Compare commits

...
4 Commits
Author SHA1 Message Date
ketle-manandClaude Sonnet 4.6 a48d9c1437 v0.3.8: Metadata tab - data9 workflow support and unclassified prompt text type
- Extended extractPromptsFromNodeSet to resolve link sources from any STRING-output
  node (ComfySwitchNode, PrimitiveStringMultiline, etc.) without type restriction
- Added fallback for SamplerCustomAdvanced (no positive/negative inputs): return
  textMap contents as texts[] instead of null
- Extended step4 PrimitiveStringMultiline and step7 final fallback to search
  subgraph nodes via collectAllNodes
- Added "text" prompt type (no POS/NEG badge) for cases where polarity cannot be
  determined; buildPromptItem and handleFile updated accordingly
- Added metaPromptText i18n key (EN/JA/ZH)
- Updated README (v0.3.8, feature description) and DEVLOG

Supported in data9: flux2_fp8, flux2_klein_4b_distilled, ernie_image,
qwen_image_edit_2511, wan2_2_14B_animate

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-16 00:32:49 +09:00
ketle-manandClaude Sonnet 4.6 48303e1452 docs: update README and DEVLOG for v0.3.7
README:
- Bump version badge to 0.3.7
- Update Metadata Tab feature section: subgraph support, expanded
  format list, updated prompt extraction description
- Add v0.3.7 changelog entry

DEVLOG:
- Append i18n/HTML text update notes to v0.3.7 entry

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-15 23:05:56 +09:00
ketle-manandClaude Sonnet 4.6 b0171add24 docs: update Metadata tab format note and help text (EN/JA/ZH)
- metaFormatTodo: replace "more formats planned" with note that
  Flux.2 / Qwen-Image / Z-Image subgraph workflows are now supported
- helpMetadata5: list Flux.2 / Qwen-Image / Z-Image explicitly
- Update static fallback text in index.html accordingly

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-15 23:03:17 +09:00
ketle-manandClaude Sonnet 4.6 7958ad1276 v0.3.7: Metadata tab support for Flux2/Qwen/Z-Image subgraph workflows
- Add extractPromptsFromNodeSet() to handle CLIPTextEncode+KSampler in
  subgraphs (definitions.subgraphs) used by official ComfyUI templates
- Add extractMarkdownNoteModels() to parse **section** → - [model](url)
  from MarkdownNote nodes as fallback model extraction
- Add PrimitiveStringMultiline prompt support (flux2-klein workflows)
- Fix extractCheckpoints/extractVAEs to use collectAllNodes() so
  subgraph loaders are included
- Restructure extractPromptsLiteGraph into 7-stage fallback chain

