Compare commits
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
255fdd3c92 |
@@ -2,6 +2,7 @@
|
||||
|
||||
Condensed version-by-version index of DEVLOG.md (each line: version + one-line summary of what changed and why). Full rationale, code details, and "How to apply" lessons live in the excluded DEVLOG.md itself; this index exists so graphify can extract a queryable semantic node per release without feeding the full log to the local Ollama model. Regenerate with `python tools/generate_doc_index.py devlog` whenever DEVLOG.md changes.
|
||||
|
||||
- **v0.7.9(2026-10-02)** — Models: CivitAIベースモデルからの自動バッジ作成
|
||||
- **v0.7.8(2026-10-01)** — Video Edit: マルチトラック化(ビデオ/オーバーレイ/テキスト/オーディオ)
|
||||
- **v0.7.7(2026-09-30)** — 意思決定モデル: Ollama 0.35+ バックエンド追加
|
||||
- **v0.7.6(2026-09-30)** — Settings: 意思決定モデル (Laya) の専用設定
|
||||
|
||||
@@ -2,6 +2,38 @@
|
||||
|
||||
---
|
||||
|
||||
## v0.7.9(2026-10-02)
|
||||
|
||||
### Models: CivitAIベースモデルからの自動バッジ作成
|
||||
|
||||
バッジ管理モーダルに「CivitAIから自動バッジ作成」を追加(`static/js/models/auto-badge.js`)。現在のタブのモデルのCivitAI `baseModel` からバッジを作る。意思決定モデルは使わず、ルール表で決定的に決める("SDXL 1.0"/"SDXL Turbo"等はSDXLにまとめ、Pony/Illustrious/NoobAI/Flux.1/Flux.2/Qwen/Wan/Z-Image/HiDream等は別バッジ。表に無い値は文字列そのまま、"Other"と空は対象外)。
|
||||
|
||||
- モデル側は既存バッジを残したまま追加(同名が既にあれば何もしない)。パレットは同名バッジの色を上書き可(チェックで切替)。「すでにバッジが付いているモデルはスキップ」オプションあり(スキップ分のバッジ名はパレットにも作らない)。
|
||||
- プレビューで件数・新規/既存を確認してから適用。
|
||||
|
||||
### CivitAI一括取得の高速化と安定化
|
||||
|
||||
参考にしたcivitai-updater/tiny-model-managerはどちらも単発GETの逐次実行で、当プラグインの一括POST(100件/回)のほうが進んでいた。遅さの原因は取得ではなくハッシュ計算側だった。
|
||||
|
||||
- **保存済みsha256の再利用**: 一括取得が保存済みハッシュを無視して毎回全件再計算していた。`sha256Size`/`sha256Mtime`を記録してファイル差し替えを検知(記録の無い旧データは信頼)。
|
||||
- **サイドカー取り込み**: `<名前>.metadata.json` のsha256とCivitAI情報を、サイズ一致かつ`files[].hashes.SHA256`一致の場合だけ採用(ハッシュ計算・通信不要)。実データでサイドカー87件中74件が利用可能、58件にCivitAI情報あり。
|
||||
- **並列ハッシュ(3並列)+ 1MB読み**、確定ごとに5秒間隔でメタデータへ保存(途中で止めても計算済み分が残る。以前は最後にまとめて保存していた)。
|
||||
- **プレビュー保存**を4並列・別スレッド化(以前はasyncハンドラ内で逐次ダウンロードしイベントループを塞いでいた)。進捗に「プレビュー保存中」を追加。
|
||||
- **未登録スキップ**: CivitAIに存在しなかったモデルに `civitaiNotFound` を記録し、ツールバーの「未登録はスキップ」で一括取得の対象外にできる(既定オフ)。通信エラーで失敗した分は記録しない。見つかれば記録は自動解除。
|
||||
|
||||
### フォルダのシンボリックリンク/ジャンクション対応
|
||||
|
||||
- モデル走査を `rglob` から自前の `_iter_files` に変更。リンクは辿りつつ、祖先フォルダへ戻る循環だけを検出して入らない(警告ログ)。
|
||||
- 一括移動の移動先チェックを `resolve()` 比較から論理パス比較に変更し、リンク先フォルダへの移動が「Destination is outside model root」で拒否される問題を修正(`..`・絶対パスでの脱出は従来通り拒否)。
|
||||
- 注意: リンク配下のモデルを削除すると、リンク先の実ファイルが消える。
|
||||
|
||||
### 実機で分かったこと
|
||||
|
||||
- TextEncoderタブでの一括取得中にComfyUIが応答しなくなる事象があった。py-spyのダンプでは、イベントループがWindowsの`closesocket`内で停止しており(作業スレッドは全て待機中でプラグインのコードはスタックに無い)、本変更が直接原因とは断定できなかった。PC再起動と動作に影響しうる他のカスタムノードの削除後は正常動作を確認。
|
||||
- この環境(管理者権限なし)ではシンボリックリンクを作れず、ジャンクションのみでテストした。
|
||||
|
||||
---
|
||||
|
||||
## v0.7.8(2026-10-01)
|
||||
|
||||
### Video Edit: マルチトラック化(ビデオ/オーバーレイ/テキスト/オーディオ)
|
||||
|
||||
+1
-1
@@ -26,7 +26,7 @@ A comprehensive workflow, asset management, and generation UI plugin for [ComfyU
|
||||
- Built-in AI tools (translation and more)
|
||||
|
||||

|
||||

|
||||

|
||||
|
||||
## Features (tab -> feature names)
|
||||
|
||||
|
||||
@@ -22,7 +22,7 @@ A comprehensive workflow, asset management, and generation UI plugin for [ComfyU
|
||||
- Built-in AI tools (translation and more)
|
||||
|
||||

|
||||

|
||||

|
||||
|
||||
## Screenshots
|
||||
|
||||
|
||||
+88
-24
@@ -147,6 +147,9 @@ async def handle_civitai_fetch(request: web.Request) -> web.Response:
|
||||
# Fetch from CivitAI
|
||||
info = await asyncio.to_thread(_civitai.fetch_by_hash, sha256)
|
||||
if not info:
|
||||
await asyncio.to_thread(
|
||||
_service.update_metadata, model_name, {"civitaiNotFound": True}
|
||||
)
|
||||
return web.json_response({
|
||||
"status": "not_found",
|
||||
"sha256": sha256,
|
||||
@@ -166,6 +169,10 @@ async def handle_civitai_fetch(request: web.Request) -> web.Response:
|
||||
if preview_saved:
|
||||
logger.info("Auto-saved preview for %s from CivitAI", model_name)
|
||||
|
||||
if model_meta.get("civitaiNotFound"):
|
||||
await asyncio.to_thread(
|
||||
_service.update_metadata, model_name, {"civitaiNotFound": False}
|
||||
)
|
||||
return web.json_response({
|
||||
"status": "ok",
|
||||
"sha256": sha256,
|
||||
@@ -244,13 +251,47 @@ async def handle_civitai_batch(request: web.Request) -> web.Response:
|
||||
def on_progress(current, total, model_name, status):
|
||||
progress_q.put({"current": current, "total": total, "model": model_name, "status": status})
|
||||
|
||||
# 保存済みsha256(size/mtime付き)を渡して再ハッシュを避ける
|
||||
import time as _time
|
||||
all_meta0 = _service.get_all_metadata()
|
||||
known = {}
|
||||
for n, _p in model_files:
|
||||
m = all_meta0.get(n) or {}
|
||||
if m.get("sha256"):
|
||||
known[n] = {"sha256": m["sha256"], "size": m.get("sha256Size"), "mtime": m.get("sha256Mtime")}
|
||||
|
||||
# ハッシュ確定ごとに溜め、5秒おきにメタデータへ書き出す(途中で止めても計算済み分が残る)
|
