Automatically update unicode data once a month.

.github/workflows/update-data.yml

@@ -0,0 +1,37 @@
+name: Update data
+
+on:
+  workflow_dispatch:
+  schedule:
+    - cron: "0 0 1 * *" # run once a month
+
+jobs:
+  resources:
+    name: Update resources
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v3
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.8'
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install -r requirements-data.txt
+
+      - name: Update data
+        run: python scripts/update_data.py
+
+      - name: Commit data
+        uses: test-room-7/action-update-file@v1
+        with:
+          file-path: |
+            fontbro/data/unicode-blocks.json
+            fontbro/data/unicode-scripts.json
+          commit-msg: "Update `unicode-blocks.json` and `unicode-scripts.json` data."
+          committer-name: "Fabio Caccamo [bot]"
+          committer-email: "[email protected]"
+          github-token: ${{ secrets.WORKFLOWS_UPDATE_DATA_TOKEN }}

requirements-data.txt

@@ -0,0 +1,4 @@
+fonttools[woff,unicode,pathops] == 4.38.0
+imagehash == 4.3.1
+pillow == 9.4.0
+python-fsutil == 0.9.3

scripts/update_data.py

@@ -0,0 +1,50 @@
+import fsutil
+from fontTools import unicodedata
+
+from fontbro import Font
+
+
+def _write_data_json(filepath, data):
+    data_filepath = fsutil.join_path(__file__, filepath)
+    fsutil.write_file_json(data_filepath, data, indent=4, sort_keys=True)
+
+
+def update_unicode_data():
+    blocks = []
+    blocks_cache = {}
+    scripts = []
+    scripts_cache = {}
+    for code in range(0, 0x110000):
+        block_name = unicodedata.block(code)
+        if block_name == "No_Block":
+            continue
+        block = {
+            "name": block_name,
+        }
+        Font._populate_unicode_items_set(blocks, blocks_cache, block)
+        script_tag = unicodedata.script(code)
+        script_name = unicodedata.script_name(script_tag)
+        script = {
+            "name": script_name,
+            "tag": script_tag,
+        }
+        Font._populate_unicode_items_set(scripts, scripts_cache, script)
+
+    for block in blocks:
+        block["characters_total"] = block["characters_count"]
+        block.pop("characters_count", None)
+
+    for script in scripts:
+        script["characters_total"] = script["characters_count"]
+        script.pop("characters_count", None)
+
+    _write_data_json(filepath="../fontbro/data/unicode-blocks.json", data=blocks)
+    _write_data_json(filepath="../fontbro/data/unicode-scripts.json", data=scripts)
+
+
+def main():
+    update_unicode_data()
+
+
+if __name__ == "__main__":
+    main()

tests/test_update_unicode_data.py

@@ -1,9 +1,4 @@
-import json
-
-import fsutil
-from fontTools import unicodedata
-
-from fontbro import Font
+from scripts.update_data import update_unicode_data
 from tests import AbstractTestCase
 
 
@@ -13,40 +8,4 @@ class UpdateUnicodeDataTestCase(AbstractTestCase):
     """
 
     def test_update_unicode_data(self):
-        blocks = []
-        blocks_cache = {}
-        scripts = []
-        scripts_cache = {}
-        for code in range(0, 0x110000):
-            block_name = unicodedata.block(code)
-            if block_name == "No_Block":
-                continue
-            block = {
-                "name": block_name,
-            }
-            Font._populate_unicode_items_set(blocks, blocks_cache, block)
-            script_tag = unicodedata.script(code)
-            script_name = unicodedata.script_name(script_tag)
-            script = {
-                "name": script_name,
-                "tag": script_tag,
-            }
-            Font._populate_unicode_items_set(scripts, scripts_cache, script)
-
-        for block in blocks:
-            block["characters_total"] = block["characters_count"]
-            block.pop("characters_count", None)
-
-        for script in scripts:
-            script["characters_total"] = script["characters_count"]
-            script.pop("characters_count", None)
-
-        data_dir = fsutil.join_path(__file__, "../fontbro/data/")
-        fsutil.write_file(
-            fsutil.join_path(data_dir, "unicode-blocks.json"),
-            json.dumps(blocks, sort_keys=True, indent=4),
-        )
-        fsutil.write_file(
-            fsutil.join_path(data_dir, "unicode-scripts.json"),
-            json.dumps(scripts, sort_keys=True, indent=4),
-        )
+        update_unicode_data()