Workflow: Publish Data

.github/workflows/publish-data.sh

@@ -0,0 +1,59 @@
+# See publish-data.yml
+
+TMP=pub/tmp
+mkdir -p $TMP
+
+COPY_YEAR=$(date -u +%Y)
+PUB_DATE=$(date -u +%Y-%m-%d)
+
+cat > $TMP/sed-readmes.txt << eof
+s/COPY_YEAR/$COPY_YEAR/
+s/PUB_DATE/$PUB_DATE/
+s/PUB_STATUS/draft/
+s/UNI_VER/$UNI_VER/
+s/EMOJI_VER/$EMOJI_VER/
+s/TR10_REV/$TR10_REV/
+s%PUBLIC_EMOJI%Public/draft/emoji%
+s%PUBLIC_UCD%Public/draft/UCD%
+eof
+
+mkdir $TMP/UCD
+cp -R unicodetools/data/ucd/dev $TMP/UCD/ucd
+mv $TMP/UCD/ucd/version-ReadMe.txt $TMP/UCD/ReadMe.txt
+rm -r $TMP/UCD/ucd/Unihan
+
+if [[ "$RELEASE_PHASE" == "Dev" ]]; then
+    rm -r $TMP/UCD/ucd/emoji
+fi
+
+if [[ "$RELEASE_PHASE" == "Alpha" || "$RELEASE_PHASE" == "Beta" ]]; then
+    cp -R unicodetools/data/emoji/dev $TMP/emoji
+
+    cp -R unicodetools/data/idna/dev $TMP/idna
+
+    mkdir $TMP/idna2008derived
+    cp unicodetools/data/idna/idna2008derived/ReadMe.txt $TMP/idna2008derived
+    cp unicodetools/data/idna/idna2008derived/Idna2008-$UNI_VER.txt $TMP/idna2008derived
+fi
+
+if [[ "$RELEASE_PHASE" == "Beta" ]]; then
+    cp -R unicodetools/data/uca/dev $TMP/UCA
+    sed -i -f $TMP/sed-readmes.txt $TMP/UCA/CollationTest.html
+
+    cp -R unicodetools/data/security/dev $TMP/security
+fi
+
+# Update the readmes in-place (-i) as set up above.
+find $TMP -name '*ReadMe.txt' | xargs sed -i -f $TMP/sed-readmes.txt
+rm $TMP/sed-readmes.txt
+
+mkdir $TMP/zipped
+mv $TMP/UCD/ucd/zipped-ReadMe.txt $TMP/zipped/ReadMe.txt
+(cd $TMP/UCD/ucd; zip -r UCD.zip *)
+mv $TMP/UCD/ucd/UCD.zip $TMP/zipped
+
+if [[ "$RELEASE_PHASE" == "Beta" ]]; then
+    (cd $TMP/UCA; zip -r CollationTest.zip CollationTest; rm -r CollationTest)
+
+    (cd $TMP/security; zip -r uts39-data-$UNI_VER.zip *)
+fi

.github/workflows/publish-data.yml

@@ -0,0 +1,43 @@
+# See https://github.com/unicode-org/unicodetools/blob/main/docs/data-workflow.md#publication
+
+# Test locally with https://github.com/nektos/act:
+#   act --workflows .github/workflows/publish-data.yml --input releasePhase=Alpha
+
+name: Publish Data
+
+run-name: "${{ github.workflow }}: ${{ inputs.releasePhase }}"
+
+on:
+  workflow_dispatch:
+    inputs:
+      releasePhase: # See ReleasePhase in https://github.com/unicode-org/unicodetools/blob/main/unicodetools/src/main/java/org/unicode/text/utility/Settings.java
+        description: Release phase
+        type: choice
+        options:
+          - Dev
+          - Alpha
+          - Beta
+        default: Dev
+
+env:
+  UNI_VER: "17.0.0"
+  EMOJI_VER: "17.0"
+  TR10_REV: "tr10-52" # UTS #10 release revision number to be used in CollationTest.html: One more than the last release revision number.
+  RELEASE_PHASE: ${{ inputs.releasePhase }}
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          sparse-checkout: |
+            .github/workflows
+            unicodetools/data/ucd/dev
+            unicodetools/data/emoji/dev
+            unicodetools/data/idna/dev
+            unicodetools/data/idna/idna2008derived
+      - run: .github/workflows/publish-data.sh
+      - uses: actions/upload-artifact@v4
+        with:
+          path: pub/tmp # See TMP in publish-data.sh

docs/data-workflow.md

@@ -82,6 +82,9 @@ https://github.com/unicode-org/unicodetools/tree/main/unicodetools/data/emoji/de
 
 ## Publication
 
+> An experimental GitHub workflow, [publish-data.yml](/.github/workflows/publish-data.yml),
+can create a dev (UCD), alpha, or beta snapshot.
+
 Certain snapshots of the .../dev/ files are copied into https://www.unicode.org/Public/draft/
 for Unicode alpha, beta, and final releases, and more as appropriate.
 *   UCD files go into https://www.unicode.org/Public/draft/ucd/