTable sliceProducing,
-      .assertion sliceProducingComplete]
+  .section "strings-and-formatting" "Strings and formatting" #[]
 
-end GroveStdlib.Std.CoreTypesAndOperations
+end GroveStdlib.Std.CoreTypesAndOperations

doc/std/grove/lake-manifest.json

@@ -5,7 +5,7 @@
    "type": "git",
    "subDir": "backend",
    "scope": "",
-   "rev": "c580a425c9b7fa2aebaec2a1d8de16b2e2283c40",
+   "rev": "3e8aabdea58c11813c5d3b7eeb187ded44ee9a34",
    "name": "grove",
    "manifestFile": "lake-manifest.json",
    "inputRev": "master",
@@ -15,10 +15,10 @@
    "type": "git",
    "subDir": null,
    "scope": "leanprover",
-   "rev": "d9fc8ae23024be37424a189982c92356e37935c8",
+   "rev": "1604206fcd0462da9a241beeac0e2df471647435",
    "name": "Cli",
    "manifestFile": "lake-manifest.json",
-   "inputRev": "nightly-testing",
+   "inputRev": "main",
    "inherited": true,
    "configFile": "lakefile.toml"}],
  "name": "grovestdlib",

doc/style.md

@@ -810,7 +810,7 @@ Docstrings for constants should have the following structure:
 
 The **short summary** should be 1–3 sentences (ideally 1) and provide
 enough information for most readers to quickly decide whether the
-constant is relevant to their task. The first (or only) sentence of
+docstring is relevant to their task. The first (or only) sentence of
 the short summary should be a *sentence fragment* in which the subject
 is implied to be the documented item, written in present tense
 indicative, or a *noun phrase* that characterizes the documented
@@ -1123,110 +1123,6 @@ infix:50 " ⇔ " => Bijection
 recommended_spelling "bij" for "⇔" in [Bijection, «term_⇔_»]
 ```
 
-#### Tactics
-
-Docstrings for tactics should have the following structure:
-
-* Short summary
-* Details
-* Variants
-* Examples
-
-Sometimes more than one declaration is needed to implement what the user
-sees as a single tactic. In that case, only one declaration should have
-the associated docstring, and the others should have the `tactic_alt`
-attribute to mark them as an implementation detail.
-
-The **short summary** should be 1–3 sentences (ideally 1) and provide
-enough information for most readers to quickly decide whether the
-tactic is relevant to their task. The first (or only) sentence of
-the short summary should be a full sentence in which the subject
-is an example invocation of the tactic, written in present tense
-indicative. If the example tactic invocation names parameters, then the
-short summary may refer to them. For the example invocation, prefer the
-simplest or most typical example. Explain more complicated forms in the
-variants section. If needed, abbreviate the invocation by naming part of
-the syntax and expanding it in the next sentence. The summary should be
-written as a single paragraph.
-
-**Details**, if needed, may be 1-3 paragraphs that describe further
-relevant information. They may insert links as needed. This section
-should fully explain the scope of the tactic: its syntax format,
-on which goals it works and what the resulting goal(s) look like. It
-should be clear whether the tactic fails if it does not close the main
-goal and whether it creates any side goals. The details may include
-explanatory examples that can’t necessarily be machine checked and
-don’t fit the format.
-
-If the tactic is extensible using `macro_rules`, mention this in the
-details, with a link to `lean-manual://section/tactic-macro-extension`
-and give a one-line example. If the tactic provides an attribute or a
-command that allows the user to extend its behavior, the documentation
-on how to extend the tactic belongs to that attribute or command. In the
-tactic docstring, use a single sentence to refer the reader to this
-further documentation.
-
-**Variants**, if needed, should be a bulleted list describing different
-options and forms of the same tactic. The reader should be able to parse
-and understand the parts of a tactic invocation they are hovering over,
-using this list. Each list item should describe an individual variant
-and take one of two formats: the **short summary** as above, or a
-**named list item**. A named list item consists of a title in bold
-followed by an indented short paragraph.
-
-Variants should be explained from the perspective of the tactic's users, not
-their implementers. A tactic that is implemented as a single Lean parser may
-have multiple variants from the perspective of users, while a tactic that is
-implemented as multiple parsers may have no variants, but merely an optional
-part of the syntax.
-
-**Examples** should start with the line `Examples:` (or `Example:` if
-there’s exactly one). The section should consist of a sequence of code
-blocks, each showing a Lean declaration (usually with the `example`
-keyword) that invokes the tactic. When the effect of the tactic is not
-clear from the code, you can use code comments to describe this. Do
-not include text between examples, because it can be unclear whether
-the text refers to the code before or after the example.
-
-##### Example
-
-````
-`rw [e]` uses the expression `e` as a rewrite rule on the main goal,
-then tries to close the goal by "cheap" (reducible) `rfl`.
-
-If `e` is a defined constant, then the equational theorems associated with `e`
-are used. This provides a convenient way to unfold `e`. If `e` has parameters,
-the tactic will try to fill these in by unification with the matching part of
-the target. Parameters are only filled in once per rule, restricting which
-later rewrites can be found. Parameters that are not filled in after
-unification will create side goals. If the `rfl` fails to close the main goal,
-no error is raised.
-
-`rw` may fail to rewrite terms "under binders", such as `∀ x, ...` or `∃ x,
-...`. `rw` can also fail with a "motive is type incorrect" error in the context
-of dependent types. In these cases, consider using `simp only`.
-
-* `rw [e₁, ... eₙ]` applies the given rules sequentially.
-* `rw [← e]` or `rw [<- e]` applies the rewrite in the reverse direction.
-* `rw [e] at l` rewrites with `e` at location(s) `l`.
-* `rw (occs := .pos L) [e]`, where `L` is a literal list of natural numbers,
-  only rewrites the given occurrences in the target. Occurrences count from 1.
-* `rw (occs := .neg L) [e]`, where `L` is a literal list of natural numbers,
-  skips rewriting the given occurrences in the target. Occurrences count from 1.
-
-Examples:
-
-```lean
-example {a b : Nat} (h : a + a = b) : (a + a) + (a + a) = b + b := by rw [h]
-```
-
-```lean
-example {f : Nat -> Nat} (h : ∀ x, f x = 1) (a b : Nat) : f a = f b := by
-  rw [h] -- `rw` instantiates `h` only once, so this is equivalent to: `rw [h a]`
-  -- goal: ⊢ 1 = f b
-  rw [h] -- equivalent to: `rw [h b]`
-```
-````
 
 
 ## Dictionary

flake.lock

@@ -2,11 +2,11 @@
   "nodes": {
     "nixpkgs": {
       "locked": {
-        "lastModified": 1769018530,
-        "narHash": "sha256-S/5RU76BdQ32bbE99a+G9gMuatpVWEvIfeSjEqyoFS4=",
-        "rev": "88d3861acdd3d2f0e361767018218e51810df8a1",
+        "lastModified": 1745636243,
+        "narHash": "sha256-kbNvlQZf8wwok3d2X1kM/TlXH/MZ+03ZNv+IPPBx+DM=",
+        "rev": "f771eb401a46846c1aebd20552521b233dd7e18b",
         "type": "tarball",
-        "url": "https://releases.nixos.org/nixos/unstable/nixos-26.05pre931542.88d3861acdd3/nixexprs.tar.xz"
+        "url": "https://releases.nixos.org/nixos/unstable/nixos-25.05pre789333.f771eb401a46/nixexprs.tar.xz"
       },
       "original": {
         "type": "tarball",

flake.nix

@@ -18,14 +18,13 @@
       # An old nixpkgs for creating releases with an old glibc
       pkgsDist-old-aarch = import inputs.nixpkgs-old { localSystem.config = "aarch64-unknown-linux-gnu"; };
 
-      llvmPackages = pkgs.llvmPackages_19;
+      llvmPackages = pkgs.llvmPackages_15;
 
       devShellWithDist = pkgsDist: pkgs.mkShell.override {
           stdenv = pkgs.overrideCC pkgs.stdenv llvmPackages.clang;
         } ({
           buildInputs = with pkgs; [
-            cmake gmp libuv ccache pkg-config openssl
-            llvmPackages.bintools  # wrapped lld
+            cmake gmp libuv ccache pkg-config
             llvmPackages.llvm  # llvm-symbolizer for asan/lsan
             gdb
             tree  # for CI

lean.code-workspace

@@ -8,15 +8,9 @@
 		},
 		{
 			"path": "tests"
-		},
-		{
-			"path": "script"
 		}
 	],
 	"settings": {
-		// Open terminal at root, not current workspace folder
-		// (there is not way to directly refer to the root folder included as `.` above)
-		"terminal.integrated.cwd": "${workspaceFolder:src}/..",
 		"files.insertFinalNewline": true,
 		"files.trimTrailingWhitespace": true,
 		"cmake.buildDirectory": "${workspaceFolder}/build/release",

releases_drafts/module-system.md

@@ -1,54 +0,0 @@
-This release introduces the Lean module system, which allows files to
-control the visibility of their contents for other files. In previous
-releases, this feature was available as a preview when the option
-`experimental.module` was set to `true`; it is now a fully supported
-feature of Lean.
-
-# Benefits
-
-Because modules reduce the amount of information exposed to other
-code, they speed up rebuilds because irrelevant changes can be
-ignored, they make it possible to be deliberate about API evolution by
-hiding details that may change from clients, they help proofs be
-checked faster by avoiding accidentally unfolding definitions, and
-they lead to smaller executable files through improved dead code
-elimination.
-
-# Visibility
-
-A source file is a module if it begins with the `module` keyword.  By
-default, declarations in a module are private; the `public` modifier
-exports them. Proofs of theorems and bodies of definitions are private
-by default even when their signatures are public; the bodies of
-definitions can be made public by adding the `@[expose]`
-attribute. Theorems and opaque constants never expose their bodies.
-
-`public section` and `@[expose] section` change the default visibility
-of declarations in the section.
-
-# Imports
-
-Modules may only import other modules. By default, `import` adds the
-public information of the imported module to the private scope of the
-current module. Adding the `public` modifier to an import places the
-imported modules's public information in the public scope of the
-current module, exposing it in turn to the current module's clients.
-
-Within a package, `import all` can be used to import another module's
-private scope into the current module; this can be used to separate
-lemmas or tests from definition modules without exposing details to
-downstream clients.
-
-# Meta Code
-
-Code used in metaprograms must be marked `meta`. This ensures that the
-code is compiled and available for execution when it is needed during
-elaboration. Meta code may only reference other meta code. A whole
-module can be made available in the meta phase using `meta import`;
-this allows code to be shared across phases by importing the module in
-each phase. Code that is reachable from public metaprograms must be
-imported via `public meta import`, while local metaprograms can use
-plain `meta import` for their dependencies.
-
-
-The module system is described in detail in [the Lean language reference](https://lean-reference-manual-review.netlify.app/find/?domain=Verso.Genre.Manual.section&name=files).

script/AnalyzeGrindAnnotations.lean

@@ -0,0 +1,132 @@
+/-
+Copyright (c) 2025 Amazon.com, Inc. or its affiliates. All Rights Reserved.
+Released under Apache 2.0 license as described in the file LICENSE.
+Authors: Leonardo de Moura
+-/
+import Lean
+
+namespace Lean.Meta.Grind.Analyzer
+
+
+/-!
+A simple E-matching annotation analyzer.
+For each theorem annotated as an E-matching candidate, it creates an artificial goal, executes `grind` and shows the
+number of instances created.
+For a theorem of the form `params -> type`, the artificial goal is of the form `params -> type -> False`.
+-/
+
+/--
+`grind` configuration for the analyzer. We disable case-splits and lookahead,
+increase the number of generations, and limit the number of instances generated.
+-/
+def config : Grind.Config := {
+  splits       := 0
+  lookahead    := false
+  mbtc         := false
+  ematch       := 20
+  instances    := 100
+  gen          := 10
+}
+
+structure Config where
+  /-- Minimum number of instantiations to trigger summary report -/
+  min : Nat := 10
+  /-- Minimum number of instantiations to trigger detailed report -/
+  detailed : Nat := 50
+
+def mkParams : MetaM Params := do
+  let params ← Grind.mkParams config
+  let ematch ← getEMatchTheorems
+  let casesTypes ← Grind.getCasesTypes
+  return { params with ematch, casesTypes }
+
+/-- Returns the total number of generated instances.  -/
+private def sum (cs : PHashMap Origin Nat) : Nat := Id.run do
+  let mut r := 0
+  for (_, c) in cs do
+    r := r + c
+  return r
+
+private def thmsToMessageData (thms : PHashMap Origin Nat) : MetaM MessageData := do
+  let data := thms.toArray.filterMap fun (origin, c) =>
+    match origin with
+    | .decl declName => some (declName, c)
+    | _ => none
+  let data := data.qsort fun (d₁, c₁) (d₂, c₂) => if c₁ == c₂ then Name.lt d₁ d₂ else c₁ > c₂
+  let data ← data.mapM fun (declName, counter) =>
+    return .trace { cls := `thm } m!"{.ofConst (← mkConstWithLevelParams declName)} ↦ {counter}" #[]
+  return .trace { cls := `thm } "instances" data
+
+/--
+Analyzes theorem `declName`. That is, creates the artificial goal based on `declName` type,
+and invokes `grind` on it.
+-/
+def analyzeEMatchTheorem (declName : Name) (c : Config) : MetaM Unit := do
+  let info ← getConstInfo declName
+  let mvarId ← forallTelescope info.type fun _ type => do
+    withLocalDeclD `h type fun _ => do
+      return (← mkFreshExprMVar (mkConst ``False)).mvarId!
+  let result ← Grind.main mvarId (← mkParams) (pure ())
+  let thms := result.counters.thm
+  let s := sum thms
+  if s > c.min then
+    IO.println s!"{declName} : {s}"
+  if s > c.detailed then
+    logInfo m!"{declName}\n{← thmsToMessageData thms}"
+
+-- Not sure why this is failing: `down_pure` perhaps has an unnecessary universe parameter?
+run_meta analyzeEMatchTheorem ``Std.Do.SPred.down_pure {}
+
+/-- Analyzes all theorems in the standard library marked as E-matching theorems. -/
+def analyzeEMatchTheorems (c : Config := {}) : MetaM Unit := do
+  let origins := (← getEMatchTheorems).getOrigins
+  let decls := origins.filterMap fun | .decl declName => some declName | _ => none
+  for declName in decls.mergeSort Name.lt do
+    try
+      analyzeEMatchTheorem declName c
+    catch e =>
+      logError m!"{declName} failed with {e.toMessageData}"
+  logInfo m!"Finished analyzing {decls.length} theorems"
+
+/-- Macro for analyzing E-match theorems with unlimited heartbeats -/
+macro "#analyzeEMatchTheorems" : command => `(
+  set_option maxHeartbeats 0 in
+  run_meta analyzeEMatchTheorems
+)
+
+#analyzeEMatchTheorems
+
+-- -- We can analyze specific theorems using commands such as
+set_option trace.grind.ematch.instance true
+
+-- 1. grind immediately sees `(#[] : Array α) = ([] : List α).toArray` but probably this should be hidden.
+-- 2. `Vector.toArray_empty` keys on `Array.mk []` rather than `#v[].toArray`
+-- I guess we could add `(#[].extract _ _).extract _ _` as a stop pattern.
+run_meta analyzeEMatchTheorem ``Array.extract_empty {}
+
+-- Neither `Option.bind_some` nor `Option.bind_fun_some` fire, because the terms appear inside
+-- lambdas. So we get crazy things like:
+-- `fun x => ((some x).bind some).bind fun x => (some x).bind fun x => (some x).bind some`
+-- We could consider replacing `filterMap_some` with
+-- `filterMap g (filterMap f xs) = filterMap (f >=> g) xs`
+-- to avoid the lambda that `grind` struggles with, but this would require more API around the fish.
+run_meta analyzeEMatchTheorem ``Array.filterMap_some {}
+
+-- Not entirely certain what is wrong here, but certainly
+-- `eq_empty_of_append_eq_empty` is firing too often.
+-- Ideally we could instantiate this is we fine `xs ++ ys` in the same equivalence class,
+-- note just as soon as we see `xs ++ ys`.
+-- I've tried removing this in https://github.com/leanprover/lean4/pull/10162
+run_meta analyzeEMatchTheorem ``Array.range'_succ {}
+
+-- Perhaps the same story here.
+run_meta analyzeEMatchTheorem ``Array.range_succ {}
+
+-- `zip_map_left` and `zip_map_right` are bad grind lemmas,
+-- checking if they can be removed in https://github.com/leanprover/lean4/pull/10163
+run_meta analyzeEMatchTheorem ``Array.zip_map {}
+
+-- It seems crazy to me that as soon as we have `0 >>> n = 0`, we instantiate based on the
+-- pattern `0 >>> n >>> m` by substituting `0` into `0 >>> n` to produce the `0 >>> n >>> n`.
+-- I don't think any forbidden subterms can help us here. I don't know what to do. :-(
+run_meta analyzeEMatchTheorem ``Int.zero_shiftRight {}

script/Modulize.lean

@@ -1,77 +0,0 @@
-import Lake.CLI.Main
-
-/-!
-
-Usage: `lean --run script/Modulize.lean [--meta] file1.lean file2.lean ...`
-
-A simple script that inserts `module` and `public section` into un-modulized files and
-bumps their imports to `public`.
-
-When `--meta` is passed, `public meta section` and `public meta import` is used instead.
--/
-
-open Lean Parser.Module
-
-def main (args : List String) : IO Unit := do
-  let mut args := args
-  let mut doMeta := false
-  while !args.isEmpty && args[0]!.startsWith "-" do
-    match args[0]! with
-    | "--meta" => doMeta := true
-    | arg => throw <| .userError s!"unknown flag '{arg}'"
-    args := args.tail
-
-  for path in args do
-    -- Parse the input file
-    let mut text ← IO.FS.readFile path
-    let inputCtx := Parser.mkInputContext text path
-    let (header, parserState, msgs) ← Parser.parseHeader inputCtx
-    if !msgs.toList.isEmpty then -- skip this file if there are parse errors
-      msgs.forM fun msg => msg.toString >>= IO.println
-      throw <| .userError "parse errors in file"
-    let `(header| $[module%$moduleTk?]? $[prelude%$preludeTk?]? $imps:import*) := header
-      | throw <| .userError s!"unexpected header syntax of {path}"
-    if moduleTk?.isSome then
-      continue
-
-    -- initial whitespace if empty header
-    let startPos := header.raw.getPos? |>.getD parserState.pos
-
-    let dummyEnv ← mkEmptyEnvironment
-    let (initCmd, parserState', msgs') :=
-      Parser.parseCommand inputCtx { env := dummyEnv, options := {} } parserState msgs
-
-    -- insert section if any trailing command (or error, which could be from an unknown command)
-    if !initCmd.isOfKind ``Parser.Command.eoi || msgs'.hasErrors then
-      let insertPos? :=
-          -- put below initial module docstring if any
-          guard (initCmd.isOfKind ``Parser.Command.moduleDoc) *> initCmd.getTailPos? <|>
-          -- else below header
-          header.raw.getTailPos?
-      let insertPos := insertPos?.getD startPos  -- empty header
-      let mut sec := if doMeta then
-        "public meta section"
-      else
-        "@[expose] public section"
-      if !imps.isEmpty then
-        sec := "\n\n" ++ sec
-      if insertPos?.isNone then
-        sec := sec ++ "\n\n"
-      let insertPos := text.pos! insertPos
-      text := text.extract text.startPos insertPos ++ sec ++ text.extract insertPos text.endPos
-
-    -- prepend each import with `public `
-    for imp in imps.reverse do
-      let insertPos := imp.raw.getPos?.get!
-      let prfx := if doMeta then "public meta " else "public "
-      let insertPos := text.pos! insertPos
-      text := text.extract text.startPos insertPos ++ prfx ++ text.extract insertPos text.endPos
-
-    -- insert `module` header
-    let mut initText := text.extract text.startPos (text.pos! startPos)
-    if !initText.trimAscii.isEmpty then
-      -- If there is a header comment, preserve it and put `module` in the line after
-      initText := initText.trimAsciiEnd.toString ++ "\n"
-    text := initText ++ "module\n\n" ++ text.extract (text.pos! startPos) text.endPos
-
-    IO.FS.writeFile path text

script/apply.lean

@@ -60,7 +60,7 @@ if (arity == fixed + {n}) \{
   for j in [n:max + 1] do
     let fs := mkFsArgs (j - n)
     let sep := if j = n then "" else ", "
-    emit s!"    case {j}: \{ obj* r = FN{j}(f)({fs}{sep}{args}); lean_free_object(f); return r; }\n"
+    emit s!"    case {j}: \{ obj* r = FN{j}(f)({fs}{sep}{args}); lean_free_small_object(f); return r; }\n"
   emit "    }
   }
   switch (arity) {\n"
@@ -162,7 +162,7 @@ static obj* fix_args(obj* f, unsigned n, obj*const* as) {
       for (unsigned i = 0; i < fixed; i++, source++, target++) {
           *target = *source;
       }
-      lean_free_object(f);
+      lean_free_small_object(f);
     }
     for (unsigned i = 0; i < n; i++, as++, target++) {
         *target = *as;

script/bench.sh

@@ -0,0 +1,96 @@
+#!/usr/bin/env bash
+set -euxo pipefail
+
+cmake --preset release 1>&2
+
+# We benchmark against stage2/bin to test new optimizations.
+timeout -s KILL 1h time make -C build/release -j$(nproc) stage3 1>&2
+export PATH=$PWD/build/release/stage2/bin:$PATH
+
+# The extra opts used to be passed to the Makefile during benchmarking only but with Lake it is
+# easier to configure them statically.
+cmake -B build/release/stage3 -S src -DLEAN_EXTRA_LAKEFILE_TOML='weakLeanArgs=["-Dprofiler=true", "-Dprofiler.threshold=9999999", "--stats"]' 1>&2
+
+(
+cd tests/bench
+timeout -s KILL 1h time temci exec --config speedcenter.yaml --in speedcenter.exec.velcom.yaml 1>&2
+temci report run_output.yaml --reporter codespeed2
+)
+
+if [ -d .git ]; then
+    DIR="$(git rev-parse @)"
+    BASE_URL="https://speed.lean-lang.org/lean4-out/$DIR"
+    {
+        cat <<'EOF'
+<!DOCTYPE html>
+<html>
+<head>
+  <meta charset="UTF-8">
+  <title>Lakeprof Report</title>
+</head>
+<h1>Lakeprof Report</h1>
+<button type="button" id="btn_fetch">View build trace in Perfetto</button>
+<script type="text/javascript">
+const ORIGIN = 'https://ui.perfetto.dev';
+
+const btnFetch = document.getElementById('btn_fetch');
+
+async function fetchAndOpen(traceUrl) {
+  const resp = await fetch(traceUrl);
+  // Error checking is left as an exercise to the reader.
+  const blob = await resp.blob();
+  const arrayBuffer = await blob.arrayBuffer();
+  openTrace(arrayBuffer, traceUrl);
+}
+
+function openTrace(arrayBuffer, traceUrl) {
+  const win = window.open(ORIGIN);
+  if (!win) {
+    btnFetch.style.background = '#f3ca63';
+  	btnFetch.onclick = () => openTrace(arrayBuffer);
+    btnFetch.innerText = 'Popups blocked, click here to open the trace file';
+    return;
+  }
+
+  const timer = setInterval(() => win.postMessage('PING', ORIGIN), 50);
+
+  const onMessageHandler = (evt) => {
+    if (evt.data !== 'PONG') return;
+
+    // We got a PONG, the UI is ready.
+    window.clearInterval(timer);
+    window.removeEventListener('message', onMessageHandler);
+
+    const reopenUrl = new URL(location.href);
+    reopenUrl.hash = `#reopen=${traceUrl}`;
+    win.postMessage({
+      perfetto: {
+        buffer: arrayBuffer,
+        title: 'Lake Build Trace',
+        url: reopenUrl.toString(),
+    }}, ORIGIN);
+  };
+
+  window.addEventListener('message', onMessageHandler);
+}
+
+// This is triggered when following the link from the Perfetto UI's sidebar.
+if (location.hash.startsWith('#reopen=')) {
+ const traceUrl = location.hash.substr(8);
+ fetchAndOpen(traceUrl);
+}
+EOF
+        cat <<EOF
+btnFetch.onclick = () => fetchAndOpen("$BASE_URL/lakeprof.trace_event");
+</script>
+EOF
+        echo "<pre><code>"
+        (cd src; lakeprof report -prc)
+        echo "</code></pre>"
+        echo "</body></html>"
+    } | tee index.html
+
+    curl -T index.html $BASE_URL/index.html
+    curl -T src/lakeprof.log $BASE_URL/lakeprof.log
+    curl -T src/lakeprof.trace_event $BASE_URL/lakeprof.trace_event
+fi

script/benchReelabRss.lean

@@ -10,16 +10,6 @@ Tests language server memory use by repeatedly re-elaborate a given file.
 NOTE: only works on Linux for now.
 -/
 
-def determineRSS (pid : UInt32) : IO Nat := do
-  let status ← IO.FS.readFile s!"/proc/{pid}/smaps_rollup"
-  let some rssLine := status.splitOn "\n" |>.find? (·.startsWith "Rss:")
-    | throw <| IO.userError "No RSS in proc status"
-  let rssLine := rssLine.dropPrefix "Rss:"
-  let rssLine := rssLine.dropWhile Char.isWhitespace
-  let some rssInKB := rssLine.takeWhile Char.isDigit |>.toNat?
-    | throw <| IO.userError "Cannot parse RSS"
-  return rssInKB
-
 def main (args : List String) : IO Unit := do
   let leanCmd :: file :: iters :: args := args | panic! "usage: script <lean> <file> <#iterations> <server-args>..."
   let file ← IO.FS.realPath file
@@ -44,14 +34,11 @@ def main (args : List String) : IO Unit := do
     let text ← IO.FS.readFile file
     let (_, headerEndPos, _) ← Elab.parseImports text
     let headerEndPos := FileMap.ofString text |>.leanPosToLspPos headerEndPos
-    let n := iters.toNat!
-    let mut lastRSS? : Option Nat := none
-    let mut totalRSSDelta : Int := 0
     let mut requestNo : Nat := 1
     let mut versionNo : Nat := 1
     Ipc.writeNotification ⟨"textDocument/didOpen", {
       textDocument := { uri := uri, languageId := "lean", version := 1, text := text } : DidOpenTextDocumentParams }⟩
-    for i in [0:n] do
+    for i in [0:iters.toNat!] do
       if i > 0 then
         versionNo := versionNo + 1
         let params : DidChangeTextDocumentParams := {
@@ -74,16 +61,9 @@ def main (args : List String) : IO Unit := do
           IO.eprintln diag.message
       requestNo := requestNo + 1
 
-      let rss ← determineRSS (← read).pid
-      -- The first `didChange` usually results in a significantly higher RSS increase than
-      -- the others, so we ignore it.
-      if i > 1 then
-        if let some lastRSS := lastRSS? then
-          totalRSSDelta := totalRSSDelta + ((rss : Int) - (lastRSS : Int))
-      lastRSS? := some rss
-
-    let avgRSSDelta := totalRSSDelta / (n - 2)
-    IO.println s!"avg-reelab-rss-delta: {avgRSSDelta}"
+      let status ← IO.FS.readFile s!"/proc/{(← read).pid}/status"
+      for line in status.splitOn "\n" |>.filter (·.startsWith "RssAnon") do
+        IO.eprintln line
 
     let _ ← Ipc.collectDiagnostics requestNo uri versionNo
     (← Ipc.stdin).writeLspMessage (Message.notification "exit" none)

script/benchReelabWatchdogRss.lean

@@ -1,89 +0,0 @@
-import Lean.Data.Lsp
-import Lean.Elab.Import
-open Lean
-open Lean.Lsp
-open Lean.JsonRpc
-
-/-!
-Tests watchdog memory use by repeatedly re-elaborate a given file.
-
-NOTE: only works on Linux for now.
--/
-
-def determineRSS (pid : UInt32) : IO Nat := do
-  let status ← IO.FS.readFile s!"/proc/{pid}/smaps_rollup"
-  let some rssLine := status.splitOn "\n" |>.find? (·.startsWith "Rss:")
-    | throw <| IO.userError "No RSS in proc status"
-  let rssLine := rssLine.dropPrefix "Rss:"
-  let rssLine := rssLine.dropWhile Char.isWhitespace
-  let some rssInKB := rssLine.takeWhile Char.isDigit |>.toNat?
-    | throw <| IO.userError "Cannot parse RSS"
-  return rssInKB
-
-def main (args : List String) : IO Unit := do
-  let leanCmd :: file :: iters :: args := args | panic! "usage: script <lean> <file> <#iterations> <server-args>..."
-  let file ← IO.FS.realPath file
-  let uri := s!"file://{file}"
-  Ipc.runWith leanCmd (#["--server", "-DstderrAsMessages=false"] ++ args ++ #[uri]) do
-    let capabilities := {
-      textDocument? := some {
-        completion? := some {
-          completionItem? := some {
-            insertReplaceSupport? := true
-          }
-        }
-      }
-    }
-    Ipc.writeRequest ⟨0, "initialize", { capabilities : InitializeParams }⟩
-    discard <| Ipc.readResponseAs 0 InitializeResult
-    Ipc.writeNotification ⟨"initialized", InitializedParams.mk⟩
-
-    let text ← IO.FS.readFile file
-    let (_, headerEndPos, _) ← Elab.parseImports text
-    let headerEndPos := FileMap.ofString text |>.leanPosToLspPos headerEndPos
-    let n := iters.toNat!
