Run crate-ci/typos in CI (#51704)

Co-authored-by: Dilum Aluthge <[email protected]>

Author: LilithHafner

.github/workflows/Typos.yml

@@ -0,0 +1,56 @@
+name: Typos
+
+permissions: {}
+
+on: [pull_request]
+
+jobs:
+  typos-check:
+    name: Check for new typos
+    runs-on: ubuntu-latest
+    timeout-minutes: 5
+    steps:
+      - name: Checkout the JuliaLang/julia repository
+        uses: actions/checkout@v4
+        with:
+          persist-credentials: false
+      - name: Check spelling with typos
+        #uses: crate-ci/typos@master
+        env:
+          GH_TOKEN: "${{ github.token }}"
+        run: |
+          git fetch --depth=1 origin ${{ github.base_ref }}
+          OLD_FILES=$(git diff-index --name-only --diff-filter=ad FETCH_HEAD)
+          NEW_FILES=$(git diff-index --name-only --diff-filter=d FETCH_HEAD)
+
+          mkdir -p "${{ runner.temp }}/typos"
+          RELEASE_ASSET_URL="$(
+            gh api /repos/crate-ci/typos/releases/latest \
+            --jq '."assets"[] | select(."name" | test("^typos-.+-x86_64-unknown-linux-musl\\.tar\\.gz$")) | ."browser_download_url"'
+          )"
+          wget --secure-protocol=TLSv1_3 --max-redirect=1 --retry-on-host-error --retry-connrefused --tries=3 \
+            --quiet --output-document=- "${RELEASE_ASSET_URL}" \
+            | tar -xz -C "${{ runner.temp }}/typos" ./typos
+          "${{ runner.temp }}/typos/typos" --version
+
+          echo -n $NEW_FILES | xargs "${{ runner.temp }}/typos/typos" --format json >> ${{ runner.temp }}/new_typos.jsonl || true
+          git checkout FETCH_HEAD -- $OLD_FILES
+          echo -n $OLD_FILES | xargs "${{ runner.temp }}/typos/typos" --format json >> ${{ runner.temp }}/old_typos.jsonl  || true
+
+          python -c '
+          import sys, json
+          old = set()
+          with open(sys.argv[1]) as old_file:
+            for line in old_file:
+              old.add(json.loads(line)["typo"])
+          clean = True
+          with open(sys.argv[2]) as new_file:
+            for line in new_file:
+              new = json.loads(line)
+              if new["typo"] not in old:
+                if len(new["typo"]) > 6: # Short typos might be false positives. Long are probably real.
+                  clean = False
+                print("::warning file={},line={},col={}::perhaps \"{}\" should be \"{}\".".format(
+                  new["path"], new["line_num"], new["byte_offset"],
+                  new["typo"], " or ".join(new["corrections"])))
+          sys.exit(1 if not clean else 0)' "${{ runner.temp }}/old_typos.jsonl" "${{ runner.temp }}/new_typos.jsonl"

NEWS.md

@@ -125,5 +125,8 @@ External dependencies
 Tooling Improvements
 --------------------
 
+* CI now performs limited automatic typo detection on all PRs. If you merge a PR with a
+  failing typo CI check, then the reported typos will be automatically ignored in future CI
+  runs on PRs that edit those same files. ([#51704])
 
 <!--- generated by NEWS-update.jl: -->

typos.toml

@@ -0,0 +1,2 @@
+[default]
+extend-ignore-words-re = ["^[a-zA-Z]?[a-zA-Z]?[a-zA-Z]?[a-zA-Z]?$"]