fix(CI): replace jq with Python script

This is temporary fix.

1 files changed, 50 insertions, 25 deletions

diff --git a/.forgejo/workflows/fetch-webmentions.yml b/.forgejo/workflows/fetch-webmentions.yml
index 3128f0e1..e29debcc 100644
--- a/.forgejo/workflows/fetch-webmentions.yml
+++ b/.forgejo/workflows/fetch-webmentions.yml
@@ -12,9 +12,6 @@ jobs:
     steps:
       - uses: actions/checkout@v4
 
-      - name: Install jq
-        run: apk add --no-cache jq
-
       - name: Fetch webmentions
         env:
           TOKEN: ${{ secrets.WEBMENTION_IO_TOKEN }}
@@ -22,28 +19,56 @@ jobs:
         run: |
           mkdir -p data
 
-          curl -s "https://webmention.io/api/mentions.jf2?domain=$DOMAIN&token=$TOKEN" \
-          | jq '
-              .children
-              | map(
-                  .["wm-target"] |= sub("^https?://[^/]+";"")
-              )
-              | sort_by(."wm-target")
-              | group_by(."wm-target")
-              | map({
-                  (.[0]."wm-target"):
-                    map({
-                      type: ."wm-property",
-                      author: .author,
-                      url: .url,
-                      source: ."wm-source",
-                      published: .published,
-                      received: ."wm-received",
-                      content: (if has("content") then .content else null end)
-                    })
-                })
-              | add
-            ' > data/webmentions.json
+          curl -s "https://webmention.io/api/mentions.jf2?domain=$DOMAIN&token=$TOKEN" > data/webmentions_raw.json
+
+          cat > /tmp/process_webmentions.py << 'PYTHON_SCRIPT'
+          import json
+          import re
+          import sys
+
+          with open('data/webmentions_raw.json', 'r') as f:
+              data = json.load(f)
+
+          children = data.get('children', [])
+
+          # Process children
+          processed = []
+          for child in children:
+              # Remove domain from wm-target
+              wm_target = child.get('wm-target', '')
+              wm_target = re.sub(r'^https?://[^/]+', '', wm_target)
+              child['wm-target'] = wm_target
+              processed.append(child)
+
+          # Sort by wm-target
+          processed.sort(key=lambda x: x.get('wm-target', ''))
+
+          # Group by wm-target
+          from itertools import groupby
+          result = {}
+          for target, group in groupby(processed, key=lambda x: x.get('wm-target', '')):
+              items = []
+              for item in group:
+                  processed_item = {
+                      'type': item.get('wm-property'),
+                      'author': item.get('author'),
+                      'url': item.get('url'),
+                      'source': item.get('wm-source'),
+                      'published': item.get('published'),
+                      'received': item.get('wm-received'),
+                      'content': item.get('content') if 'content' in item else None
+                  }
+                  items.append(processed_item)
+              result[target] = items
+
+          with open('data/webmentions.json', 'w') as f:
+              json.dump(result, f, indent=2)
+
+          import os
+          os.remove('data/webmentions_raw.json')
+          PYTHON_SCRIPT
+
+          python3 /tmp/process_webmentions.py
 
       - name: Commit if changed
         run: |