-    let mut lastRSS? : Option Nat := none
-    let mut totalRSSDelta : Int := 0
-    let mut requestNo : Nat := 1
-    let mut versionNo : Nat := 1
-    Ipc.writeNotification ⟨"textDocument/didOpen", {
-      textDocument := { uri := uri, languageId := "lean", version := 1, text := text } : DidOpenTextDocumentParams }⟩
-    for i in [0:iters.toNat!] do
-      if i > 0 then
-        versionNo := versionNo + 1
-        let params : DidChangeTextDocumentParams := {
-          textDocument := {
-            uri      := uri
-            version? := versionNo
-          }
-          contentChanges := #[TextDocumentContentChangeEvent.rangeChange {
-            start := headerEndPos
-            «end» := headerEndPos
-          } " "]
-        }
-        let params := toJson params
-        Ipc.writeNotification ⟨"textDocument/didChange", params⟩
-        requestNo := requestNo + 1
-
-      let diags ← Ipc.collectDiagnostics requestNo uri versionNo
-      if let some diags := diags then
-        for diag in diags.param.diagnostics do
-          IO.eprintln diag.message
-      requestNo := requestNo + 1
-
-      Ipc.waitForILeans requestNo uri versionNo
-
-      let rss ← determineRSS (← read).pid
-      -- The first `didChange` usually results in a significantly higher RSS increase than
-      -- the others, so we ignore it.
-      if i > 1 then
-        if let some lastRSS := lastRSS? then
-          totalRSSDelta := totalRSSDelta + ((rss : Int) - (lastRSS : Int))
-      lastRSS? := some rss
-
-    let avgRSSDelta := totalRSSDelta / (n - 2)
-    IO.println s!"avg-reelab-rss-delta: {avgRSSDelta}"
-
-    let _ ← Ipc.collectDiagnostics requestNo uri versionNo
-    Ipc.shutdown requestNo
-    discard <| Ipc.waitForExit

script/build_artifact.py

@@ -1,441 +0,0 @@
-#!/usr/bin/env python3
-"""
-build_artifact.py: Download pre-built CI artifacts for a Lean commit.
-
-Usage:
-    build_artifact.py                     # Download artifact for current HEAD
-    build_artifact.py --sha abc1234       # Download artifact for specific commit
-    build_artifact.py --clear-cache       # Clear artifact cache
-
-This script downloads pre-built binaries from GitHub Actions CI runs,
-which is much faster than building from source (~30s vs 2-5min).
-
-Artifacts are cached in ~/.cache/lean_build_artifact/ for reuse.
-"""
-
-import argparse
-import json
-import os
-import platform
-import shutil
-import subprocess
-import sys
-import urllib.request
-import urllib.error
-from pathlib import Path
-from typing import Optional
-
-# Constants
-GITHUB_API_BASE = "https://api.github.com"
-LEAN4_REPO = "leanprover/lean4"
-
-# CI artifact cache
-CACHE_DIR = Path.home() / '.cache' / 'lean_build_artifact'
-ARTIFACT_CACHE = CACHE_DIR
-
-# Sentinel value indicating CI failed (don't bother building locally)
-CI_FAILED = object()
-
-# ANSI colors for terminal output
-class Colors:
-    RED = '\033[91m'
-    GREEN = '\033[92m'
-    YELLOW = '\033[93m'
-    BLUE = '\033[94m'
-    BOLD = '\033[1m'
-    RESET = '\033[0m'
-
-def color(text: str, c: str) -> str:
-    """Apply color to text if stdout is a tty."""
-    if sys.stdout.isatty():
-        return f"{c}{text}{Colors.RESET}"
-    return text
-
-def error(msg: str) -> None:
-    """Print error message and exit."""
-    print(color(f"Error: {msg}", Colors.RED), file=sys.stderr)
-    sys.exit(1)
-
-def warn(msg: str) -> None:
-    """Print warning message."""
-    print(color(f"Warning: {msg}", Colors.YELLOW), file=sys.stderr)
-
-def info(msg: str) -> None:
-    """Print info message."""
-    print(color(msg, Colors.BLUE), file=sys.stderr)
-
-def success(msg: str) -> None:
-    """Print success message."""
-    print(color(msg, Colors.GREEN), file=sys.stderr)
-
-# -----------------------------------------------------------------------------
-# Platform detection
-# -----------------------------------------------------------------------------
-
-def get_artifact_name() -> Optional[str]:
-    """Get CI artifact name for current platform."""
-    system = platform.system()
-    machine = platform.machine()
-
-    if system == 'Darwin':
-        if machine == 'arm64':
-            return 'build-macOS aarch64'
-        return 'build-macOS'  # Intel
-    elif system == 'Linux':
-        if machine == 'aarch64':
-            return 'build-Linux aarch64'
-        return 'build-Linux release'
-    # Windows not supported for CI artifact download
-    return None
-
-# -----------------------------------------------------------------------------
-# GitHub API helpers
-# -----------------------------------------------------------------------------
-
-_github_token_warning_shown = False
-
-def get_github_token() -> Optional[str]:
-    """Get GitHub token from environment or gh CLI."""
-    global _github_token_warning_shown
-
-    # Check environment variable first
-    token = os.environ.get('GITHUB_TOKEN')
-    if token:
-        return token
-
-    # Try to get token from gh CLI
-    try:
-        result = subprocess.run(
-            ['gh', 'auth', 'token'],
-            capture_output=True,
-            text=True,
-            timeout=5
-        )
-        if result.returncode == 0 and result.stdout.strip():
-            return result.stdout.strip()
-    except (FileNotFoundError, subprocess.TimeoutExpired):
-        pass
-
-    # Warn once if no token available
-    if not _github_token_warning_shown:
-        _github_token_warning_shown = True
-        warn("No GitHub authentication found. API rate limits may apply.")
-        warn("Run 'gh auth login' or set GITHUB_TOKEN to avoid rate limiting.")
-
-    return None
-
-def github_api_request(url: str) -> dict:
-    """Make a GitHub API request and return JSON response."""
-    headers = {
-        'Accept': 'application/vnd.github.v3+json',
-        'User-Agent': 'build-artifact'
-    }
-
-    token = get_github_token()
-    if token:
-        headers['Authorization'] = f'token {token}'
-
-    req = urllib.request.Request(url, headers=headers)
-    try:
-        with urllib.request.urlopen(req, timeout=30) as response:
-            return json.loads(response.read().decode())
-    except urllib.error.HTTPError as e:
-        if e.code == 403:
-            error(f"GitHub API rate limit exceeded. Set GITHUB_TOKEN environment variable to increase limit.")
-        elif e.code == 404:
-            error(f"GitHub resource not found: {url}")
-        else:
-            error(f"GitHub API error: {e.code} {e.reason}")
-    except urllib.error.URLError as e:
-        error(f"Network error accessing GitHub API: {e.reason}")
-
-# -----------------------------------------------------------------------------
-# CI artifact cache functions
-# -----------------------------------------------------------------------------
-
-def get_cache_path(sha: str) -> Path:
-    """Get cache directory for a commit's artifact."""
-    return ARTIFACT_CACHE / sha[:12]
-
-def is_cached(sha: str) -> bool:
-    """Check if artifact for this commit is already cached and valid."""
-    cache_path = get_cache_path(sha)
-    return cache_path.exists() and (cache_path / 'bin' / 'lean').exists()
-
-def check_zstd_support() -> bool:
-    """Check if tar supports zstd compression."""
-    try:
-        result = subprocess.run(
-            ['tar', '--zstd', '--version'],
-            capture_output=True,
-            timeout=5
-        )
-        return result.returncode == 0
-    except (subprocess.TimeoutExpired, FileNotFoundError):
-        return False
-
-def check_gh_available() -> bool:
-    """Check if gh CLI is available and authenticated."""
-    try:
-        result = subprocess.run(
-            ['gh', 'auth', 'status'],
-            capture_output=True,
-            timeout=10
-        )
-        return result.returncode == 0
-    except (subprocess.TimeoutExpired, FileNotFoundError):
-        return False
-
-def download_ci_artifact(sha: str, quiet: bool = False):
-    """
-    Try to download CI artifact for a commit.
-    Returns:
-      - Path to extracted toolchain directory if available
-      - CI_FAILED sentinel if CI run failed (don't bother building locally)
-      - None if no artifact available but local build might work
-    """
-    # Check cache first
-    if is_cached(sha):
-        return get_cache_path(sha)
-
-    artifact_name = get_artifact_name()
-    if artifact_name is None:
-        return None  # Unsupported platform
-
-    cache_path = get_cache_path(sha)
-
-    try:
-        # Query for CI workflow run for this commit, including status
-        # Note: Query parameters must be in the URL for GET requests
-        result = subprocess.run(
-            ['gh', 'api', f'repos/{LEAN4_REPO}/actions/runs?head_sha={sha}&per_page=100',
-             '--jq', r'.workflow_runs[] | select(.name == "CI") | "\(.id) \(.conclusion // "null")"'],
-            capture_output=True,
-            text=True,
-            timeout=30
-        )
-        if result.returncode != 0 or not result.stdout.strip():
-            return None  # No CI run found (old commit?)
-
-        # Parse "run_id conclusion" format
-        line = result.stdout.strip().split('\n')[0]
-        parts = line.split(' ', 1)
-        run_id = parts[0]
-        conclusion = parts[1] if len(parts) > 1 else "null"
-
-        # Check if the desired artifact exists for this run
-        result = subprocess.run(
-            ['gh', 'api', f'repos/{LEAN4_REPO}/actions/runs/{run_id}/artifacts',
-             '--jq', f'.artifacts[] | select(.name == "{artifact_name}") | .id'],
-            capture_output=True,
-            text=True,
-            timeout=30
-        )
-        if result.returncode != 0 or not result.stdout.strip():
-            # No artifact available
-            # If CI failed and no artifact, the build itself likely failed - skip
-            if conclusion == "failure":
-                return CI_FAILED
-            # Otherwise (in progress, expired, etc.) - fall back to local build
-            return None
-
-        # Download artifact
-        cache_path.mkdir(parents=True, exist_ok=True)
-        if not quiet:
-            print("downloading CI artifact... ", end='', flush=True)
-
-        result = subprocess.run(
-            ['gh', 'run', 'download', run_id,
-             '-n', artifact_name,
-             '-R', LEAN4_REPO,
-             '-D', str(cache_path)],
-            capture_output=True,
-            text=True,
-            timeout=600  # 10 minutes for large downloads
-        )
-
-        if result.returncode != 0:
-            shutil.rmtree(cache_path, ignore_errors=True)
-            return None
-
-        # Extract tar.zst - find the file (name varies by platform/version)
-        tar_files = list(cache_path.glob('*.tar.zst'))
-        if not tar_files:
-            shutil.rmtree(cache_path, ignore_errors=True)
-            return None
-
-        tar_file = tar_files[0]
-        if not quiet:
-            print("extracting... ", end='', flush=True)
-
-        result = subprocess.run(
-            ['tar', '--zstd', '-xf', tar_file.name],
-            cwd=cache_path,
-            capture_output=True,
-            timeout=300
-        )
-
-        if result.returncode != 0:
-            shutil.rmtree(cache_path, ignore_errors=True)
-            return None
-
-        # Move contents up from lean-VERSION-PLATFORM/ to cache_path/
-        # The extracted directory name varies (e.g., lean-4.15.0-linux, lean-4.15.0-darwin_aarch64)
-        extracted_dirs = [d for d in cache_path.iterdir() if d.is_dir() and d.name.startswith('lean-')]
-        if extracted_dirs:
-            extracted = extracted_dirs[0]
-            for item in extracted.iterdir():
-                dest = cache_path / item.name
-                if dest.exists():
-                    if dest.is_dir():
-                        shutil.rmtree(dest)
-                    else:
-                        dest.unlink()
-                shutil.move(str(item), str(cache_path / item.name))
-            extracted.rmdir()
-
-        # Clean up tar file
-        tar_file.unlink()
-
-        # Verify the extraction worked
-        if not (cache_path / 'bin' / 'lean').exists():
-            shutil.rmtree(cache_path, ignore_errors=True)
-            return None
-
-        return cache_path
-
-    except (subprocess.TimeoutExpired, FileNotFoundError):
-        shutil.rmtree(cache_path, ignore_errors=True)
-        return None
-
-# -----------------------------------------------------------------------------
-# Git helpers
-# -----------------------------------------------------------------------------
-
-def get_current_commit() -> str:
-    """Get the current git HEAD commit SHA."""
-    try:
-        result = subprocess.run(
-            ['git', 'rev-parse', 'HEAD'],
-            capture_output=True,
-            text=True,
-            timeout=5
-        )
-        if result.returncode == 0:
-            return result.stdout.strip()
-        error(f"Failed to get current commit: {result.stderr.strip()}")
-    except subprocess.TimeoutExpired:
-        error("Timeout getting current commit")
-    except FileNotFoundError:
-        error("git not found")
-
-def resolve_sha(short_sha: str) -> str:
-    """Resolve a (possibly short) SHA to full 40-character SHA using git rev-parse."""
-    if len(short_sha) == 40:
-        return short_sha
-    try:
-        result = subprocess.run(
-            ['git', 'rev-parse', short_sha],
-            capture_output=True,
-            text=True,
-            timeout=5
-        )
-        if result.returncode == 0:
-            full_sha = result.stdout.strip()
-            if len(full_sha) == 40:
-                return full_sha
-        error(f"Cannot resolve SHA '{short_sha}': {result.stderr.strip() or 'not found in repository'}")
-    except subprocess.TimeoutExpired:
-        error(f"Timeout resolving SHA '{short_sha}'")
-    except FileNotFoundError:
-        error("git not found - required for SHA resolution")
-
-# -----------------------------------------------------------------------------
-# Main
-# -----------------------------------------------------------------------------
-
-def main():
-    parser = argparse.ArgumentParser(
-        description='Download pre-built CI artifacts for a Lean commit.',
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-        epilog="""
-This script downloads pre-built binaries from GitHub Actions CI runs,
-which is much faster than building from source (~30s vs 2-5min).
-
-Artifacts are cached in ~/.cache/lean_build_artifact/ for reuse.
-
-Examples:
-  build_artifact.py                   # Download for current HEAD
-  build_artifact.py --sha abc1234     # Download for specific commit
-  build_artifact.py --clear-cache     # Clear cache to free disk space
-"""
-    )
-
-    parser.add_argument('--sha', metavar='SHA',
-                        help='Commit SHA to download artifact for (default: current HEAD)')
-    parser.add_argument('--clear-cache', action='store_true',
-                        help='Clear artifact cache and exit')
-    parser.add_argument('--quiet', '-q', action='store_true',
-                        help='Suppress progress messages (still prints result path)')
-
-    args = parser.parse_args()
-
-    # Handle cache clearing
-    if args.clear_cache:
-        if ARTIFACT_CACHE.exists():
-            size = sum(f.stat().st_size for f in ARTIFACT_CACHE.rglob('*') if f.is_file())
-            shutil.rmtree(ARTIFACT_CACHE)
-            info(f"Cleared cache at {ARTIFACT_CACHE} ({size / 1024 / 1024:.1f} MB)")
-        else:
-            info(f"Cache directory does not exist: {ARTIFACT_CACHE}")
-        return
-
-    # Get commit SHA
-    if args.sha:
-        sha = resolve_sha(args.sha)
-    else:
-        sha = get_current_commit()
-
-    if not args.quiet:
-        info(f"Commit: {sha[:12]}")
-
-    # Check prerequisites
-    if not check_gh_available():
-        error("gh CLI not available or not authenticated. Run 'gh auth login' first.")
-
-    if not check_zstd_support():
-        error("tar does not support zstd compression. Install zstd or a newer tar.")
-
-    artifact_name = get_artifact_name()
-    if artifact_name is None:
-        error(f"No CI artifacts available for this platform ({platform.system()} {platform.machine()})")
-
-    if not args.quiet:
-        info(f"Platform: {artifact_name}")
-
-    # Check cache
-    if is_cached(sha):
-        path = get_cache_path(sha)
-        if not args.quiet:
-            success("Using cached artifact")
-        print(path)
-        return
-
-    # Download artifact
-    result = download_ci_artifact(sha, quiet=args.quiet)
-
-    if result is CI_FAILED:
-        if not args.quiet:
-            print()  # End the "downloading..." line
-        error(f"CI build failed for commit {sha[:12]}")
-    elif result is None:
-        if not args.quiet:
-            print()  # End the "downloading..." line
-        error(f"No CI artifact available for commit {sha[:12]}")
-    else:
-        if not args.quiet:
-            print(color("done", Colors.GREEN))
-        print(result)
-
-if __name__ == '__main__':
-    main()

script/fmt

@@ -1,13 +0,0 @@
-#!/usr/bin/env bash
-set -euo pipefail
-
-# This script expects to be run from the repo root.
-
-# Format cmake files
-find -regex '.*/CMakeLists\.txt\(\.in\)?\|.*\.cmake\(\.in\)?' \
-  ! -path './build/*' \
-  ! -path "./stage0/*" \
-  -exec \
-    uvx gersemi --in-place --line-length 120 --indent 2 \
-    --definitions src/cmake/Modules/ src/CMakeLists.txt \
-    -- {} +

script/lakefile.toml

@@ -1,5 +0,0 @@
-name = "scripts"
-
-[[lean_exe]]
-name = "modulize"
-root = "Modulize"

script/lean-bisect-test.lean

@@ -1,307 +0,0 @@
-/-
-  Copyright Strata Contributors
-
-  SPDX-License-Identifier: Apache-2.0 OR MIT
--/
-
-namespace Strata
-namespace Python
-
-/-
-Parser and translator for some basic regular expression patterns supported by
-Python's `re` library
-Ref.: https://docs.python.org/3/library/re.html
-
-Also see
-https://github.com/python/cpython/blob/759a048d4bea522fda2fe929be0fba1650c62b0e/Lib/re/_parser.py
-for a reference implementation.
--/
-
--------------------------------------------------------------------------------
-
-inductive ParseError where
-  /--
-    `patternError` is raised when Python's `re.patternError` exception is
-    raised.
-    [Reference: Python's re exceptions](https://docs.python.org/3/library/re.html#exceptions):
-
-    "Exception raised when a string passed to one of the functions here is not a
-    valid regular expression (for example, it might contain unmatched
-    parentheses) or when some other error occurs during compilation or matching.
-    It is never an error if a string contains no match for a pattern."
-  -/
-  | patternError  (message : String) (pattern : String) (pos : String.Pos.Raw)
-  /--
-  `unimplemented` is raised whenever we don't support some regex operations
-  (e.g., lookahead assertions).
-  -/
-  | unimplemented (message : String) (pattern : String) (pos : String.Pos.Raw)
-  deriving Repr
-
-def ParseError.toString : ParseError → String
-  | .patternError msg pat pos => s!"Pattern error at position {pos.byteIdx}: {msg} in pattern '{pat}'"
-  | .unimplemented msg pat pos => s!"Unimplemented at position {pos.byteIdx}: {msg} in pattern '{pat}'"
-
-instance : ToString ParseError where
-  toString := ParseError.toString
-
--------------------------------------------------------------------------------
-
-/--
-Regular Expression Nodes
--/
-inductive RegexAST where
-  /-- Single literal character: `a` -/
-  | char : Char → RegexAST
-  /-- Character range: `[a-z]` -/
-  | range : Char → Char → RegexAST
-  /-- Alternation: `a|b` -/
-  | union : RegexAST → RegexAST → RegexAST
-  /-- Concatenation: `ab` -/
-  | concat : RegexAST → RegexAST → RegexAST
-  /-- Any character: `.` -/
-  | anychar : RegexAST
-  /-- Zero or more: `a*` -/
-  | star : RegexAST → RegexAST
-  /-- One or more: `a+` -/
-  | plus : RegexAST → RegexAST
-  /-- Zero or one: `a?` -/
-  | optional : RegexAST → RegexAST
-  /-- Bounded repetition: `a{n,m}` -/
-  | loop : RegexAST → Nat → Nat → RegexAST
-  /-- Start of string: `^` -/
-  | anchor_start : RegexAST
-  /-- End of string: `$` -/
-  | anchor_end : RegexAST
-  /-- Grouping: `(abc)` -/
-  | group : RegexAST → RegexAST
-  /-- Empty string: `()` or `""` -/
-  | empty : RegexAST
-  /-- Complement: `[^a-z]` -/
-  | complement : RegexAST → RegexAST
-  deriving Inhabited, Repr
-
--------------------------------------------------------------------------------
-
-/-- Parse character class like [a-z], [0-9], etc. into union of ranges and
-  chars. Note that this parses `|` as a character. -/
-def parseCharClass (s : String) (pos : String.Pos.Raw) : Except ParseError (RegexAST × String.Pos.Raw) := do
-  if pos.get? s != some '[' then throw (.patternError "Expected '[' at start of character class" s pos)
-  let mut i := pos.next s
-
-  -- Check for complement (negation) with leading ^
-  let isComplement := !i.atEnd s && i.get? s == some '^'
-  if isComplement then
-    i := i.next s
-
-  let mut result : Option RegexAST := none
-
-  -- Process each element in the character class.
-  while !i.atEnd s && i.get? s != some ']' do
-    -- Uncommenting this makes the code stop
-    --dbg_trace "Working" (pure ())
-    let some c1 := i.get? s | throw (.patternError "Invalid character in class" s i)
-    let i1 := i.next s
-    -- Check for range pattern: c1-c2.
-    if !i1.atEnd s && i1.get? s == some '-' then
-      let i2 := i1.next s
-      if !i2.atEnd s && i2.get? s != some ']' then
-        let some c2 := i2.get? s | throw (.patternError "Invalid character in range" s i2)
-        if c1 > c2 then
-          throw (.patternError s!"Invalid character range [{c1}-{c2}]: \
-                                  start character '{c1}' is greater than end character '{c2}'" s i)
-        let r := RegexAST.range c1 c2
-        -- Union with previous elements.
-        result := some (match result with | none => r | some prev => RegexAST.union prev r)
-        i := i2.next s
-        continue
-    -- Single character.
-    let r := RegexAST.char c1
-    result := some (match result with | none => r | some prev => RegexAST.union prev r)
-    i := i.next s
-
-  let some ast := result | throw (.patternError "Unterminated character set" s pos)
-  let finalAst := if isComplement then RegexAST.complement ast else ast
-  pure (finalAst, i.next s)
-
--------------------------------------------------------------------------------
-
-/-- Parse numeric repeats like `{10}` or `{1,10}` into min and max bounds. -/
-def parseBounds (s : String) (pos : String.Pos.Raw) : Except ParseError (Nat × Nat × String.Pos.Raw) := do
-  if pos.get? s != some '{' then throw (.patternError "Expected '{' at start of bounds" s pos)
-  let mut i := pos.next s
-  let mut numStr := ""
-
-  -- Parse first number.
-  while !i.atEnd s && (i.get? s).any Char.isDigit do
-    numStr := numStr.push ((i.get? s).get!)
-    i := i.next s
-
-  let some n := numStr.toNat? | throw (.patternError "Invalid minimum bound" s pos)
-
-  -- Check for comma (range) or closing brace (exact count).
-  match i.get? s with
-  | some '}' => pure (n, n, i.next s)  -- {n} means exactly n times.
-  | some ',' =>
-    i := i.next s
-    -- Parse maximum bound
-    numStr := ""
-    while !i.atEnd s && (i.get? s).any Char.isDigit do
-      numStr := numStr.push ((i.get? s).get!)
-      i := i.next s
-    let some max := numStr.toNat? | throw (.patternError "Invalid maximum bound" s i)
-    if i.get? s != some '}' then throw (.patternError "Expected '}' at end of bounds" s i)
-    -- Validate bounds order
-    if max < n then
-      throw (.patternError s!"Invalid repeat bounds \{{n},{max}}: \
-                              maximum {max} is less than minimum {n}" s pos)
-    pure (n, max, i.next s)
-  | _ => throw (.patternError "Invalid bounds syntax" s i)
-
--------------------------------------------------------------------------------
-
-mutual
-/--
-Parse atom: single element (char, class, anchor, group) with optional
-quantifier. Stops at the first `|`.
--/
-partial def parseAtom (s : String) (pos : String.Pos.Raw) : Except ParseError (RegexAST × String.Pos.Raw) := do
-  if pos.atEnd s then throw (.patternError "Unexpected end of regex" s pos)
-
-  let some c := pos.get? s | throw (.patternError "Invalid position" s pos)
-
-  -- Detect invalid quantifier at start
-  if c == '*' || c == '+' || c == '{' || c == '?' then
-    throw (.patternError s!"Quantifier '{c}' at position {pos} has nothing to quantify" s pos)
-
-  -- Detect unbalanced closing parenthesis
-  if c == ')' then
-    throw (.patternError "Unbalanced parenthesis" s pos)
-
-  -- Parse base element (anchor, char class, group, anychar, escape, or single char).
-  let (base, nextPos) ← match c with
-    | '^' => pure (RegexAST.anchor_start, pos.next s)
-    | '$' => pure (RegexAST.anchor_end, pos.next s)
-    | '[' => parseCharClass s pos
-    | '(' => parseExplicitGroup s pos
-    | '.' => pure (RegexAST.anychar, pos.next s)
-    | '\\' =>
-      -- Handle escape sequence.
-      -- Note: Python uses a single backslash as an escape character, but Lean
-      -- strings need to escape that. After DDMification, we will see two
-      -- backslashes in Strata for every Python backslash.
-      let nextPos := pos.next s
-      if nextPos.atEnd s then throw (.patternError "Incomplete escape sequence at end of regex" s pos)
-      let some escapedChar := nextPos.get? s | throw (.patternError "Invalid escape position" s nextPos)
-      -- Check for special sequences (unsupported right now).
-      match escapedChar with
-      | 'A' | 'b' | 'B' | 'd' | 'D' | 's' | 'S' | 'w' | 'W' | 'z' | 'Z' =>
-        throw (.unimplemented s!"Special sequence \\{escapedChar} is not supported" s pos)
-      | 'a' | 'f' | 'n' | 'N' | 'r' | 't' | 'u' | 'U' | 'v' | 'x' =>
-        throw (.unimplemented s!"Escape sequence \\{escapedChar} is not supported" s pos)
-      | c =>
-        if c.isDigit then
-          throw (.unimplemented s!"Backreference \\{c} is not supported" s pos)
-        else
-          pure (RegexAST.char escapedChar, nextPos.next s)
-    | _ => pure (RegexAST.char c, pos.next s)
-
-  -- Check for numeric repeat suffix on base element (but not on anchors)
-  match base with
-  | .anchor_start | .anchor_end => pure (base, nextPos)
-  | _ =>
-    if !nextPos.atEnd s then
-      match nextPos.get? s with
-      | some '{' =>
-        let (min, max, finalPos) ← parseBounds s nextPos
-        pure (RegexAST.loop base min max, finalPos)
-      | some '*' =>
-        let afterStar := nextPos.next s
-        if !afterStar.atEnd s then
-          match afterStar.get? s with
-          | some '?' => throw (.unimplemented "Non-greedy quantifier *? is not supported" s nextPos)
-          | some '+' => throw (.unimplemented "Possessive quantifier *+ is not supported" s nextPos)
-          | _ => pure (RegexAST.star base, afterStar)
-        else pure (RegexAST.star base, afterStar)
-      | some '+' =>
-        let afterPlus := nextPos.next s
-        if !afterPlus.atEnd s then
-          match afterPlus.get? s with
-          | some '?' => throw (.unimplemented "Non-greedy quantifier +? is not supported" s nextPos)
-          | some '+' => throw (.unimplemented "Possessive quantifier ++ is not supported" s nextPos)
-          | _ => pure (RegexAST.plus base, afterPlus)
-        else pure (RegexAST.plus base, afterPlus)
-      | some '?' =>
-        let afterQuestion := nextPos.next s
-        if !afterQuestion.atEnd s then
-          match afterQuestion.get? s with
-          | some '?' => throw (.unimplemented "Non-greedy quantifier ?? is not supported" s nextPos)
-          | some '+' => throw (.unimplemented "Possessive quantifier ?+ is not supported" s nextPos)
-          | _ => pure (RegexAST.optional base, afterQuestion)
-        else pure (RegexAST.optional base, afterQuestion)
-      | _ => pure (base, nextPos)
-    else
-      pure (base, nextPos)
-
-/-- Parse explicit group with parentheses. -/
-partial def parseExplicitGroup (s : String) (pos : String.Pos.Raw) : Except ParseError (RegexAST × String.Pos.Raw) := do
-  if pos.get? s != some '(' then throw (.patternError "Expected '(' at start of group" s pos)
-  let mut i := pos.next s
-
-  -- Check for extension notation (?...
-  if !i.atEnd s && i.get? s == some '?' then
-    let i1 := i.next s
-    if !i1.atEnd s then
-      match i1.get? s with
-      | some '=' => throw (.unimplemented "Positive lookahead (?=...) is not supported" s pos)
-      | some '!' => throw (.unimplemented "Negative lookahead (?!...) is not supported" s pos)
-      | _ => throw (.unimplemented "Extension notation (?...) is not supported" s pos)
-
-  let (inner, finalPos) ← parseGroup s i (some ')')
-  pure (.group inner, finalPos)
-
-/-- Parse group: handles alternation and concatenation at current scope. -/
-partial def parseGroup (s : String) (pos : String.Pos.Raw) (endChar : Option Char) :
-    Except ParseError (RegexAST × String.Pos.Raw) := do
-  let mut alternatives : List (List RegexAST) := [[]]
-  let mut i := pos
-
-  -- Parse until end of string or `endChar`.
