mirror of
https://github.com/willmiao/ComfyUI-Lora-Manager.git
synced 2026-03-24 06:32:12 -03:00
fix(autocomplete): improve tag search ranking with popularity-based sorting
- Add LOG10(post_count) weighting to BM25 score for better relevance ranking - Prioritize tag_name prefix matches above alias matches using CASE statement - Remove frontend re-scoring logic to trust backend排序 results - Fix pagination consistency: page N+1 scores <= page N minimum score Key improvements: - '1girl' (6M posts) now ranks #1 instead of #149 for search '1' - tag_name prefix matches always appear before alias matches - Popular tags rank higher than obscure ones with same prefix - Consistent ordering across pagination boundaries Test coverage: - Add test_search_tag_name_prefix_match_priority - Add test_search_ranks_popular_tags_higher - Add test_search_pagination_ordering_consistency - Add test_search_rank_score_includes_popularity_weight - Update test data with 15 tags starting with '1' Fixes issues with autocomplete dropdown showing inconsistent results when scrolling through paginated search results.
This commit is contained in:
Will Miao
@@ -449,6 +449,11 @@ class TagFTSIndex:
|
||||
Supports alias search: if the query matches an alias rather than
|
||||
the tag_name, the result will include a "matched_alias" field.
|
||||
|
||||
Ranking is based on a combination of:
|
||||
1. FTS5 bm25 relevance score (how well the text matches)
|
||||
2. Post count (popularity)
|
||||
3. Exact prefix match boost (tag_name starts with query)
|
||||
|
||||
Args:
|
||||
query: The search query string.
|
||||
categories: Optional list of category IDs to filter by.
|
||||
@@ -457,7 +462,7 @@ class TagFTSIndex:
|
||||
|
||||
Returns:
|
||||
List of dictionaries with tag_name, category, post_count,
|
||||
and optionally matched_alias.
|
||||
rank_score, and optionally matched_alias.
|
||||
"""
|
||||
# Ensure index is ready (lazy initialization)
|
||||
if not self.ensure_ready():
|
||||
@@ -473,35 +478,67 @@ class TagFTSIndex:
|
||||
if not fts_query:
|
||||
return []
|
||||
|
||||
query_lower = query.lower().strip()
|
||||
|
||||
try:
|
||||
with self._lock:
|
||||
conn = self._connect(readonly=True)
|
||||
try:
|
||||
# Build the SQL query - now also fetch aliases for matched_alias detection
|
||||
# Use subquery for category filter to ensure FTS is evaluated first
|
||||
# Build the SQL query with bm25 ranking
|
||||
# FTS5 bm25() returns negative scores, lower is better
|
||||
# We use -bm25() to get higher=better scores
|
||||
# Weights: -100.0 for exact matches, 1.0 for others
|
||||
# Add LOG10(post_count) weighting to boost popular tags
|
||||
# Use CASE to boost tag_name prefix matches above alias matches
|
||||
if categories:
|
||||
placeholders = ",".join("?" * len(categories))
|
||||
sql = f"""
|
||||
SELECT t.tag_name, t.category, t.post_count, t.aliases
|
||||
FROM tags t
|
||||
WHERE t.rowid IN (
|
||||
SELECT rowid FROM tag_fts WHERE searchable_text MATCH ?
|
||||
)
|
||||
SELECT t.tag_name, t.category, t.post_count, t.aliases,
|
||||
CASE
|
||||
WHEN t.tag_name LIKE ? ESCAPE '\\' THEN 1
|
||||
ELSE 0
|
||||
END AS is_tag_name_match,
|
||||
bm25(tag_fts, -100.0, 1.0, 1.0) + LOG10(t.post_count + 1) * 10.0 AS rank_score
|
||||
FROM tag_fts
|
||||
JOIN tags t ON tag_fts.rowid = t.rowid
|
||||
WHERE tag_fts.searchable_text MATCH ?
|
||||
AND t.category IN ({placeholders})
|
||||
ORDER BY t.post_count DESC
|
||||
ORDER BY is_tag_name_match DESC, rank_score DESC
|
||||
LIMIT ? OFFSET ?
|
||||
"""
|
||||
params = [fts_query] + categories + [limit, offset]
|
||||
# Escape special LIKE characters and add wildcard
|
||||
query_escaped = (
|
||||
query_lower.lstrip("/")
|
||||
.replace("\\", "\\\\")
|
||||
.replace("%", "\\%")
|
||||
.replace("_", "\\_")
|
||||
)
|
||||
params = (
|
||||
[query_escaped + "%", fts_query]
|
||||
+ categories
|
||||
+ [limit, offset]
|
||||
)
|
||||
else:
|
||||
sql = """
|
||||
SELECT t.tag_name, t.category, t.post_count, t.aliases
|
||||
FROM tag_fts f
|
||||
JOIN tags t ON f.rowid = t.rowid
|
||||
WHERE f.searchable_text MATCH ?
|
||||
ORDER BY t.post_count DESC
|
||||
SELECT t.tag_name, t.category, t.post_count, t.aliases,
|
||||
CASE
|
||||
WHEN t.tag_name LIKE ? ESCAPE '\\' THEN 1
|
||||
ELSE 0
|
||||
END AS is_tag_name_match,
|
||||
bm25(tag_fts, -100.0, 1.0, 1.0) + LOG10(t.post_count + 1) * 10.0 AS rank_score
|
||||
FROM tag_fts
|
||||
JOIN tags t ON tag_fts.rowid = t.rowid
|
||||
WHERE tag_fts.searchable_text MATCH ?
|
||||
ORDER BY is_tag_name_match DESC, rank_score DESC
|
||||
LIMIT ? OFFSET ?
|
||||
"""
|
||||
params = [fts_query, limit, offset]
|
||||
query_escaped = (
|
||||
query_lower.lstrip("/")
|
||||
.replace("\\", "\\\\")
|
||||
.replace("%", "\\%")
|
||||
.replace("_", "\\_")
|
||||
)
|
||||
params = [query_escaped + "%", fts_query, limit, offset]
|
||||
|
||||
cursor = conn.execute(sql, params)
|
||||
results = []
|
||||
@@ -510,8 +547,17 @@ class TagFTSIndex:
|
||||
"tag_name": row[0],
|
||||
"category": row[1],
|
||||
"post_count": row[2],
|
||||
"is_tag_name_match": row[4] == 1,
|
||||
"rank_score": row[5],
|
||||
}
|
||||
|
||||
# Set is_exact_prefix based on tag_name match
|
||||
tag_name = row[0]
|
||||
if tag_name.lower().startswith(query_lower.lstrip("/")):
|
||||
result["is_exact_prefix"] = True
|
||||
else:
|
||||
result["is_exact_prefix"] = result["is_tag_name_match"]
|
||||
|
||||
# Check if search matched an alias rather than the tag_name
|
||||
matched_alias = self._find_matched_alias(query, row[0], row[3])
|
||||
if matched_alias:
|
||||
|
||||
Reference in New Issue
Block a user