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-15 22:09:06 +09:00
7 changed files with 311 additions and 99 deletions
+113
View File
@@ -1,5 +1,118 @@
# DEVLOG - ComfyUI-Workflow-Studio
## 2026-05-16: v0.3.8 Metadata タブ: data9 対応 / プロンプト判別不能テキスト
### 概要
data9 の新規ワークフロー(Flux.2 Dev fp8、Flux.2 Klein 4B Distilled、Ernie Image、
Qwen-Image-Edit 2511、WAN2.2 14B Animate)の Metadata タブ表示に対応。
あわせて、ポジティブ / ネガティブの判別ができないプロンプトを **Text(バッジなし)** として
区別して表示する仕様に変更した。
### 問題と原因
| ファイル | 問題 |
|---|---|
| image_flux2_fp8.json | `SamplerCustomAdvanced` に positive/negative 入力ポートがなく、リンク解決でプロンプトが空になる |
| image_ernie_image.json | `PrimitiveStringMultiline` がサブグラフ内にあり、ステップ4(トップのみ対象)で見つからない |
| image_qwen_image_edit_2511.json | `TextEncodeQwenImageEditPlus` → `FluxKontextMultiReferenceLatentMethod` → KSampler という中間ノード経由で textMap にプロンプトが登録されない |
| video_wan2_2_14B_animate.json | トップレベルの CLIPTextEncode とサブグラフ内の KSampler がクロスレベル接続され、単一レベル解析では取得できない |
### 変更内容
#### `static/js/metadata-tab.js`
**`extractPromptsFromNodeSet`**
- CLIPTextEncode のテキスト入力リンクのソースノードのタイプ制限を撤廃
- 旧: `isPromptStylerNode` または `WFS_PromptText` のみ
- 新: PromptStyler 以外は任意ノードの `widgets_values[originSlot]`(`ComfySwitchNode`、`PrimitiveStringMultiline` 等を包含)
- テキスト入力名に `"prompt"` を追加(`TextEncodeQwenImageEditPlus` 対応)
- `foundSampler = true` かつ pos / neg が両方空の場合(`SamplerCustomAdvanced` 等)、
textMap のテキストを `{ positives: [], negatives: [], texts: [...] }` として返すよう変更
(旧: `return null`)
**`extractPromptsLiteGraph`**
- ステップ4 `PrimitiveStringMultiline` — `nodes`(トップのみ)→ `collectAllNodes(wf)`(サブグラフ含む)に変更
- 戻り値: `{ positives: primTexts, negatives: [] }` → `{ positives: [], negatives: [], texts: primTexts }`
- ステップ7 最終フォールバック — `nodes`(トップのみ)→ `collectAllNodes(wf)`(サブグラフ含む)に変更
- 戻り値: `{ positives: all, negatives: all }` → `{ positives: [], negatives: [], texts: all }`
**`extractPromptsAPI`**
- フォールバック時の戻り値: `{ positives: all, negatives: all }` → `{ positives: [], negatives: [], texts: all }`
**UI**
- `buildPromptItem`: `type === "text"` のときバッジ(POS/NEG)を表示しない
- `buildPromptItem`: fullLabel に `t("metaPromptText")` を使用
- `handleFile`: `(meta.texts ?? []).map(p => ({ type: "text", text: p }))` を allPrompts に追加
#### `static/js/i18n.js`
- `metaPromptText` を EN / JA / ZH に追加("Text" / "テキスト" / "文本")
### 対応済みワークフロー(data/data9)
| ファイル | 適用ステップ | プロンプト結果 |
|---|---|---|
| image_flux2_fp8.json | step5(サブグラフ + SamplerCustomAdvanced fallback) | texts(バッジなし) |
| image_flux2_klein_image_edit_4b_distilled.json | step5(サブグラフ CLIPTextEncode + KSampler) | texts(バッジなし) |
| image_ernie_image.json | step4(PrimitiveStringMultiline in subgraph) | texts(バッジなし) |
| image_qwen_image_edit_2511.json | step5(TextEncodeQwenImageEditPlus fallback) | texts(バッジなし) |
| video_wan2_2_14B_animate.json | step7(全ノード fallback) | texts(バッジなし) |
---
## 2026-05-15: v0.3.7 Metadata タブ: Flux2 / Qwen / Z-Image 対応
### 概要
ComfyUI 公式テンプレート(Flux.2 Dev/Klein、Qwen-Image-Edit、Z-Image)で使われる
**サブグラフ(subgraph)形式**ワークフローの Metadata タブ読み込みに対応。
- `definitions.subgraphs` 内の `UNETLoader` / `CLIPLoader` / `VAELoader` / `LoraLoaderModelOnly` からモデルを抽出
- `MarkdownNote` ノードの `**section** → - [model](url)` パターンからモデルを補完(フォールバック)
- サブグラフ内の `CLIPTextEncode + KSampler` リンクを辿ってプロンプトを取得
- `PrimitiveStringMultiline` ノード(flux2-klein など)からプロンプトを取得
### 変更内容
#### `static/js/metadata-tab.js`
- **`extractPromptsFromNodeSet(nodes, links)`** 新規追加
CLIPTextEncode + KSampler によるプロンプト抽出ロジックを汎用化。トップレベル・サブグラフの両方で再利用
- **`extractMarkdownNoteModels(wf)`** 新規追加
`MarkdownNote` の `**section**` → `- [name](url)` を正規表現で解析し、text_encoders / diffusion_models / vae / loras を抽出
- **`extractCheckpoints` / `extractVAEs`** を `collectAllNodes()` 使用に修正
サブグラフ内のノードも走査対象に
- **`extractPromptsLiteGraph`** を 7 段階フォールバック構成に再設計