-  while !i.atEnd s && (endChar.isNone || i.get? s != endChar) do
-    if i.get? s == some '|' then
-      -- Push a new scope to `alternatives`.
-      alternatives := [] :: alternatives
-      i := i.next s
-    else
-      let (ast, nextPos) ← parseAtom s i
-      alternatives := match alternatives with
-        | [] => [[ast]]
-        | head :: tail => (ast :: head) :: tail
-      i := nextPos
-
-  -- Check for expected end character.
-  if let some ec := endChar then
-    if i.get? s != some ec then
-      throw (.patternError s!"Expected '{ec}'" s i)
-    i := i.next s
-
-  -- Build result: concatenate each alternative, then union them.
-  let concatAlts := alternatives.reverse.filterMap fun alt =>
-    match alt.reverse with
-    | [] => -- Empty regex.
-      some (.empty)
-    | [single] => some single
-    | head :: tail => some (tail.foldl RegexAST.concat head)
-
-  match concatAlts with
-  | [] => pure (.empty, i)
-  | [single] => pure (single, i)
-  | head :: tail => pure (tail.foldl RegexAST.union head, i)
-end
-
-/-- info: Except.ok (Strata.Python.RegexAST.range 'A' 'z', { byteIdx := 5 }) -/
-#guard_msgs in
-#eval parseCharClass "[A-z]" ⟨0⟩
-
--- Test code: Print done
-#print "Done!"

script/lean-toolchain

@@ -1 +0,0 @@
-lean4

script/prepare-llvm-linux.sh

@@ -58,11 +58,7 @@ OPTIONS=()
 # We build cadical using the custom toolchain on Linux to avoid glibc versioning issues
 echo -n " -DLEAN_STANDALONE=ON -DCADICAL_USE_CUSTOM_CXX=ON"
 echo -n " -DCMAKE_CXX_COMPILER=$PWD/llvm-host/bin/clang++ -DLEAN_CXX_STDLIB='-Wl,-Bstatic -lc++ -lc++abi -Wl,-Bdynamic'"
-# these should also be used for cadical, so do not use `LEAN_EXTRA_CXX_FLAGS` here
-echo -n " -DCMAKE_CXX_FLAGS='--sysroot $PWD/llvm -idirafter $GLIBC_DEV/include ${EXTRA_FLAGS:-}'"
-# the above does not include linker flags which will be added below based on context, so skip the
-# generic check by cmake
-echo -n " -DCMAKE_C_COMPILER_WORKS=1 -DCMAKE_CXX_COMPILER_WORKS=1"
+echo -n " -DLEAN_EXTRA_CXX_FLAGS='--sysroot $PWD/llvm -idirafter $GLIBC_DEV/include ${EXTRA_FLAGS:-}'"
 # use target compiler directly when not cross-compiling
 if [[ -L llvm-host ]]; then
   echo -n " -DCMAKE_C_COMPILER=$PWD/stage1/bin/clang"

script/release_checklist.py

@@ -1,60 +1,5 @@
 #!/usr/bin/env python3
 
-"""
-Release Checklist for Lean4 and Downstream Repositories
-
-This script validates the status of a Lean4 release across all dependent repositories.
-It checks whether repositories are ready for release and identifies missing steps.
-
-IMPORTANT: Keep this documentation up-to-date when modifying the script's behavior!
-
-What this script does:
-1. Validates preliminary Lean4 release infrastructure:
-   - Checks that the release branch (releases/vX.Y.0) exists
-   - Verifies CMake version settings are correct
-   - Confirms the release tag exists
-   - Validates the release page exists on GitHub (created automatically by CI after tag push)
-   - Checks the release notes page on lean-lang.org (updated while bumping the `reference-manual` repository)
-
-   **IMPORTANT: If the release page doesn't exist, the script will skip checking
-   downstream repositories and the master branch configuration. The preliminary
-   infrastructure must be in place before the release process can proceed.**
-
-   **NOTE: The GitHub release page is created AUTOMATICALLY by CI after the tag is pushed.
-   DO NOT create it manually. Wait for CI to complete after pushing the tag.**
-
-2. For each downstream repository (batteries, mathlib4, etc.):
-   - Checks if dependencies are ready (e.g., mathlib4 depends on batteries)
-   - Verifies the main branch is on the target toolchain (or newer)
-   - Checks if a PR exists to bump the toolchain (if not yet updated)
-   - Validates tags exist for the release version
-   - Ensures tags are merged into stable branches (for non-RC releases)
-   - Verifies bump branches exist and are configured correctly
-   - Special handling for ProofWidgets4 release tags
-   - For mathlib4: runs verify_version_tags.py to validate the release tag
-     (checks git/GitHub consistency, toolchain, elan, cache, and build)
-
-3. Optionally automates missing steps (when not in --dry-run mode):
-   - Creates missing release tags using push_repo_release_tag.py
-   - Merges tags into stable branches using merge_remote.py
-
-Usage:
-    ./release_checklist.py v4.24.0           # Check release status
-    ./release_checklist.py v4.24.0 --verbose # Show detailed debug info
-    ./release_checklist.py v4.24.0 --dry-run # Check only, don't execute fixes
-
-For automated release management with Claude Code:
-    /release v4.24.0                 # Run full release process with Claude
-
-The script reads repository configurations from release_repos.yml and reports:
-- ✅ for completed requirements
-- ❌ for missing requirements (with instructions to fix)
-- 🟡 for repositories waiting on dependencies
-- ⮕ for automated actions being taken
-
-This script is idempotent and safe to rerun multiple times.
-"""
-
 import argparse
 import yaml
 import requests
@@ -131,39 +76,6 @@ def release_page_exists(repo_url, tag_name, github_token):
     response = requests.get(api_url, headers=headers)
     return response.status_code == 200
 
-def get_tag_workflow_status(repo_url, tag_name, github_token):
-    """Get the status of CI workflows running for a specific tag."""
-    api_base = repo_url.replace("https://github.com/", "https://api.github.com/repos/")
-    headers = {'Authorization': f'token {github_token}'} if github_token else {}
-
-    # Get workflow runs for the tag
-    # GitHub's workflow runs API uses the branch/tag name in the 'head_branch' field
-    api_url = f"{api_base}/actions/runs?event=push&head_branch={tag_name}"
-    response = requests.get(api_url, headers=headers)
-
-    if response.status_code != 200:
-        return None
-
-    data = response.json()
-    workflow_runs = data.get('workflow_runs', [])
-
-    if not workflow_runs:
-        return None
-
-    # Get the most recent workflow run for this tag
-    run = workflow_runs[0]
-    status = run.get('status')
-    conclusion = run.get('conclusion')
-    workflow_name = run.get('name', 'CI')
-    run_id = run.get('id')
-
-    return {
-        'status': status,
-        'conclusion': conclusion,
-        'workflow_name': workflow_name,
-        'run_id': run_id
-    }
-
 def get_release_notes(tag_name):
     """Fetch release notes page title from lean-lang.org."""
     # Strip -rcX suffix if present for the URL
@@ -172,43 +84,22 @@ def get_release_notes(tag_name):
     try:
         response = requests.get(reference_url)
         response.raise_for_status() # Raise HTTPError for bad responses (4xx or 5xx)
-
+        
         # Extract title using regex
         match = re.search(r"<title>(.*?)</title>", response.text, re.IGNORECASE | re.DOTALL)
         if match:
             return match.group(1).strip()
         else:
+            print(f"  ⚠️ Could not find <title> tag in {reference_url}")
             return None
-
-    except requests.exceptions.RequestException:
+            
+    except requests.exceptions.RequestException as e:
+        print(f"  ❌ Error fetching release notes from {reference_url}: {e}")
         return None
-    except Exception:
+    except Exception as e:
+        print(f"  ❌ An unexpected error occurred while processing release notes: {e}")
         return None
 
-def check_release_notes_file_exists(toolchain, github_token):
-    """Check if the release notes file exists in the reference-manual repository.
-
-    For -rc1 releases, this checks that the release notes have been created.
-    For subsequent RCs and stable releases, release notes should already exist.
-
-    Returns tuple (exists: bool, is_rc1: bool) where is_rc1 indicates if this is
-    the first release candidate (when release notes need to be written).
-    """
-    # Determine the release notes file path
-    # e.g., v4.28.0-rc1 -> Manual/Releases/v4_28_0.lean
-    base_version = strip_rc_suffix(toolchain.lstrip('v'))  # "4.28.0"
-    file_name = f"v{base_version.replace('.', '_')}.lean"  # "v4_28_0.lean"
-    file_path = f"Manual/Releases/{file_name}"
-
-    is_rc1 = toolchain.endswith("-rc1")
-
-    repo_url = "https://github.com/leanprover/reference-manual"
-
-    # Check if the file exists on main branch
-    content = get_branch_content(repo_url, "main", file_path, github_token)
-
-    return (content is not None, is_rc1)
-
 def get_branch_content(repo_url, branch, file_path, github_token):
     api_url = repo_url.replace("https://github.com/", "https://api.github.com/repos/") + f"/contents/{file_path}?ref={branch}"
     headers = {'Authorization': f'token {github_token}'} if github_token else {}
@@ -395,68 +286,6 @@ def check_bump_branch_toolchain(url, bump_branch, github_token):
     print(f"  ✅ Bump branch correctly uses toolchain: {content}")
     return True
 
-def get_pr_ci_status(repo_url, pr_number, github_token):
-    """Get the CI status for a pull request."""
-    api_base = repo_url.replace("https://github.com/", "https://api.github.com/repos/")
-    headers = {'Authorization': f'token {github_token}'} if github_token else {}
-
-    # Get PR details to find the head SHA
-    pr_response = requests.get(f"{api_base}/pulls/{pr_number}", headers=headers)
-    if pr_response.status_code != 200:
-        return "unknown", "Could not fetch PR details"
-
-    pr_data = pr_response.json()
-    head_sha = pr_data['head']['sha']
-
-    # Get check runs for the commit
-    check_runs_response = requests.get(
-        f"{api_base}/commits/{head_sha}/check-runs",
-        headers=headers
-    )
-
-    if check_runs_response.status_code != 200:
-        return "unknown", "Could not fetch check runs"
-
-    check_runs_data = check_runs_response.json()
-    check_runs = check_runs_data.get('check_runs', [])
-
-    if not check_runs:
-        # No check runs, check for status checks (legacy)
-        status_response = requests.get(
-            f"{api_base}/commits/{head_sha}/status",
-            headers=headers
-        )
-        if status_response.status_code == 200:
-            status_data = status_response.json()
-            state = status_data.get('state', 'unknown')
-            if state == 'success':
-                return "success", "All status checks passed"
-            elif state == 'failure':
-                return "failure", "Some status checks failed"
-            elif state == 'pending':
-                return "pending", "Status checks in progress"
-        return "unknown", "No CI checks found"
-
-    # Analyze check runs
-    conclusions = [run['conclusion'] for run in check_runs if run.get('status') == 'completed']
-    in_progress = [run for run in check_runs if run.get('status') in ['queued', 'in_progress']]
-
-    if in_progress:
-        return "pending", f"{len(in_progress)} check(s) in progress"
-
-    if not conclusions:
-        return "pending", "Checks queued"
-
-    if all(c == 'success' for c in conclusions):
-        return "success", f"All {len(conclusions)} checks passed"
-
-    failed = sum(1 for c in conclusions if c in ['failure', 'timed_out', 'action_required'])
-    if failed > 0:
-        return "failure", f"{failed} check(s) failed"
-
-    # Some checks are cancelled, skipped, or neutral
-    return "warning", f"Some checks did not complete normally"
-
 def pr_exists_with_title(repo_url, title, github_token):
     api_url = repo_url.replace("https://github.com/", "https://api.github.com/repos/") + "/pulls"
     headers = {'Authorization': f'token {github_token}'} if github_token else {}
@@ -525,127 +354,6 @@ def check_proofwidgets4_release(repo_url, target_toolchain, github_token):
     print(f"     You will need to create and push a tag v0.0.{next_version}")
     return False
 
-def check_reference_manual_release_title(repo_url, toolchain, pr_branch, github_token):
-    """Check if the reference-manual release notes title matches the release type.
-
-    For RC releases (e.g., v4.27.0-rc1), the title should contain the exact RC suffix.
-    For final releases (e.g., v4.27.0), the title should NOT contain any "-rc".
-
-    Returns True if check passes or is not applicable, False if title needs updating.
-    """
-    is_rc = is_release_candidate(toolchain)
-
-    # For RC releases, get the base version and RC suffix
-    # e.g., "v4.27.0-rc1" -> version="4.27.0", rc_suffix="-rc1"
-    if is_rc:
-        parts = toolchain.lstrip('v').split('-', 1)
-        version = parts[0]
-        rc_suffix = '-' + parts[1] if len(parts) > 1 else ''
-    else:
-        version = toolchain.lstrip('v')
-        rc_suffix = ''
-
-    # Construct the release notes file path (e.g., Manual/Releases/v4_27_0.lean for v4.27.0)
-    file_name = f"v{version.replace('.', '_')}.lean"  # "v4_27_0.lean"
-    file_path = f"Manual/Releases/{file_name}"
-
-    # Try to get the file from the PR branch first, then fall back to main branch
-    content = get_branch_content(repo_url, pr_branch, file_path, github_token)
-    if content is None:
-        # Try the default branch
-        content = get_branch_content(repo_url, "main", file_path, github_token)
-
-    if content is None:
-        print(f"  ⚠️  Could not check release notes file: {file_path}")
-        return True  # Don't block on this
-
-    # Look for the #doc line with the title
-    for line in content.splitlines():
-        if line.strip().startswith('#doc') and 'Manual' in line:
-            has_rc_in_title = '-rc' in line.lower()
-
-            if is_rc:
-                # For RC releases, title should contain the exact RC suffix (e.g., "-rc1")
-                # Use regex to match exact suffix followed by non-digit (to avoid -rc1 matching -rc10)
-                # Pattern matches the RC suffix followed by a non-digit or end-of-string context
-                # e.g., "-rc1" followed by space, quote, paren, or similar
-                exact_match = re.search(rf'{re.escape(rc_suffix)}(?![0-9])', line, re.IGNORECASE)
-                if exact_match:
-                    print(f"  ✅ Release notes title correctly shows {rc_suffix}")
-                    return True
-                elif has_rc_in_title:
-                    print(f"  ❌ Release notes title shows wrong RC version (expected {rc_suffix})")
-                    print(f"     Update {file_path} to use '{rc_suffix}' in the title")
-                    return False
-                else:
-                    print(f"  ❌ Release notes title missing RC suffix")
-                    print(f"     Update {file_path} to include '{rc_suffix}' in the title")
-                    return False
-            else:
-                # For final releases, title should NOT contain -rc
-                if has_rc_in_title:
-                    print(f"  ❌ Release notes title still shows RC version")
-                    print(f"     Update {file_path} to remove '-rcN' from the title")
-                    return False
-                else:
-                    print(f"  ✅ Release notes title is updated for final release")
-                    return True
-
-    # If we didn't find the #doc line, don't block
-    print(f"  ⚠️  Could not find release notes title in {file_path}")
-    return True
-
-def run_mathlib_verify_version_tags(toolchain, verbose=False):
-    """Run mathlib4's verify_version_tags.py script to validate the release tag.
-
-    This clones mathlib4 to a temp directory and runs the verification script.
-    Returns True if verification passes, False otherwise.
-    """
-    import tempfile
-
-    print(f"  ... Running mathlib4 verify_version_tags.py {toolchain}")
-
-    with tempfile.TemporaryDirectory() as tmpdir:
-        # Clone mathlib4 (shallow clone is sufficient for running the script)
-        clone_result = subprocess.run(
-            ['git', 'clone', '--depth', '1', 'https://github.com/leanprover-community/mathlib4.git', tmpdir],
-            capture_output=True,
-            text=True
-        )
-        if clone_result.returncode != 0:
-            print(f"  ❌ Failed to clone mathlib4: {clone_result.stderr.strip()[:200]}")
-            return False
-
-        # Run the verification script
-        script_path = os.path.join(tmpdir, 'scripts', 'verify_version_tags.py')
-        if not os.path.exists(script_path):
-            print(f"  ❌ verify_version_tags.py not found in mathlib4 (expected at scripts/verify_version_tags.py)")
-            return False
-
-        # Run from the mathlib4 directory so git operations work
-        result = subprocess.run(
-            ['python3', script_path, toolchain],
-            cwd=tmpdir,
-            capture_output=True,
-            text=True,
-            timeout=900  # 15 minutes timeout for cache download etc.
-        )
-
-        # Print output with indentation
-        if result.stdout:
-            for line in result.stdout.strip().split('\n'):
-                print(f"     {line}")
-        if result.stderr:
-            for line in result.stderr.strip().split('\n'):
-                print(f"     {line}")
-
-        if result.returncode != 0:
-            print(f"  ❌ mathlib4 verify_version_tags.py failed")
-            return False
-
-        print(f"  ✅ mathlib4 verify_version_tags.py passed")
-        return True
-
 def main():
     parser = argparse.ArgumentParser(description="Check release status of Lean4 repositories")
     parser.add_argument("toolchain", help="The toolchain version to check (e.g., v4.6.0)")
@@ -696,78 +404,30 @@ def main():
             print(f"  ❌ Short commit hash {commit_hash[:SHORT_HASH_LENGTH]} is numeric and starts with 0, causing issues for version parsing. Try regenerating the last commit to get a new hash.")
             lean4_success = False
 
-    release_page_ready = release_page_exists(lean_repo_url, toolchain, github_token)
-    if not release_page_ready:
-        print(f"  ❌ Release page for {toolchain} does not exist (This will be created by CI.)")
-
-        # Check CI workflow status
-        workflow_status = get_tag_workflow_status(lean_repo_url, toolchain, github_token)
-        if workflow_status:
-            status = workflow_status['status']
-            conclusion = workflow_status['conclusion']
-            workflow_name = workflow_status['workflow_name']
-            run_id = workflow_status['run_id']
-            workflow_url = f"{lean_repo_url}/actions/runs/{run_id}"
-
-            if status == 'in_progress' or status == 'queued':
-                print(f"     🔄 {workflow_name} workflow is {status}: {workflow_url}")
-            elif status == 'completed':
-                if conclusion == 'success':
-                    print(f"     ✅ {workflow_name} workflow completed successfully: {workflow_url}")
-                elif conclusion == 'failure':
-                    print(f"     ❌ {workflow_name} workflow failed: {workflow_url}")
-                else:
-                    print(f"     ⚠️  {workflow_name} workflow completed with status: {conclusion}: {workflow_url}")
-            else:
-                print(f"     ℹ️  {workflow_name} workflow status: {status}: {workflow_url}")
-
+    if not release_page_exists(lean_repo_url, toolchain, github_token):
+        print(f"  ❌ Release page for {toolchain} does not exist")
         lean4_success = False
     else:
         print(f"  ✅ Release page for {toolchain} exists")
-
-    # Check the actual release notes page title (informational only - does not block)
+        
+    # Check the actual release notes page title
     actual_title = get_release_notes(toolchain)
     expected_title_prefix = f"Lean {toolchain.lstrip('v')}" # e.g., "Lean 4.19.0" or "Lean 4.19.0-rc1"
-    base_tag = toolchain.split('-')[0]
-    release_notes_url = f"https://lean-lang.org/doc/reference/latest/releases/{base_tag}/"
 
     if actual_title is None:
-        print(f"  ⚠️  Release notes not found at {release_notes_url} (this will be fixed while updating the reference-manual repository)")
+        # Error already printed by get_release_notes
+        lean4_success = False
     elif not actual_title.startswith(expected_title_prefix):
-        print(f"  ⚠️  Release notes page title mismatch. Expected prefix '{expected_title_prefix}', got '{actual_title}'. Check {release_notes_url}")
+        # Construct URL for the error message (using the base tag)
+        base_tag = toolchain.split('-')[0]
+        check_url = f"https://lean-lang.org/doc/reference/latest/releases/{base_tag}/"
+        print(f"  ❌ Release notes page title mismatch. Expected prefix '{expected_title_prefix}', got '{actual_title}'. Check {check_url}")
+        lean4_success = False
     else:
         print(f"  ✅ Release notes page title looks good ('{actual_title}').")
 
-    # Check if release notes file exists in reference-manual repository
-    # For -rc1 releases, this is when release notes need to be written
-    # For subsequent RCs and stable releases, they should already exist
-    release_notes_exists, is_rc1 = check_release_notes_file_exists(toolchain, github_token)
-    base_version = strip_rc_suffix(toolchain.lstrip('v'))
-    release_notes_file = f"Manual/Releases/v{base_version.replace('.', '_')}.lean"
-
-    if not release_notes_exists:
-        if is_rc1:
-            print(f"  ❌ Release notes file not found: {release_notes_file}")
-            print(f"     This is an -rc1 release, so release notes need to be written.")
-            print(f"     Run `script/release_notes.py --since <previous_version>` to generate them.")
-            print(f"     See doc/dev/release_checklist.md section 'Writing the release notes' for details.")
-            lean4_success = False
-        else:
-            print(f"  ❌ Release notes file not found: {release_notes_file}")
-            print(f"     Release notes should have been created for -rc1. Check the reference-manual repository.")
-            lean4_success = False
-    else:
-        print(f"  ✅ Release notes file exists: {release_notes_file}")
-
     repo_status["lean4"] = lean4_success
 
-    # If the release page doesn't exist, skip repository checks and master branch checks
-    # The preliminary infrastructure must be in place first
-    if not release_page_exists(lean_repo_url, toolchain, github_token):
-        print("\n⚠️  Release process blocked: preliminary Lean4 infrastructure incomplete.")
-        print("   Complete the steps above, then rerun this script to proceed with downstream repositories.")
-        return
-
     # Load repositories and perform further checks
     print("\nChecking repositories...")
 
@@ -811,24 +471,6 @@ def main():
             if pr_info:
                 pr_number, pr_url = pr_info
                 print(f"  ✅ PR with title '{pr_title}' exists: #{pr_number} ({pr_url})")
-
-                # Check CI status
-                ci_status, ci_message = get_pr_ci_status(url, pr_number, github_token)
-                if ci_status == "success":
-                    print(f"     ✅ CI: {ci_message}")
-                elif ci_status == "failure":
-                    print(f"     ❌ CI: {ci_message}")
-                elif ci_status == "pending":
-                    print(f"     🔄 CI: {ci_message}")
-                elif ci_status == "warning":
-                    print(f"     ⚠️  CI: {ci_message}")
-                else:
-                    print(f"     ❓ CI: {ci_message}")
-
-                # For reference-manual, check that the release notes title has been updated
-                if name == "reference-manual":
-                    pr_branch = f"bump_to_{toolchain}"
-                    check_reference_manual_release_title(url, toolchain, pr_branch, github_token)
             else:
                 print(f"  ❌ PR with title '{pr_title}' does not exist")
                 print(f"     Run `script/release_steps.py {toolchain} {name}` to create it")
@@ -936,12 +578,6 @@ def main():
                 repo_status[name] = False
                 continue
 
-        # For mathlib4, run verify_version_tags.py to validate the release tag
-        if name == "mathlib4":
-            if not run_mathlib_verify_version_tags(toolchain, verbose):
-                repo_status[name] = False
-                continue
-
         repo_status[name] = success
 
     # Final check for lean4 master branch

script/release_repos.yml

@@ -14,6 +14,13 @@ repositories:
     bump-branch: true
     dependencies: []
 
+  - name: verso
+    url: https://github.com/leanprover/verso
+    toolchain-tag: true
+    stable-branch: false
+    branch: main
+    dependencies: []
+
   - name: lean4checker
     url: https://github.com/leanprover/lean4checker
     toolchain-tag: true
@@ -35,14 +42,6 @@ repositories:
     branch: main
     dependencies: []
 
-  - name: verso
-    url: https://github.com/leanprover/verso
-    toolchain-tag: true
-    stable-branch: false
-    branch: main
-    dependencies:
-      - plausible
-
   - name: import-graph
     url: https://github.com/leanprover-community/import-graph
     toolchain-tag: true
@@ -51,26 +50,12 @@ repositories:
     dependencies:
       - lean4-cli
 
-  - name: lean4-unicode-basic
-    url: https://github.com/fgdorais/lean4-unicode-basic
-    toolchain-tag: true
-    stable-branch: false
-    branch: main
-    dependencies: []
-
-  - name: BibtexQuery
-    url: https://github.com/dupuisf/BibtexQuery
-    toolchain-tag: true
-    stable-branch: false
-    branch: master
-    dependencies: [lean4-unicode-basic]
-
   - name: doc-gen4
     url: https://github.com/leanprover/doc-gen4
     toolchain-tag: true
     stable-branch: false
     branch: main
-    dependencies: [lean4-cli, BibtexQuery]
+    dependencies: [lean4-cli]
 
   - name: reference-manual
     url: https://github.com/leanprover/reference-manual
@@ -128,30 +113,10 @@ repositories:
     dependencies:
       - mathlib4
 
-  - name: verso-web-components
-    url: https://github.com/leanprover/verso-web-components
-    toolchain-tag: true
-    stable-branch: false
-    branch: main
-    dependencies:
-      - verso
-
   - name: lean-fro.org
     url: https://github.com/leanprover/lean-fro.org
     toolchain-tag: false
     stable-branch: false
     branch: master
     dependencies:
-      - verso-web-components
-
-  - name: comparator
-    url: https://github.com/leanprover/comparator
-    toolchain-tag: true
-    stable-branch: false
-    branch: master
-
-  - name: lean4export
-    url: https://github.com/leanprover/lean4export
-    toolchain-tag: true
-    stable-branch: false
-    branch: master
+      - verso

script/release_steps.py

@@ -1,54 +1,30 @@
 #!/usr/bin/env python3
 
 """
-Execute Release Steps for Lean4 Downstream Repositories
+Execute release steps for Lean4 repositories.
 
-This script automates the process of updating a downstream repository to a new Lean4 release.
-It handles creating branches, updating toolchains, merging changes, building, testing, and
-creating pull requests.
-
-IMPORTANT: Keep this documentation up-to-date when modifying the script's behavior!
-
-What this script does:
-1. Sets up the downstream_releases/ directory for cloning repositories
-
-2. Clones or updates the target repository
-
-3. Creates a branch named bump_to_{version} for the changes
-
-4. Updates the lean-toolchain file to the target version
-
-5. Handles repository-specific variations:
-   - Different dependency update mechanisms
-   - Special merging strategies for repositories with nightly-testing branches
-   - Safety checks for repositories using bump branches
-   - Custom build and test procedures
-   - lean-fro.org: runs scripts/update.sh to regenerate site content
-
-6. Commits the changes with message "chore: bump toolchain to {version}"
-
-7. Builds the project (with a clean .lake cache)
-
-8. Runs tests if available
-
-9. Pushes the branch to GitHub
-
-10. Creates a pull request (or reports if one already exists)
+This script helps automate the release process for Lean4 and its dependent repositories
+by actually executing the step-by-step instructions for updating toolchains, creating tags,
+and managing branches.
 
 Usage:
-    ./release_steps.py v4.24.0 batteries    # Update batteries to v4.24.0
-    ./release_steps.py v4.24.0-rc1 mathlib4 # Update mathlib4 to v4.24.0-rc1
+    python3 release_steps.py <version> <repo>
 
-The script reads repository configurations from release_repos.yml.
-Each repository has specific handling for merging, dependencies, and testing.
+Arguments:
+    version: The version to set in the lean-toolchain file (e.g., v4.6.0)
+    repo: The repository name as specified in release_repos.yml
 
-This script is idempotent - it's safe to rerun if it fails partway through.
-Existing branches, commits, and PRs will be reused rather than duplicated.
+Example:
+    python3 release_steps.py v4.6.0 mathlib4
+    python3 release_steps.py v4.6.0 batteries
 
-Error handling:
-- If build or tests fail, the script continues to create the PR anyway
-- Manual conflicts must be resolved by the user
-- Network issues during push/PR creation are reported with manual instructions
+The script reads repository configurations from release_repos.yml in the same directory.
+Each repository may have specific requirements for:
+- Branch management
+- Toolchain updates
+- Dependency updates
+- Tagging conventions
+- Stable branch handling
 """
 
 import argparse
@@ -413,14 +389,20 @@ def execute_release_steps(repo, version, config):
         run_command("lake update", cwd=repo_path, stream_output=True)
         print(blue("Running `lake update` in examples/hero..."))
         run_command("lake update", cwd=repo_path / "examples" / "hero", stream_output=True)
-
-        # Run scripts/update.sh to regenerate content
-        print(blue("Running `scripts/update.sh` to regenerate content..."))
-        run_command("scripts/update.sh", cwd=repo_path, stream_output=True)
-        print(green("Content regenerated successfully"))
     elif repo_name == "cslib":
         print(blue("Updating lakefile.toml..."))
         run_command(f'perl -pi -e \'s/"v4\\.[0-9]+(\\.[0-9]+)?(-rc[0-9]+)?"/"' + version + '"/g\' lakefile.*', cwd=repo_path)
+        
+        print(blue("Updating docs/lakefile.toml..."))
+        run_command(f'perl -pi -e \'s/"v4\\.[0-9]+(\\.[0-9]+)?(-rc[0-9]+)?"/"' + version + '"/g\' lakefile.*', cwd=repo_path / "docs")
+
+        # Update lean-toolchain in docs
+        print(blue("Updating docs/lean-toolchain..."))
