tools: automate icu-small update

marco-ippolito · targos · commit f6ff318d4c43 · 2023-05-12T11:57:20.000+02:00
PR-URL: #47727 Refs: nodejs/security-wg#828 Reviewed-By: Steven R Loomis <srl295@gmail.com> Reviewed-By: Rafael Gonzaga <rafael.nunu@hotmail.com>
diff --git a/.github/workflows/tools.yml b/.github/workflows/tools.yml
@@ -24,6 +24,7 @@ on:
           - doc
           - eslint
           - googletest
+          - icu
           - libuv
           - lint-md-dependencies
           - llhttp
@@ -36,6 +37,9 @@ on:
           - undici
           - uvwasi
 
+env:
+  PYTHON_VERSION: '3.11'
+
 permissions:
   contents: read
 
@@ -252,11 +256,24 @@ jobs:
               cat temp-output
               tail -n1 temp-output | grep "NEW_VERSION=" >> "$GITHUB_ENV" || true
               rm temp-output
+          - id: icu
+            subsystem: deps
+            label: dependencies, test
+            run: |
+              ./tools/dep_updaters/update-icu.sh > temp-output
+              cat temp-output
+              tail -n1 temp-output | grep "NEW_VERSION=" >> "$GITHUB_ENV" || true
+              rm temp-output
     steps:
       - uses: actions/checkout@8e5e7e5ab8b370d6c329ec480221332ada57f0ab  # v3.5.2
         if: github.event_name == 'schedule' || inputs.id == 'all' || inputs.id == matrix.id
         with:
           persist-credentials: false
+      - name: Set up Python ${{ env.PYTHON_VERSION }}
+        if: matrix.id == 'icu' && (github.event_name == 'schedule' || inputs.id == 'all' || inputs.id == matrix.id)
+        uses: actions/setup-python@d27e3f3d7c64b4bbf8e4abfb9b63b83e846e0435  # v4.5.0
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
       - run: ${{ matrix.run }}
         if: github.event_name == 'schedule' || inputs.id == 'all' || inputs.id == matrix.id
         env:
diff --git a/doc/contributing/maintaining/maintaining-icu.md b/doc/contributing/maintaining/maintaining-icu.md
@@ -101,6 +101,9 @@ Node.js is built.
 
 ## How to upgrade ICU
 
+> The script `tools/dep_updaters/update-icu.sh` automates
+> this process.
+
 * Make sure your Node.js workspace is clean (`git status`
   should be sufficient).
 * Configure Node.js with the specific [ICU version](http://site.icu-project.org/download)
diff --git a/tools/dep_updaters/update-icu.sh b/tools/dep_updaters/update-icu.sh
@@ -0,0 +1,75 @@
+#!/bin/sh
+set -e
+# Shell script to update icu in the source tree to a specific version
+
+BASE_DIR=$(cd "$(dirname "$0")/../.." && pwd)
+DEPS_DIR="$BASE_DIR/deps"
+TOOLS_DIR="$BASE_DIR/tools"
+
+[ -z "$NODE" ] && NODE="$BASE_DIR/out/Release/node"
+[ -x "$NODE" ] || NODE=$(command -v node)
+
+NEW_VERSION="$("$NODE" --input-type=module <<'EOF'
+const res = await fetch('https://api.github.com/repos/unicode-org/icu/releases/latest');
+if (!res.ok) throw new Error(`FetchError: ${res.status} ${res.statusText}`, { cause: res });
+const { tag_name } = await res.json();
+console.log(tag_name.replace('release-', '').replace('-','.'));
+EOF
+)"
+
+ICU_VERSION_H="$DEPS_DIR/icu-small/source/common/unicode/uvernum.h"
+
+CURRENT_VERSION="$(grep "#define U_ICU_VERSION " "$ICU_VERSION_H" | cut -d'"' -f2)"
+
+echo "Comparing $NEW_VERSION with $CURRENT_VERSION"
+
+if [ "$NEW_VERSION" = "$CURRENT_VERSION" ]; then
+  echo "Skipped because icu is on the latest version."
+  exit 0
+fi
+
+DASHED_NEW_VERSION=$(echo "$NEW_VERSION" | sed 's/\./-/g')
+
+LOW_DASHED_NEW_VERSION=$(echo "$NEW_VERSION" | sed 's/\./_/g')
+
+NEW_VERSION_TGZ="icu4c-${LOW_DASHED_NEW_VERSION}-src.tgz"
+
+NEW_VERSION_TGZ_URL="https://github.com/unicode-org/icu/releases/download/release-${DASHED_NEW_VERSION}/$NEW_VERSION_TGZ"
+
+NEW_VERSION_MD5="https://github.com/unicode-org/icu/releases/download/release-${DASHED_NEW_VERSION}/icu4c-${LOW_DASHED_NEW_VERSION}-src.md5"
+
+./configure --with-intl=full-icu --with-icu-source="$NEW_VERSION_TGZ_URL"
+
+"$TOOLS_DIR/icu/shrink-icu-src.py"
+
+rm -rf "$DEPS_DIR/icu"
+
+CHECKSUM=$(curl -sL "$NEW_VERSION_MD5" | grep "$NEW_VERSION_TGZ" | grep -v "\.asc$" | awk '{print $1}')
+
+GENERATED_CHECKSUM=$( curl -sL "$NEW_VERSION_TGZ_URL" | md5sum | cut -d ' ' -f1)
+
+echo "Comparing checksums: deposited $CHECKSUM with $GENERATED_CHECKSUM"
+
+if [ "$CHECKSUM" != "$GENERATED_CHECKSUM" ]; then
+  echo "Skipped because checksums do not match."
+  exit 0
+fi
+
+sed -i '' -e "s|\"url\": \"\(.*\)\".*|\"url\": \"$NEW_VERSION_TGZ_URL\",|" "$TOOLS_DIR/icu/current_ver.dep"
+
+sed -i '' -e "s|\"md5\": \"\(.*\)\".*|\"md5\": \"$CHECKSUM\"|" "$TOOLS_DIR/icu/current_ver.dep" 
+
+rm -rf out "$DEPS_DIR/icu" "$DEPS_DIR/icu4c*"
+
+echo "All done!"
+echo ""
+echo "Please git add icu, commit the new version:"
+echo ""
+echo "$ git add -A deps/icu-small"
+echo "$ git add tools/icu/current_ver.dep"
+echo "$ git commit -m \"deps: update icu to $NEW_VERSION\""
+echo ""
+
+# The last line of the script should always print the new version,
+# as we need to add it to $GITHUB_ENV variable.
+echo "NEW_VERSION=$NEW_VERSION"
diff --git a/tools/icu/shrink-icu-src.py b/tools/icu/shrink-icu-src.py