||||
pending_hashes = {}
|
||||
last_flush = [_time.monotonic()]
|
||||
|
||||
def flush_hashes():
|
||||
if not pending_hashes:
|
||||
return
|
||||
try:
|
||||
all_meta = _service.get_all_metadata()
|
||||
for mname, (sha, size, mtime) in list(pending_hashes.items()):
|
||||
entry = all_meta.setdefault(mname, {"tags": [], "favorite": False, "memo": ""})
|
||||
entry["sha256"] = sha
|
||||
entry["sha256Size"] = size
|
||||
entry["sha256Mtime"] = mtime
|
||||
_service._save_metadata(all_meta)
|
||||
pending_hashes.clear()
|
||||
except Exception as e:
|
||||
logger.warning("Failed to flush hashes: %s", e)
|
||||
last_flush[0] = _time.monotonic()
|
||||
|
||||
def on_hash(name, sha, size, mtime):
|
||||
pending_hashes[name] = (sha, size, mtime)
|
||||
if _time.monotonic() - last_flush[0] > 5:
|
||||
flush_hashes()
|
||||
|
||||
# Start batch in background thread
|
||||
loop = asyncio.get_event_loop()
|
||||
fetch_task = loop.run_in_executor(None, _civitai.batch_fetch, model_files, on_progress)
|
||||
fetch_task = loop.run_in_executor(
|
||||
None, lambda: _civitai.batch_fetch(model_files, on_progress, known, on_hash)
|
||||
)
|
||||
|
||||
# Stream progress events
|
||||
import time as _time
|
||||
|
||||
while not fetch_task.done():
|
||||
# Drain queue
|
||||
while not progress_q.empty():
|
||||
@@ -270,43 +311,66 @@ async def handle_civitai_batch(request: web.Request) -> web.Response:
|
||||
break
|
||||
|
||||
# Get results and update metadata
|
||||
results = fetch_task.result()
|
||||
try:
|
||||
results = fetch_task.result()
|
||||
finally:
|
||||
await asyncio.to_thread(flush_hashes)
|
||||
meta_updates = {}
|
||||
for model_name, result in results.items():
|
||||
sha256 = result.get("sha256")
|
||||
if sha256:
|
||||
meta_updates[model_name] = sha256
|
||||
|
||||
# Bulk update sha256 in metadata
|
||||
# Bulk update sha256 (+ CivitAI not-found marker) in metadata
|
||||
if meta_updates:
|
||||
all_meta = _service.get_all_metadata()
|
||||
for mname, sha in meta_updates.items():
|
||||
if mname not in all_meta:
|
||||
all_meta[mname] = {"tags": [], "favorite": False, "memo": ""}
|
||||
all_meta[mname]["sha256"] = sha
|
||||
r = results[mname]
|
||||
if r.get("civitai"):
|
||||
all_meta[mname].pop("civitaiNotFound", None)
|
||||
elif not r.get("fetch_failed") and not r.get("error"):
|
||||
all_meta[mname]["civitaiNotFound"] = _service._now_iso()
|
||||
_service._save_metadata(all_meta)
|
||||
|
||||
# Auto-set previews for models without preview image
|
||||
# (ネットワークI/Oなのでイベントループを塞がないようスレッドで並列実行)
|
||||
await response.write(send_sse("progress", {
|
||||
"current": len(model_files), "total": len(model_files), "model": "", "status": "previews",
|
||||
}))
|
||||
dirs = _get_model_dirs(model_type)
|
||||
preview_saved_count = 0
|
||||
for model_name, result in results.items():
|
||||
civitai = result.get("civitai")
|
||||
if not civitai or not civitai.get("images"):
|
||||
continue
|
||||
if _service.find_preview_image(model_type, model_name):
|
||||
continue
|
||||
model_file = None
|
||||
for d in dirs:
|
||||
p = d / model_name
|
||||
if p.is_file():
|
||||
model_file = p
|
||||
break
|
||||
if not model_file:
|
||||
continue
|
||||
preview_path = model_file.parent / (model_file.stem + ".preview.png")
|
||||
if CivitaiService.download_image(civitai["images"][0], preview_path):
|
||||
preview_saved_count += 1
|
||||
logger.info("Auto-saved preview for %s from CivitAI", model_name)
|
||||
|
||||
def _save_previews():
|
||||
jobs = []
|
||||
for model_name, result in results.items():
|
||||
civitai = result.get("civitai")
|
||||
if not civitai or not civitai.get("images"):
|
||||
continue
|
||||
if _service.find_preview_image(model_type, model_name):
|
||||
continue
|
||||
model_file = None
|
||||
for d in dirs:
|
||||
p = d / model_name
|
||||
if p.is_file():
|
||||
model_file = p
|
||||
break
|
||||
if not model_file:
|
||||
continue
|
||||
jobs.append((model_name, civitai["images"][0],
|
||||
model_file.parent / (model_file.stem + ".preview.png")))
|
||||
if not jobs:
|
||||
return 0
|
||||
from concurrent.futures import ThreadPoolExecutor
|
||||
with ThreadPoolExecutor(max_workers=4) as pool:
|
||||
oks = list(pool.map(lambda j: CivitaiService.download_image(j[1], j[2]), jobs))
|
||||
for (mname, _u, _p), ok in zip(jobs, oks):
|
||||
if ok:
|
||||
logger.info("Auto-saved preview for %s from CivitAI", mname)
|
||||
return sum(1 for ok in oks if ok)
|
||||
|
||||
preview_saved_count = await asyncio.to_thread(_save_previews)
|
||||
|
||||
# Send final result
|
||||
summary = {
|
||||
|
||||
+144
-28
@@ -117,7 +117,7 @@ class CivitaiService:
|
||||
# ── ハッシュ計算 ──────────────────────────────────────────
|
||||
|
||||
@staticmethod
|
||||
def calculate_sha256(file_path, chunk_size=65536):
|
||||
def calculate_sha256(file_path, chunk_size=1024 * 1024):
|
||||
"""Calculate SHA256 hash of a file."""