+        docs_toolchain = repo_path / "docs" / "lean-toolchain"
+        with open(docs_toolchain, "w") as f:
+            f.write(f"leanprover/lean4:{version}\n")
+        print(green(f"Updated docs/lean-toolchain to leanprover/lean4:{version}"))
+
         run_command("lake update", cwd=repo_path, stream_output=True)
     elif dependencies:
         run_command(f'perl -pi -e \'s/"v4\\.[0-9]+(\\.[0-9]+)?(-rc[0-9]+)?"/"' + version + '"/g\' lakefile.*', cwd=repo_path)
@@ -584,19 +566,8 @@ def execute_release_steps(repo, version, config):
         
         # Clean lake cache for a fresh build
         print(blue("Cleaning lake cache..."))
-        run_command("lake clean", cwd=repo_path)
-
-        # Check if downstream of Mathlib and get cache if so
-        mathlib_package_dir = repo_path / ".lake" / "packages" / "mathlib"
-        if mathlib_package_dir.exists():
-            print(blue("Project is downstream of Mathlib, fetching cache..."))
-            try:
-                run_command("lake exe cache get", cwd=repo_path, stream_output=True)
-                print(green("Cache fetched successfully"))
-            except subprocess.CalledProcessError as e:
-                print(yellow("Failed to fetch cache, continuing anyway..."))
-                print(yellow(f"Cache fetch error: {e}"))
-
+        run_command("rm -rf .lake", cwd=repo_path)
+        
         try:
             run_command("lake build", cwd=repo_path, stream_output=True)
             print(green("Build completed successfully"))

src/Init.lean

@@ -4,6 +4,7 @@ Released under Apache 2.0 license as described in the file LICENSE.
 Authors: Leonardo de Moura
 -/
 module
+
 prelude
 public import Init.Prelude
 public import Init.Notation
@@ -13,8 +14,8 @@ public import Init.ByCases
 public import Init.RCases
 public import Init.Core
 public import Init.Control
+public import Init.Data.Basic
 public import Init.WF
-public import Init.WFComputable
 public import Init.WFTactics
 public import Init.Data
 public import Init.System
@@ -37,12 +38,11 @@ public import Init.Omega
 public import Init.MacroTrace
 public import Init.Grind
 public import Init.GrindInstances
-public import Init.Sym
 public import Init.While
 public import Init.Syntax
 public import Init.Internal
 public import Init.Try
-public meta import Init.Try  -- shake: keep (make sure `Try.Config` can be evaluated anywhere)
+public meta import Init.Try  -- make sure `Try.Config` can be evaluated anywhere
 public import Init.BinderNameHint
 public import Init.Task
 public import Init.MethodSpecsSimp

src/Init/BinderNameHint.lean

@@ -8,7 +8,7 @@ module
 
 prelude
 public import Init.Prelude
-import Init.Tactics
+public import Init.Tactics
 
 public section
 

src/Init/BinderPredicates.lean

@@ -6,10 +6,7 @@ Authors: Gabriel Ebner
 module
 
 prelude
-public meta import Init.Grind.Tactics
-public import Init.Notation
-import Init.Meta.Defs
-import Init.NotationExtra
+public import Init.NotationExtra
 
 public section
 

src/Init/ByCases.lean

@@ -6,9 +6,7 @@ Authors: Leonardo de Moura, Mario Carneiro
 module
 
 prelude
-public meta import Init.Grind.Tactics
-public import Init.Grind.Tactics
-import Init.SimpLemmas
+public import Init.Classical
 
 public section
 
@@ -46,10 +44,3 @@ theorem apply_ite (f : α → β) (P : Prop) [Decidable P] (x y : α) :
 /-- A `dite` whose results do not actually depend on the condition may be reduced to an `ite`. -/
 @[simp] theorem dite_eq_ite [Decidable P] :
   (dite P (fun _ => a) (fun _ => b)) = ite P a b := rfl
-
--- Remark: dite and ite are "defally equal" when we ignore the proofs.
-@[deprecated dite_eq_ite (since := "2025-10-29")]
-theorem dif_eq_if (c : Prop) {h : Decidable c} {α : Sort u} (t : α) (e : α) : dite c (fun _ => t) (fun _ => e) = ite c t e :=
-  match h with
-  | isTrue _    => rfl
-  | isFalse _   => rfl

src/Init/Classical.lean

@@ -8,7 +8,7 @@ module
 prelude
 public import Init.PropLemmas
 
-@[expose] public section
+public section
 
 universe u v
 
@@ -102,7 +102,7 @@ noncomputable def strongIndefiniteDescription {α : Sort u} (p : α → Prop) (h
       ⟨xp.val, fun _ => xp.property⟩)
     (fun hp => ⟨choice h, fun h => absurd h hp⟩)
 
-/-- The Hilbert epsilon function. -/
+/-- the Hilbert epsilon Function -/
 noncomputable def epsilon {α : Sort u} [h : Nonempty α] (p : α → Prop) : α :=
   (strongIndefiniteDescription p h).val
 
@@ -142,7 +142,6 @@ is classically true but not constructively. -/
 
 /-- Transfer decidability of `¬ p` to decidability of `p`. -/
 -- This can not be an instance as it would be tried everywhere.
-@[instance_reducible]
 def decidable_of_decidable_not (p : Prop) [h : Decidable (¬ p)] : Decidable p :=
   match h with
   | isFalse h => isTrue (Classical.not_not.mp h)
@@ -182,6 +181,9 @@ theorem not_imp_iff_and_not : ¬(a → b) ↔ a ∧ ¬b := Decidable.not_imp_iff
 
 theorem not_and_iff_not_or_not : ¬(a ∧ b) ↔ ¬a ∨ ¬b := Decidable.not_and_iff_not_or_not
 
+@[deprecated not_and_iff_not_or_not (since := "2025-03-18")]
+abbrev not_and_iff_or_not_not := @not_and_iff_not_or_not
+
 theorem not_iff : ¬(a ↔ b) ↔ (¬a ↔ b) := Decidable.not_iff
 
 @[simp] theorem imp_iff_left_iff  : (b ↔ a → b) ↔ a ∨ b := Decidable.imp_iff_left_iff
@@ -206,5 +208,3 @@ export Classical (imp_iff_right_iff imp_and_neg_imp_iff and_or_imp not_imp)
 
 /-- Show that an element extracted from `P : ∃ a, p a` using `P.choose` satisfies `p`. -/
 theorem Exists.choose_spec {p : α → Prop} (P : ∃ a, p a) : p P.choose := Classical.choose_spec P
-
-grind_pattern Exists.choose_spec => P.choose

src/Init/Coe.lean

@@ -116,7 +116,7 @@ On top of these instances this file defines several auxiliary type classes:
   * `CoeOTC := CoeOut* Coe*`
   * `CoeHTC := CoeHead? CoeOut* Coe*`
   * `CoeHTCT := CoeHead? CoeOut* Coe* CoeTail?`
-  * `CoeT := CoeHead? CoeOut* Coe* CoeTail? | CoeDep`
+  * `CoeDep := CoeHead? CoeOut* Coe* CoeTail? | CoeDep`
 
 -/
 

src/Init/Control.lean

@@ -16,5 +16,5 @@ public import Init.Control.Option
 public import Init.Control.Lawful
 public import Init.Control.StateCps
 public import Init.Control.ExceptCps
-public import Init.Control.MonadAttach
-public import Init.Control.EState
+
+public section

src/Init/Control/Basic.lean

@@ -25,11 +25,11 @@ instances are provided for the same type.
 instance (priority := 500) instForInOfForIn' [ForIn' m ρ α d] : ForIn m ρ α where
   forIn x b f := forIn' x b fun a _ => f a
 
-@[simp] theorem forIn'_eq_forIn [d : Membership α ρ] [ForIn' m ρ α d] {β} (x : ρ) (b : β)
+@[simp] theorem forIn'_eq_forIn [d : Membership α ρ] [ForIn' m ρ α d] {β} [Monad m] (x : ρ) (b : β)
     (f : (a : α) → a ∈ x → β → m (ForInStep β)) (g : (a : α) → β → m (ForInStep β))
     (h : ∀ a m b, f a m b = g a b) :
     forIn' x b f = forIn x b g := by
-  simp [forIn]
+  simp [instForInOfForIn']
   congr
   apply funext
   intro a
@@ -40,11 +40,14 @@ instance (priority := 500) instForInOfForIn' [ForIn' m ρ α d] : ForIn m ρ α
   simp [h]
   rfl
 
-@[wf_preprocess] theorem forIn_eq_forIn' [d : Membership α ρ] [ForIn' m ρ α d] {β}
+@[wf_preprocess] theorem forIn_eq_forIn' [d : Membership α ρ] [ForIn' m ρ α d] {β} [Monad m]
     (x : ρ) (b : β) (f : (a : α) → β → m (ForInStep β)) :
     forIn x b f = forIn' x b (fun x h => binderNameHint x f <| binderNameHint h () <| f x) := by
   rfl
 
+@[deprecated forIn_eq_forIn' (since := "2025-04-04")]
+abbrev forIn_eq_forin' := @forIn_eq_forIn'
+
 /--
 Extracts the value from a `ForInStep`, ignoring whether it is `ForInStep.done` or `ForInStep.yield`.
 -/
@@ -144,7 +147,7 @@ instance : ToBool Bool where
 Converts the result of the monadic action `x` to a `Bool`. If it is `true`, returns it and ignores
 `y`; otherwise, runs `y` and returns its result.
 
-This is a monadic counterpart to the short-circuiting `||` operator, usually accessed via the `<||>`
+This a monadic counterpart to the short-circuiting `||` operator, usually accessed via the `<||>`
 operator.
 -/
 @[macro_inline] def orM {m : Type u → Type v} {β : Type u} [Monad m] [ToBool β] (x y : m β) : m β := do
@@ -161,7 +164,7 @@ recommended_spelling "orM" for "<||>" in [orM, «term_<||>_»]
 Converts the result of the monadic action `x` to a `Bool`. If it is `true`, returns `y`; otherwise,
 returns the original result of `x`.
 
-This is a monadic counterpart to the short-circuiting `&&` operator, usually accessed via the `<&&>`
+This a monadic counterpart to the short-circuiting `&&` operator, usually accessed via the `<&&>`
 operator.
 -/
 @[macro_inline] def andM {m : Type u → Type v} {β : Type u} [Monad m] [ToBool β] (x y : m β) : m β := do
@@ -322,8 +325,6 @@ class MonadControl (m : semiOutParam (Type u → Type v)) (n : Type u → Type w
   -/
   restoreM : {α : Type u} → m (stM α) → n α
 
-attribute [reducible] MonadControl.stM
-
 /--
 A way to lift a computation from one monad to another while providing the lifted computation with a
 means of interpreting computations from the outer monad. This provides a means of lifting
@@ -351,8 +352,6 @@ class MonadControlT (m : Type u → Type v) (n : Type u → Type w) where
   -/
   restoreM {α : Type u} : stM α → n α
 
-attribute [reducible] MonadControlT.stM
-
 export MonadControlT (stM liftWith restoreM)
 
 @[always_inline]
@@ -407,7 +406,7 @@ class ForM (m : Type u → Type v) (γ : Type w₁) (α : outParam (Type w₂))
   /--
   Runs the monadic action `f` on each element of the collection `coll`.
   -/
-  forM (coll : γ) (f : α → m PUnit) : m PUnit
+  forM [Monad m] (coll : γ) (f : α → m PUnit) : m PUnit
 
 export ForM (forM)
 

src/Init/Control/EState.lean

@@ -6,8 +6,9 @@ Authors: Leonardo de Moura
 module
 
 prelude
-public import Init.Data.ToString.Basic
 public import Init.Control.State
+public import Init.Control.Except
+public import Init.Data.ToString.Basic
 
 public section
 universe u v
@@ -26,12 +27,6 @@ instance [Repr ε] [Repr α] : Repr (Result ε σ α) where
     | Result.error e _, prec => Repr.addAppParen ("EStateM.Result.error " ++ reprArg e) prec
     | Result.ok a _, prec    => Repr.addAppParen ("EStateM.Result.ok " ++ reprArg a) prec
 
-instance : MonadAttach (EStateM ε σ) where
-  CanReturn x a := Exists fun s => Exists fun s' => x.run s = .ok a s'
-  attach x s := match h : x s with
-    | .ok a s' => .ok ⟨a, s, s', h⟩ s'
-    | .error e s' => .error e s'
-
 end EStateM
 
 namespace EStateM

src/Init/Control/Except.lean

@@ -10,6 +10,7 @@ module
 prelude
 public import Init.Control.Basic
 public import Init.Control.Id
+public import Init.Coe
 
 @[expose] public section
 
@@ -148,23 +149,6 @@ This is the inverse of `ExceptT.mk`.
 @[always_inline, inline, expose]
 def ExceptT.run {ε : Type u} {m : Type u → Type v} {α : Type u} (x : ExceptT ε m α) : m (Except ε α) := x
 
-/--
-Use a monadic action that may throw an exception by providing explicit success and failure
-continuations.
--/
-@[always_inline, inline, expose]
-def ExceptT.runK [Monad m] (x : ExceptT ε m α) (ok : α → m β) (error : ε → m β) : m β :=
-  x.run >>= (·.casesOn error ok)
-
-/--
-Returns the value of a computation, forgetting whether it was an exception or a success.
-
-This corresponds to early return.
--/
-@[always_inline, inline, expose]
-def ExceptT.runCatch [Monad m] (x : ExceptT α m α) : m α :=
-  x.runK pure pure
-
 namespace ExceptT
 
 variable {ε : Type u} {m : Type u → Type v} [Monad m]
@@ -329,8 +313,3 @@ instance ExceptT.finally {m : Type u → Type v} {ε : Type u} [MonadFinally m]
     | (.ok a,    .ok b)    => pure (.ok (a, b))
     | (_,        .error e) => pure (.error e)  -- second error has precedence
     | (.error e, _)        => pure (.error e)
-
-instance [Monad m] [MonadAttach m] : MonadAttach (ExceptT ε m) where
-  CanReturn x a := MonadAttach.CanReturn (m := m) x (.ok a)
-  attach x := show m (Except ε _) from
-      (fun ⟨a, h⟩ => match a with | .ok a => .ok ⟨a, h⟩ | .error e => .error e) <$> MonadAttach.attach (m := m) x

src/Init/Control/ExceptCps.lean

@@ -7,7 +7,6 @@ module
 
 prelude
 public import Init.Control.Lawful.Basic
-import Init.SimpLemmas
 
 public section
 
@@ -76,13 +75,6 @@ instance [Monad m] : MonadLift m (ExceptCpsT σ m) where
 instance [Inhabited ε] : Inhabited (ExceptCpsT ε m α) where
   default := fun _ _ k₂ => k₂ default
 
-/--
-For continuation monads, it is not possible to provide a computable `MonadAttach` instance that
-actually adds information about the return value. Therefore, this instance always attaches a proof
-of `True`.
--/
-instance : MonadAttach (ExceptCpsT ε m) := .trivial
-
 @[simp] theorem run_pure [Monad m] : run (pure x : ExceptCpsT ε m α) = pure (Except.ok x) := rfl
 
 @[simp] theorem run_lift {α ε : Type u} [Monad m] (x : m α) : run (ExceptCpsT.lift x : ExceptCpsT ε m α) = (x >>= fun a => pure (Except.ok a) : m (Except ε α)) := rfl

src/Init/Control/Id.lean

@@ -8,7 +8,7 @@ The identity Monad.
 module
 
 prelude
-public import Init.Control.MonadAttach
+public import Init.Core
 
 public section
 
@@ -67,15 +67,4 @@ instance [OfNat α n] : OfNat (Id α) n :=
 instance {m : Type u → Type v} [Pure m] : MonadLiftT Id m where
   monadLift x := pure x.run
 
-instance : MonadAttach Id where
-  CanReturn x a := x.run = a
-  attach x := pure ⟨x.run, rfl⟩
-
-instance : LawfulMonadAttach Id where
-  map_attach := rfl
-  canReturn_map_imp := by
-    intro _ _ x _ h
-    cases h
-    exact x.run.2
-
 end Id

src/Init/Control/Lawful.lean

@@ -10,4 +10,5 @@ public import Init.Control.Lawful.Basic
 public import Init.Control.Lawful.Instances
 public import Init.Control.Lawful.Lemmas
 public import Init.Control.Lawful.MonadLift
-public import Init.Control.Lawful.MonadAttach
+
+public section

src/Init/Control/Lawful/Basic.lean

@@ -6,9 +6,9 @@ Authors: Sebastian Ullrich, Leonardo de Moura, Mario Carneiro
 module
 
 prelude
-public import Init.Control.Id
-public import Init.Grind.Tactics
-import Init.Ext
+public import Init.Ext
+public import Init.SimpLemmas
+public import Init.Meta
 
 public section
 
@@ -172,7 +172,6 @@ theorem bind_pure_unit [Monad m] [LawfulMonad m] {x : m PUnit} : (x >>= fun _ =>
 theorem map_congr [Functor m] {x : m α} {f g : α → β} (h : ∀ a, f a = g a) : (f <$> x : m β) = g <$> x := by
   simp [funext h]
 
-@[deprecated seq_eq_bind_map (since := "2025-10-26")]
 theorem seq_eq_bind {α β : Type u} [Monad m] [LawfulMonad m] (mf : m (α → β)) (x : m α) : mf <*> x = mf >>= fun f => f <$> x := by
   rw [bind_map]
 
@@ -250,12 +249,27 @@ namespace Id
 instance : LawfulMonad Id := by
   refine LawfulMonad.mk' _ ?_ ?_ ?_ <;> intros <;> rfl
 
-@[simp, grind =] theorem run_map (x : Id α) (f : α → β) : (f <$> x).run = f x.run := rfl
-@[simp, grind =] theorem run_bind (x : Id α) (f : α → Id β) : (x >>= f).run = (f x.run).run := rfl
-@[simp, grind =] theorem run_pure (a : α) : (pure a : Id α).run = a := rfl
-@[simp, grind =] theorem pure_run (a : Id α) : pure a.run = a := rfl
+@[simp] theorem run_map (x : Id α) (f : α → β) : (f <$> x).run = f x.run := rfl
+@[simp] theorem run_bind (x : Id α) (f : α → Id β) : (x >>= f).run = (f x.run).run := rfl
+@[simp] theorem run_pure (a : α) : (pure a : Id α).run = a := rfl
 @[simp] theorem run_seqRight (x y : Id α) : (x *> y).run = y.run := rfl
 @[simp] theorem run_seqLeft (x y : Id α) : (x <* y).run = x.run := rfl
 @[simp] theorem run_seq (f : Id (α → β)) (x : Id α) : (f <*> x).run = f.run x.run := rfl
 
+-- These lemmas are bad as they abuse the defeq of `Id α` and `α`
+@[deprecated run_map (since := "2025-03-05")] theorem map_eq (x : Id α) (f : α → β) : f <$> x = f x := rfl
+@[deprecated run_bind (since := "2025-03-05")] theorem bind_eq (x : Id α) (f : α → id β) : x >>= f = f x := rfl
+@[deprecated run_pure (since := "2025-03-05")] theorem pure_eq (a : α) : (pure a : Id α) = a := rfl
+
 end Id
+
+/-! # Option -/
+
+instance : LawfulMonad Option := LawfulMonad.mk'
+  (id_map := fun x => by cases x <;> rfl)
+  (pure_bind := fun _ _ => rfl)
+  (bind_assoc := fun x _ _ => by cases x <;> rfl)
+  (bind_pure_comp := fun _ x => by cases x <;> rfl)
+
+instance : LawfulApplicative Option := inferInstance
+instance : LawfulFunctor Option := inferInstance

src/Init/Control/Lawful/Instances.lean

@@ -7,21 +7,17 @@ module
 
 prelude
 public import Init.Control.Lawful.Basic
+public import Init.Control.Except
 import all Init.Control.Except
-public import Init.Control.Option
-import all Init.Control.Option
+public import Init.Control.State
 import all Init.Control.State
 public import Init.Control.StateRef
-public import Init.Control.State
 public import Init.Ext
 
 public section
 
 open Function
 
-@[simp, grind =] theorem monadMap_refl {m : Type _ → Type _} {α} (f : ∀ {α}, m α → m α) :
-    monadMap @f = @f α := rfl
-
 /-! # ExceptT -/
 
 namespace ExceptT
@@ -30,8 +26,6 @@ namespace ExceptT
   simp [run] at h
   assumption
 
-@[simp, grind =] theorem run_mk (x : m (Except ε α)) : run (mk x : ExceptT ε m α) = x := rfl
-
 @[simp, grind =] theorem run_pure [Monad m] (x : α) : run (pure x : ExceptT ε m α) = pure (Except.ok x) := rfl
 
 @[simp, grind =] theorem run_lift  [Monad.{u, v} m] (x : m α) : run (ExceptT.lift x : ExceptT ε m α) = (Except.ok <$> x : m (Except ε α)) := rfl
@@ -62,9 +56,6 @@ theorem run_bind [Monad m] (x : ExceptT ε m α) (f : α → ExceptT ε m β)
   apply bind_congr
   intro a; cases a <;> simp [Except.map]
 
-@[simp, grind =] theorem run_monadMap [MonadFunctorT n m] (f : {β : Type u} → n β → n β) (x : ExceptT ε m α)
-    : (monadMap @f x : ExceptT ε m α).run = monadMap @f (x.run) := rfl
-
 protected theorem seq_eq {α β ε : Type u} [Monad m] (mf : ExceptT ε m (α → β)) (x : ExceptT ε m α) : mf <*> x = mf >>= fun f => f <$> x :=
   rfl
 
@@ -104,30 +95,9 @@ instance [Monad m] [LawfulMonad m] : LawfulMonad (ExceptT ε m) where
 
 @[simp] theorem map_throw [Monad m] [LawfulMonad m] {α β : Type _} (f : α → β) (e : ε) :
     f <$> (throw e : ExceptT ε m α) = (throw e : ExceptT ε m β) := by
-  simp only [Functor.map, ExceptT.map, ExceptT.mk, throw, throwThe, MonadExceptOf.throw,
+  simp only [ExceptT.instMonad, ExceptT.map, ExceptT.mk, throw, throwThe, MonadExceptOf.throw,
     pure_bind]
 
-/-! Note that the `MonadControl` instance for `ExceptT` is not monad-generic. -/
-
-@[simp] theorem run_restoreM [Monad m] (x : stM m (ExceptT ε m) α) :
-    ExceptT.run (restoreM x) = pure x := rfl
-
-@[simp] theorem run_liftWith [Monad m] (f : ({β : Type u} → ExceptT ε m β → m (stM m (ExceptT ε m) β)) → m α) :
-    ExceptT.run (liftWith f) = Except.ok <$> (f fun x => x.run) :=
-  rfl
-
-@[simp] theorem run_controlAt [Monad m] [LawfulMonad m] (f : ({β : Type u} → ExceptT ε m β → m (stM m (ExceptT ε m) β)) → m (stM m (ExceptT ε m) α)) :
-    ExceptT.run (controlAt m f) = f fun x => x.run := by
-  simp [controlAt, run_bind, bind_map_left]
-
-@[simp] theorem run_control [Monad m] [LawfulMonad m] (f : ({β : Type u} → ExceptT ε m β → m (stM m (ExceptT ε m) β)) → m (stM m (ExceptT ε m) α)) :
-    ExceptT.run (control f) = f fun x => x.run := run_controlAt f
-
-@[simp, grind =]
-theorem run_adapt [Monad m] (f : ε → ε') (x : ExceptT ε m α)
-    : run (ExceptT.adapt f x : ExceptT ε' m α) = Except.mapError f <$> run x :=
-  rfl
-
 end ExceptT
 
 /-! # Except -/
@@ -140,142 +110,6 @@ instance : LawfulMonad (Except ε) := LawfulMonad.mk'
 instance : LawfulApplicative (Except ε) := inferInstance
 instance : LawfulFunctor (Except ε) := inferInstance
 
-/-! # OptionT -/
-
-namespace OptionT
-
-@[ext] theorem ext {x y : OptionT m α} (h : x.run = y.run) : x = y := by
-  simp [run] at h
-  assumption
-
-@[simp, grind =] theorem run_mk {m : Type u → Type v} (x : m (Option α)) :
-    OptionT.run (OptionT.mk x) = x := by rfl
-
-@[simp, grind =] theorem run_pure [Monad m] (x : α) : run (pure x : OptionT m α) = pure (some x) := by
-  simp [run, pure, OptionT.pure, OptionT.mk]
-
-@[simp, grind =] theorem run_lift  [Monad.{u, v} m] (x : m α) : run (OptionT.lift x : OptionT m α) = (return some (← x) : m (Option α)) := by
-  simp [run, OptionT.lift, OptionT.mk]
-
-@[simp, grind =] theorem run_throw [Monad m] : run (throw e : OptionT m β) = pure none := by
-  simp [run, throw, throwThe, MonadExceptOf.throw, OptionT.fail, OptionT.mk]
-
-@[simp, grind =] theorem run_bind_lift [Monad m] [LawfulMonad m] (x : m α) (f : α → OptionT m β) : run (OptionT.lift x >>= f : OptionT m β) = x >>= fun a => run (f a) := by
-  simp [OptionT.run, OptionT.lift, bind, OptionT.bind, OptionT.mk]
-
-@[simp, grind =] theorem bind_throw [Monad m] [LawfulMonad m] (f : α → OptionT m β) : (throw e >>= f) = throw e := by
-  simp [throw, throwThe, MonadExceptOf.throw, bind, OptionT.bind, OptionT.mk, OptionT.fail]
-
-@[simp, grind =] theorem run_bind (f : α → OptionT m β) [Monad m] :
-    (x >>= f).run = Option.elimM x.run (pure none) (fun x => (f x).run) := by
-  change x.run >>= _ = _
-  simp [Option.elimM]
-  exact bind_congr fun |some _ => rfl | none => rfl
-
-@[simp, grind =] theorem lift_pure [Monad m] [LawfulMonad m] {α : Type u} (a : α) : OptionT.lift (pure a : m α) = pure a := by
-  simp only [OptionT.lift, OptionT.mk, bind_pure_comp, map_pure, pure, OptionT.pure]
-
-@[simp, grind =] theorem run_map [Monad m] [LawfulMonad m] (f : α → β) (x : OptionT m α)
-    : (f <$> x).run = Option.map f <$> x.run := by
-  simp [Functor.map, Option.map, ←bind_pure_comp]
-  apply bind_congr
-  intro a; cases a <;> simp [OptionT.pure, OptionT.mk]
-
-@[simp, grind =] theorem run_monadMap [MonadFunctorT n m] (f : {β : Type u} → n β → n β) (x : OptionT m α)
-    : (monadMap @f x : OptionT m α).run = monadMap @f (x.run) := rfl
-
-protected theorem seq_eq {α β : Type u} [Monad m] (mf : OptionT m (α → β)) (x : OptionT m α) : mf <*> x = mf >>= fun f => f <$> x :=
-  rfl
-
-protected theorem bind_pure_comp [Monad m] (f : α → β) (x : OptionT m α) : x >>= pure ∘ f = f <$> x := by
-  intros; rfl
-
-protected theorem seqLeft_eq {α β : Type u} {m : Type u → Type v} [Monad m] [LawfulMonad m] (x : OptionT m α) (y : OptionT m β) : x <* y = const β <$> x <*> y := by
-  change (x >>= fun a => y >>= fun _ => pure a) = (const (α := α) β <$> x) >>= fun f => f <$> y
-  rw [← OptionT.bind_pure_comp]
-  apply ext
-  simp [Option.elimM, Option.elim]
-  apply bind_congr
-  intro
-  | none => simp
-  | some _ =>
-    simp [←bind_pure_comp]; apply bind_congr; intro b;
-    cases b <;> simp [const]
-
-protected theorem seqRight_eq [Monad m] [LawfulMonad m] (x : OptionT m α) (y : OptionT m β) : x *> y = const α id <$> x <*> y := by
-  change (x >>= fun _ => y) = (const α id <$> x) >>= fun f => f <$> y
-  rw [← OptionT.bind_pure_comp]
-  apply ext
-  simp [Option.elimM, Option.elim]
-  apply bind_congr
-  intro a; cases a <;> simp
-
-instance [Monad m] [LawfulMonad m] : LawfulMonad (OptionT m) where
-  id_map         := by intros; apply ext; simp
-  map_const      := by intros; rfl
-  seqLeft_eq     := OptionT.seqLeft_eq
-  seqRight_eq    := OptionT.seqRight_eq
-  pure_seq       := by intros; apply ext; simp [OptionT.seq_eq, Option.elimM, Option.elim]
-  bind_pure_comp := OptionT.bind_pure_comp
-  bind_map       := by intros; rfl
-  pure_bind      := by intros; apply ext; simp [Option.elimM, Option.elim]