1. ImageMetadataPromptLoader → 2. WFS_PromptText → 3. トップレベルCLIPTextEncode+KSampler →
4. PrimitiveStringMultiline → **5. サブグラフCLIPTextEncode+KSampler(新規)** →
6. PromptStyler → 7. 全CLIPTextEncodeテキスト
- **`fromWorkflow`** に MarkdownNote フォールバック補完を追加
### 対応済みワークフロー(data/data8)
| ファイル | モデル抽出 | プロンプト抽出 |
|---|---|---|
| image_flux2_fp8.json | UNETLoader + CLIPLoader + VAELoader + LoraLoader (subgraph) | CLIPTextEncode (subgraph) |
| image_flux2_klein_text_to_image.json | UNETLoader + CLIPLoader + VAELoader (subgraph) | PrimitiveStringMultiline |
| image_qwen_image_edit.json | UNETLoader + CLIPLoader + VAELoader + LoraLoader (subgraph) | なし(画像入力のみ) |
| image_qwen_image_edit_2511.json | 同上 | なし |
| image_qwen_image_layered.json | UNETLoader + CLIPLoader + VAELoader (subgraph) | PrimitiveStringMultiline |
| image_z_image.json | UNETLoader + CLIPLoader + VAELoader (subgraph) | CLIPTextEncode (subgraph) |
| image_z_image_turbo.json | 同上 | CLIPTextEncode (subgraph) |
#### `static/js/i18n.js` + `templates/index.html`
- `metaFormatTodo` を「Flux.2 / Qwen-Image / Z-Image サブグラフ形式に対応済み」に更新(EN/JA/ZH)
- `helpMetadata5` の対応フォーマット一覧を更新し Flux.2 / Qwen-Image / Z-Image を明記(EN/JA/ZH)
- index.html フォールバックテキストも同内容に更新
---
## 2026-05-15: v0.3.6 Metadata タブ追加
### 概要
+19 -7
View File
@@ -5,7 +5,7 @@ A comprehensive workflow management and generation UI plugin for [ComfyUI](https
Browse, organize, and execute workflows directly from a dedicated studio interface — without switching between windows or manually editing JSON.
![Workflow Studio](https://img.shields.io/badge/ComfyUI-Custom_Node-blue)
![Version](https://img.shields.io/badge/version-0.3.6-green)
![Version](https://img.shields.io/badge/version-0.3.8-green)
## Screenshots
@@ -92,15 +92,15 @@ Requires the **[comfyui-image-feeder](https://github.com/ketle-man/comfyui-image
- **Wildcard input toolbar** — one-click buttons to insert `{|}`, `{n$|}`, `__|__`, `<lora::1:LBW=;>` and other wildcard syntax; wraps selected text when applicable
- **Wildcard file manager** — create, view, and edit `.txt` / `.yaml` wildcard files stored in `user/default/Workflow-Studio/wildcard/`; click a filename in the file picker to insert `__filename__` at cursor
### Metadata Tab (v0.3.6)
### Metadata Tab (v0.3.8)
- **3-column layout** — Drop zone (left) | Model info (center) | LoRA + Prompt (right)
- **File drop** — drop a ComfyUI-generated PNG / WebP or workflow JSON onto the drop zone (or click to open a file picker); PNG/WebP images are shown as a preview
- **Model extraction** — automatically extracts Checkpoint, VAE, Diffusion Model, and Text Encoder names from the embedded workflow metadata
- **Model extraction** — automatically extracts Checkpoint, VAE, Diffusion Model, and Text Encoder names from the workflow; supports both standard and subgraph-based workflows (Flux.2 Dev/Klein, Qwen-Image-Edit/2511/Layered, Z-Image Base/Turbo, Ernie Image, WAN2.2)
- **LoRA extraction** — lists all LoRA models with `strength_model / strength_clip` values
- **Prompt extraction** — lists positive and negative prompts with POS/NEG badges; click any entry to view the full text below
- **Prompt extraction** — lists prompts with POS / NEG badges when positive/negative can be determined; when distinction is not possible (e.g. `SamplerCustomAdvanced`, intermediate nodes, cross-level connections), prompts are shown without a badge as plain **Text**; click any entry to view the full text below
- **Prompt actions** — Copy to clipboard, **GenUI:P/N** (set GenerateUI positive/negative prompt), **Prompt:P/N** (set Prompt tab preset positive/negative)
- **Format support** — ComfyUI (PNG/WebP/JSON), SD WebUI, SD Forge, Fooocus; support for additional model types beyond SD1.5/SDXL/Illustrious is planned