|
||||
h = hashlib.sha256()
|
||||
path = Path(file_path)
|
||||
@@ -198,6 +198,7 @@ class CivitaiService:
|
||||
url = f"{CIVITAI_API_BASE}/model-versions/by-hash"
|
||||
results = {h.lower(): None for h in sha256_hashes}
|
||||
cache = self._load_cache()
|
||||
failed = set()
|
||||
|
||||
for chunk_start in range(0, len(sha256_hashes), _BATCH_CHUNK_SIZE):
|
||||
chunk = sha256_hashes[chunk_start:chunk_start + _BATCH_CHUNK_SIZE]
|
||||
@@ -239,6 +240,7 @@ class CivitaiService:
|
||||
time.sleep(wait)
|
||||
continue
|
||||
logger.warning("CivitAI batch POST error: %s %s", e.code, e.reason)
|
||||
failed.update(chunk_lower)
|
||||
break
|
||||
except (URLError, Exception) as e:
|
||||
if attempt < _MAX_RETRIES - 1:
|
||||
@@ -247,8 +249,11 @@ class CivitaiService:
|
||||
time.sleep(wait)
|
||||
continue
|
||||
logger.warning("CivitAI batch POST failed: %s", e)
|
||||
failed.update(chunk_lower)
|
||||
break
|
||||
|
||||
# 通信失敗したチャンクは「CivitAIに存在しない」と区別するため記録する
|
||||
self.last_failed_hashes = failed
|
||||
self._save_cache()
|
||||
return results
|
||||
|
||||
@@ -338,64 +343,175 @@ class CivitaiService:
|
||||
|
||||
# ── バッチフェッチ ────────────────────────────────────────
|
||||
|
||||
def batch_fetch(self, model_files, progress_callback=None):
|
||||
@staticmethod
|
||||
def _known_hash(known, st):
|
||||
"""メタデータに保存済みのsha256が、現在のファイルに対して有効ならそれを返す。
|
||||
|
||||
size/mtime が記録されていれば一致を確認する(ファイル差し替え検知)。
|
||||
記録の無い旧データは信頼する。
|
||||
"""
|
||||
sha = (known or {}).get("sha256")
|
||||
if not sha:
|
||||
return None
|
||||
try:
|
||||
ks, km = known.get("size"), known.get("mtime")
|
||||
if ks is not None and int(ks) != st.st_size:
|
||||
return None
|
||||
if km is not None and abs(float(km) - st.st_mtime) > 2:
|
||||
return None
|
||||
except (TypeError, ValueError):
|
||||
return None
|
||||
return str(sha).lower()
|
||||
|
||||
def _read_sidecar(self, file_path, size):
|
||||
"""`<stem>.metadata.json` サイドカーからsha256とCivitAI情報を読む(ハッシュ計算・通信不要)。
|
||||
|
||||
サイドカーのsizeがファイルサイズと一致する場合のみ採用する。
|
||||
Returns: (sha256_lower or None, extracted_info or None)
|
||||
"""
|
||||
sc = file_path.with_name(file_path.stem + ".metadata.json")
|
||||
try:
|
||||
with open(sc, "r", encoding="utf-8") as f:
|
||||
d = json.load(f)
|
||||
sha = str(d.get("sha256", "")).lower()
|
||||
if len(sha) != 64 or any(c not in "0123456789abcdef" for c in sha):
|
||||
return None, None
|
||||
if d.get("size") is None or int(float(d["size"])) != size:
|
||||
return None, None
|
||||
except Exception:
|
||||
return None, None
|
||||
|
||||
info = None
|
||||
civ = d.get("civitai")
|
||||
if isinstance(civ, dict) and civ.get("id"):
|
||||
# 別ファイルの情報を取り込まないよう、files[].hashes.SHA256 に一致するものだけ採用
|
||||
matched = any(
|
||||
str(f.get("hashes", {}).get("SHA256", "")).lower() == sha
|
||||
for f in civ.get("files", []) if isinstance(f, dict)
|
||||
)
|
||||
if matched:
|
||||
try:
|
||||
info = self._extract_info(civ)
|
||||
except Exception:
|
||||
info = None
|
||||
return sha, info
|
||||
|
||||
def batch_fetch(self, model_files, progress_callback=None, known=None, on_hash=None, workers=3):
|
||||
"""Batch fetch CivitAI info for multiple model files.
|
||||
|
||||
Phase 1: SHA256 計算("hashing")
|
||||
Phase 0: 保存済みsha256 / サイドカーから解決(ハッシュ計算・通信なし)
|
||||
Phase 1: 残りのSHA256を並列計算("hashing")
|
||||
Phase 2: POST で一括取得("fetching")— キャッシュ済みはスキップ
|
||||
|
||||
Args:
|
||||
model_files: list of (model_name, file_path) tuples
|
||||
progress_callback: fn(current, total, model_name, status) called per model
|
||||
known: { model_name: {"sha256", "size", "mtime"} } メタデータ保存済みハッシュ
|
||||
on_hash: fn(model_name, sha256, size, mtime) ハッシュが新たに確定するたびに呼ばれる
|
||||
workers: ハッシュ計算の並列数
|
||||
|
||||
Returns: dict of { model_name: { sha256, civitai_info_or_none } }
|
||||
"""
|
||||
results = {}
|
||||
total = len(model_files)
|
||||
cache = self._load_cache()
|
||||
known = known or {}
|
||||
finished = 0
|
||||
cache_dirty = False
|
||||
|
||||
# Phase 1: ハッシュ計算
|
||||
hashes_needed = [] # [(model_name, sha256_lower, original_index)]
|
||||
hashes_needed = [] # [(model_name, sha256_lower)]
|
||||
to_hash = [] # [(model_name, file_path, stat)]
|
||||
|
||||
for i, (model_name, file_path) in enumerate(model_files):
|
||||
if progress_callback:
|
||||
progress_callback(i, total, model_name, "hashing")
|
||||
|
||||
sha256 = self.calculate_sha256(file_path)
|
||||
if not sha256:
|
||||
results[model_name] = {"sha256": None, "civitai": None, "error": "hash_failed"}
|
||||
def _resolved(name, sha, st, new_hash):
|
||||
"""sha256確定後の共通処理。キャッシュにあれば完了、無ければ取得待ちへ。"""
|
||||
nonlocal finished
|
||||
results[name] = {"sha256": sha, "civitai": None}
|
||||
if new_hash and on_hash:
|
||||
on_hash(name, sha, st.st_size, st.st_mtime)
|
||||
if sha in cache:
|
||||
results[name]["civitai"] = cache[sha]
|
||||
finished += 1
|
||||
if progress_callback:
|
||||
progress_callback(i + 1, total, model_name, "not_found")
|
||||
progress_callback(finished, total, name, "cached")
|
||||
else:
|
||||
hashes_needed.append((name, sha))
|
||||
|
||||
# Phase 0: ハッシュ計算なしで解決できるものを先に処理
|
||||
for model_name, file_path in model_files:
|
||||
try:
|
||||
st = Path(file_path).stat()
|
||||
except OSError:
|
||||
results[model_name] = {"sha256": None, "civitai": None, "error": "hash_failed"}
|
||||
finished += 1
|
||||
if progress_callback:
|
||||
progress_callback(finished, total, model_name, "not_found")
|
||||
continue
|
||||
|
||||
sha256_lower = sha256.lower()
|
||||
results[model_name] = {"sha256": sha256_lower, "civitai": None}
|
||||
sha = self._known_hash(known.get(model_name), st)
|
||||
if sha:
|
||||
_resolved(model_name, sha, st, new_hash=False)
|
||||
continue
|
||||
|
||||
if sha256_lower in cache:
|
||||
results[model_name]["civitai"] = cache[sha256_lower]
|
||||
if progress_callback:
|
||||
progress_callback(i + 1, total, model_name, "cached")
|
||||
else:
|
||||
hashes_needed.append((model_name, sha256_lower, i))
|
||||
sha, info = self._read_sidecar(Path(file_path), st.st_size)
|
||||
if sha:
|
||||
if info and sha not in cache:
|
||||
cache[sha] = info
|
||||
cache_dirty = True
|
||||
_resolved(model_name, sha, st, new_hash=True)
|
||||
continue
|
||||
|
||||
to_hash.append((model_name, file_path, st))
|
||||
|
||||
if cache_dirty:
|
||||
self._save_cache()
|
||||
|
||||
# Phase 1: 残りを並列でハッシュ計算(hashlibは大きなチャンクでGILを解放する)
|
||||
if to_hash:
|
||||