-  bind_assoc     := by intros; apply ext; simp [Option.elimM, Option.elim]; apply bind_congr; intro a; cases a <;> simp
-
-@[simp] theorem run_seq [Monad m] [LawfulMonad m] (f : OptionT m (α → β)) (x : OptionT m α) :
-    (f <*> x).run = Option.elimM f.run (pure none) (fun f => Option.map f <$> x.run) := by
-  simp [seq_eq_bind_map, Option.elimM, Option.elim]
-
-@[simp] theorem run_seqLeft [Monad m] [LawfulMonad m] (x : OptionT m α) (y : OptionT m β) :
-    (x <* y).run = Option.elimM x.run (pure none)
-      (fun x => Option.map (Function.const β x) <$> y.run) := by
-  simp [seqLeft_eq, seq_eq_bind_map, Option.elimM, OptionT.run_bind]
-
-@[simp] theorem run_seqRight [Monad m] [LawfulMonad m] (x : OptionT m α) (y : OptionT m β) :
-    (x *> y).run = Option.elimM x.run (pure none) (Function.const α y.run) := by
-  simp only [seqRight_eq, run_seq, Option.elimM, run_map, Option.elim, bind_map_left]
-  refine bind_congr (fun | some _ => by simp | none => by simp)
-
-@[simp, grind =] theorem run_failure [Monad m] : (failure : OptionT m α).run = pure none := by rfl
-
-@[simp] theorem map_failure [Monad m] [LawfulMonad m] {α β : Type _} (f : α → β) :
-    f <$> (failure : OptionT m α) = (failure : OptionT m β) := by
-  simp [OptionT.mk, Functor.map, Alternative.failure, OptionT.fail, OptionT.bind]
-
-@[simp] theorem run_orElse [Monad m] (x : OptionT m α) (y : OptionT m α) :
-    (x <|> y).run = Option.elimM x.run y.run (fun x => pure (some x)) :=
-  bind_congr fun | some _ => by rfl | none => by rfl
-
-/-! Note that the `MonadControl` instance for `OptionT` is not monad-generic. -/
-
-@[simp] theorem run_restoreM [Monad m] (x : stM m (OptionT m) α) :
-    OptionT.run (restoreM x) = pure x := rfl
-
-@[simp] theorem run_liftWith [Monad m] [LawfulMonad m] (f : ({β : Type u} → OptionT m β → m (stM m (OptionT m) β)) → m α) :
-    OptionT.run (liftWith f) = Option.some <$> (f fun x => x.run) := by
-  dsimp [liftWith]
-  rw [← bind_pure_comp]
-  rfl
-
-@[simp] theorem run_controlAt [Monad m] [LawfulMonad m] (f : ({β : Type u} → OptionT m β → m (stM m (OptionT m) β)) → m (stM m (OptionT m) α)) :
-    OptionT.run (controlAt m f) = f fun x => x.run := by
-  simp [controlAt, Option.elimM, Option.elim]
-
-@[simp] theorem run_control [Monad m] [LawfulMonad m] (f : ({β : Type u} → OptionT m β → m (stM m (OptionT m) β)) → m (stM m (OptionT m) α)) :
-    OptionT.run (control f) = f fun x => x.run := run_controlAt f
-
-end OptionT
-
-/-! # Option -/
-
-instance : LawfulMonad Option := LawfulMonad.mk'
-  (id_map := fun x => by cases x <;> rfl)
-  (pure_bind := fun _ _ => by rfl)
-  (bind_assoc := fun a _ _ => by cases a <;> rfl)
-  (bind_pure_comp := fun _ x => by cases x <;> rfl)
-
-instance : LawfulApplicative Option := inferInstance
-instance : LawfulFunctor Option := inferInstance
-
 /-! # ReaderT -/
 
 namespace ReaderT
@@ -284,9 +118,6 @@ namespace ReaderT
   simp [run] at h
   exact funext h
 
-@[simp, grind =] theorem run_mk (x : ρ → m α) (ctx : ρ) : run (.mk x : ReaderT ρ m α) ctx = x ctx :=
-  rfl
-
 @[simp, grind =] theorem run_pure [Monad m] (a : α) (ctx : ρ) : (pure a : ReaderT ρ m α).run ctx = pure a := rfl
 
 @[simp, grind =] theorem run_bind [Monad m] (x : ReaderT ρ m α) (f : α → ReaderT ρ m β) (ctx : ρ)
@@ -334,22 +165,6 @@ instance [Monad m] [LawfulMonad m] : LawfulMonad (ReaderT ρ m) where
   pure_bind      := by intros; apply ext; intros; simp
   bind_assoc     := by intros; apply ext; intros; simp
 
-/-! Note that the `MonadControl` instance for `ReaderT` is not monad-generic. -/
-
-@[simp] theorem run_restoreM [Monad m] (x : stM m (ReaderT ρ m) α) (ctx : ρ) :
-    ReaderT.run (restoreM x) ctx = pure x := rfl
-
-@[simp] theorem run_liftWith [Monad m] (f : ({β : Type u} → ReaderT ρ m β → m (stM m (ReaderT ρ m) β)) → m α) (ctx : ρ) :
-    ReaderT.run (liftWith f) ctx = (f fun x => x.run ctx) :=
-  rfl
-
-@[simp] theorem run_controlAt [Monad m] [LawfulMonad m] (f : ({β : Type u} → ReaderT ρ m β → m (stM m (ReaderT ρ m) β)) → m (stM m (ReaderT ρ m) α)) (ctx : ρ) :
-    ReaderT.run (controlAt m f) ctx = f fun x => x.run ctx := by
-  simp [controlAt]
-
-@[simp] theorem run_control [Monad m] [LawfulMonad m] (f : ({β : Type u} → ReaderT ρ m β → m (stM m (ReaderT ρ m) β)) → m (stM m (ReaderT ρ m) α)) (ctx : ρ) :
-    ReaderT.run (control f) ctx = f fun x => x.run ctx := run_controlAt f ctx
-
 end ReaderT
 
 /-! # StateRefT -/
@@ -364,20 +179,17 @@ namespace StateT
 @[ext, grind ext] theorem ext {x y : StateT σ m α} (h : ∀ s, x.run s = y.run s) : x = y :=
   funext h
 
-@[simp, grind =] theorem run_mk [Monad m] (x : σ → m (α × σ)) (s : σ) : run (.mk x) s = x s :=
-  rfl
-
 @[simp, grind =] theorem run'_eq [Monad m] (x : StateT σ m α) (s : σ) : run' x s = (·.1) <$> run x s :=
   rfl
 
 @[simp, grind =] theorem run_pure [Monad m] (a : α) (s : σ) : (pure a : StateT σ m α).run s = pure (a, s) := rfl
 
 @[simp, grind =] theorem run_bind [Monad m] (x : StateT σ m α) (f : α → StateT σ m β) (s : σ)
-    : (x >>= f).run s = x.run s >>= λ p => (f p.1).run p.2 := rfl
+    : (x >>= f).run s = x.run s >>= λ p => (f p.1).run p.2 := by
+  simp [bind, StateT.bind, run]
 
 @[simp, grind =] theorem run_map {α β σ : Type u} [Monad m] [LawfulMonad m] (f : α → β) (x : StateT σ m α) (s : σ) : (f <$> x).run s = (fun (p : α × σ) => (f p.1, p.2)) <$> x.run s := by
-  rw [← bind_pure_comp (m := m)]
-  rfl
+  simp [Functor.map, StateT.map, run, ←bind_pure_comp]
 
 @[simp, grind =] theorem run_get [Monad m] (s : σ)    : (get : StateT σ m σ).run s = pure (s, s) := rfl
 
@@ -386,13 +198,13 @@ namespace StateT
 @[simp, grind =] theorem run_modify [Monad m] (f : σ → σ) (s : σ) : (modify f : StateT σ m PUnit).run s = pure (⟨⟩, f s) := rfl
 
 @[simp, grind =] theorem run_modifyGet [Monad m] (f : σ → α × σ) (s : σ) : (modifyGet f : StateT σ m α).run s = pure ((f s).1, (f s).2) := by
-  rfl
+  simp [modifyGet, MonadStateOf.modifyGet, StateT.modifyGet, run]
 
 @[simp, grind =] theorem run_lift {α σ : Type u} [Monad m] (x : m α) (s : σ) : (StateT.lift x : StateT σ m α).run s = x >>= fun a => pure (a, s) := rfl
 
 @[grind =]
 theorem run_bind_lift {α σ : Type u} [Monad m] [LawfulMonad m] (x : m α) (f : α → StateT σ m β) (s : σ) : (StateT.lift x >>= f).run s = x >>= fun a => (f a).run s := by
-  simp
+  simp [StateT.lift, StateT.run, bind, StateT.bind]
 
 @[simp, grind =] theorem run_monadLift {α σ : Type u} [Monad m] [MonadLiftT n m] (x : n α) (s : σ) : (monadLift x : StateT σ m α).run s = (monadLift x : m α) >>= fun a => pure (a, s) := rfl
 
@@ -432,78 +244,20 @@ instance [Monad m] [LawfulMonad m] : LawfulMonad (StateT σ m) where
   pure_bind      := by intros; apply ext; intros; simp
   bind_assoc     := by intros; apply ext; intros; simp
 
-/-! Note that the `MonadControl` instance for `StateT` is not monad-generic. -/
-
-@[simp] theorem run_restoreM [Monad m] [LawfulMonad m] (x : stM m (StateT σ m) α) (s : σ) :
-    StateT.run (restoreM x) s = pure x := by
-  simp [restoreM, MonadControl.restoreM]
-  rfl
-
-@[simp] theorem run_liftWith [Monad m] [LawfulMonad m] (f : ({β : Type u} → StateT σ m β → m (stM m (StateT σ m) β)) → m α) (s : σ) :
-    StateT.run (liftWith f) s = ((·, s) <$> f fun x => x.run s) := by
-  simp [liftWith, MonadControl.liftWith, Function.comp_def]
-
-@[simp] theorem run_controlAt [Monad m] [LawfulMonad m] (f : ({β : Type u} → StateT σ m β → m (stM m (StateT σ m) β)) → m (stM m (StateT σ m) α)) (s : σ) :
-    StateT.run (controlAt m f) s = f fun x => x.run s := by
-  simp [controlAt]
-
-@[simp] theorem run_control [Monad m] [LawfulMonad m] (f : ({β : Type u} → StateT σ m β → m (stM m (StateT σ m) β)) → m (stM m (StateT σ m) α)) (s : σ) :
-    StateT.run (control f) s = f fun x => x.run s := run_controlAt f s
-
 end StateT
 
 /-! # EStateM -/
 
-namespace EStateM
-
-@[simp, grind =] theorem run_pure (a : α) (s : σ) :
-    EStateM.run (pure a : EStateM ε σ α) s = .ok a s := rfl
-
-@[simp, grind =] theorem run_get (s : σ) :
-    EStateM.run (get : EStateM ε σ σ) s = .ok s s := rfl
-
-@[simp, grind =] theorem run_set (s₁ s₂ : σ) :
-    EStateM.run (set s₁ : EStateM ε σ PUnit) s₂ = .ok .unit s₁ := rfl
-
-@[simp, grind =] theorem run_modify (f : σ → σ) (s : σ) :
-    EStateM.run (modify f : EStateM ε σ PUnit) s = .ok .unit (f s) := rfl
-
-@[simp, grind =] theorem run_modifyGet (f : σ → α × σ) (s : σ) :
-    EStateM.run (modifyGet f : EStateM ε σ α) s = .ok (f s).1 (f s).2 := rfl
-
-@[simp, grind =] theorem run_throw (e : ε) (s : σ):
-    EStateM.run (throw e : EStateM ε σ PUnit) s = .error e s := rfl
-
-@[simp, grind =] theorem run_bind (x : EStateM ε σ α) (f : α → EStateM ε σ β)
-    : EStateM.run (x >>= f : EStateM ε σ β) s
-      =
-      match EStateM.run x s with
-      | .ok x s => EStateM.run (f x) s
-      | .error e s => .error e s :=
-  rfl
-
-@[simp, grind =]
-theorem run_adaptExcept (f : ε → ε') (x : EStateM ε σ α) (s : σ)
-    : EStateM.run (EStateM.adaptExcept f x : EStateM ε' σ α) s
-     =
-     match EStateM.run x s with
-     | .ok x s => .ok x s
-     | .error e s => .error (f e) s := by
-  simp only [EStateM.run, EStateM.adaptExcept]
-  cases (x s) <;> rfl
-
 instance : LawfulMonad (EStateM ε σ) := .mk'
   (id_map := fun x => funext <| fun s => by
-    simp only [Functor.map, EStateM.map]
+    dsimp only [EStateM.instMonad, EStateM.map]
     match x s with
     | .ok _ _ => rfl
     | .error _ _ => rfl)
   (pure_bind := fun _ _ => by rfl)
   (bind_assoc := fun x _ _ => funext <| fun s => by
-    simp only [bind, EStateM.bind]
+    dsimp only [EStateM.instMonad, EStateM.bind]
     match x s with
     | .ok _ _ => rfl
     | .error _ _ => rfl)
   (map_const := fun _ _ => rfl)
-
-end EStateM

src/Init/Control/Lawful/Lemmas.lean

@@ -7,9 +7,8 @@ module
 
 prelude
 public import Init.Control.Lawful.Basic
-public import Init.Classical
-public import Init.Ext
-import Init.ByCases
+public import Init.RCases
+public import Init.ByCases
 
 public section
 

src/Init/Control/Lawful/MonadAttach.lean

@@ -1,10 +0,0 @@
-/-
-Copyright (c) 2025 Lean FRO, LLC. All rights reserved.
-Released under Apache 2.0 license as described in the file LICENSE.
-Authors: Paul Reichert
--/
-module
-
-prelude
-public import Init.Control.Lawful.MonadAttach.Lemmas
-public import Init.Control.Lawful.MonadAttach.Instances

src/Init/Control/Lawful/MonadAttach/Instances.lean

@@ -1,88 +0,0 @@
-/-
-Copyright (c) 2025 Lean FRO, LLC. All rights reserved.
-Released under Apache 2.0 license as described in the file LICENSE.
-Authors: Paul Reichert
--/
-module
-
-prelude
-import Init.Control.Lawful.MonadAttach.Lemmas
-public import Init.Control.Lawful.Basic
-public import Init.Control.State
-public import Init.Control.StateRef
-public import Init.Ext
-
-public instance [Monad m] [LawfulMonad m] [MonadAttach m] [WeaklyLawfulMonadAttach m] :
-    WeaklyLawfulMonadAttach (ReaderT ρ m) where
-  map_attach := by
-    simp only [Functor.map, MonadAttach.attach, Functor.map_map, WeaklyLawfulMonadAttach.map_attach]
-    intros; rfl
-
-public instance [Monad m] [LawfulMonad m] [MonadAttach m] [LawfulMonadAttach m] :
-    LawfulMonadAttach (ReaderT ρ m) where
-  canReturn_map_imp := by
-    simp only [Functor.map, MonadAttach.CanReturn, ReaderT.run]
-    rintro _ _ x a ⟨r, h⟩
-    apply LawfulMonadAttach.canReturn_map_imp h
-
-public instance [Monad m] [LawfulMonad m] [MonadAttach m] [WeaklyLawfulMonadAttach m] :
-    WeaklyLawfulMonadAttach (StateT σ m) where
-  map_attach := by
-    intro α x
-    simp only [Functor.map, StateT, funext_iff, StateT.map, bind_pure_comp, MonadAttach.attach,
-      Functor.map_map]
-    exact fun s => WeaklyLawfulMonadAttach.map_attach
-
-public instance [Monad m] [LawfulMonad m] [MonadAttach m] [LawfulMonadAttach m] :
-    LawfulMonadAttach (StateT σ m) where
-  canReturn_map_imp := by
-    simp only [Functor.map, MonadAttach.CanReturn, StateT.run, StateT.map, bind_pure_comp]
-    rintro _ _ x a ⟨s, s', h⟩
-    obtain ⟨a, h, h'⟩ := LawfulMonadAttach.canReturn_map_imp' h
-    cases h'
-    exact a.1.2
-
-public instance [Monad m] [LawfulMonad m] [MonadAttach m] [WeaklyLawfulMonadAttach m] :
-    WeaklyLawfulMonadAttach (ExceptT ε m) where
-  map_attach {α} x := by
-    simp only [Functor.map, MonadAttach.attach, ExceptT.map]
-    simp
-    conv => rhs; rw [← WeaklyLawfulMonadAttach.map_attach (m := m) (x := x)]
-    simp only [map_eq_pure_bind]
-    apply bind_congr; intro a
-    match a with
-    | ⟨.ok _, _⟩ => simp
-    | ⟨.error _, _⟩ => simp
-
-public instance [Monad m] [LawfulMonad m] [MonadAttach m] [LawfulMonadAttach m] :
-    LawfulMonadAttach (ExceptT ε m) where
-  canReturn_map_imp {α P x a} := by
-    simp only [Functor.map, MonadAttach.CanReturn, ExceptT.map, ExceptT.mk]
-    let x' := (fun a => show Subtype (fun a : Except _ _ => match a with | .ok a => P a | .error e => True) from ⟨match a with | .ok a => .ok a.1 | .error e => .error e, by cases a <;> simp [Subtype.property]⟩) <$> show m _ from x
-    have := LawfulMonadAttach.canReturn_map_imp (m := m) (x := x') (a := .ok a)
-    simp only at this
-    intro h
-    apply this
-    simp only [x', map_eq_pure_bind, bind_assoc]
-    refine cast ?_ h
-    congr 1
-    apply bind_congr; intro a
-    split <;> simp
-
-public instance [Monad m] [MonadAttach m] [LawfulMonad m] [WeaklyLawfulMonadAttach m] :
-    WeaklyLawfulMonadAttach (StateRefT' ω σ m) :=
-  inferInstanceAs (WeaklyLawfulMonadAttach (ReaderT _ _))
-
-public instance [Monad m] [MonadAttach m] [LawfulMonad m] [LawfulMonadAttach m] :
-    LawfulMonadAttach (StateRefT' ω σ m) :=
-  inferInstanceAs (LawfulMonadAttach (ReaderT _ _))
-
-section
-
-attribute [local instance] MonadAttach.trivial
-
-public instance [Monad m] [LawfulMonad m] :
-    WeaklyLawfulMonadAttach m where
-  map_attach := by simp [MonadAttach.attach]
-
-end

src/Init/Control/Lawful/MonadAttach/Lemmas.lean

@@ -1,92 +0,0 @@
-/-
-Copyright (c) 2025 Lean FRO, LLC. All rights reserved.
-Released under Apache 2.0 license as described in the file LICENSE.
-Authors: Paul Reichert
--/
-module
-
-prelude
-import all Init.Control.MonadAttach
-public import Init.Classical
-public import Init.Control.Lawful.Basic
-public import Init.Control.Lawful.MonadLift.Basic
-import Init.Control.Lawful.MonadLift.Lemmas
-import Init.RCases
-
-public theorem LawfulMonadAttach.canReturn_bind_imp' [Monad m] [LawfulMonad m]
-    [MonadAttach m] [LawfulMonadAttach m]
-    {x : m α} {f : α → m β} :
-    MonadAttach.CanReturn (x >>= f) b → Exists fun a => MonadAttach.CanReturn x a ∧ MonadAttach.CanReturn (f a) b := by
-  intro h
-  let P (b : β) := Exists fun a => MonadAttach.CanReturn x a ∧ MonadAttach.CanReturn (f a) b
-  have h' : (x >>= f) = Subtype.val <$> (MonadAttach.attach x >>= (fun a => (do
-      let b ← MonadAttach.attach (f a)
-      return ⟨b.1, a.1, a.2, b.2⟩ : m (Subtype P)))) := by
-    simp only [map_bind, map_pure]
-    simp only [bind_pure_comp, WeaklyLawfulMonadAttach.map_attach]
-    rw (occs := [1]) [← WeaklyLawfulMonadAttach.map_attach (x := x)]
-    simp
-  rw [h'] at h
-  have := LawfulMonadAttach.canReturn_map_imp h
-  exact this
-
-public theorem LawfulMonadAttach.eq_of_canReturn_pure [Monad m] [MonadAttach m]
-    [LawfulMonad m] [LawfulMonadAttach m] {a b : α}
-    (h : MonadAttach.CanReturn (m := m) (pure a) b) :
-    a = b := by
-  let x : m (Subtype (a = ·)) := pure ⟨a, rfl⟩
-  have : pure a = Subtype.val <$> x := by simp [x]
-  rw [this] at h
-  exact LawfulMonadAttach.canReturn_map_imp h
-
-public theorem LawfulMonadAttach.canReturn_map_imp' [Monad m] [LawfulMonad m]
-    [MonadAttach m] [LawfulMonadAttach m]
-    {x : m α} {f : α → β} :
-    MonadAttach.CanReturn (f <$> x) b → Exists fun a => MonadAttach.CanReturn x a ∧ f a = b := by
-  rw [map_eq_pure_bind]
-  intro h
-  obtain ⟨a, h, h'⟩ := canReturn_bind_imp' h
-  exact ⟨a, h, eq_of_canReturn_pure h'⟩
-
-public theorem LawfulMonadAttach.canReturn_liftM_imp'
-    [Monad m] [MonadAttach m] [LawfulMonad m] [LawfulMonadAttach m]
-    [Monad n] [MonadAttach n] [LawfulMonad n] [LawfulMonadAttach n]
-    [MonadLiftT m n] [LawfulMonadLiftT m n] {x : m α} {a : α} :
-    MonadAttach.CanReturn (liftM (n := n) x) a → MonadAttach.CanReturn x a := by
-  intro h
-  simp only [← WeaklyLawfulMonadAttach.map_attach (x := x), liftM_map] at h
-  exact canReturn_map_imp h
-
-public theorem WeaklyLawfulMonadAttach.attach_bind_val
-    [Monad m] [MonadAttach m] [LawfulMonad m] [WeaklyLawfulMonadAttach m]
-    {x : m α} {f : α → m β} :
-    MonadAttach.attach x >>= (fun a => f a.val) = x >>= f := by
-  conv => rhs; simp only [← map_attach (x := x), bind_map_left]
-
-public theorem WeaklyLawfulMonadAttach.bind_attach_of_nonempty
-    [Monad m] [MonadAttach m] [LawfulMonad m] [WeaklyLawfulMonadAttach m] [Nonempty (m β)]
-    {x : m α} {f : Subtype (MonadAttach.CanReturn x) → m β} :
-    open scoped Classical in
-    MonadAttach.attach x >>= f = x >>= (fun a => if ha : MonadAttach.CanReturn x a then f ⟨a, ha⟩ else Classical.ofNonempty) := by
-  conv => rhs; simp +singlePass only [← map_attach (x := x)]
-  simp [Subtype.property]
-
-public theorem MonadAttach.attach_bind_eq_pbind
-    [Monad m] [MonadAttach m]
-    {x : m α} {f : Subtype (MonadAttach.CanReturn x) → m β} :
-    MonadAttach.attach x >>= f = MonadAttach.pbind x (fun a ha => f ⟨a, ha⟩) := by
-  simp [MonadAttach.pbind]
-
-public theorem WeaklyLawfulMonadAttach.pbind_eq_bind
-    [Monad m] [MonadAttach m] [LawfulMonad m] [WeaklyLawfulMonadAttach m]
-    {x : m α} {f : α → m β} :
-    MonadAttach.pbind x (fun a _ => f a) = x >>= f := by
-  conv => rhs; rw [← map_attach (x := x)]
-  simp [MonadAttach.pbind]
-
-public theorem WeaklyLawfulMonadAttach.pbind_eq_bind'
-    [Monad m] [MonadAttach m] [LawfulMonad m] [WeaklyLawfulMonadAttach m]
-    {x : m α} {f : α → m β} :
-    MonadAttach.pbind x (fun a _ => f a) = x >>= f := by
-  conv => rhs; rw [← map_attach (x := x)]
-  simp [MonadAttach.pbind]

src/Init/Control/Lawful/MonadLift.lean

@@ -9,3 +9,5 @@ prelude
 public import Init.Control.Lawful.MonadLift.Basic
 public import Init.Control.Lawful.MonadLift.Lemmas
 public import Init.Control.Lawful.MonadLift.Instances
+
+public section

src/Init/Control/Lawful/MonadLift/Basic.lean

@@ -6,7 +6,7 @@ Authors: Quang Dao
 module
 
 prelude
-public import Init.Notation
+public import Init.Control.Basic
 
 public section
 

src/Init/Control/Lawful/MonadLift/Instances.lean

@@ -6,20 +6,20 @@ Authors: Quang Dao, Paul Reichert
 module
 
 prelude
+public import Init.Control.Option
 import all Init.Control.Option
+public import Init.Control.Except
 import all Init.Control.Except
 public import Init.Control.ExceptCps
 import all Init.Control.ExceptCps
+public import Init.Control.StateRef
 import all Init.Control.StateRef
 public import Init.Control.StateCps
 import all Init.Control.StateCps
+public import Init.Control.Id
 import all Init.Control.Id
-public import Init.Control.Lawful.MonadLift.Basic
-public import Init.Control.Option
-public import Init.Control.State
-public import Init.Control.StateRef
-import Init.Control.Lawful.Instances
-import Init.Control.Lawful.MonadLift.Lemmas
+public import Init.Control.Lawful.MonadLift.Lemmas
+public import Init.Control.Lawful.Instances
 
 public section
 
@@ -64,10 +64,14 @@ namespace OptionT
 
 variable [Monad m] [LawfulMonad m]
 
+@[simp]
+theorem lift_pure {α : Type u} (a : α) : OptionT.lift (pure a : m α) = pure a := by
+  simp only [OptionT.lift, OptionT.mk, bind_pure_comp, map_pure, pure, OptionT.pure]
+
 @[simp]
 theorem lift_bind {α β : Type u} (ma : m α) (f : α → m β) :
     OptionT.lift (ma >>= f) = OptionT.lift ma >>= (fun a => OptionT.lift (f a)) := by
-  simp only [bind, OptionT.bind, OptionT.mk, OptionT.lift, bind_pure_comp, bind_map_left,
+  simp only [instMonad, OptionT.bind, OptionT.mk, OptionT.lift, bind_pure_comp, bind_map_left,
     map_bind]
 
 instance : LawfulMonadLift m (OptionT m) where
@@ -83,7 +87,7 @@ variable [Monad m] [LawfulMonad m]
 @[simp]
 theorem lift_bind {α β ε : Type u} (ma : m α) (f : α → m β) :
     ExceptT.lift (ε := ε) (ma >>= f) = ExceptT.lift ma >>= (fun a => ExceptT.lift (f a)) := by
-  simp only [bind, ExceptT.bind, mk, ExceptT.lift, bind_map_left, ExceptT.bindCont, map_bind]
+  simp only [instMonad, ExceptT.bind, mk, ExceptT.lift, bind_map_left, ExceptT.bindCont, map_bind]
 
 instance : LawfulMonadLift m (ExceptT ε m) where
   monadLift_pure := lift_pure
@@ -93,7 +97,8 @@ instance : LawfulMonadLift (Except ε) (ExceptT ε m) where
   monadLift_pure _ := by
     simp only [MonadLift.monadLift, mk, pure, Except.pure, ExceptT.pure]
   monadLift_bind ma _ := by
-    simp only [bind, ExceptT.bind, mk, MonadLift.monadLift, pure_bind, ExceptT.bindCont, Except.bind]
+    simp only [instMonad, ExceptT.bind, mk, MonadLift.monadLift, pure_bind, ExceptT.bindCont,
+      Except.instMonad, Except.bind]
     rcases ma with _ | _ <;> simp
 
 end ExceptT

src/Init/Control/Lawful/MonadLift/Lemmas.lean

@@ -8,20 +8,11 @@ module
 prelude
 public import Init.Control.Lawful.Basic
 public import Init.Control.Lawful.MonadLift.Basic
-import Init.Ext
 
 public section
 
 universe u v w
 
-theorem instMonadLiftTOfMonadLift_instMonadLiftTOfPure [Monad m] [Monad n] {_ : MonadLift m n}
-    [LawfulMonadLift m n] : instMonadLiftTOfMonadLift Id m n = Id.instMonadLiftTOfPure := by
-  have hext {a b : MonadLiftT Id n} (h : @a.monadLift = @b.monadLift) : a = b := by
-    cases a; cases b; simp [monadLift] at h; simp [h]
-  apply hext
-  ext α x
-  simp [monadLift, LawfulMonadLift.monadLift_pure]
-
 variable {m : Type u → Type v} {n : Type u → Type w} [Monad m] [Monad n] [MonadLiftT m n]
   [LawfulMonadLiftT m n] {α β : Type u}
 
@@ -32,7 +23,7 @@ theorem monadLift_map [LawfulMonad m] [LawfulMonad n] (f : α → β) (ma : m α
 
 theorem monadLift_seq [LawfulMonad m] [LawfulMonad n] (mf : m (α → β)) (ma : m α) :
     monadLift (mf <*> ma) = monadLift mf <*> (monadLift ma : n α) := by
-  simp only [seq_eq_bind_map, monadLift_map, monadLift_bind]
+  simp only [seq_eq_bind, monadLift_map, monadLift_bind]
 
 theorem monadLift_seqLeft [LawfulMonad m] [LawfulMonad n] (x : m α) (y : m β) :
     monadLift (x <* y) = (monadLift x : n α) <* (monadLift y : n β) := by

src/Init/Control/MonadAttach.lean

@@ -1,126 +0,0 @@
-/-
-Copyright (c) 2025 Lean FRO, LLC. All rights reserved.
-Released under Apache 2.0 license as described in the file LICENSE.
-Authors: Paul Reichert
--/
-module
-
-prelude
-public import Init.Core
-
-set_option linter.all true
-
-set_option doc.verso true
-
-/-!
-# {name (scope := "Init.Control.MonadAttach")}`MonadAttach`
-
-This module provides a mechanism for attaching proofs to the return values of monadic computations,
-producing a new monadic computation returning a {name}`Subtype`.
-
-This function is primarily used to allow definitions by [well-founded
-recursion](lean-manual://section/well-founded-recursion) that sequence computations using
-{name}`Bind.bind` (`>>=`) to prove properties about the return values of prior computations when
-a recursive call happens.
-This allows the well-founded recursion mechanism to prove that the function terminates.
