Merge photo highlights by name and capture year

photo-upload/highlights.py

@@ -0,0 +1,56 @@
+"""Keep one gallery highlight link per name and capture year."""
+
+from datetime import datetime, timezone
+import json
+from pathlib import Path
+import re
+
+
+def field(content, name):
+    match = re.search(rf"^{re.escape(name)}: (.+)$", content.split("---", 2)[1], re.MULTILINE)
+    if not match:
+        return None
+    value = match.group(1).strip()
+    return json.loads(value) if value.startswith('"') else value
+
+
+def normalize_highlights(art):
+    """Place each link on the newest photo in its named, dated group."""
+    groups = {}
+    for path in Path(art).glob("*.md"):
+        content = path.read_text()
+        name = field(content, "highlight_name") or field(content, "anchor")
+        date = field(content, "date")
+        if not name or not date:
+            continue
+        taken = datetime.fromisoformat(str(date))
+        if taken.tzinfo is None:
+            taken = taken.replace(tzinfo=timezone.utc)
+        groups.setdefault((name, taken.year), []).append((taken, path, content))
+
+    changed = []
+    for (name, _), entries in groups.items():
+        # An old single anchor is already the only link; a newer named photo
+        # moves the link without changing any photo's capture date or position.
+        newest = max(entries, key=lambda entry: (entry[0], entry[1].name))[1]
+        if len(entries) == 1 and field(entries[0][2], "anchor") == name:
+            continue
+        for _, path, content in entries:
+            old = content
+            had_anchor = field(content, "anchor") is not None
+            if (path == newest and had_anchor and
+                    all(field(other[2], "anchor") is None for other in entries if other[1] != path)):
+                continue
+            if had_anchor:
+                content = re.sub(r"^anchor: .+\n", "", content, count=1, flags=re.MULTILINE)
+            if path == newest:
+                content = content.replace("categories: art\n", "categories: art\n" +
+                                          f"anchor: {json.dumps(name, ensure_ascii=False)}\n", 1)
+            elif had_anchor:
+                title = field(content, "title")
+                if title and title.startswith("photo-") and f'id="{title}"' not in content:
+                    content = content.replace("---\n\n", f'---\n\n<div id="{title}"></div>\n\n', 1)
+            if content != old:
+                path.write_text(content)
+                changed.append(path)
+    return changed