from concurrent.futures import ThreadPoolExecutor, as_completed
|
||||
|
||||
def _hash(item):
|
||||
return self.calculate_sha256(item[1])
|
||||
|
||||
with ThreadPoolExecutor(max_workers=max(1, workers)) as pool:
|
||||
futures = {pool.submit(_hash, item): item for item in to_hash}
|
||||
for fut in as_completed(futures):
|
||||
model_name, _fp, st = futures[fut]
|
||||
try:
|
||||
sha = fut.result()
|
||||
except Exception:
|
||||
sha = None
|
||||
if not sha:
|
||||
results[model_name] = {"sha256": None, "civitai": None, "error": "hash_failed"}
|
||||
finished += 1
|
||||
if progress_callback:
|
||||
progress_callback(finished, total, model_name, "not_found")
|
||||
continue
|
||||
if progress_callback:
|
||||
progress_callback(finished, total, model_name, "hashing")
|
||||
_resolved(model_name, sha.lower(), st, new_hash=True)
|
||||
|
||||
# Phase 2: POST で一括取得(未キャッシュ分)
|
||||
if hashes_needed:
|
||||
if progress_callback:
|
||||
progress_callback(len(results), total, "", "fetching")
|
||||
progress_callback(finished, total, "", "fetching")
|
||||
|
||||
# 同一ハッシュが複数モデルに対応する場合を考慮
|
||||
sha256_to_entries: dict[str, list] = {}
|
||||
for name, sha256_lower, idx in hashes_needed:
|
||||
sha256_to_entries.setdefault(sha256_lower, []).append((name, idx))
|
||||
sha256_to_names: dict[str, list] = {}
|
||||
for name, sha256_lower in hashes_needed:
|
||||
sha256_to_names.setdefault(sha256_lower, []).append(name)
|
||||
|
||||
batch_results = self._batch_fetch_post(list(sha256_to_entries.keys()))
|
||||
batch_results = self._batch_fetch_post(list(sha256_to_names.keys()))
|
||||
failed = getattr(self, "last_failed_hashes", set())
|
||||
|
||||
for sha256_lower, info in batch_results.items():
|
||||
for name, idx in sha256_to_entries.get(sha256_lower, []):
|
||||
for name in sha256_to_names.get(sha256_lower, []):
|
||||
results[name]["civitai"] = info
|
||||
if not info and sha256_lower in failed:
|
||||
results[name]["fetch_failed"] = True
|
||||
finished += 1
|
||||
if progress_callback:
|
||||
status = "found" if info else "not_found"
|
||||
progress_callback(idx + 1, total, name, status)
|
||||
progress_callback(finished, total, name, status)
|
||||
|
||||
if progress_callback:
|
||||
progress_callback(total, total, "", "done")
|
||||
|
||||
@@ -2,6 +2,7 @@
|
||||
|
||||
import json
|
||||
import logging
|
||||
import os
|
||||
import shutil
|
||||
import time
|
||||
from datetime import datetime, timezone
|
||||
@@ -41,6 +42,43 @@ MODEL_TYPE_FOLDER_KEYS = {
|
||||
}
|
||||
|
||||
|
||||
def _iter_files(root):
|
||||
"""root 配下の全ファイルを返す。フォルダのシンボリックリンク/ジャンクションも辿る。
|
||||
|
||||
リンクが祖先フォルダを指していて循環する場合のみ、その枝には入らない
|
||||
(rglob だとパス長上限まで再帰し続ける恐れがある)。
|
||||
"""
|
||||
def walk(d, ancestors):
|
||||
try:
|
||||
entries = list(os.scandir(d))
|
||||
except OSError:
|
||||
return
|
||||
for e in entries:
|
||||
try:
|
||||
if e.is_dir(): # リンクは辿る
|
||||
real = os.path.realpath(e.path)
|
||||
if real in ancestors:
|
||||
logger.warning("Skipping looping folder link: %s -> %s", e.path, real)
|
||||
continue
|
||||
yield from walk(e.path, ancestors | {real})
|
||||
elif e.is_file():
|
||||
yield Path(e.path)
|
||||
except OSError:
|
||||
continue
|
||||
|
||||
root = Path(root)
|
||||
yield from walk(root, frozenset([os.path.realpath(root)]))
|
||||
|
||||
|
||||
def _is_within(path, root):
|
||||
"""path が root 配下か。`..` は正規化するが、リンクは解決しない(リンク先フォルダへの移動を許可)。"""
|
||||
try:
|
||||
a, r = os.path.abspath(path), os.path.abspath(root)
|
||||
return os.path.commonpath([os.path.normcase(a), os.path.normcase(r)]) == os.path.normcase(r)
|
||||
except ValueError: # 別ドライブなど
|
||||
return False
|
||||
|
||||
|
||||
def _get_model_dirs(model_type):
|
||||
"""Get all model directories for a type using ComfyUI's folder_paths.
|
||||
|
||||
@@ -111,6 +149,12 @@ class ModelsService:
|
||||
entry["sha256"] = updates["sha256"]
|
||||
if "badges" in updates:
|
||||
entry["badges"] = updates["badges"]
|
||||
if "civitaiNotFound" in updates:
|
||||
# CivitAIに存在しなかった記録(一括取得のスキップ用)。None/False で解除
|
||||
if updates["civitaiNotFound"]:
|
||||
entry["civitaiNotFound"] = self._now_iso()
|
||||
else:
|
||||
entry.pop("civitaiNotFound", None)
|
||||
entry["updatedAt"] = self._now_iso()
|
||||
data[model_name] = entry
|
||||
self._save_metadata(data)
|
||||
@@ -131,9 +175,7 @@ class ModelsService:
|
||||
for d in dirs:
|
||||
if not d.is_dir():
|
||||
continue
|
||||
for f in d.rglob("*"):
|
||||
if not f.is_file():
|
||||
continue
|
||||
for f in _iter_files(d):
|
||||
try:
|
||||
rel = str(f.relative_to(d)).replace("\\", "/")
|
||||
if rel.endswith(_DISABLED_SUFFIX):
|
||||
@@ -153,9 +195,7 @@ class ModelsService:
|
||||
for d in dirs:
|
||||
if not d.is_dir():
|
||||
continue
|
||||
for f in d.rglob("*"):
|
||||
if not f.is_file():
|
||||
continue
|
||||
for f in _iter_files(d):
|
||||
name = f.name
|
||||
if name.endswith(_DISABLED_SUFFIX):
|
||||
name = name[: -len(_DISABLED_SUFFIX)]
|
||||
@@ -280,8 +320,8 @@ class ModelsService:
|
||||
for d in dirs:
|
||||
if not d.is_dir():
|
||||
continue
|
||||
for f in d.rglob("*"):
|
||||
if f.is_file() and f.name.endswith(_DISABLED_SUFFIX):
|
||||
for f in _iter_files(d):
|
||||
if f.name.endswith(_DISABLED_SUFFIX):
|
||||
try:
|
||||
rel = str(f.relative_to(d))
|
||||
if rel.endswith(_DISABLED_SUFFIX):
|
||||
@@ -436,11 +476,10 @@ class ModelsService:
|
||||
errors.append({"model": model_name, "error": "Cannot determine root directory"})
|
||||
continue
|
||||
|
||||
# Destination directory — verify it stays within root_dir after resolution
|
||||
# Destination directory — verify it stays within root_dir (`..`/絶対パスは拒否。
|
||||
# フォルダのシンボリックリンク/ジャンクション配下への移動は許可する)
|
||||
dest_dir = root_dir / dest_subdir if dest_subdir else root_dir
|
||||
try:
|
||||
dest_dir.resolve().relative_to(root_dir.resolve())
|
||||
except ValueError:
|
||||
if not _is_within(dest_dir, root_dir):
|
||||
errors.append({"model": model_name, "error": "Destination is outside model root"})
|
||||
continue
|
||||
dest_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
+1
-1
@@ -1,7 +1,7 @@
|
||||
[project]
|
||||
name = "comfyui-workflow-studio"
|
||||
description = "A comprehensive workflow, asset management, and generation UI plugin for ComfyUI. Manage workflows, models, images, and prompts; batch generation and Image Feeder in the GenerateUI tab; library side panel with canvas drag & drop and AI tools."