--/
-
--- verso docstring is added below
-set_option linter.missingDocs false in
-public class MonadAttach (m : Type u → Type v) where
-  /--
-  A predicate that can be assumed to be true for all return values {name}`a` of actions {name}`x`
-  in {name}`m`, in all situations.
-  -/
-  CanReturn {α : Type u} : (x : m α) → (a : α) → Prop
-  /--
-  Attaches a proof of {name}`MonadAttach.CanReturn` to the return value of {name}`x`. This proof
-  can be used to prove the termination of well-founded recursive functions.
-  -/
-  attach {α : Type u} (x : m α) : m (Subtype (CanReturn x))
-
--- verso docstring is added below
-set_option linter.missingDocs false in
-public class WeaklyLawfulMonadAttach (m : Type u → Type v) [Monad m] [MonadAttach m] where
-  map_attach {α : Type u} {x : m α} : Subtype.val <$> MonadAttach.attach x = x
-
-/--
-This type class ensures that {name}`MonadAttach.CanReturn` is the unique strongest possible
-postcondition.
--/
-public class LawfulMonadAttach (m : Type u → Type v) [Monad m] [MonadAttach m] extends
-    WeaklyLawfulMonadAttach m where
-  canReturn_map_imp {α : Type u} {P : α → Prop} {x : m (Subtype P)} {a : α} :
-      MonadAttach.CanReturn (Subtype.val <$> x) a → P a
-
-/--
-Like {name}`Bind.bind`, {name}`pbind` sequences two computations {lean}`x : m α` and {lean}`f`,
-allowing the second to depend on the value computed by the first.
-But other than with {name}`Bind.bind`, the second computation can also depend on a proof that
-the return value {given}`a` of {name}`x` satisfies {lean}`MonadAttach.CanReturn x a`.
--/
-public def MonadAttach.pbind [Monad m] [MonadAttach m]
-    (x : m α) (f : (a : α) → MonadAttach.CanReturn x a → m β) : m β :=
-  MonadAttach.attach x >>= (fun ⟨a, ha⟩ => f a ha)
-
-/--
-A {lean}`MonadAttach` instance where all return values are possible and {name}`attach` adds no
-information to the return value, except a trivial proof of {name}`True`.
-
-This instance is used whenever no more useful {name}`MonadAttach` instance can be implemented.
-It always has a {name}`WeaklyLawfulMonadAttach`, but usually no {name}`LawfulMonadAttach` instance.
--/
-@[expose, instance_reducible]
-public protected def MonadAttach.trivial {m : Type u → Type v} [Monad m] : MonadAttach m where
-  CanReturn _ _ := True
-  attach x := (⟨·, .intro⟩) <$> x
-
-section
-
-variable (α : Type u) [∀ m, Monad m] [∀ m, MonadAttach m]
-
-set_option doc.verso true
-
-/--
-For every {given}`x : m α`, this type class provides a predicate {lean}`MonadAttach.CanReturn x`
-and a way to attach a proof of this predicate to the return values of {name}`x` by providing
-an element {lean}`MonadAttach.attach x` of {lean}`m { a : α // MonadAttach.CanReturn x a }`.
-
-Instances should abide the law {lean}`Subtype.val <$> MonadAttach.attach x = x`, which is encoded by
-the {name}`WeaklyLawfulMonadAttach` type class. The stronger type class {name}`LawfulMonadAttach`
-ensures that {lean}`MonadAttach.CanReturn x` is the _unique_ strongest possible predicate.
-
-Similarly to {name (scope := "Init.Data.List.Attach")}`List.attach`, the purpose of
-{name}`MonadAttach` is to attach proof terms necessary for well-founded termination proofs.
-The iterator library relies on {name}`MonadAttach` for combinators such as
-{name (scope := "Init.Data.Iterators")}`Std.Iter.filterM` in order to automatically attach
-information about the monadic predicate's behavior that could be relevant for the termination
-behavior of the iterator.
-
-*Limitations*:
-
-For many monads, there is a strongly lawful {lean}`MonadAttach` instance, but there are exceptions.
-For example, there is no way to provide a computable {lean}`MonadAttach` instance for the CPS monad
-transformers
-{name (scope := "Init.Control.StateCps")}`StateCpsT` and
-{name (scope := "Init.Control.StateCps")}`ExceptCpsT` with a predicate that is not always
-{name}`True`. Therefore, such CPS monads only provide the trivial {lean}`MonadAttach` instance
-{lean}`MonadAttach.trivial` together with {name}`WeaklyLawfulMonadAttach`, but without
-{name}`LawfulMonadAttach`.
-
-For most monads with side effects, {lean}`MonadAttach` is too weak to fully capture the behavior of
-computations because the postcondition represented by {name}`MonadAttach.CanReturn` neither depends
-on the prior internal state of the monad, nor does it contain information about how the state of the
-monad changes with the computation.
--/
-add_decl_doc MonadAttach
-
-/--
-This type class ensures that every monadic action {given}`x : m α` can be recovered by stripping the
-proof component from the subtypes returned by
-{lean}`(MonadAttach.attach x) : m { a : α // MonadAttach.CanReturn x a }` . In other words,
-the type class ensures that {lean}`Subtype.val <$> MonadAttach.attach x = x`.
--/
-add_decl_doc WeaklyLawfulMonadAttach
-
-end

src/Init/Control/Option.lean

@@ -7,7 +7,8 @@ module
 
 prelude
 public import Init.Data.Option.Basic
-public import Init.Control.MonadAttach
+public import Init.Control.Basic
+public import Init.Control.Except
 
 public section
 
@@ -27,7 +28,7 @@ failure occurred.
 /--
 Executes an action that might fail in the underlying monad `m`, returning `none` in case of failure.
 -/
-@[always_inline, inline, expose]
+@[always_inline, inline]
 def OptionT.run {m : Type u → Type v} {α : Type u} (x : OptionT m α) : m (Option α) :=
   x
 
@@ -38,14 +39,13 @@ variable {m : Type u → Type v} [Monad m] {α β : Type u}
 Converts an action that returns an `Option` into one that might fail, with `none` indicating
 failure.
 -/
-@[always_inline, inline, expose]
 protected def mk (x : m (Option α)) : OptionT m α :=
   x
 
 /--
 Sequences two potentially-failing actions. The second action is run only if the first succeeds.
 -/
-@[always_inline, inline, expose]
+@[always_inline, inline]
 protected def bind (x : OptionT m α) (f : α → OptionT m β) : OptionT m β := OptionT.mk do
   match (← x) with
   | some a => f a
@@ -54,7 +54,7 @@ protected def bind (x : OptionT m α) (f : α → OptionT m β) : OptionT m β :
 /--
 Succeeds with the provided value.
 -/
-@[always_inline, inline, expose]
+@[always_inline, inline]
 protected def pure (a : α) : OptionT m α := OptionT.mk do
   pure (some a)
 
@@ -69,7 +69,7 @@ instance {m : Type u → Type v} [Pure m] : Inhabited (OptionT m α) where
 /--
 Recovers from failures. Typically used via the `<|>` operator.
 -/
-@[always_inline, inline, expose] protected def orElse (x : OptionT m α) (y : Unit → OptionT m α) : OptionT m α := OptionT.mk do
+@[always_inline, inline] protected def orElse (x : OptionT m α) (y : Unit → OptionT m α) : OptionT m α := OptionT.mk do
   match (← x) with
   | some a => pure (some a)
   | _      => y ()
@@ -77,7 +77,7 @@ Recovers from failures. Typically used via the `<|>` operator.
 /--
 A recoverable failure.
 -/
-@[always_inline, inline, expose] protected def fail : OptionT m α := OptionT.mk do
+@[always_inline, inline] protected def fail : OptionT m α := OptionT.mk do
   pure none
 
 instance : Alternative (OptionT m) where
@@ -90,7 +90,7 @@ Converts a computation from the underlying monad into one that could fail, even
 This function is typically implicitly accessed via a `MonadLiftT` instance as part of [automatic
 lifting](lean-manual://section/monad-lifting).
 -/
-@[always_inline, inline, expose] protected def lift (x : m α) : OptionT m α := OptionT.mk do
+@[always_inline, inline] protected def lift (x : m α) : OptionT m α := OptionT.mk do
   return some (← x)
 
 instance : MonadLift m (OptionT m) := ⟨OptionT.lift⟩
@@ -100,11 +100,11 @@ instance : MonadFunctor m (OptionT m) := ⟨fun f x => f x⟩
 /--
 Handles failures by treating them as exceptions of type `Unit`.
 -/
-@[always_inline, inline, expose] protected def tryCatch (x : OptionT m α) (handle : PUnit → OptionT m α) : OptionT m α := OptionT.mk do
-  let some a ← x | handle ⟨⟩
+@[always_inline, inline] protected def tryCatch (x : OptionT m α) (handle : Unit → OptionT m α) : OptionT m α := OptionT.mk do
+  let some a ← x | handle ()
   pure <| some a
 
-instance : MonadExceptOf PUnit (OptionT m) where
+instance : MonadExceptOf Unit (OptionT m) where
   throw    := fun _ => OptionT.fail
   tryCatch := OptionT.tryCatch
 
@@ -112,12 +112,6 @@ instance (ε : Type u) [MonadExceptOf ε m] : MonadExceptOf ε (OptionT m) where
   throw e           := OptionT.mk <| throwThe ε e
   tryCatch x handle := OptionT.mk <| tryCatchThe ε x handle
 
-instance [MonadAttach m] : MonadAttach (OptionT m) where
-  CanReturn x a := MonadAttach.CanReturn x.run (some a)
-  attach x := .mk ((fun
-      | ⟨some a, h⟩ => some ⟨a, h⟩
-      | ⟨none, _⟩ => none) <$> MonadAttach.attach x.run)
-
 end OptionT
 
 instance [Monad m] : MonadControl m (OptionT m) where

src/Init/Control/Reader.lean

@@ -8,6 +8,8 @@ The Reader monad transformer for passing immutable State.
 module
 
 prelude
+public import Init.Control.Basic
+public import Init.Control.Id
 public import Init.Control.Except
 
 public section
@@ -51,7 +53,3 @@ A monad with access to a read-only value of type `ρ`. The value can be locally
 `withReader`, but it cannot be mutated.
 -/
 abbrev ReaderM (ρ : Type u) := ReaderT ρ Id
-
-instance [Monad m] [MonadAttach m] : MonadAttach (ReaderT ρ m) where
-  CanReturn x a := Exists (fun r => MonadAttach.CanReturn (x.run r) a)
-  attach x := fun r => (fun ⟨a, h⟩ => ⟨a, r, h⟩) <$> MonadAttach.attach (x.run r)

src/Init/Control/State.lean

@@ -8,6 +8,8 @@ The State monad transformer.
 module
 
 prelude
+public import Init.Control.Basic
+public import Init.Control.Id
 public import Init.Control.Except
 
 public section
@@ -25,12 +27,6 @@ of a value and a state.
 @[expose] def StateT (σ : Type u) (m : Type u → Type v) (α : Type u) : Type (max u v) :=
   σ → m (α × σ)
 
-/--
-Interpret `σ → m (α × σ)` as an element of `StateT σ m α`.
--/
-@[always_inline, inline, expose]
-def StateT.mk {σ : Type u} {m : Type u → Type v} {α : Type u} (x : σ → m (α × σ)) : StateT σ m α := x
-
 /--
 Executes an action from a monad with added state in the underlying monad `m`. Given an initial
 state, it returns a value paired with the final state.
@@ -204,7 +200,3 @@ instance StateT.tryFinally {m : Type u → Type v} {σ : Type u} [MonadFinally m
       | some (a, s') => h (some a) s'
       | none         => h none s
     pure ((a, b), s'')
-
-instance [Monad m] [MonadAttach m] : MonadAttach (StateT σ m) where
-  CanReturn x a := Exists fun s => Exists fun s' => MonadAttach.CanReturn (x.run s) (a, s')
-  attach x := fun s => (fun ⟨⟨a, s'⟩, h⟩ => ⟨⟨a, s, s', h⟩, s'⟩) <$> MonadAttach.attach (x.run s)

src/Init/Control/StateCps.lean

@@ -7,7 +7,6 @@ module
 
 prelude
 public import Init.Control.Lawful.Basic
-public import Init.Ext
 
 public section
 
@@ -69,13 +68,6 @@ instance : MonadStateOf σ (StateCpsT σ m) where
   set s := fun _ _ k => k ⟨⟩ s
   modifyGet f := fun _ s k => let (a, s) := f s; k a s
 
-/--
-For continuation monads, it is not possible to provide a computable `MonadAttach` instance that
-actually adds information about the return value. Therefore, this instance always attaches a proof
-of `True`.
--/
-instance : MonadAttach (StateCpsT ε m) := .trivial
-
 /--
 Runs an action from the underlying monad in the monad with state. The state is not modified.
 

src/Init/Control/StateRef.lean

@@ -9,7 +9,6 @@ module
 
 prelude
 public import Init.System.ST
-public import Init.Control.Reader
 
 public section
 
@@ -65,7 +64,6 @@ instance [Monad m] : Monad (StateRefT' ω σ m) := inferInstanceAs (Monad (Reade
 instance : MonadLift m (StateRefT' ω σ m) := ⟨StateRefT'.lift⟩
 instance (σ m) : MonadFunctor m (StateRefT' ω σ m) := inferInstanceAs (MonadFunctor m (ReaderT _ _))
 instance [Alternative m] [Monad m] : Alternative (StateRefT' ω σ m) := inferInstanceAs (Alternative (ReaderT _ _))
-instance [Monad m] [MonadAttach m] : MonadAttach (StateRefT' ω σ m) := inferInstanceAs (MonadAttach (ReaderT _ _))
 
 /--
 Retrieves the current value of the monad's mutable state.

src/Init/Conv.lean

@@ -51,21 +51,6 @@ scoped syntax (name := withAnnotateState)
 /-- `skip` does nothing. -/
 syntax (name := skip) "skip" : conv
 
-/--
-`cbv` performs simplification that closely mimics call-by-value evaluation.
-It reduces the target term by unfolding definitions using their defining equations and
-applying matcher equations. The unfolding is propositional, so `cbv` also works
-with functions defined via well-founded recursion or partial fixpoints.
-
-The proofs produced by `cbv` only use the three standard axioms.
-In particular, they do not require trust in the correctness of the code
-generator.
-
-This tactic is experimental and its behavior is likely to change in upcoming
-releases of Lean.
--/
-syntax (name := cbv) "cbv" : conv
-
 /--
 Traverses into the left subterm of a binary operator.
 

src/Init/Core.lean

@@ -8,16 +8,12 @@ notation, basic datatypes and type classes
 module
 
 prelude
+public meta import Init.Prelude
 public import Init.SizeOf
-public import Init.Tactics
 
 public section
 set_option linter.missingDocs true -- keep it documented
 
--- BEq instance for Option defined here so it's available early in the import chain
--- (before Init.Grind.Config and Init.MetaTypes which need BEq (Option Nat))
-deriving instance BEq for Option
-
 @[expose] section
 
 universe u v w
@@ -148,9 +144,8 @@ Computed values are cached, so the value is not recomputed.
   x.fn ()
 
 -- Ensure `Thunk.fn` is still computable even if it shouldn't be accessed directly.
-/-- Implementation detail. -/
-@[inline] def Thunk.fnImpl (x : Thunk α) : Unit → α := fun _ => x.get
-@[csimp] theorem Thunk.fn_eq_fnImpl : @Thunk.fn = @Thunk.fnImpl := rfl
+@[inline] private def Thunk.fnImpl (x : Thunk α) : Unit → α := fun _ => x.get
+@[csimp] private theorem Thunk.fn_eq_fnImpl : @Thunk.fn = @Thunk.fnImpl := rfl
 
 /--
 Constructs a new thunk that forces `x` and then applies `x` to the result. Upon forcing, the result
@@ -206,7 +201,6 @@ An element of `α ⊕ β` is either an `a : α` wrapped in `Sum.inl` or a `b :
 indication of which of the two types was chosen. The union of a singleton set with itself contains
 one element, while `Unit ⊕ Unit` contains distinct values `inl ()` and `inr ()`.
 -/
-@[suggest_for Either]
 inductive Sum (α : Type u) (β : Type v) where
   /-- Left injection into the sum type `α ⊕ β`. -/
   | inl (val : α) : Sum α β
@@ -342,7 +336,7 @@ inductive Exists {α : Sort u} (p : α → Prop) : Prop where
 An indication of whether a loop's body terminated early that's used to compile the `for x in xs`
 notation.
 
-A collection's `ForIn` or `ForIn'` instance describes how to iterate over its elements. The monadic
+A collection's `ForIn` or `ForIn'` instance describe's how to iterate over its elements. The monadic
 action that represents the body of the loop returns a `ForInStep α`, where `α` is the local state
 used to implement features such as `let mut`.
 -/
@@ -383,7 +377,7 @@ class ForIn (m : Type u₁ → Type u₂) (ρ : Type u) (α : outParam (Type v))
   More information about the translation of `for` loops into `ForIn.forIn` is available in [the Lean
   reference manual](lean-manual://section/monad-iteration-syntax).
   -/
-  forIn {β} (xs : ρ) (b : β) (f : α → β → m (ForInStep β)) : m β
+  forIn {β} [Monad m] (xs : ρ) (b : β) (f : α → β → m (ForInStep β)) : m β
 
 export ForIn (forIn)
 
@@ -411,7 +405,7 @@ class ForIn' (m : Type u₁ → Type u₂) (ρ : Type u) (α : outParam (Type v)
   More information about the translation of `for` loops into `ForIn'.forIn'` is available in [the
   Lean reference manual](lean-manual://section/monad-iteration-syntax).
   -/
-  forIn' {β} (x : ρ) (b : β) (f : (a : α) → a ∈ x → β → m (ForInStep β)) : m β
+  forIn' {β} [Monad m] (x : ρ) (b : β) (f : (a : α) → a ∈ x → β → m (ForInStep β)) : m β
 
 export ForIn' (forIn')
 
@@ -489,8 +483,6 @@ class HasEquiv (α : Sort u) where
   the notion of equivalence is type-dependent. -/
   Equiv : α → α → Sort v
 
-attribute [reducible] HasEquiv.Equiv
-
 @[inherit_doc] infix:50 " ≈ "  => HasEquiv.Equiv
 
 recommended_spelling "equiv" for "≈" in [HasEquiv.Equiv, «term_≈_»]
@@ -517,12 +509,12 @@ abbrev SSuperset [HasSSubset α] (a b : α) := SSubset b a
 
 /-- Notation type class for the union operation `∪`. -/
 class Union (α : Type u) where
-  /-- `a ∪ b` is the union of `a` and `b`. -/
+  /-- `a ∪ b` is the union of`a` and `b`. -/
   union : α → α → α
 
 /-- Notation type class for the intersection operation `∩`. -/
 class Inter (α : Type u) where
-  /-- `a ∩ b` is the intersection of `a` and `b`. -/
+  /-- `a ∩ b` is the intersection of`a` and `b`. -/
   inter : α → α → α
 
 /-- Notation type class for the set difference `\`. -/
@@ -545,10 +537,10 @@ infix:50 " ⊇ " => Superset
 /-- Strict superset relation: `a ⊃ b`  -/
 infix:50 " ⊃ " => SSuperset
 
-/-- `a ∪ b` is the union of `a` and `b`. -/
+/-- `a ∪ b` is the union of`a` and `b`. -/
 infixl:65 " ∪ " => Union.union
 
-/-- `a ∩ b` is the intersection of `a` and `b`. -/
+/-- `a ∩ b` is the intersection of`a` and `b`. -/
 infixl:70 " ∩ " => Inter.inter
 
 /--
@@ -608,6 +600,17 @@ export LawfulSingleton (insert_empty_eq)
 
 attribute [simp] insert_empty_eq
 
+@[deprecated insert_empty_eq (since := "2025-03-12")]
+theorem insert_emptyc_eq [EmptyCollection β] [Insert α β] [Singleton α β]
+    [LawfulSingleton α β] (x : α) : (insert x ∅ : β) = singleton x :=
+  insert_empty_eq _
+
+@[deprecated insert_empty_eq (since := "2025-03-12")]
+theorem LawfulSingleton.insert_emptyc_eq [EmptyCollection β] [Insert α β] [Singleton α β]
+    [LawfulSingleton α β] (x : α) : (insert x ∅ : β) = singleton x :=
+  insert_empty_eq _
+
+
 /-- Type class used to implement the notation `{ a ∈ c | p a }` -/
 class Sep (α : outParam <| Type u) (γ : Type v) where
   /-- Computes `{ a ∈ c | p a }`. -/
@@ -935,14 +938,6 @@ noncomputable def HEq.ndrec.{u1, u2} {α : Sort u2} {a : α} {motive : {β : Sor
 noncomputable def HEq.ndrecOn.{u1, u2} {α : Sort u2} {a : α} {motive : {β : Sort u2} → β → Sort u1} {β : Sort u2} {b : β} (h : a ≍ b) (m : motive a) : motive b :=
   h.rec m
 
-/-- `HEq.ndrec` specialized to homogeneous heterogeneous equality -/
-noncomputable def HEq.homo_ndrec.{u1, u2} {α : Sort u2} {a : α} {motive : α → Sort u1} (m : motive a) {b : α} (h : a ≍ b) : motive b :=
-  (eq_of_heq h).ndrec m
-
-/-- `HEq.ndrec` specialized to homogeneous heterogeneous equality, symmetric variant -/
-noncomputable def HEq.homo_ndrec_symm.{u1, u2} {α : Sort u2} {a : α} {motive : α → Sort u1} (m : motive a) {b : α} (h : b ≍ a) : motive b :=
-  (eq_of_heq h).ndrec_symm m
-
 /-- `HEq.ndrec` variant -/
 noncomputable def HEq.elim {α : Sort u} {a : α} {p : α → Sort v} {b : α} (h₁ : a ≍ b) (h₂ : p a) : p b :=
   eq_of_heq h₁ ▸ h₂
@@ -955,7 +950,9 @@ theorem HEq.subst {p : (T : Sort u) → T → Prop} (h₁ : a ≍ b) (h₂ : p
 @[symm] theorem HEq.symm (h : a ≍ b) : b ≍ a :=
   h.rec (HEq.refl a)
 
-
+/-- Propositionally equal terms are also heterogeneously equal. -/
+theorem heq_of_eq (h : a = a') : a ≍ a' :=
+  Eq.subst h (HEq.refl a)
 
 /-- Heterogeneous equality is transitive. -/
 theorem HEq.trans (h₁ : a ≍ b) (h₂ : b ≍ c) : a ≍ c :=
@@ -1098,6 +1095,14 @@ theorem of_toBoolUsing_eq_true {p : Prop} {d : Decidable p} (h : toBoolUsing d =
 theorem of_toBoolUsing_eq_false {p : Prop} {d : Decidable p} (h : toBoolUsing d = false) : ¬p :=
   of_decide_eq_false h
 
+set_option linter.missingDocs false in
+@[deprecated of_toBoolUsing_eq_true (since := "2025-04-04")]
+abbrev ofBoolUsing_eq_true := @of_toBoolUsing_eq_true
+
+set_option linter.missingDocs false in
+@[deprecated of_toBoolUsing_eq_false (since := "2025-04-04")]
+abbrev ofBoolUsing_eq_false := @of_toBoolUsing_eq_false
+
 instance : Decidable True :=
   isTrue trivial
 
@@ -1160,7 +1165,6 @@ end
     else isFalse (fun h => absurd (h hp) hq)
   else isTrue (fun h => absurd h hp)
 
-@[inline]
 instance {p q} [Decidable p] [Decidable q] : Decidable (p ↔ q) :=
   if hp : p then
     if hq : q then
@@ -1202,13 +1206,17 @@ theorem dif_neg {c : Prop} {h : Decidable c} (hnc : ¬c) {α : Sort u} {t : c
   | isTrue hc   => absurd hc hnc
   | isFalse _   => rfl
 
-@[macro_inline]
+-- Remark: dite and ite are "defally equal" when we ignore the proofs.
+theorem dif_eq_if (c : Prop) {h : Decidable c} {α : Sort u} (t : α) (e : α) : dite c (fun _ => t) (fun _ => e) = ite c t e :=
+  match h with
+  | isTrue _    => rfl
+  | isFalse _   => rfl
+
 instance {c t e : Prop} [dC : Decidable c] [dT : Decidable t] [dE : Decidable e] : Decidable (if c then t else e) :=
   match dC with
   | isTrue _   => dT
   | isFalse _  => dE
 
-@[inline]
 instance {c : Prop} {t : c → Prop} {e : ¬c → Prop} [dC : Decidable c] [dT : ∀ h, Decidable (t h)] [dE : ∀ h, Decidable (e h)] : Decidable (if h : c then t h else e h) :=
   match dC with
   | isTrue hc  => dT hc
@@ -1359,12 +1367,12 @@ namespace Subtype
 theorem exists_of_subtype {α : Type u} {p : α → Prop} : { x // p x } → Exists (fun x => p x)
   | ⟨a, h⟩ => ⟨a, h⟩
 
-variable {α : Sort u} {p : α → Prop}
+set_option linter.missingDocs false in
+@[deprecated exists_of_subtype (since := "2025-04-04")]
+abbrev existsOfSubtype := @exists_of_subtype
 
-protected theorem ext : ∀ {a1 a2 : {x // p x}}, val a1 = val a2 → a1 = a2
-  | ⟨_, _⟩, ⟨_, _⟩, rfl => rfl
+variable {α : Type u} {p : α → Prop}
 
-@[deprecated Subtype.ext (since := "2025-10-26")]
 protected theorem eq : ∀ {a1 a2 : {x // p x}}, val a1 = val a2 → a1 = a2
   | ⟨_, _⟩, ⟨_, _⟩, rfl => rfl
 
@@ -1379,12 +1387,12 @@ instance {α : Type u} {p : α → Prop} [BEq α] [ReflBEq α] : ReflBEq {x : α
   rfl {x} := BEq.refl x.1
 
 instance {α : Type u} {p : α → Prop} [BEq α] [LawfulBEq α] : LawfulBEq {x : α // p x} where
-  eq_of_beq h := Subtype.ext (eq_of_beq h)
+  eq_of_beq h := Subtype.eq (eq_of_beq h)
 
-instance {α : Sort u} {p : α → Prop} [DecidableEq α] : DecidableEq {x : α // p x} :=
+instance {α : Type u} {p : α → Prop} [DecidableEq α] : DecidableEq {x : α // p x} :=
   fun ⟨a, h₁⟩ ⟨b, h₂⟩ =>
     if h : a = b then isTrue (by subst h; exact rfl)
-    else isFalse (fun h' => Subtype.noConfusion rfl .rfl (heq_of_eq h') (fun h' => absurd (eq_of_heq h') h))
+    else isFalse (fun h' => Subtype.noConfusion h' (fun h' => absurd h' h))
 
 end Subtype
 
@@ -1443,8 +1451,8 @@ instance [DecidableEq α] [DecidableEq β] : DecidableEq (α × β) :=
     | isTrue e₁ =>
       match decEq b b' with
       | isTrue e₂  => isTrue (e₁ ▸ e₂ ▸ rfl)
-      | isFalse n₂ => isFalse fun h => Prod.noConfusion rfl rfl (heq_of_eq h) fun _   e₂' => absurd (eq_of_heq e₂') n₂
-    | isFalse n₁ => isFalse fun h => Prod.noConfusion rfl rfl (heq_of_eq h) fun e₁' _   => absurd (eq_of_heq e₁') n₁
+      | isFalse n₂ => isFalse fun h => Prod.noConfusion h fun _   e₂' => absurd e₂' n₂
+    | isFalse n₁ => isFalse fun h => Prod.noConfusion h fun e₁' _   => absurd e₁' n₁
 
 instance [BEq α] [BEq β] : BEq (α × β) where
   beq := fun (a₁, b₁) (a₂, b₂) => a₁ == a₂ && b₁ == b₂
@@ -1482,36 +1490,11 @@ def Prod.map {α₁ : Type u₁} {α₂ : Type u₂} {β₁ : Type v₁} {β₂
 
 @[simp] theorem Prod.map_apply (f : α → β) (g : γ → δ) (x) (y) :
     Prod.map f g (x, y) = (f x, g y) := rfl
-
--- We add `@[grind =]` to these in `Init.Data.Prod`.