- **Format note** — supported formats are always shown in the left column; planned formats are noted in amber
- **Format support** — ComfyUI PNG/WebP/JSON (standard + Flux.2 / Qwen-Image / Z-Image / Ernie Image / WAN2.2 subgraph workflows), SD WebUI, SD Forge, Fooocus
- **Format note** — supported formats and covered model types are always shown in the left column
### Settings Tab
- **2-column layout** — left column for all settings; right column shows the Theme panel fixed in place (sticky)
@@ -245,11 +245,23 @@ Click the **camera icon** (next to the W button) in ComfyUI's top bar to capture
## Changelog
### v0.3.8
- **Metadata Tab: extended subgraph support** — added extraction for Flux.2 Dev fp8 (`SamplerCustomAdvanced`), Flux.2 Klein 4B Distilled (image edit), Ernie Image, Qwen-Image-Edit 2511 (`TextEncodeQwenImageEditPlus`), and WAN2.2 14B Animate (top-level CLIPTextEncode + subgraph KSampler)
- **Prompt type: Text (no badge)** — when positive/negative polarity cannot be determined (indirect sampler connections, cross-level links, `PrimitiveStringMultiline` etc.), prompts are listed without a POS/NEG badge; action buttons still work normally
- **Prompt extraction improvements** — `extractPromptsFromNodeSet` now resolves link sources from any STRING-output node (e.g. `ComfySwitchNode`, `PrimitiveStringMultiline`) without type restriction; `PrimitiveStringMultiline` search and final text-encoder fallback both extended to include subgraph nodes via `collectAllNodes`
- **i18n** — added `metaPromptText` key (EN: "Text" / JA: "テキスト" / ZH: "文本")
### v0.3.7
- **Metadata Tab: subgraph workflow support** — Flux.2 (Dev fp8 / Klein), Qwen-Image-Edit (/ 2511 / Layered), and Z-Image (Base / Turbo) official templates now correctly extract models and prompts; these use ComfyUI's `definitions.subgraphs` format with `UNETLoader`, `CLIPLoader`, and `VAELoader` inside the subgraph
- **Prompt extraction improvements** — 7-stage fallback chain: ImageMetadataPromptLoader → WFS_PromptText → top-level CLIPTextEncode+KSampler → `PrimitiveStringMultiline` (flux2-klein) → subgraph CLIPTextEncode+KSampler → PromptStyler → all CLIPTextEncode text
- **MarkdownNote model fallback** — `extractMarkdownNoteModels()` parses `**section** → - [model](url)` patterns from MarkdownNote nodes as a supplemental model source
- **Format note updated** — "Flux.2 / Qwen-Image / Z-Image subgraph workflows supported" (EN/JA/ZH)
### v0.3.6
- **Metadata Tab** — new tab (between Prompt and Gallery) that extracts model and prompt information from ComfyUI-generated PNG/WebP images and workflow JSON files; 3-column layout: drop zone with image preview (left), Checkpoint/VAE/Diffusion Model/Text Encoder lists (center), LoRA list with strength values + Prompt list with full-text viewer (right)
- **Prompt action buttons** — Copy, GenUI:P/N (set GenerateUI prompts), Prompt:P/N (set Prompt tab preset prompts) directly from the extracted prompt text
- **SDXL Prompt Styler support** — improved prompt extraction resolves `PromptStyler → CLIPTextEncode` link chains in both LiteGraph and API workflow formats
- **Format support** — ComfyUI (PNG/WebP/JSON), SD WebUI, SD Forge, Fooocus; format note in the left column indicates planned support for additional model types
- **Format support** — ComfyUI (PNG/WebP/JSON), SD WebUI, SD Forge, Fooocus
### v0.3.5
- **Feeder subtab** — new subtab in GenerateUI for controlling [comfyui-image-feeder](https://github.com/ketle-man/comfyui-image-feeder) nodes; left pane: node selector + all ImageFeeder parameters + presets + Run/Stop controls; center pane: folder tree, image grid with checkbox selection, and preview panel
Binary file not shown.