|
||||
version = "0.7.8"
|
||||
version = "0.7.9"
|
||||
license = { file = "LICENSE" }
|
||||
requires-python = ">= 3.10"
|
||||
dependencies = ["jinja2"]
|
||||
|
||||
+42
-6
@@ -41,6 +41,16 @@ const LANGUAGES = {
|
||||
badgeNone: "No badges defined yet.",
|
||||
badgeNoneHint: "No badges defined. Click \u2699 Manage Badges to add.",
|
||||
badgeNewLabel: "Badge label...",
|
||||
autoBadgeTitle: "Auto badges from CivitAI",
|
||||
autoBadgeDesc: "Create badges from CivitAI base model info of the current tab's models. Existing badges on models are kept; badges are added.",
|
||||
autoBadgeNoData: "No CivitAI base model info found. Fetch CivitAI info first.",
|
||||
autoBadgeExisting: "existing (color overwritten)",
|
||||
autoBadgeNew: "new",
|
||||
autoBadgeNoInfo: "Models without base model info (skipped)",
|
||||
autoBadgeOverwrite: "Overwrite colors of existing badges with the same name",
|
||||
autoBadgeSkipExisting: "Skip models that already have any badge",
|
||||
autoBadgeApply: "Apply",
|
||||
autoBadgeDone: "badges added to models",
|
||||
badgeColorHint: "Change color",
|
||||
defaultView: "Default View",
|
||||
viewThumbnail: "Thumbnail",
|
||||
@@ -548,6 +558,8 @@ const LANGUAGES = {
|
||||
civitaiWordCopied: "Trigger word copied",
|
||||
civitaiRefresh: "Refresh",
|
||||
civitaiBatchAllCached: "All models already have CivitAI info",
|
||||
civitaiSkipNotFound: "Skip not found",
|
||||
civitaiSavingPreviews: "Saving previews...",
|
||||
civitaiHashing2: "Hashing...",
|
||||
civitaiFetching: "Fetching...",
|
||||
civitaiBatchDone: (found, notFound) => `CivitAI: ${found} found, ${notFound} not found`,
|
||||
@@ -947,11 +959,11 @@ const LANGUAGES = {
|
||||
helpModels2: "Thumbnail and Table view modes with pagination; Table view shows Type and Base Model columns sourced from CivitAI data alongside filename, subdirectory, and extension; column headers are clickable to sort — click once for ascending (▲), again for descending (▼), a third time to clear the sort; the E/D column sorts by enable/disable status",
|
||||
helpModels3: "Search, tag filter, group filter, favorites filter, and status filter (All / Enabled only / Disabled only)",
|
||||
helpModels4: "Side panel: Info (file path, tags, memo), Groups management, CivitAI integration",
|
||||
helpModels5: "CivitAI integration: fetch model info by SHA256 hash, batch fetch per model type; preview image is automatically downloaded and saved if none exists",
|
||||
helpModels5: "CivitAI integration: fetch model info by SHA256 hash, batch fetch per model type; preview image is automatically downloaded and saved if none exists. The batch fetch reuses saved hashes, imports hash and CivitAI info from existing <name>.metadata.json sidecars (no hashing or network needed when the size matches), hashes the rest in parallel and saves progress as it goes. Tick 'Skip not found' to skip models already known to be missing on CivitAI",
|
||||
helpModels6: "Detail modal: preview image, CivitAI info (base model, trained words, tags), thumbnail change — upload any image format; changes take effect immediately; broken/empty preview files are automatically skipped",
|
||||
helpModels7: "Group management: create, rename, delete groups and assign/remove models",
|
||||
helpModels8: "GenUI Model button: applies the selected model to the workflow node (Checkpoint / LoRA / VAE / ControlNet / UNET / TextEncoder / Hypernetwork); LoRA additionally switches to the Single tab, disables all Stack models, and shows LORA SYNTAX / TRIGGER WORDS; Embedding tab shows GenUI PP / GenUI NP buttons instead — appends (embedding:Name:1.0) to the Positive or Negative prompt",
|
||||
helpModels9: "User-defined badges: assign free-label badges with custom colors shared with the Workflow tab",
|
||||
helpModels9: "User-defined badges: assign free-label badges with custom colors shared with the Workflow tab. In Manage Badges, 'Auto badges from CivitAI' creates badges from the CivitAI base model of the current tab's models (variants such as SDXL 1.0 / Turbo are grouped into one family): badges are added to models while their existing badges are kept, colors of same-named badges can be overwritten, and models that already have badges can be skipped",
|
||||
helpModels10: "Enable/Disable models: toggle visibility in ComfyUI by renaming file extension (.disabled suffix). Disable per model (⏸ button on each card) or per group (Enable All / Disable All in Group panel). Changes take effect after ComfyUI refresh.",
|
||||
helpModels11: "Multi-select & bulk operations: click Select to enter selection mode, check multiple models, then use the bulk action bar to add/remove from groups or permanently delete files (model file + preview images + sidecar files such as .json/.info).",
|
||||
helpModels12: "After moving or deleting models, ComfyUI may fail to load them because it caches the old file paths. Perform a Hard Reset from the ComfyUI menu (Queue → Hard Reset) or restart ComfyUI to refresh the model cache.",
|
||||
@@ -1563,6 +1575,16 @@ const LANGUAGES = {
|
||||
badgeNone: "バッジが未定義です。",
|
||||
badgeNoneHint: "バッジが未定義です。⚙ バッジ管理から追加できます。",
|
||||
badgeNewLabel: "バッジ名...",
|
||||
autoBadgeTitle: "CivitAIから自動バッジ作成",
|
||||
autoBadgeDesc: "現在のタブのモデルのCivitAIベースモデル情報からバッジを作成します。モデルに付いている既存バッジは残したまま追加します。",
|
||||
autoBadgeNoData: "CivitAIのベースモデル情報がありません。先にCivitAI情報を取得してください。",
|
||||
autoBadgeExisting: "既存(色を上書き)",
|
||||
autoBadgeNew: "新規",
|
||||
autoBadgeNoInfo: "ベースモデル情報なし(対象外)",
|
||||
autoBadgeOverwrite: "同名の既存バッジの色を上書きする",
|
||||
autoBadgeSkipExisting: "すでにバッジが付いているモデルはスキップ",
|
||||
autoBadgeApply: "適用",
|
||||
autoBadgeDone: "件のバッジを追加しました",
|
||||
badgeColorHint: "色を変更",
|
||||
defaultView: "デフォルト表示",
|
||||
viewThumbnail: "サムネイル",
|
||||
@@ -2070,6 +2092,8 @@ const LANGUAGES = {
|
||||
civitaiWordCopied: "トリガーワードをコピーしました",
|
||||
civitaiRefresh: "更新",
|
||||
civitaiBatchAllCached: "すべてのモデルのCivitAI情報は取得済みです",
|
||||
civitaiSkipNotFound: "未登録はスキップ",
|
||||
civitaiSavingPreviews: "プレビュー保存中...",
|
||||
civitaiHashing2: "ハッシュ計算中...",
|