 @[simp] theorem Prod.map_fst (f : α → β) (g : γ → δ) (x) : (Prod.map f g x).1 = f x.1 := rfl
 @[simp] theorem Prod.map_snd (f : α → β) (g : γ → δ) (x) : (Prod.map f g x).2 = g x.2 := rfl
 
 /-! # Dependent products -/
 
-instance {α : Type u} {β : α → Type v} [h₁ : DecidableEq α] [h₂ : ∀ a, DecidableEq (β a)] :
-    DecidableEq (Sigma β)
-  | ⟨a₁, b₁⟩, ⟨a₂, b₂⟩ =>
-    match a₁, b₁, a₂, b₂, h₁ a₁ a₂ with
-    | _, b₁, _, b₂, isTrue (Eq.refl _) =>
-      match b₁, b₂, h₂ _ b₁ b₂ with
-      | _, _, isTrue (Eq.refl _) => isTrue rfl
-      | _, _, isFalse n => isFalse fun h ↦
-        Sigma.noConfusion rfl .rfl (heq_of_eq h) fun _ e₂ ↦ n (eq_of_heq e₂)
-    | _, _, _, _, isFalse n => isFalse fun h ↦
-      Sigma.noConfusion rfl .rfl (heq_of_eq h) fun e₁ _ ↦ n (eq_of_heq e₁)
-
-instance {α : Sort u} {β : α → Sort v} [h₁ : DecidableEq α] [h₂ : ∀ a, DecidableEq (β a)] : DecidableEq (PSigma β)
-  | ⟨a₁, b₁⟩, ⟨a₂, b₂⟩ =>
-    match a₁, b₁, a₂, b₂, h₁ a₁ a₂ with
-    | _, b₁, _, b₂, isTrue (Eq.refl _) =>
-      match b₁, b₂, h₂ _ b₁ b₂ with
-      | _, _, isTrue (Eq.refl _) => isTrue rfl
-      | _, _, isFalse n => isFalse fun h ↦
-        PSigma.noConfusion rfl .rfl (heq_of_eq h) fun _ e₂ ↦ n (eq_of_heq e₂)
-    | _, _, _, _, isFalse n => isFalse fun h ↦
-      PSigma.noConfusion rfl .rfl (heq_of_eq h) fun e₁ _ ↦ n (eq_of_heq e₁)
-
 theorem Exists.of_psigma_prop {α : Sort u} {p : α → Prop} : (PSigma (fun x => p x)) → Exists (fun x => p x)
   | ⟨x, hx⟩ => ⟨x, hx⟩
 
@@ -1523,24 +1506,20 @@ protected theorem PSigma.eta {α : Sort u} {β : α → Sort v} {a₁ a₂ : α}
 
 /-! # Universe polymorphic unit -/
 
-theorem PUnit.ext (a b : PUnit) : a = b := by
-  cases a; cases b; exact rfl
-
-@[deprecated PUnit.ext (since := "2025-10-26")]
 theorem PUnit.subsingleton (a b : PUnit) : a = b := by
   cases a; cases b; exact rfl
 
 theorem PUnit.eq_punit (a : PUnit) : a = ⟨⟩ :=
-  PUnit.ext a ⟨⟩
+  PUnit.subsingleton a ⟨⟩
 
 instance : Subsingleton PUnit :=
-  Subsingleton.intro PUnit.ext
+  Subsingleton.intro PUnit.subsingleton
 
 instance : Inhabited PUnit where
   default := ⟨⟩
 
 instance : DecidableEq PUnit :=
-  fun a b => isTrue (PUnit.ext a b)
+  fun a b => isTrue (PUnit.subsingleton a b)
 
 /-! # Setoid -/
 
@@ -1599,10 +1578,6 @@ instance {p q : Prop} [d : Decidable (p ↔ q)] : Decidable (p = q) :=
   | isTrue h => isTrue (propext h)
   | isFalse h => isFalse fun heq => h (heq ▸ Iff.rfl)
 
-/-- Helper theorem for proving injectivity theorems -/
-theorem Lean.injEq_helper {P Q R : Prop} :
-  (P → Q → R) → (P ∧ Q → R) := by intro h ⟨h₁,h₂⟩; exact h h₁ h₂
-
 gen_injective_theorems% Array
 gen_injective_theorems% BitVec
 gen_injective_theorems% ByteArray
@@ -1630,8 +1605,8 @@ gen_injective_theorems% PSigma
 gen_injective_theorems% PSum
 gen_injective_theorems% Sigma
 gen_injective_theorems% String
-gen_injective_theorems% String.Pos.Raw
-gen_injective_theorems% Substring.Raw
+gen_injective_theorems% String.Pos
+gen_injective_theorems% Substring
 gen_injective_theorems% Subtype
 gen_injective_theorems% Sum
 gen_injective_theorems% Task
@@ -2363,10 +2338,8 @@ namespace Lean
 /--
 Depends on the correctness of the Lean compiler, interpreter, and all `[implemented_by ...]` and `[extern ...]` annotations.
 -/
-@[deprecated "in-kernel native reduction is deprecated; assert native evaluations with axioms instead" (since := "2026-02-01")]
 axiom trustCompiler : True
 
-set_option linter.deprecated false in
 /--
 When the kernel tries to reduce a term `Lean.reduceBool c`, it will invoke the Lean interpreter to evaluate `c`.
 The kernel will not use the interpreter if `c` is not a constant.
@@ -2386,13 +2359,11 @@ Recall that the compiler trusts the correctness of all `[implemented_by ...]` an
 If an extern function is executed, then the trusted code base will also include the implementation of the associated
 foreign function.
 -/
-@[deprecated "in-kernel native reduction is deprecated; assert native evaluations with axioms instead" (since := "2026-02-01")]
 opaque reduceBool (b : Bool) : Bool :=
   -- This ensures that `#print axioms` will track use of `reduceBool`.
   have := trustCompiler
   b
 
-set_option linter.deprecated false in
 /--
 Similar to `Lean.reduceBool` for closed `Nat` terms.
 
@@ -2400,14 +2371,12 @@ Remark: we do not have plans for supporting a generic `reduceValue {α} (a : α)
 The main issue is that it is non-trivial to convert an arbitrary runtime object back into a Lean expression.
 We believe `Lean.reduceBool` enables most interesting applications (e.g., proof by reflection).
 -/
-@[deprecated "in-kernel native reduction is deprecated; assert native evaluations with axioms instead" (since := "2026-02-01")]
 opaque reduceNat (n : Nat) : Nat :=
   -- This ensures that `#print axioms` will track use of `reduceNat`.
   have := trustCompiler
   n
 
 
-set_option linter.deprecated false in
 /--
 The axiom `ofReduceBool` is used to perform proofs by reflection. See `reduceBool`.
 
@@ -2421,10 +2390,8 @@ external type checkers that do not implement this feature.
 Keep in mind that if you are using Lean as programming language, you are already trusting the Lean compiler and interpreter.
 So, you are mainly losing the capability of type checking your development using external checkers.
 -/
-@[deprecated "in-kernel native reduction is deprecated; assert native evaluations with axioms instead" (since := "2026-02-01")]
 axiom ofReduceBool (a b : Bool) (h : reduceBool a = b) : a = b
 
-set_option linter.deprecated false in
 /--
 The axiom `ofReduceNat` is used to perform proofs by reflection. See `reduceBool`.
 
@@ -2434,7 +2401,6 @@ external type checkers that do not implement this feature.
 Keep in mind that if you are using Lean as programming language, you are already trusting the Lean compiler and interpreter.
 So, you are mainly losing the capability of type checking your development using external checkers.
 -/
-@[deprecated "in-kernel native reduction is deprecated; assert native evaluations with axioms instead" (since := "2026-02-01")]
 axiom ofReduceNat (a b : Nat) (h : reduceNat a = b) : a = b
 
 
@@ -2485,7 +2451,7 @@ class IdempotentOp (op : α → α → α) : Prop where
   idempotent : (x : α) → op x x = x
 
 /--
-`LeftIdentity op o` indicates `o` is a left identity of `op`.
+`LeftIdentify op o` indicates `o` is a left identity of `op`.
 
 This class does not require a proof that `o` is an identity, and
 is used primarily for inferring the identity using class resolution.
@@ -2493,7 +2459,7 @@ is used primarily for inferring the identity using class resolution.
 class LeftIdentity (op : α → β → β) (o : outParam α) : Prop
 
 /--
-`LawfulLeftIdentity op o` indicates `o` is a verified left identity of
+`LawfulLeftIdentify op o` indicates `o` is a verified left identity of
 `op`.
 -/
 class LawfulLeftIdentity (op : α → β → β) (o : outParam α) : Prop extends LeftIdentity op o where
@@ -2501,7 +2467,7 @@ class LawfulLeftIdentity (op : α → β → β) (o : outParam α) : Prop extend
   left_id : ∀ a, op o a = a
 
 /--
-`RightIdentity op o` indicates `o` is a right identity `o` of `op`.
+`RightIdentify op o` indicates `o` is a right identity `o` of `op`.
 
 This class does not require a proof that `o` is an identity, and is used
 primarily for inferring the identity using class resolution.
@@ -2509,7 +2475,7 @@ primarily for inferring the identity using class resolution.
 class RightIdentity (op : α → β → α) (o : outParam β) : Prop
 
 /--
-`LawfulRightIdentity op o` indicates `o` is a verified right identity of
+`LawfulRightIdentify op o` indicates `o` is a verified right identity of
 `op`.
 -/
 class LawfulRightIdentity (op : α → β → α) (o : outParam β) : Prop extends RightIdentity op o where
@@ -2557,7 +2523,8 @@ class Antisymm (r : α → α → Prop) : Prop where
   /-- An antisymmetric relation `r` satisfies `r a b → r b a → a = b`. -/
   antisymm (a b : α) : r a b → r b a → a = b
 
-/-- `Asymm r` means that the binary relation `r` is asymmetric, that is, `r a b → ¬ r b a`. -/
+/-- `Asymm r` means that the binary relation `r` is asymmetric, that is,
+`r a b → ¬ r b a`. -/
 class Asymm (r : α → α → Prop) : Prop where
   /-- An asymmetric relation satisfies `r a b → ¬ r b a`. -/
   asymm : ∀ a b, r a b → ¬r b a
@@ -2567,19 +2534,16 @@ class Symm (r : α → α → Prop) : Prop where
   /-- A symmetric relation satisfies `r a b → r b a`. -/
   symm : ∀ a b, r a b → r b a
 
-/-- `Total X r` means that the binary relation `r` on `X` is total, that is, `r a b` or `r b a`. -/
+/-- `Total X r` means that the binary relation `r` on `X` is total, that is, that for any
+`x y : X` we have `r x y` or `r y x`. -/
 class Total (r : α → α → Prop) : Prop where
-  /-- A total relation satisfies `r a b` or `r b a`. -/
+  /-- A total relation satisfies `r a b ∨ r b a`. -/
   total : ∀ a b, r a b ∨ r b a
 
-/-- `Irrefl r` means the binary relation `r` is irreflexive, that is, `r x x` never holds. -/
+/-- `Irrefl r` means the binary relation `r` is irreflexive, that is, `r x x` never
+holds. -/
 class Irrefl (r : α → α → Prop) : Prop where
   /-- An irreflexive relation satisfies `¬ r a a`. -/
   irrefl : ∀ a, ¬r a a
 
-/-- `Trichotomous r` says that `r` is trichotomous, that is, `¬ r a b → ¬ r b a → a = b`. -/
-class Trichotomous (r : α → α → Prop) : Prop where
-  /-- An trichotomous relation `r` satisfies `¬ r a b → ¬ r b a → a = b`. -/
-  trichotomous (a b : α) : ¬ r a b → ¬ r b a → a = b
-
 end Std

src/Init/Data.lean

@@ -6,6 +6,7 @@ Authors: Leonardo de Moura
 module
 
 prelude
+public import Init.Data.Basic
 public import Init.Data.Nat
 public import Init.Data.Bool
 public import Init.Data.BitVec
@@ -29,7 +30,6 @@ public import Init.Data.Random
 public import Init.Data.ToString
 public import Init.Data.Range
 public import Init.Data.Hashable
-public import Init.Data.LawfulHashable
 public import Init.Data.OfScientific
 public import Init.Data.Format
 public import Init.Data.Stream
@@ -52,3 +52,5 @@ public import Init.Data.Slice
 public import Init.Data.Order
 public import Init.Data.Rat
 public import Init.Data.Dyadic
+
+public section

src/Init/Data/AC.lean

@@ -7,9 +7,8 @@ Authors: Dany Fabian
 module
 
 prelude
-public import Init.GetElem
-import Init.ByCases
-import Init.PropLemmas
+public import Init.Classical
+public import Init.ByCases
 
 @[expose] public section
 

src/Init/Data/Array.lean

@@ -30,7 +30,5 @@ public import Init.Data.Array.Erase
 public import Init.Data.Array.Zip
 public import Init.Data.Array.InsertIdx
 public import Init.Data.Array.Extract
-public import Init.Data.Array.MinMax
-public import Init.Data.Array.Nat
-public import Init.Data.Array.Int
-public import Init.Data.Array.Count
+
+public section

src/Init/Data/Array/Attach.lean

@@ -6,10 +6,11 @@ Authors: Joachim Breitner, Mario Carneiro
 module
 
 prelude
-import all Init.Data.List.Attach
+public import Init.Data.Array.Mem
 public import Init.Data.Array.Lemmas
-import Init.Data.Array.Bootstrap
-import Init.Data.Array.Count
+public import Init.Data.Array.Count
+public import Init.Data.List.Attach
+import all Init.Data.List.Attach
 
 public section
 
@@ -83,10 +84,10 @@ well-founded recursion mechanism to prove that the function terminates.
   simp [pmap]
 
 /-- Implementation of `pmap` using the zero-copy version of `attach`. -/
-@[inline] def pmapImpl {P : α → Prop} (f : ∀ a, P a → β) (xs : Array α) (H : ∀ a ∈ xs, P a) :
+@[inline] private def pmapImpl {P : α → Prop} (f : ∀ a, P a → β) (xs : Array α) (H : ∀ a ∈ xs, P a) :
     Array β := (xs.attachWith _ H).map fun ⟨x, h'⟩ => f x h'
 
-@[csimp] theorem pmap_eq_pmapImpl : @pmap = @pmapImpl := by
+@[csimp] private theorem pmap_eq_pmapImpl : @pmap = @pmapImpl := by
   funext α β p f xs H
   cases xs
   simp only [pmap, pmapImpl, List.attachWith_toArray, List.map_toArray, mk.injEq, List.map_attachWith_eq_pmap]
@@ -94,16 +95,16 @@ well-founded recursion mechanism to prove that the function terminates.
   intro a m h₁ h₂
   congr
 
-@[simp] theorem pmap_empty {P : α → Prop} (f : ∀ a, P a → β) : pmap f #[] (by simp) = #[] := rfl
+@[simp, grind =] theorem pmap_empty {P : α → Prop} (f : ∀ a, P a → β) : pmap f #[] (by simp) = #[] := rfl
 
-@[simp] theorem pmap_push {P : α → Prop} (f : ∀ a, P a → β) (a : α) (xs : Array α) (h : ∀ b ∈ xs.push a, P b) :
+@[simp, grind =] theorem pmap_push {P : α → Prop} (f : ∀ a, P a → β) (a : α) (xs : Array α) (h : ∀ b ∈ xs.push a, P b) :
     pmap f (xs.push a) h =
       (pmap f xs (fun a m => by simp at h; exact h a (.inl m))).push (f a (h a (by simp))) := by
   simp [pmap]
 
-@[simp] theorem attach_empty : (#[] : Array α).attach = #[] := rfl
+@[simp, grind =] theorem attach_empty : (#[] : Array α).attach = #[] := rfl
 
-@[simp] theorem attachWith_empty {P : α → Prop} (H : ∀ x ∈ #[], P x) : (#[] : Array α).attachWith P H = #[] := rfl
+@[simp, grind =] theorem attachWith_empty {P : α → Prop} (H : ∀ x ∈ #[], P x) : (#[] : Array α).attachWith P H = #[] := rfl
 
 @[simp] theorem _root_.List.attachWith_mem_toArray {l : List α} :
     l.attachWith (fun x => x ∈ l.toArray) (fun x h => by simpa using h) =
@@ -124,11 +125,13 @@ theorem pmap_congr_left {p q : α → Prop} {f : ∀ a, p a → β} {g : ∀ a,
   simp only [List.pmap_toArray, mk.injEq]
   rw [List.pmap_congr_left _ h]
 
+@[grind =]
 theorem map_pmap {p : α → Prop} {g : β → γ} {f : ∀ a, p a → β} {xs : Array α} (H) :
     map g (pmap f xs H) = pmap (fun a h => g (f a h)) xs H := by
   cases xs
   simp [List.map_pmap]
 
+@[grind =]
 theorem pmap_map {p : β → Prop} {g : ∀ b, p b → γ} {f : α → β} {xs : Array α} (H) :
     pmap g (map f xs) H = pmap (fun a h => g (f a) h) xs fun _ h => H _ (mem_map_of_mem h) := by
   cases xs
@@ -144,14 +147,14 @@ theorem attachWith_congr {xs ys : Array α} (w : xs = ys) {P : α → Prop} {H :
   subst w
   simp
 
-@[simp] theorem attach_push {a : α} {xs : Array α} :
+@[simp, grind =] theorem attach_push {a : α} {xs : Array α} :
     (xs.push a).attach =
       (xs.attach.map (fun ⟨x, h⟩ => ⟨x, mem_push_of_mem a h⟩)).push ⟨a, by simp⟩ := by
   cases xs
   rw [attach_congr (List.push_toArray _ _)]
   simp [Function.comp_def]
 
-@[simp] theorem attachWith_push {a : α} {xs : Array α} {P : α → Prop} {H : ∀ x ∈ xs.push a, P x} :
+@[simp, grind =] theorem attachWith_push {a : α} {xs : Array α} {P : α → Prop} {H : ∀ x ∈ xs.push a, P x} :
     (xs.push a).attachWith P H =
       (xs.attachWith P (fun x h => by simp at H; exact H x (.inl h))).push ⟨a, H a (by simp)⟩ := by
   cases xs
@@ -173,6 +176,9 @@ theorem attach_map_val (xs : Array α) (f : α → β) :
   cases xs
   simp
 
+@[deprecated attach_map_val (since := "2025-02-17")]
+abbrev attach_map_coe := @attach_map_val
+
 -- The argument `xs : Array α` is explicit to allow rewriting from right to left.
 theorem attach_map_subtype_val (xs : Array α) : xs.attach.map Subtype.val = xs := by
   cases xs; simp
@@ -181,6 +187,9 @@ theorem attachWith_map_val {p : α → Prop} {f : α → β} {xs : Array α} (H
     ((xs.attachWith p H).map fun (i : { i // p i}) => f i) = xs.map f := by
   cases xs; simp
 
+@[deprecated attachWith_map_val (since := "2025-02-17")]
+abbrev attachWith_map_coe := @attachWith_map_val
+
 theorem attachWith_map_subtype_val {p : α → Prop} {xs : Array α} (H : ∀ a ∈ xs, p a) :
     (xs.attachWith p H).map Subtype.val = xs := by
   cases xs; simp
@@ -285,23 +294,25 @@ theorem getElem_attach {xs : Array α} {i : Nat} (h : i < xs.attach.size) :
     xs.attach[i] = ⟨xs[i]'(by simpa using h), getElem_mem (by simpa using h)⟩ :=
   getElem_attachWith h
 
-@[simp] theorem pmap_attach {xs : Array α} {p : {x // x ∈ xs} → Prop} {f : ∀ a, p a → β} (H) :
+@[simp, grind =] theorem pmap_attach {xs : Array α} {p : {x // x ∈ xs} → Prop} {f : ∀ a, p a → β} (H) :
     pmap f xs.attach H =
       xs.pmap (P := fun a => ∃ h : a ∈ xs, p ⟨a, h⟩)
         (fun a h => f ⟨a, h.1⟩ h.2) (fun a h => ⟨h, H ⟨a, h⟩ (by simp)⟩) := by
   ext <;> simp
 
-@[simp] theorem pmap_attachWith {xs : Array α} {p : {x // q x} → Prop} {f : ∀ a, p a → β} (H₁ H₂) :
+@[simp, grind =] theorem pmap_attachWith {xs : Array α} {p : {x // q x} → Prop} {f : ∀ a, p a → β} (H₁ H₂) :
     pmap f (xs.attachWith q H₁) H₂ =
       xs.pmap (P := fun a => ∃ h : q a, p ⟨a, h⟩)
         (fun a h => f ⟨a, h.1⟩ h.2) (fun a h => ⟨H₁ _ h, H₂ ⟨a, H₁ _ h⟩ (by simpa)⟩) := by
   ext <;> simp
 
+@[grind =]
 theorem foldl_pmap {xs : Array α} {P : α → Prop} {f : (a : α) → P a → β}
     (H : ∀ (a : α), a ∈ xs → P a) (g : γ → β → γ) (x : γ) :
     (xs.pmap f H).foldl g x = xs.attach.foldl (fun acc a => g acc (f a.1 (H _ a.2))) x := by
   rw [pmap_eq_map_attach, foldl_map]
 
+@[grind =]
 theorem foldr_pmap {xs : Array α} {P : α → Prop} {f : (a : α) → P a → β}
     (H : ∀ (a : α), a ∈ xs → P a) (g : β → γ → γ) (x : γ) :
     (xs.pmap f H).foldr g x = xs.attach.foldr (fun a acc => g (f a.1 (H _ a.2)) acc) x := by
@@ -359,18 +370,20 @@ theorem foldr_attach {xs : Array α} {f : α → β → β} {b : β} :
   ext
   simpa using fun a => List.mem_of_getElem? a
 
+@[grind =]
 theorem attach_map {xs : Array α} {f : α → β} :
     (xs.map f).attach = xs.attach.map (fun ⟨x, h⟩ => ⟨f x, mem_map_of_mem h⟩) := by
   cases xs
   ext <;> simp
 
+@[grind =]
 theorem attachWith_map {xs : Array α} {f : α → β} {P : β → Prop} (H : ∀ (b : β), b ∈ xs.map f → P b) :
     (xs.map f).attachWith P H = (xs.attachWith (P ∘ f) (fun _ h => H _ (mem_map_of_mem h))).map
       fun ⟨x, h⟩ => ⟨f x, h⟩ := by
   cases xs
   simp [List.attachWith_map]
 
-@[simp] theorem map_attachWith {xs : Array α} {P : α → Prop} {H : ∀ (a : α), a ∈ xs → P a}
+@[simp, grind =] theorem map_attachWith {xs : Array α} {P : α → Prop} {H : ∀ (a : α), a ∈ xs → P a}
     {f : { x // P x } → β} :
     (xs.attachWith P H).map f = xs.attach.map fun ⟨x, h⟩ => f ⟨x, H _ h⟩ := by
   cases xs <;> simp_all
@@ -388,6 +401,9 @@ theorem map_attach_eq_pmap {xs : Array α} {f : { x // x ∈ xs } → β} :
   cases xs
   ext <;> simp
 
+@[deprecated map_attach_eq_pmap (since := "2025-02-09")]
+abbrev map_attach := @map_attach_eq_pmap
+
 @[grind =]
 theorem attach_filterMap {xs : Array α} {f : α → Option β} :
     (xs.filterMap f).attach = xs.attach.filterMap
@@ -423,6 +439,7 @@ theorem filter_attachWith {q : α → Prop} {xs : Array α} {p : {x // q x} →
   cases xs
   simp [Function.comp_def, List.filter_map]
 
+@[grind =]
 theorem pmap_pmap {p : α → Prop} {q : β → Prop} {g : ∀ a, p a → β} {f : ∀ b, q b → γ} {xs} (H₁ H₂) :
     pmap f (pmap g xs H₁) H₂ =
       pmap (α := { x // x ∈ xs }) (fun a h => f (g a h) (H₂ (g a h) (mem_pmap_of_mem a.2))) xs.attach
@@ -430,7 +447,7 @@ theorem pmap_pmap {p : α → Prop} {q : β → Prop} {g : ∀ a, p a → β} {f
   cases xs
   simp [List.pmap_pmap, List.pmap_map]
 
-@[simp] theorem pmap_append {p : ι → Prop} {f : ∀ a : ι, p a → α} {xs ys : Array ι}
+@[simp, grind =] theorem pmap_append {p : ι → Prop} {f : ∀ a : ι, p a → α} {xs ys : Array ι}
     (h : ∀ a ∈ xs ++ ys, p a) :
     (xs ++ ys).pmap f h =
       (xs.pmap f fun a ha => h a (mem_append_left ys ha)) ++
@@ -445,7 +462,7 @@ theorem pmap_append' {p : α → Prop} {f : ∀ a : α, p a → β} {xs ys : Arr
       xs.pmap f h₁ ++ ys.pmap f h₂ :=
   pmap_append _
 
-@[simp] theorem attach_append {xs ys : Array α} :
+@[simp, grind =] theorem attach_append {xs ys : Array α} :
     (xs ++ ys).attach = xs.attach.map (fun ⟨x, h⟩ => ⟨x, mem_append_left ys h⟩) ++
       ys.attach.map fun ⟨x, h⟩ => ⟨x, mem_append_right xs h⟩ := by
   cases xs
@@ -453,59 +470,62 @@ theorem pmap_append' {p : α → Prop} {f : ∀ a : α, p a → β} {xs ys : Arr
   rw [attach_congr (List.append_toArray _ _)]
   simp [List.attach_append, Function.comp_def]
 
-@[simp] theorem attachWith_append {P : α → Prop} {xs ys : Array α}
+@[simp, grind =] theorem attachWith_append {P : α → Prop} {xs ys : Array α}
     {H : ∀ (a : α), a ∈ xs ++ ys → P a} :
     (xs ++ ys).attachWith P H = xs.attachWith P (fun a h => H a (mem_append_left ys h)) ++
       ys.attachWith P (fun a h => H a (mem_append_right xs h)) := by
   simp [attachWith]
 
-@[simp] theorem pmap_reverse {P : α → Prop} {f : (a : α) → P a → β} {xs : Array α}
+@[simp, grind =] theorem pmap_reverse {P : α → Prop} {f : (a : α) → P a → β} {xs : Array α}
     (H : ∀ (a : α), a ∈ xs.reverse → P a) :
     xs.reverse.pmap f H = (xs.pmap f (fun a h => H a (by simpa using h))).reverse := by
   induction xs <;> simp_all
 
+@[grind =]
 theorem reverse_pmap {P : α → Prop} {f : (a : α) → P a → β} {xs : Array α}
     (H : ∀ (a : α), a ∈ xs → P a) :
     (xs.pmap f H).reverse = xs.reverse.pmap f (fun a h => H a (by simpa using h)) := by
   rw [pmap_reverse]
 
-@[simp] theorem attachWith_reverse {P : α → Prop} {xs : Array α}
+@[simp, grind =] theorem attachWith_reverse {P : α → Prop} {xs : Array α}
     {H : ∀ (a : α), a ∈ xs.reverse → P a} :
     xs.reverse.attachWith P H =
       (xs.attachWith P (fun a h => H a (by simpa using h))).reverse := by
   cases xs
   simp
 
+@[grind =]
 theorem reverse_attachWith {P : α → Prop} {xs : Array α}
     {H : ∀ (a : α), a ∈ xs → P a} :
     (xs.attachWith P H).reverse = (xs.reverse.attachWith P (fun a h => H a (by simpa using h))) := by
   cases xs
   simp
 
-@[simp] theorem attach_reverse {xs : Array α} :
+@[simp, grind =] theorem attach_reverse {xs : Array α} :
     xs.reverse.attach = xs.attach.reverse.map fun ⟨x, h⟩ => ⟨x, by simpa using h⟩ := by
   cases xs
   rw [attach_congr List.reverse_toArray]
   simp
 
+@[grind =]
 theorem reverse_attach {xs : Array α} :
     xs.attach.reverse = xs.reverse.attach.map fun ⟨x, h⟩ => ⟨x, by simpa using h⟩ := by
   cases xs
   simp
 
-@[simp] theorem back?_pmap {P : α → Prop} {f : (a : α) → P a → β} {xs : Array α}
+@[simp, grind =] theorem back?_pmap {P : α → Prop} {f : (a : α) → P a → β} {xs : Array α}
     (H : ∀ (a : α), a ∈ xs → P a) :
     (xs.pmap f H).back? = xs.attach.back?.map fun ⟨a, m⟩ => f a (H a m) := by
   cases xs
   simp
 
-@[simp] theorem back?_attachWith {P : α → Prop} {xs : Array α}
+@[simp, grind =] theorem back?_attachWith {P : α → Prop} {xs : Array α}
     {H : ∀ (a : α), a ∈ xs → P a} :
     (xs.attachWith P H).back? = xs.back?.pbind (fun a h => some ⟨a, H _ (mem_of_back? h)⟩) := by
   cases xs
   simp
 
-@[simp]
+@[simp, grind =]
 theorem back?_attach {xs : Array α} :
     xs.attach.back? = xs.back?.pbind fun a h => some ⟨a, mem_of_back? h⟩ := by
   cases xs
@@ -574,6 +594,9 @@ def unattach {α : Type _} {p : α → Prop} (xs : Array { x // p x }) : Array
 @[simp] theorem unattach_empty {p : α → Prop} : (#[] : Array { x // p x }).unattach = #[] := by
   simp [unattach]
 
+@[deprecated unattach_empty (since := "2025-05-26")]
+abbrev unattach_nil := @unattach_empty
+
 @[simp] theorem unattach_push {p : α → Prop} {a : { x // p x }} {xs : Array { x // p x }} :
     (xs.push a).unattach = xs.unattach.push a.1 := by
   simp only [unattach, Array.map_push]
@@ -748,6 +771,9 @@ and simplifies these to the function directly taking the value.
