chore: update GBK encoding table with corrected character mappings (#371)

Author: bjohansebas

.github/workflows/generator.yml

@@ -0,0 +1,35 @@
+name: Update Encodings
+
+on:
+    workflow_dispatch:
+    # schedule:
+    #     - cron:  '0 8 * * *'  # Uncomment this when it is in the master branch.
+
+permissions:
+    pull-requests: write
+    contents: write
+
+jobs:
+    update-encodings:
+        runs-on: ubuntu-latest
+        steps:
+            - uses: actions/checkout@v6
+
+            - name: Set up Node.js
+              uses: actions/setup-node@v6
+
+            - name: Run scripts
+              run: npm run generate
+
+            - name: Create Pull Request
+              uses: gr2m/create-or-update-pull-request-action@v1
+              env:
+                GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+              with: 
+                commit-message: 'chore: update encodings'
+                title: 'chore: update encodings'
+                body: >
+                  This auto-generated PR updates encodings to the latest version. 
+                  
+                  cc: @pillarjs/iconv-lite-collaborators 
+                branch: bot/update-encodings

encodings/tables/gbk-added.json

@@ -18,9 +18,9 @@
 ["a640","",62],
 ["a680","",32],
 ["a6b9","",7],
-["a6d9","",6],
-["a6ec",""],
-["a6f3",""],
+["a6d9","︐︒︑︓︔︕︖"],
+["a6ec","︗︘"],
+["a6f3","︙"],
 ["a6f6","",8],
 ["a740","",62],
 ["a780","",32],
@@ -50,7 +50,7 @@
 ["fba1","",93],
 ["fca1","",93],
 ["fda1","",93],
-["fe50","⺁⺄㑳㑇⺈⺋㖞㘚㘎⺌⺗㥮㤘㧏㧟㩳㧐㭎㱮㳠⺧⺪䁖䅟⺮䌷⺳⺶⺷䎱䎬⺻䏝䓖䙡䙌"],
-["fe80","䜣䜩䝼䞍⻊䥇䥺䥽䦂䦃䦅䦆䦟䦛䦷䦶䲣䲟䲠䲡䱷䲢䴓",6,"䶮",93],
+["fe50","⺁⺄㑳㑇⺈⺋龴㖞㘚㘎⺌⺗㥮㤘龵㧏㧟㩳㧐龶龷㭎㱮㳠⺧龸⺪䁖䅟⺮䌷⺳⺶⺷䎱䎬⺻䏝䓖䙡䙌龹"],
+["fe80","䜣䜩䝼䞍⻊䥇䥺䥽䦂䦃䦅䦆䦟䦛䦷䦶龺䲣䲟䲠䲡䱷䲢䴓",6,"䶮龻",93],
 ["8135f437",""]
 ]

generation/gen-dbcs.js

@@ -99,7 +99,7 @@ async.parallel({
     }
 
     // GB18030:2005 addition
-    gbk2005add = [['8135f437', '']];
+    var gbk2005add = [['8135f437', '']];
 
     utils.writeTable("gbk-added", utils.generateTable(gbkadd).concat(gbk2005add));
 

generation/package.json

@@ -2,6 +2,11 @@
     "name": "iconv-lite-generators",
     "private": true,
     "version": "1.0.0",
+    "scripts": {
+        "generate": "npm run generate:dbcs && npm run generate:sbcs",
+        "generate:dbcs": "node gen-dbcs.js",
+        "generate:sbcs": "node gen-sbcs.js"
+    },
     "devDependencies": {
         "async": "^3.2.0",
         "errto": "^0.2.1",

package.json

@@ -28,6 +28,7 @@
     "scripts": {
         "lint": "eslint",
         "lint:fix": "eslint --fix",
+        "generate": "npm run generate --workspaces --if-present",
         "test": "mocha --reporter spec --check-leaks --grep .",
         "test:ci": "nyc --exclude test --reporter=lcovonly --reporter=text npm test",
         "test:cov": "nyc --exclude test --reporter=html --reporter=text npm test",
@@ -57,5 +58,8 @@
         "unorm": "^1.6.0",
         "typescript": "~5.9.3"
     },
-    "dependencies": {}
+    "dependencies": {},
+    "workspaces": [
+        "generation"
+    ]
 }

test/dbcs-test.js

@@ -501,6 +501,24 @@ var iconvCannotDecode = { // Characters that we can decode, but iconv cannot. En
     fe9d: "䴘",
     fe9e: "䴙",
     fe9f: "䶮",
+    a6d9: "︐",
+    a6da: "︒",
+    a6db: "︑",
+    a6dc: "︓",
+    a6dd: "︔",
+    a6de: "︕",
+    a6df: "︖",
+    a6ec: "︗",
+    a6ed: "︘",
+    a6f3: "︙",
+    fe59: "龴",
+    fe61: "龵",
+    fe66: "龶",
+    fe67: "龷",
+    fe6d: "龸",
+    fe7e: "龹",
+    fe90: "龺",
+    fea0: "龻",
 
     // iconv and ICU are mapping "a3 a0" -> U+E5E5. However, WebKit/Chrome maps it to U+3000 noting compatibility with older websites.
     // Encoding Standard stands on the side of WebKit, so we are too.