||||
civitaiFetching: "取得中...",
|
||||
civitaiBatchDone: (found, notFound) => `CivitAI: ${found}件取得, ${notFound}件未登録`,
|
||||
@@ -2469,11 +2493,11 @@ const LANGUAGES = {
|
||||
helpModels2: "サムネイル / テーブル表示モード切り替え(ページネーション付き)。テーブル表示ではファイル名・サブフォルダ・拡張子に加えCivitAI由来のType / Base Model列を表示。列ヘッダーをクリックでソート — 1回で昇順(▲)、2回で降順(▼)、3回でソート解除。E/D列は有効/無効状態でソート",
|
||||
helpModels3: "検索、タグフィルタ、グループフィルタ、お気に入りフィルタ、ステータスフィルタ(すべて / 有効のみ / 無効のみ)",
|
||||
helpModels4: "サイドパネル:基本情報(ファイルパス、タグ、メモ)、グループ管理、CivitAI連携",
|
||||
helpModels5: "CivitAI連携:SHA256ハッシュでモデル情報取得、モデルタイプごとの一括取得;プレビュー画像がない場合は自動でダウンロード・保存",
|
||||
helpModels5: "CivitAI連携:SHA256ハッシュでモデル情報取得、モデルタイプごとの一括取得;プレビュー画像がない場合は自動でダウンロード・保存。一括取得は保存済みハッシュを再利用し、既存の <名前>.metadata.json サイドカーからハッシュとCivitAI情報を取り込み(サイズ一致時はハッシュ計算・通信不要)、残りは並列でハッシュ計算して途中経過も保存します。「未登録はスキップ」をオンにするとCivitAIに存在しないと判明済みのモデルを対象外にできます",
|
||||
helpModels6: "詳細モーダル:プレビュー画像、CivitAI情報(ベースモデル、トリガーワード、タグ)、サムネイル変更 — 任意の画像形式をアップロード可能;ComfyUI再起動なしで即時反映;破損/空のプレビューファイルは自動スキップ",
|
||||
helpModels7: "グループ管理:グループの作成・名前変更・削除、モデルの割り当て・解除",
|
||||
helpModels8: "GenUI Modelボタン:選択モデルをワークフローノードに適用(Checkpoint / LoRA / VAE / ControlNet / UNET / TextEncoder / Hypernetwork);LoRAはSingleタブへ切り替え・Stack全無効化・LORA SYNTAX/TRIGGER WORDSをLoRAペインに表示;EmbeddingタイプはGenUI PP / GenUI NPボタンを表示 — (embedding:Name:1.0)形式でPositive/Negativeプロンプトの末尾に追加",
|
||||
helpModels9: "ユーザー定義バッジ:自由ラベルのバッジをモデルに付与(Workflowタブと色パレットを共有)",
|
||||
helpModels9: "ユーザー定義バッジ:自由ラベルのバッジをモデルに付与(Workflowタブと色パレットを共有)。バッジ管理の「CivitAIから自動バッジ作成」で、現在のタブのモデルのCivitAIベースモデルからバッジを作成できます(SDXL 1.0 / Turbo などの派生は1つのファミリーにまとめます)。既存バッジは残したまま追加、同名バッジの色は上書き可、すでにバッジが付いているモデルのスキップも可能です",
|
||||
helpModels10: "モデルの有効/無効切り替え:拡張子リネーム(.disabledサフィックス)でComfyUIへの表示を制御。各カードの⏸ボタンで単体操作、グループパネルの「全て有効化 / 全て無効化」でグループ一括操作。変更はComfyUIリフレッシュ後に反映。",
|
||||
helpModels11: "複数選択と一括操作:「選択」ボタンで選択モードに入り複数モデルにチェックを入れると、一括操作バーからグループへの追加/削除、またはファイルの完全削除(モデルファイル+プレビュー画像+.json/.infoなどのサイドカーファイル)が行えます。",
|
||||
helpModels12: "モデルの移動・削除後はComfyUIが古いパスをキャッシュしているためモデル読み込みエラーが発生することがあります。ComfyUIメニューのHard Reset(Queue → Hard Reset)またはComfyUIを再起動してモデルキャッシュを更新してください。",
|
||||
@@ -3085,6 +3109,16 @@ const LANGUAGES = {
|
||||
badgeNone: "尚未定义标签。",
|
||||
badgeNoneHint: "尚未定义标签。点击 ⚙ 管理标签来添加。",
|
||||
badgeNewLabel: "标签名称...",
|
||||
autoBadgeTitle: "根据CivitAI自动创建徽章",
|
||||
autoBadgeDesc: "根据当前标签页模型的CivitAI基础模型信息创建徽章。模型上已有的徽章会保留,仅追加。",
|
||||
autoBadgeNoData: "未找到CivitAI基础模型信息,请先获取CivitAI信息。",
|
||||
autoBadgeExisting: "已存在(覆盖颜色)",
|
||||
autoBadgeNew: "新建",
|
||||
autoBadgeNoInfo: "无基础模型信息(已跳过)",
|
||||
autoBadgeOverwrite: "覆盖同名现有徽章的颜色",
|
||||
autoBadgeSkipExisting: "跳过已有徽章的模型",
|
||||
autoBadgeApply: "应用",
|
||||
autoBadgeDone: "个徽章已添加",
|
||||
badgeColorHint: "更改颜色",
|
||||
defaultView: "默认视图",
|
||||
viewThumbnail: "缩略图",
|
||||
@@ -3592,6 +3626,8 @@ const LANGUAGES = {
|
||||
civitaiWordCopied: "触发词已复制",
|
||||
civitaiRefresh: "刷新",
|
||||
civitaiBatchAllCached: "所有模型的CivitAI信息已获取",
|
||||
civitaiSkipNotFound: "跳过未收录",
|
||||
civitaiSavingPreviews: "正在保存预览...",
|
||||
civitaiHashing2: "计算哈希中...",
|
||||
civitaiFetching: "获取中...",
|
||||
civitaiBatchDone: (found, notFound) => `CivitAI: ${found}个已找到, ${notFound}个未找到`,
|
||||
@@ -3991,11 +4027,11 @@ const LANGUAGES = {
|
||||
helpModels2: "缩略图和表格视图模式切换(带分页)。表格视图除文件名、子文件夹和扩展名外,还显示来自CivitAI数据的Type / Base Model列。点击列标题可排序 — 第一次升序(▲),第二次降序(▼),第三次取消排序。E/D列按启用/禁用状态排序",
|
||||
helpModels3: "搜索、标签筛选、分组筛选、收藏筛选、状态筛选(全部 / 仅启用 / 仅禁用)",
|
||||
helpModels4: "侧面板:基本信息(文件路径、标签、备注)、分组管理、CivitAI集成",
|
||||
helpModels5: "CivitAI集成:通过SHA256哈希获取模型信息,按模型类型批量获取;无预览图时自动下载并保存",
|
||||
helpModels5: "CivitAI集成:通过SHA256哈希获取模型信息,按模型类型批量获取;无预览图时自动下载并保存。批量获取会复用已保存的哈希,从现有的 <名称>.metadata.json 附属文件导入哈希和CivitAI信息(大小一致时无需计算哈希或联网),其余模型并行计算哈希并实时保存进度。勾选“跳过未收录”可跳过已确认CivitAI上不存在的模型",
|
||||
helpModels6: "详情弹窗:预览图片、CivitAI信息(基础模型、触发词、标签)、缩略图更改 — 可上传任意格式图片;无需重启ComfyUI立即生效;损坏/空预览文件自动跳过",
|
||||
helpModels7: "分组管理:创建、重命名、删除分组,分配/移除模型",
|
||||
helpModels8: "GenUI Model按钮:将选中模型应用到工作流节点(Checkpoint / LoRA / VAE / ControlNet / UNET / TextEncoder / Hypernetwork);LoRA还会切换到Single标签、禁用所有Stack模型并在LoRA面板显示LORA SYNTAX/TRIGGER WORDS;Embedding标签改为显示GenUI PP / GenUI NP按钮 — 以(embedding:Name:1.0)格式追加到Positive或Negative提示词末尾",
|
||||
helpModels9: "用户自定义徽章:为模型添加自由标签徽章(颜色调色板与Workflow标签共享)",
|
||||
helpModels9: "用户自定义徽章:为模型添加自由标签徽章(颜色调色板与Workflow标签共享)。在徽章管理中,“根据CivitAI自动创建徽章”可根据当前标签页模型的CivitAI基础模型创建徽章(SDXL 1.0 / Turbo 等变体归为同一系列)。保留模型已有徽章并追加,可覆盖同名徽章的颜色,也可跳过已有徽章的模型",
|
||||
helpModels10: "模型启用/禁用:通过重命名文件扩展名(添加.disabled后缀)控制ComfyUI中的显示。点击每个卡片上的⏸按钮单独操作,或在分组面板中使用「全部启用/全部禁用」进行批量操作。更改在刷新ComfyUI后生效。",
|
||||
helpModels11: "多选与批量操作:点击「选择」按钮进入选择模式,勾选多个模型后,可通过批量操作栏将其添加/移出分组,或永久删除文件(模型文件+预览图+.json/.info等附属文件)。",
|
||||
helpModels12: "移动或删除模型后,ComfyUI可能因缓存旧路径而无法加载模型。请通过ComfyUI菜单执行Hard Reset(Queue → Hard Reset)或重启ComfyUI以刷新模型缓存。",
|
||||
|
||||
@@ -550,6 +550,16 @@ export function initModelsTab() {
|
||||
document.getElementById("wfm-models-civitai-batch-btn")?.addEventListener("click", () => {
|
||||
batchFetchCivitai();
|
||||
});
|
||||
// Skip models already known to be missing on CivitAI (persisted per browser)
|
||||
const skipNfEl = document.getElementById("wfm-models-civitai-skip-nf");
|
||||
const skipNfLabel = document.getElementById("wfm-models-civitai-skip-nf-label");
|
||||
if (skipNfLabel) skipNfLabel.textContent = t("civitaiSkipNotFound");
|
||||
if (skipNfEl) {
|
||||
skipNfEl.checked = localStorage.getItem("wfm_models_civitai_skip_nf") === "1";
|
||||
skipNfEl.addEventListener("change", () => {
|
||||
localStorage.setItem("wfm_models_civitai_skip_nf", skipNfEl.checked ? "1" : "0");
|
||||
});
|
||||
}
|
||||
|
||||
// Side panel tab switching
|
||||
document.querySelectorAll(".wfm-models-side-tab-btn").forEach((btn) => {
|
||||
|
||||
@@ -0,0 +1,136 @@
|
||||
/**
|
||||
* Models Tab - Auto badges from CivitAI base model info
|
||||
*
|
||||
* CivitAI's `baseModel` is already a categorical string, so labels are derived
|
||||
* with a deterministic rule table (variants such as "SDXL 1.0" / "SDXL Turbo"
|
||||
* collapse into one family badge). Unknown strings fall back to the raw value.