     (Array.replicate n x).unattach = Array.replicate n x.1 := by
   simp [unattach]
 
+@[deprecated unattach_replicate (since := "2025-03-18")]
+abbrev unattach_mkArray := @unattach_replicate
+
 /-! ### Well-founded recursion preprocessing setup -/
 
 @[wf_preprocess] theorem map_wfParam {xs : Array α} {f : α → β} :

src/Init/Data/Array/Basic.lean

@@ -6,14 +6,15 @@ Authors: Leonardo de Moura
 module
 
 prelude
+public import Init.WFTactics
+public import Init.Data.Nat.Basic
+public import Init.Data.Fin.Basic
+public import Init.Data.UInt.BasicAux
 public import Init.GetElem
 public import Init.Data.List.ToArrayImpl
 import all Init.Data.List.ToArrayImpl
 public import Init.Data.Array.Set
 import all Init.Data.Array.Set
-public import Init.WF
-meta import Init.MetaTypes
-import Init.WFTactics
 
 public section
 
@@ -128,11 +129,20 @@ end Array
 
 namespace List
 
+@[deprecated Array.toArray_toList (since := "2025-02-17")]
+abbrev toArray_toList := @Array.toArray_toList
+
 -- This does not need to be a simp lemma, as already after the `whnfR` the right hand side is `as`.
 theorem toList_toArray {as : List α} : as.toArray.toList = as := rfl
 
+@[deprecated toList_toArray (since := "2025-02-17")]
+abbrev _root_.Array.toList_toArray := @List.toList_toArray
+
 @[simp, grind =] theorem size_toArray {as : List α} : as.toArray.size = as.length := by simp [Array.size]
 
+@[deprecated size_toArray (since := "2025-02-17")]
+abbrev _root_.Array.size_toArray := @List.size_toArray
+
 @[simp, grind =] theorem getElem_toArray {xs : List α} {i : Nat} (h : i < xs.toArray.size) :
     xs.toArray[i] = xs[i]'(by simpa using h) := rfl
 
@@ -212,6 +222,20 @@ Examples:
 def replicate {α : Type u} (n : Nat) (v : α) : Array α where
   toList := List.replicate n v
 
+/--
+Creates an array that contains `n` repetitions of `v`.
+
+The corresponding `List` function is `List.replicate`.
+
+Examples:
+ * `Array.mkArray 2 true = #[true, true]`
+ * `Array.mkArray 3 () = #[(), (), ()]`
+ * `Array.mkArray 0 "anything" = #[]`
+-/
+@[extern "lean_mk_array", deprecated replicate (since := "2025-03-18")]
+def mkArray {α : Type u} (n : Nat) (v : α) : Array α where
+  toList := List.replicate n v
+
 /--
 Swaps two elements of an array. The modification is performed in-place when the reference to the
 array is unique.
@@ -229,7 +253,7 @@ def swap (xs : Array α) (i j : @& Nat) (hi : i < xs.size := by get_elem_tactic)
   let xs'  := xs.set i v₂
   xs'.set j v₁ (Nat.lt_of_lt_of_eq hj (size_set _).symm)
 
-@[simp, grind =] theorem size_swap {xs : Array α} {i j : Nat} {hi hj} : (xs.swap i j hi hj).size = xs.size := by
+@[simp] theorem size_swap {xs : Array α} {i j : Nat} {hi hj} : (xs.swap i j hi hj).size = xs.size := by
   change ((xs.set i xs[j]).set j xs[i]
     (Nat.lt_of_lt_of_eq hj (size_set _).symm)).size = xs.size
   rw [size_set, size_set]
@@ -245,7 +269,7 @@ Examples:
 * `#["red", "green", "blue", "brown"].swapIfInBounds 0 4 = #["red", "green", "blue", "brown"]`
 * `#["red", "green", "blue", "brown"].swapIfInBounds 9 2 = #["red", "green", "blue", "brown"]`
 -/
-@[extern "lean_array_swap", expose]
+@[extern "lean_array_swap", grind]
 def swapIfInBounds (xs : Array α) (i j : @& Nat) : Array α :=
   if h₁ : i < xs.size then
   if h₂ : j < xs.size then swap xs i j
@@ -388,6 +412,10 @@ that requires a proof the array is non-empty.
 def back? (xs : Array α) : Option α :=
   xs[xs.size - 1]?
 
+@[deprecated "Use `a[i]?` instead." (since := "2025-02-12"), expose]
+def get? (xs : Array α) (i : Nat) : Option α :=
+  if h : i < xs.size then some xs[i] else none
+
 /--
 Swaps a new element with the element at the given index.
 
@@ -451,7 +479,7 @@ Examples:
 -/
 abbrev take (xs : Array α) (i : Nat) : Array α := extract xs 0 i
 
-@[simp, grind =] theorem take_eq_extract {xs : Array α} {i : Nat} : xs.take i = xs.extract 0 i := rfl
+@[simp] theorem take_eq_extract {xs : Array α} {i : Nat} : xs.take i = xs.extract 0 i := rfl
 
 /--
 Removes the first `i` elements of `xs`. If `xs` has fewer than `i` elements, the new array is empty.
@@ -465,7 +493,7 @@ Examples:
 -/
 abbrev drop (xs : Array α) (i : Nat) : Array α := extract xs i xs.size
 
-@[simp, grind =] theorem drop_eq_extract {xs : Array α} {i : Nat} : xs.drop i = xs.extract i xs.size := rfl
+@[simp] theorem drop_eq_extract {xs : Array α} {i : Nat} : xs.drop i = xs.extract i xs.size := rfl
 
 @[inline]
 unsafe def modifyMUnsafe [Monad m] (xs : Array α) (i : Nat) (f : α → m α) : m (Array α) := do
@@ -573,7 +601,7 @@ protected def forIn' {α : Type u} {β : Type v} {m : Type v → Type w} [Monad
       | ForInStep.yield b => loop i (Nat.le_of_lt h') b
   loop as.size (Nat.le_refl _) b
 
-instance [Monad m] : ForIn' m (Array α) α inferInstance where
+instance : ForIn' m (Array α) α inferInstance where
   forIn' := Array.forIn'
 
 -- No separate `ForIn` instance is required because it can be derived from `ForIn'`.
@@ -592,8 +620,6 @@ unsafe def foldlMUnsafe {α : Type u} {β : Type v} {m : Type v → Type w} [Mon
   if start < stop then
     if stop ≤ as.size then
       fold (USize.ofNat start) (USize.ofNat stop) init
-    else if start < as.size then
-      fold (USize.ofNat start) (USize.ofNat as.size) init
     else
       pure init
   else
@@ -739,7 +765,8 @@ of results.
 def mapM {α : Type u} {β : Type v} {m : Type v → Type w} [Monad m] (f : α → m β) (as : Array α) : m (Array β) :=
   -- Note: we cannot use `foldlM` here for the reference implementation because this calls
   -- `bind` and `pure` too many times. (We are not assuming `m` is a `LawfulMonad`)
-  let rec map (i : Nat) (bs : Array β) : m (Array β) := do
+  let rec @[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
+    map (i : Nat) (bs : Array β) : m (Array β) := do
       if hlt : i < as.size then
         map (i+1) (bs.push (← f as[i]))
       else
@@ -899,7 +926,8 @@ entire array is checked.
 @[implemented_by anyMUnsafe, expose]
 def anyM {α : Type u} {m : Type → Type w} [Monad m] (p : α → m Bool) (as : Array α) (start := 0) (stop := as.size) : m Bool :=
   let any (stop : Nat) (h : stop ≤ as.size) :=
-    let rec loop (j : Nat) : m Bool := do
+    let rec @[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
+    loop (j : Nat) : m Bool := do
       if hlt : j < stop then
         have : j < as.size := Nat.lt_of_lt_of_le hlt h
         if (← p as[j]) then
@@ -1006,7 +1034,7 @@ unless `start < stop`. By default, the entire array is used.
 protected def forM {α : Type u} {m : Type v → Type w} [Monad m] (f : α → m PUnit) (as : Array α) (start := 0) (stop := as.size) : m PUnit :=
   as.foldlM (fun _ => f) ⟨⟩ start stop
 
-instance [Monad m] : ForM m (Array α) α where
+instance : ForM m (Array α) α where
   forM xs f := Array.forM f xs
 
 -- We simplify `Array.forM` to `forM`.
@@ -1237,7 +1265,8 @@ Examples:
 -/
 @[inline, expose]
 def findIdx? {α : Type u} (p : α → Bool) (as : Array α) : Option Nat :=
-  let rec loop (j : Nat) :=
+  let rec @[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
+  loop (j : Nat) :=
     if h : j < as.size then
       if p as[j] then some j else loop (j + 1)
     else none
@@ -1254,7 +1283,8 @@ Examples:
 -/
 @[inline]
 def findFinIdx? {α : Type u} (p : α → Bool) (as : Array α) : Option (Fin as.size) :=
-  let rec loop (j : Nat) :=
+  let rec @[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
+  loop (j : Nat) :=
     if h : j < as.size then
       if p as[j] then some ⟨j, h⟩ else loop (j + 1)
     else none
@@ -1290,6 +1320,7 @@ Examples:
 @[inline, expose]
 def findIdx (p : α → Bool) (as : Array α) : Nat := (as.findIdx? p).getD as.size
 
+@[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
 def idxOfAux [BEq α] (xs : Array α) (v : α) (i : Nat) : Option (Fin xs.size) :=
   if h : i < xs.size then
     if xs[i] == v then some ⟨i, h⟩
@@ -1300,7 +1331,7 @@ decreasing_by simp_wf; decreasing_trivial_pre_omega
 
 
 /--
-Returns the index of the first element equal to `a`, or `none` if no element is equal
+Returns the index of the first element equal to `a`, or the size of the array if no element is equal
 to `a`. The index is returned as a `Fin`, which guarantees that it is in bounds.
 
 Examples:
@@ -1353,7 +1384,7 @@ Examples:
 * `#[2, 4, 5, 6].any (· % 2 = 0) = true`
 * `#[2, 4, 5, 6].any (· % 2 = 1) = true`
 -/
-@[inline, expose, suggest_for Array.some]
+@[inline, expose]
 def any (as : Array α) (p : α → Bool) (start := 0) (stop := as.size) : Bool :=
   Id.run <| as.anyM (pure <| p ·) start stop
 
@@ -1371,7 +1402,7 @@ Examples:
 * `#[2, 4, 6].all (· % 2 = 0) = true`
 * `#[2, 4, 5, 6].all (· % 2 = 0) = false`
 -/
-@[inline, suggest_for Array.every]
+@[inline]
 def all (as : Array α) (p : α → Bool) (start := 0) (stop := as.size) : Bool :=
   Id.run <| as.allM (pure <| p ·) start stop
 
@@ -1699,6 +1730,7 @@ Examples:
  * `#[3, 2, 3, 4].popWhile (· > 2) = #[3, 2]`
  * `(#[] : Array Nat).popWhile (· > 2) = #[]`
 -/
+@[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
 def popWhile (p : α → Bool) (as : Array α) : Array α :=
   if h : as.size > 0 then
     if p (as[as.size - 1]'(Nat.sub_lt h (by decide))) then
@@ -1709,7 +1741,7 @@ def popWhile (p : α → Bool) (as : Array α) : Array α :=
     as
 decreasing_by simp_wf; decreasing_trivial_pre_omega
 
-@[simp, grind =] theorem popWhile_empty {p : α → Bool} :
+@[simp] theorem popWhile_empty {p : α → Bool} :
     popWhile p #[] = #[] := by
   simp [popWhile]
 
@@ -1723,7 +1755,8 @@ Examples:
  * `#[0, 1, 2, 3, 2, 1].takeWhile (· < 0) = #[]`
 -/
 def takeWhile (p : α → Bool) (as : Array α) : Array α :=
-  let rec go (i : Nat) (acc : Array α) : Array α :=
+  let rec @[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
+  go (i : Nat) (acc : Array α) : Array α :=
     if h : i < as.size then
       let a := as[i]
       if p a then
@@ -1746,6 +1779,7 @@ Examples:
 * `#["apple", "pear", "orange"].eraseIdx 1 = #["apple", "orange"]`
 * `#["apple", "pear", "orange"].eraseIdx 2 = #["apple", "pear"]`
 -/
+@[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
 def eraseIdx (xs : Array α) (i : Nat) (h : i < xs.size := by get_elem_tactic) : Array α :=
   if h' : i + 1 < xs.size then
     let xs' := xs.swap (i + 1) i
@@ -1756,8 +1790,7 @@ termination_by xs.size - i
 decreasing_by simp_wf; exact Nat.sub_succ_lt_self _ _ h
 
 -- This is required in `Lean.Data.PersistentHashMap`.
-@[simp, grind =]
-theorem size_eraseIdx {xs : Array α} (i : Nat) (h) : (xs.eraseIdx i h).size = xs.size - 1 := by
+@[simp] theorem size_eraseIdx {xs : Array α} (i : Nat) (h) : (xs.eraseIdx i h).size = xs.size - 1 := by
   induction xs, i, h using Array.eraseIdx.induct with
   | @case1 xs i h h' xs' ih =>
     unfold eraseIdx
@@ -1779,6 +1812,7 @@ Examples:
 * `#["apple", "pear", "orange"].eraseIdxIfInBounds 3 = #["apple", "pear", "orange"]`
 * `#["apple", "pear", "orange"].eraseIdxIfInBounds 5 = #["apple", "pear", "orange"]`
 -/
+@[grind]
 def eraseIdxIfInBounds (xs : Array α) (i : Nat) : Array α :=
   if h : i < xs.size then xs.eraseIdx i h else xs
 
@@ -1841,7 +1875,8 @@ Examples:
  * `#["tues", "thur", "sat"].insertIdx 3 "wed" = #["tues", "thur", "sat", "wed"]`
 -/
 @[inline] def insertIdx (as : Array α) (i : Nat) (a : α) (_ : i ≤ as.size := by get_elem_tactic) : Array α :=
-  let rec loop (as : Array α) (j : Fin as.size) :=
+  let rec @[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
+  loop (as : Array α) (j : Fin as.size) :=
     if i < j then
       let j' : Fin as.size := ⟨j-1, Nat.lt_of_le_of_lt (Nat.pred_le _) j.2⟩
       let as := as.swap j' j
@@ -1895,6 +1930,7 @@ def insertIdxIfInBounds (as : Array α) (i : Nat) (a : α) : Array α :=
   else
     as
 
+@[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
 def isPrefixOfAux [BEq α] (as bs : Array α) (hle : as.size ≤ bs.size) (i : Nat) : Bool :=
   if h : i < as.size then
     let a := as[i]
@@ -1923,7 +1959,7 @@ def isPrefixOf [BEq α] (as bs : Array α) : Bool :=
   else
     false
 
-@[specialize]
+@[semireducible, specialize] -- This is otherwise irreducible because it uses well-founded recursion.
 def zipWithMAux {m : Type v → Type w} [Monad m] (as : Array α) (bs : Array β) (f : α → β → m γ) (i : Nat) (cs : Array γ) : m (Array γ) := do
   if h : i < as.size then
     let a := as[i]
@@ -2086,6 +2122,7 @@ private def allDiffAuxAux [BEq α] (as : Array α) (a : α) : forall (i : Nat),
     have : i < as.size := Nat.lt_trans (Nat.lt_succ_self _) h;
     a != as[i] && allDiffAuxAux as a i this
 
+@[semireducible] -- This is otherwise irreducible because it uses well-founded recursion.
 private def allDiffAux [BEq α] (as : Array α) (i : Nat) : Bool :=
   if h : i < as.size then
     allDiffAuxAux as as[i] i h && allDiffAux as (i+1)
@@ -2125,7 +2162,7 @@ Examples:
 
 /-! ### Repr and ToString -/
 
-protected def repr {α : Type u} [Repr α] (xs : Array α) : Std.Format :=
+protected def Array.repr {α : Type u} [Repr α] (xs : Array α) : Std.Format :=
   let _ : Std.ToFormat α := ⟨repr⟩
   if xs.size == 0 then
     "#[]"
@@ -2139,3 +2176,5 @@ instance [ToString α] : ToString (Array α) where
   toString xs := String.Internal.append "#" (toString xs.toList)
 
 end Array
+
+export Array (mkArray)

src/Init/Data/Array/BasicAux.lean

@@ -6,10 +6,10 @@ Authors: Leonardo de Moura
 module
 
 prelude
+public import Init.Data.Array.Basic
 import all Init.Data.Array.Basic
-public import Init.Data.Array.Set
-public import Init.Util
-import Init.Data.Nat.Linear
+public import Init.Data.Nat.Linear
+public import Init.NotationExtra
 
 public section
 

src/Init/Data/Array/BinSearch.lean

@@ -7,9 +7,8 @@ module
 
 prelude
 public import Init.Data.Array.Basic
-import Init.Data.Bool
-import Init.Omega
-import Init.WFTactics
+public import Init.Data.Int.DivMod.Lemmas
+public import Init.Omega
 
 public section
 universe u v

src/Init/Data/Array/Bootstrap.lean

@@ -7,10 +7,9 @@ Authors: Mario Carneiro
 module
 
 prelude
+public import Init.Data.List.TakeDrop
+public import Init.Data.Array.Basic
 import all Init.Data.Array.Basic
-public import Init.Data.List.Control
-import Init.Data.List.Lemmas
-import Init.Data.List.TakeDrop
 
 public section
 
@@ -25,6 +24,29 @@ set_option linter.indexVariables true -- Enforce naming conventions for index va
 
 namespace Array
 
+/--
+Use the indexing notation `a[i]` instead.
+
+Access an element from an array without needing a runtime bounds checks,
+using a `Nat` index and a proof that it is in bounds.
+
+This function does not use `get_elem_tactic` to automatically find the proof that
+the index is in bounds. This is because the tactic itself needs to look up values in
+arrays.
+-/
+@[deprecated "Use indexing notation `as[i]` instead" (since := "2025-02-17")]
+def get {α : Type u} (xs : @& Array α) (i : @& Nat) (h : LT.lt i xs.size) : α :=
+  xs.toList.get ⟨i, h⟩
+
+/--
+Use the indexing notation `a[i]!` instead.
+
+Access an element from an array, or panic if the index is out of bounds.
+-/
+@[deprecated "Use indexing notation `as[i]!` instead" (since := "2025-02-17"), expose]
+def get! {α : Type u} [Inhabited α] (xs : @& Array α) (i : @& Nat) : α :=
+  Array.getD xs i default
+
 theorem foldlM_toList.aux [Monad m]
     {f : β → α → m β} {xs : Array α} {i j} (H : xs.size ≤ i + j) {b} :
     foldlM.loop f xs xs.size (Nat.le_refl _) i j b = (xs.toList.drop j).foldlM f b := by
@@ -33,7 +55,7 @@ theorem foldlM_toList.aux [Monad m]
   · cases Nat.not_le_of_gt ‹_› (Nat.zero_add _ ▸ H)
   · rename_i i; rw [Nat.succ_add] at H
     simp [foldlM_toList.aux (j := j+1) H]
-    rw (occs := [2]) [← List.getElem_cons_drop ‹_›]
+    rw (occs := [2]) [← List.getElem_cons_drop_succ_eq_drop ‹_›]
     simp
   · rw [List.drop_of_length_le (Nat.ge_of_not_lt ‹_›)]; simp
 
@@ -75,6 +97,9 @@ theorem foldrM_eq_reverse_foldlM_toList [Monad m] {f : α → β → m β} {init
   rcases xs with ⟨xs⟩
   simp [push, List.concat_eq_append]
 
+@[deprecated toList_push (since := "2025-05-26")]
+abbrev push_toList := @toList_push
+
 @[simp, grind =] theorem toListAppend_eq {xs : Array α} {l : List α} : xs.toListAppend l = xs.toList ++ l := by
   simp [toListAppend, ← foldr_toList]
 
@@ -83,6 +108,9 @@ theorem foldrM_eq_reverse_foldlM_toList [Monad m] {f : α → β → m β} {init
 
 @[simp, grind =] theorem toList_pop {xs : Array α} : xs.pop.toList = xs.toList.dropLast := rfl
 
+@[deprecated toList_pop (since := "2025-02-17")]
+abbrev pop_toList := @Array.toList_pop
+
 @[simp] theorem append_eq_append {xs ys : Array α} : xs.append ys = xs ++ ys := rfl
 
 @[simp, grind =] theorem toList_append {xs ys : Array α} :
@@ -99,15 +127,9 @@ theorem foldrM_eq_reverse_foldlM_toList [Monad m] {f : α → β → m β} {init
 @[simp, grind =] theorem empty_append {xs : Array α} : #[] ++ xs = xs := by
   apply ext'; simp only [toList_append, List.nil_append]
 
-@[simp] theorem append_assoc {xs ys zs : Array α} : xs ++ ys ++ zs = xs ++ (ys ++ zs) := by
+@[simp, grind _=_] theorem append_assoc {xs ys zs : Array α} : xs ++ ys ++ zs = xs ++ (ys ++ zs) := by
   apply ext'; simp only [toList_append, List.append_assoc]
 
-grind_pattern append_assoc => (xs ++ ys) ++ zs where
-  xs =/= #[]; ys =/= #[]; zs =/= #[]
-
-grind_pattern append_assoc => xs ++ (ys ++ zs) where
-  xs =/= #[]; ys =/= #[]; zs =/= #[]
-
 @[simp] theorem appendList_eq_append {xs : Array α} {l : List α} : xs.appendList l = xs ++ l := rfl
 
 @[simp, grind =] theorem toList_appendList {xs : Array α} {l : List α} :
@@ -115,4 +137,6 @@ grind_pattern append_assoc => xs ++ (ys ++ zs) where
   rw [← appendList_eq_append]; unfold Array.appendList
   induction l generalizing xs <;> simp [*]
 
+
+
 end Array

src/Init/Data/Array/Count.lean

@@ -6,15 +6,10 @@ Authors: Kim Morrison
 module
 
 prelude
+public import Init.Data.Array.Basic
 import all Init.Data.Array.Basic
-import Init.Grind.Util  -- shake: keep (`@[grind]` dependency)
-public import Init.BinderPredicates
-public import Init.Ext
-public import Init.NotationExtra
-import Init.Data.Array.Lemmas
-import Init.Data.Bool
-import Init.Data.List.Count
-import Init.Data.List.Nat.Count
+public import Init.Data.Array.Lemmas
+public import Init.Data.List.Nat.Count
 
 public section
 
@@ -68,12 +63,12 @@ theorem size_eq_countP_add_countP {xs : Array α} : xs.size = countP p xs + coun
   rcases xs with ⟨xs⟩
   simp [List.length_eq_countP_add_countP (p := p)]
 
+@[grind _=_]
 theorem countP_eq_size_filter {xs : Array α} : countP p xs = (filter p xs).size := by
   rcases xs with ⟨xs⟩
   simp [List.countP_eq_length_filter]
 
-grind_pattern countP_eq_size_filter => xs.countP p, xs.filter p
-
+@[grind =]
 theorem countP_eq_size_filter' : countP p = size ∘ filter p := by
   funext xs
   apply countP_eq_size_filter
@@ -98,18 +93,6 @@ theorem countP_le_size : countP p xs ≤ xs.size := by
   rcases xs with ⟨xs⟩
   simp
 
-/-- This lemma is only relevant for `grind`. -/
-@[grind ←=]
-theorem _root_.Std.Internal.Array.countP_eq_zero_of_forall {xs : Array α} (h : ∀ x ∈ xs, ¬ p x) : xs.countP p = 0 :=
-  countP_eq_zero.mpr h
-
-/-- This lemma is only relevant for `grind`. -/
-theorem _root_.Std.Internal.Array.not_of_countP_eq_zero_of_mem {xs : Array α} (h : xs.countP p = 0) (h' : x ∈ xs) : ¬ p x :=
-   countP_eq_zero.mp h _ h'
-
-grind_pattern Std.Internal.Array.not_of_countP_eq_zero_of_mem => xs.countP p, x ∈ xs where
-  guard xs.countP p = 0
-
 @[simp] theorem countP_eq_size {p} : countP p xs = xs.size ↔ ∀ a ∈ xs, p a := by
   rcases xs with ⟨xs⟩
   simp
@@ -117,6 +100,9 @@ grind_pattern Std.Internal.Array.not_of_countP_eq_zero_of_mem => xs.countP p, x
 theorem countP_replicate {a : α} {n : Nat} : countP p (replicate n a) = if p a then n else 0 := by
   simp [← List.toArray_replicate, List.countP_replicate]
 
+@[deprecated countP_replicate (since := "2025-03-18")]
+abbrev countP_mkArray := @countP_replicate
+
 theorem boole_getElem_le_countP {xs : Array α} {i : Nat} (h : i < xs.size) :
     (if p xs[i] then 1 else 0) ≤ xs.countP p := by
   rcases xs with ⟨xs⟩
@@ -277,9 +263,15 @@ theorem count_eq_size {xs : Array α} : count a xs = xs.size ↔ ∀ b ∈ xs, a
 @[simp] theorem count_replicate_self {a : α} {n : Nat} : count a (replicate n a) = n := by
   simp [← List.toArray_replicate]
 
+@[deprecated count_replicate_self (since := "2025-03-18")]
+abbrev count_mkArray_self := @count_replicate_self
+
 theorem count_replicate {a b : α} {n : Nat} : count a (replicate n b) = if b == a then n else 0 := by
   simp [← List.toArray_replicate, List.count_replicate]
 
+@[deprecated count_replicate (since := "2025-03-18")]
+abbrev count_mkArray := @count_replicate
+
 theorem filter_beq {xs : Array α} (a : α) : xs.filter (· == a) = replicate (count a xs) a := by
   rcases xs with ⟨xs⟩
   simp [List.filter_beq]
@@ -293,6 +285,9 @@ theorem replicate_count_eq_of_count_eq_size {xs : Array α} (h : count a xs = xs
   rw [← toList_inj]
   simp [List.replicate_count_eq_of_count_eq_length (by simpa using h)]
 
+@[deprecated replicate_count_eq_of_count_eq_size (since := "2025-03-18")]
+abbrev mkArray_count_eq_of_count_eq_size := @replicate_count_eq_of_count_eq_size
+
 @[simp] theorem count_filter {xs : Array α} (h : p a) : count a (filter p xs) = count a xs := by
   rcases xs with ⟨xs⟩
   simp [List.count_filter, h]

src/Init/Data/Array/DecidableEq.lean

@@ -6,15 +6,11 @@ Authors: Leonardo de Moura
 module
 
 prelude
-import all Init.Data.Array.Basic
 public import Init.Data.Array.Basic
-public import Init.Data.Nat.Lemmas
-import Init.ByCases
-import Init.Classical
-import Init.Data.BEq
-import Init.Data.Bool
-import Init.Data.List.Nat.BEq
-import Init.RCases
+import all Init.Data.Array.Basic
+public import Init.Data.BEq
+public import Init.Data.List.Nat.BEq
+public import Init.ByCases
 
 public section
 
@@ -95,57 +91,11 @@ theorem isEqv_self_beq [BEq α] [ReflBEq α] (xs : Array α) : Array.isEqv xs xs
 theorem isEqv_self [DecidableEq α] (xs : Array α) : Array.isEqv xs xs (· = ·) = true := by
   simp [isEqv, isEqvAux_self]
 
-def instDecidableEqImpl [DecidableEq α] : DecidableEq (Array α) := fun xs ys =>
-  match h:isEqv xs ys (fun a b => a = b) with
-  | true  => isTrue (eq_of_isEqv xs ys h)
-  | false => isFalse (by subst ·; rw [isEqv_self] at h; contradiction)
-
-instance instDecidableEq [DecidableEq α] : DecidableEq (Array α) := fun xs ys =>
-  match xs with
-  | ⟨[]⟩ =>
-    match ys with
-    | ⟨[]⟩ => isTrue rfl
-    | ⟨_ :: _⟩ => isFalse (fun h => Array.noConfusion rfl (heq_of_eq h) (fun h => List.noConfusion rfl h))
-  | ⟨a :: as⟩ =>
-    match ys with
-    | ⟨[]⟩ => isFalse (fun h => Array.noConfusion rfl (heq_of_eq h) (fun h => List.noConfusion rfl h))
-    | ⟨b :: bs⟩ => instDecidableEqImpl ⟨a :: as⟩ ⟨b :: bs⟩
-
-@[csimp]
-theorem instDecidableEq_csimp : @instDecidableEq = @instDecidableEqImpl :=
-  Subsingleton.allEq _ _
-
-/--
-Equality with `#[]` is decidable even if the underlying type does not have decidable equality.
--/
-instance instDecidableEqEmp (xs : Array α) : Decidable (xs = #[]) :=
-  match xs with
-  | ⟨[]⟩ => isTrue rfl
-  | ⟨_ :: _⟩ => isFalse (fun h => Array.noConfusion rfl (heq_of_eq h) (fun h => List.noConfusion rfl h))
-
-/--
-Equality with `#[]` is decidable even if the underlying type does not have decidable equality.
--/
-instance instDecidableEmpEq (ys : Array α) : Decidable (#[] = ys) :=
-  match ys with
-  | ⟨[]⟩ => isTrue rfl
-  | ⟨_ :: _⟩ => isFalse (fun h => Array.noConfusion rfl (heq_of_eq h) (fun h => List.noConfusion rfl h))
-
-@[inline]
-def instDecidableEqEmpImpl (xs : Array α) : Decidable (xs = #[]) :=
-  decidable_of_iff xs.isEmpty <| by rcases xs with ⟨⟨⟩⟩ <;> simp [Array.isEmpty]
-
-@[inline]
-def instDecidableEmpEqImpl (xs : Array α) : Decidable (#[] = xs) :=
-  decidable_of_iff xs.isEmpty <| by rcases xs with ⟨⟨⟩⟩ <;> simp [Array.isEmpty]
-
-@[csimp]
-theorem instDecidableEqEmp_csimp : @instDecidableEqEmp = @instDecidableEqEmpImpl :=
-  Subsingleton.allEq _ _
-
-@[csimp]
-theorem instDecidableEmpEq_csimp : @instDecidableEmpEq = @instDecidableEmpEqImpl :=
-  Subsingleton.allEq _ _
+instance [DecidableEq α] : DecidableEq (Array α) :=
+  fun xs ys =>
+    match h:isEqv xs ys (fun a b => a = b) with
+    | true  => isTrue (eq_of_isEqv xs ys h)
+    | false => isFalse fun h' => by subst h'; rw [isEqv_self] at h; contradiction
 
 theorem beq_eq_decide [BEq α] (xs ys : Array α) :
     (xs == ys) = if h : xs.size = ys.size then