Refactor: Rewrite in Python with headers sorting and scheme formatting

.github/workflows/sort-format-json.yml

@@ -1,4 +1,4 @@
-name: Sort and Format wmn-data.json
+name: Sort and Format JSON Files
 
 on:
   push:
@@ -7,88 +7,25 @@ on:
       - 'wmn-data-schema.json'
 
 jobs:
-  sort-json:
+  sort-and-format-json:
+    name: Sort and Format JSON Files
     runs-on: ubuntu-latest
 
     steps:
       - uses: actions/checkout@v3
 
-      - name: Set up Node.js
-        uses: actions/setup-node@v3
+      - name: Set up Python
+        uses: actions/setup-python@v5
         with:
-          node-version: '20'
+          python-version: '3.11'
 
-      - name: Create sorting script
-        run: |
-          cat << 'EOF' > sort-wmn-data.js
-          const fs = require('fs');
-
-          const dataPath = 'wmn-data.json';
-          const schemaPath = 'wmn-data-schema.json';
-
-          const original = fs.readFileSync(dataPath, 'utf8');
-          const data = JSON.parse(original);
-          const schema = JSON.parse(fs.readFileSync(schemaPath, 'utf8'));
-
-          // ---------- Sorting Helpers ----------
-          function sortArrayAlphabetically(arr) {
-            return arr.sort((a, b) => a.localeCompare(b));
-          }
-
-          function reorderObjectKeys(obj, keyOrder) {
-            const reordered = {};
-            keyOrder.forEach(key => {
-              if (key in obj) reordered[key] = obj[key];
-            });
-            Object.keys(obj).forEach(key => {
-              if (!keyOrder.includes(key)) reordered[key] = obj[key];
-            });
-            return reordered;
-          }
-
-          // ---------- Extract Schema Key Order ----------
-          const siteSchema = schema.properties?.sites?.items;
-          const keyOrder = siteSchema?.properties ? Object.keys(siteSchema.properties) : [];
-
-          // ---------- Sort authors and categories ----------
-          if (Array.isArray(data.authors)) {
-            data.authors = sortArrayAlphabetically(data.authors);
-          }
-
-          if (Array.isArray(data.categories)) {
-            data.categories = sortArrayAlphabetically(data.categories);
-          }
-
-          // ---------- Sort and Reorder Sites ----------
-          if (Array.isArray(data.sites)) {
-            data.sites.sort((a, b) => a.name.localeCompare(b.name));
-            data.sites = data.sites.map(site => reorderObjectKeys(site, keyOrder));
-          }
-
-          // Convert to JSON with 2-space indentation
-          const updated = JSON.stringify(data, null, 2);
-
-          // Only write if content has changed
-          if (original !== updated) {
-            fs.writeFileSync(dataPath, updated);
-            console.log('wmn-data.json updated and reordered.');
-          } else {
-            console.log('wmn-data.json is already sorted and formatted.');
-          }
-          EOF
-
-      - name: Install Prettier
-        run: npm install --global prettier
-
-      - name: Run sorting script and format
-        run: |
-          node sort-wmn-data.js
-          prettier --write wmn-data.json
+      - name: Run Python sorting script
+        run: python scripts/sort_wmn_data.py
 
       - name: Commit if changed
         run: |
           git config --global user.name "github-actions"
           git config --global user.email "github-actions@github.com"
-          git add wmn-data.json
-          git diff --cached --quiet || git commit -m "chore: auto-sort wmn-data.json by schema"
+          git add wmn-data.json wmn-data-schema.json
+          git diff --cached --quiet || git commit -m "chore: auto-sort and format JSON files"
           git push

scripts/sort_format_json.py

@@ -0,0 +1,71 @@
+import json
+
+data_path = 'wmn-data.json'
+schema_path = 'wmn-data-schema.json'
+
+def sort_array_alphabetically(arr):
+    return sorted(arr, key=str.lower)
+
+def reorder_object_keys(obj, key_order):
+    reordered = {k: obj[k] for k in key_order if k in obj}
+    for k in obj:
+        if k not in key_order:
+            reordered[k] = obj[k]
+    return reordered
+
+def sort_headers(site):
+    headers = site.get("headers")
+    if isinstance(headers, dict):
+        site["headers"] = dict(sorted(headers.items(), key=lambda item: item[0].lower()))
+
+def load_and_format_json(path):
+    with open(path, 'r', encoding='utf-8') as f:
+        raw_content = f.read()
+        data = json.loads(raw_content)
+    formatted = json.dumps(data, indent=2, ensure_ascii=False)
+    return data, raw_content, formatted
+
+data, data_raw, data_formatted = load_and_format_json(data_path)
+schema, schema_raw, schema_formatted = load_and_format_json(schema_path)
+
+changed = False
+
+# Sort authors and categories
+if isinstance(data.get('authors'), list):
+    data['authors'] = sort_array_alphabetically(data['authors'])
+
+if isinstance(data.get('categories'), list):
+    data['categories'] = sort_array_alphabetically(data['categories'])
+
+# Sort and reorder sites
+site_schema = schema.get('properties', {}).get('sites', {}).get('items', {})
+key_order = list(site_schema.get('properties', {}).keys())
+
+if isinstance(data.get('sites'), list):
+    data['sites'].sort(key=lambda site: site.get('name', '').lower())
+    for site in data['sites']:
+        sort_headers(site)
+    data['sites'] = [reorder_object_keys(site, key_order) for site in data['sites']]
+
+updated_data_formatted = json.dumps(data, indent=2, ensure_ascii=False)
+
+# Write wmn-data.json if changed
+if data_raw.strip() != updated_data_formatted.strip():
+    with open(data_path, 'w', encoding='utf-8') as f:
+        f.write(updated_data_formatted)
+    print("Updated and sorted wmn-data.json.")
+    changed = True
+else:
+    print("wmn-data.json already formatted.")
+
+# Write formatted wmn-data-schema.json if changed
+if schema_raw.strip() != schema_formatted.strip():
+    with open(schema_path, 'w', encoding='utf-8') as f:
+        f.write(schema_formatted)
+    print("Formatted wmn-data-schema.json.")
+    changed = True
+else:
+    print("wmn-data-schema.json already formatted.")
+
+if not changed:
+    print("No changes made.")