|
||||
* Apply semantics: model badges are appended (existing badges are kept);
|
||||
* a label already in the palette gets its color overwritten (optional).
|
||||
*/
|
||||
|
||||
import { openModal, closeModal, showToast } from "../app.js";
|
||||
import { t } from "../i18n.js";
|
||||
import { escapeHtml } from "../util.js";
|
||||
import { state } from "./state.js";
|
||||
import { getCurrentModels } from "./filters.js";
|
||||
import { getBadgePalette, saveBadgePalette, renderBadgeFilter } from "./badges.js";
|
||||
import { renderModelGrid } from "./grid-view.js";
|
||||
import { saveModelMetadata } from "../models-tab.js";
|
||||
|
||||
// [pattern, label, color] — first match wins, so specific families come first.
|
||||
const FAMILY_RULES = [
|
||||
[/^pony/i, "Pony", "#d946ef"],
|
||||
[/^illustrious/i, "Illustrious", "#f59e0b"],
|
||||
[/^noobai/i, "NoobAI", "#14b8a6"],
|
||||
[/^sdxl|^sd ?xl/i, "SDXL", "#3b82f6"],
|
||||
[/^sd ?1\.[45]/i, "SD1.5", "#22c55e"],
|
||||
[/^sd ?2/i, "SD2", "#84cc16"],
|
||||
[/^sd ?3/i, "SD3", "#06b6d4"],
|
||||
[/^flux\.?1|^flux$/i, "Flux.1", "#ef4444"],
|
||||
[/^flux\.?2/i, "Flux.2", "#dc2626"],
|
||||
[/^qwen/i, "Qwen", "#8b5cf6"],
|
||||
[/^wan/i, "Wan", "#0ea5e9"],
|
||||
[/^z-?image/i, "Z-Image", "#ec4899"],
|
||||
[/^hidream/i, "HiDream", "#f97316"],
|
||||
[/^hunyuan/i, "Hunyuan", "#6366f1"],
|
||||
[/^chroma/i, "Chroma", "#a855f7"],
|
||||
[/^lumina/i, "Lumina", "#10b981"],
|
||||
[/^auraflow/i, "AuraFlow", "#64748b"],
|
||||
[/^kolors/i, "Kolors", "#e11d48"],
|
||||
];
|
||||
const SKIP_VALUES = new Set(["", "other", "unknown"]);
|
||||
const FALLBACK_COLORS = ["#0891b2", "#7c3aed", "#be185d", "#b45309", "#4d7c0f", "#475569"];
|
||||
|
||||
export function baseModelToBadge(baseModel) {
|
||||
const raw = (baseModel || "").trim();
|
||||
if (SKIP_VALUES.has(raw.toLowerCase())) return null;
|
||||
for (const [re, label, color] of FAMILY_RULES) {
|
||||
if (re.test(raw)) return { label, color };
|
||||
}
|
||||
let h = 0;
|
||||
for (const ch of raw) h = (h * 31 + ch.charCodeAt(0)) >>> 0;
|
||||
return { label: raw, color: FALLBACK_COLORS[h % FALLBACK_COLORS.length] };
|
||||
}
|
||||
|
||||
/** Scan current-type models → { assignments: [{model, label}], labels: Map(label → {color, count}) } */
|
||||
function buildPlan() {
|
||||
const assignments = [];
|
||||
const labels = new Map();
|
||||
let noInfo = 0;
|
||||
for (const m of getCurrentModels()) {
|
||||
const sha = state.modelMetadata[m]?.sha256;
|
||||
const civ = sha && state.civitaiCache[sha];
|
||||
const badge = civ ? baseModelToBadge(civ.baseModel) : null;
|
||||
if (!badge) { noInfo++; continue; }
|
||||
assignments.push({ model: m, label: badge.label });
|
||||
const entry = labels.get(badge.label) || { color: badge.color, count: 0 };
|
||||
entry.count++;
|
||||
labels.set(badge.label, entry);
|
||||
}
|
||||
return { assignments, labels, noInfo };
|
||||
}
|
||||
|
||||
export function openAutoBadgeModal(onDone = null) {
|
||||
const { assignments, labels, noInfo } = buildPlan();
|
||||
if (assignments.length === 0) {
|
||||
showToast(t("autoBadgeNoData"), "warning");
|
||||
return;
|
||||
}
|
||||
const palette = getBadgePalette();
|
||||
const rows = [...labels.entries()].sort((a, b) => b[1].count - a[1].count).map(([label, info]) => {
|
||||
const exists = label in palette;
|
||||
return `<div class="wfm-badge-color-row">
|
||||
<span class="wfm-badge wfm-badge-model" style="background:${info.color};color:#fff;">${escapeHtml(label)}</span>
|
||||
<span style="font-size:12px;">${info.count}</span>
|
||||
<span style="font-size:11px;color:var(--wfm-text-secondary);">${exists ? t("autoBadgeExisting") : t("autoBadgeNew")}</span>
|
||||
</div>`;
|
||||
}).join("");
|
||||
|
||||
openModal(t("autoBadgeTitle"), `
|
||||
<div style="min-width:340px;">
|
||||
<p style="font-size:12px;color:var(--wfm-text-secondary);margin:0 0 8px;">${t("autoBadgeDesc")}</p>
|
||||
${rows}
|
||||
<p style="font-size:11px;color:var(--wfm-text-secondary);">${t("autoBadgeNoInfo")}: ${noInfo}</p>
|
||||
<label style="font-size:12px;display:block;margin:8px 0;">
|
||||
<input type="checkbox" id="wfm-auto-badge-overwrite" checked> ${t("autoBadgeOverwrite")}
|
||||
</label>
|
||||
<label style="font-size:12px;display:block;margin:8px 0;">
|
||||
<input type="checkbox" id="wfm-auto-badge-skip"> ${t("autoBadgeSkipExisting")}
|
||||
</label>
|
||||
<button class="wfm-btn wfm-btn-sm wfm-btn-primary" id="wfm-auto-badge-apply">${t("autoBadgeApply")} (${assignments.length})</button>
|
||||
</div>`);
|
||||
|
||||
const hasBadges = (m) => (state.modelMetadata[m]?.badges || []).length > 0;
|
||||
const targets = () => document.getElementById("wfm-auto-badge-skip")?.checked
|
||||
? assignments.filter(({ model }) => !hasBadges(model)) : assignments;
|
||||
document.getElementById("wfm-auto-badge-skip")?.addEventListener("change", () => {
|
||||
const btn = document.getElementById("wfm-auto-badge-apply");
|
||||
if (btn) btn.textContent = `${t("autoBadgeApply")} (${targets().length})`;
|
||||
});
|
||||
|
||||
document.getElementById("wfm-auto-badge-apply")?.addEventListener("click", async (e) => {
|
||||
e.target.disabled = true;
|
||||
const overwrite = document.getElementById("wfm-auto-badge-overwrite")?.checked;
|
||||
const target = targets();
|
||||
const pal = getBadgePalette();
|
||||
const usedLabels = new Set(target.map((a) => a.label));
|
||||
for (const [label, info] of labels) {
|
||||
if (!usedLabels.has(label)) continue;
|
||||
if (!(label in pal) || overwrite) pal[label] = info.color;
|
||||
}
|
||||
saveBadgePalette(pal);
|
||||
|
||||
const todo = target.filter(({ model, label }) => !(state.modelMetadata[model]?.badges || []).includes(label));
|
||||
for (let i = 0; i < todo.length; i += 8) {
|
||||
await Promise.all(todo.slice(i, i + 8).map(({ model, label }) => {
|
||||
const badges = [...(state.modelMetadata[model]?.badges || []), label];
|
||||
return saveModelMetadata(model, { badges });
|
||||
}));
|
||||
}
|
||||
renderBadgeFilter();
|
||||
renderModelGrid();
|
||||
showToast(`${todo.length} ${t("autoBadgeDone")}`, "success");
|
||||
closeModal();
|
||||
if (onDone) onDone();
|
||||
});
|
||||
}
|
||||
@@ -6,6 +6,7 @@ import { openModal } from "../app.js";
|
||||
import { t } from "../i18n.js";
|
||||
import { escapeHtml, readJsonStorage } from "../util.js";
|
||||
import { state } from "./state.js";
|
||||
import { openAutoBadgeModal } from "./auto-badge.js";
|
||||
|
||||
// Grid re-render is owned by models-tab.js (grid rendering hasn't been split
|
||||
// out yet); registered once from initModelsTab() to avoid a circular import.
|
||||
@@ -61,6 +62,9 @@ export function openBadgeEditModal(onPaletteChange = null) {
|
||||
<input type="color" id="wfm-badge-new-color" value="#6366f1" style="width:36px;height:28px;padding:1px;border-radius:4px;cursor:pointer;">
|
||||
<button class="wfm-btn wfm-btn-sm wfm-btn-primary" id="wfm-badge-add-btn">${t("badgeAdd")}</button>
|
||||
</div>
|
||||
<div style="margin-top:8px;">
|
||||
<button class="wfm-btn wfm-btn-sm" id="wfm-badge-auto-btn" title="${t("autoBadgeDesc")}">✨ ${t("autoBadgeTitle")}</button>
|
||||
</div>
|
||||
</div>`;
|
||||
|
||||
openModal(t("badgeManage"), html);
|
||||
@@ -93,6 +97,10 @@ function bindBadgeModalEvents(onPaletteChange = null) {
|
||||
|
||||
bindBadgeRowEvents(refreshList, afterChange);
|
||||
|
||||
document.getElementById("wfm-badge-auto-btn")?.addEventListener("click", () => {
|
||||
openAutoBadgeModal(onPaletteChange);
|
||||
});
|
||||
|
||||
document.getElementById("wfm-badge-add-btn")?.addEventListener("click", () => {
|
||||
const labelInput = document.getElementById("wfm-badge-new-label");
|
||||
const colorInput = document.getElementById("wfm-badge-new-color");
|
||||
|
||||
@@ -815,10 +815,13 @@ export async function batchFetchCivitai() {
|
||||
}
|
||||
|
||||
// Filter out models that already have CivitAI data
|
||||
// Optionally also skip models previously recorded as "not found on CivitAI"
|
||||
const skipNotFound = document.getElementById("wfm-models-civitai-skip-nf")?.checked;
|
||||
const meta = state.modelMetadata;
|
||||
const uncached = models.filter((m) => {
|
||||
const sha = meta[m]?.sha256;
|
||||
return !sha || !state.civitaiCache[sha];
|
||||
if (sha && state.civitaiCache[sha]) return false;
|
||||
return !(skipNotFound && meta[m]?.civitaiNotFound);
|
||||
});
|
||||
|
||||
if (uncached.length === 0) {
|
||||
@@ -860,6 +863,7 @@ export async function batchFetchCivitai() {
|
||||
const pct = data.total > 0 ? Math.round((data.current / data.total) * 100) : 0;
|
||||
const statusText = data.status === "hashing" ? t("civitaiHashing2")
|
||||
: data.status === "fetching" ? t("civitaiFetching")
|
||||
: data.status === "previews" ? t("civitaiSavingPreviews")
|
||||
: data.status === "cached" ? "✓"
|
||||
: data.status === "found" ? "✓"
|
||||
: data.status === "not_found" ? "—"
|
||||
|
||||
@@ -247,6 +247,9 @@
|
||||
<span id="wfm-models-count" class="wfm-nodes-count"></span>
|
||||
<button id="wfm-models-refresh-btn" class="wfm-btn wfm-btn-sm">Refresh</button>
|
||||
<button id="wfm-models-civitai-batch-btn" class="wfm-btn wfm-btn-sm" title="Fetch CivitAI info for all models">CivitAI</button>
|
||||
<label style="font-size:11px;color:var(--wfm-text-secondary);white-space:nowrap;cursor:pointer;">
|
||||
<input type="checkbox" id="wfm-models-civitai-skip-nf"> <span id="wfm-models-civitai-skip-nf-label">Skip not found</span>
|
||||
</label>
|
||||
<span id="wfm-models-civitai-progress" style="font-size:12px;color:var(--wfm-text-secondary);"></span>
|
||||
<button id="wfm-models-badge-settings-btn" class="wfm-btn wfm-btn-sm" title="Badge settings">⚙ Badge</button>
|
||||
<button id="wfm-models-select-btn" class="wfm-btn wfm-btn-sm wfm-btn-success" title="Multi-select models">Select</button>
|
||||
@@ -1824,7 +1827,7 @@
|
||||
<h3 id="wfm-help-models-card-civitai-title">Side Panel & CivitAI</h3>
|
||||
<ul>
|
||||
<li id="wfm-help-models-4">Side panel: opens to CivitAI tab by default when a model is selected; Info (file path, tags, memo), Groups management, CivitAI integration</li>
|
||||
<li id="wfm-help-models-5">CivitAI integration: fetch model info by SHA256 hash, batch fetch per model type</li>
|
||||
<li id="wfm-help-models-5">CivitAI integration: fetch model info by SHA256 hash, batch fetch per model type; the batch fetch reuses saved hashes, imports from existing .metadata.json sidecars and hashes the rest in parallel; 'Skip not found' skips models known to be missing on CivitAI</li>
|
||||
</ul>
|
||||
</div>
|
||||
<div class="wfm-help-card">
|
||||
@@ -1838,7 +1841,7 @@
|
||||
<h3 id="wfm-help-models-card-groups-title">Groups & Badges</h3>
|
||||
<ul>
|
||||
<li id="wfm-help-models-7">Group management: create, rename, delete groups and assign/remove models; Checkpoint and LoRA tabs support B (Batch) and S (Stack) quick-assign buttons per model in both card and table views</li>
|
||||
<li id="wfm-help-models-9">User-defined badges: assign free-label badges with custom colors shared with the Workflow tab</li>
|
||||
<li id="wfm-help-models-9">User-defined badges: assign free-label badges with custom colors shared with the Workflow tab; 'Auto badges from CivitAI' in Manage Badges creates badges from CivitAI base models</li>
|
||||
</ul>
|
||||
</div>
|
||||
<div class="wfm-help-card">
|
||||
|
||||
Reference in New Issue
Block a user