After

Width:  |  Height:  |  Size: 141 KiB

+1 -1
View File
@@ -1,7 +1,7 @@
[project]
name = "comfyui-workflow-studio"
description = "A comprehensive workflow management and generation UI plugin for ComfyUI. Browse, organize, and execute workflows from a dedicated studio interface."
version = "0.3.6"
version = "0.3.7"
license = "MIT"
requires-python = ">= 3.10"
dependencies = ["jinja2"]
+9 -6
View File
@@ -558,19 +558,20 @@ const LANGUAGES = {
metaParseError: "Failed to parse metadata",
metaPromptPositive: "Positive",
metaPromptNegative: "Negative",
metaPromptText: "Text",
metaSelectPrompt: "Select a prompt above to view full text...",
metaFormatNoteTitle: "Supported formats",
metaFmtComfyui: "ComfyUI — PNG / WebP / JSON",
metaFmtSdwebui: "SD WebUI / SD Forge — PNG (parameters)",
metaFmtFooocus: "Fooocus — PNG (fooocus_scheme)",
metaFormatTodo: "More formats planned (non-SD1.5/SDXL/Illustrious models)",
metaFormatTodo: "Flux.2 / Qwen-Image / Z-Image subgraph workflows supported",
helpMetadataTitle: "Metadata Tab",
helpMetadataDesc: "Drop a ComfyUI-generated PNG/WebP or workflow JSON to instantly extract and view the embedded model and prompt information.",
helpMetadata1: "Left column: drop zone (image preview for PNG/WebP) and file info; click or drop to load",
helpMetadata2: "Center column: Checkpoint, VAE, Diffusion Model, and Text Encoder lists extracted from the workflow",
helpMetadata3: "Right column: LoRA list (with strength values) and Prompt list (click to view full text below)",
helpMetadata4: "Prompt actions — Copy: copy to clipboard; GenUI:P/N: set GenerateUI positive/negative prompt; Prompt:P/N: set Prompt tab preset positive/negative",
helpMetadata5: "Supported formats: ComfyUI (PNG/WebP/JSON), SD WebUI, SD Forge, Fooocus. Support for additional model types (beyond SD1.5/SDXL/Illustrious) is planned.",
helpMetadata5: "Supported: ComfyUI PNG/WebP/JSON (standard + Flux.2 / Qwen-Image / Z-Image subgraph workflows), SD WebUI, SD Forge, Fooocus.",
},
ja: {
@@ -1127,19 +1128,20 @@ const LANGUAGES = {
metaParseError: "メタデータの解析に失敗しました",
metaPromptPositive: "ポジティブ",
metaPromptNegative: "ネガティブ",
metaPromptText: "テキスト",
metaSelectPrompt: "上のプロンプトを選択すると全文を表示します...",
metaFormatNoteTitle: "対応フォーマット",
metaFmtComfyui: "ComfyUI — PNG / WebP / JSON",
metaFmtSdwebui: "SD WebUI / SD Forge — PNG(parameters チャンク)",
metaFmtFooocus: "Fooocus — PNG(fooocus_scheme チャンク)",
metaFormatTodo: "対応予定: SD1.5 / SDXL / Illustrious 以外のモデル形式も順次対応",
metaFormatTodo: "Flux.2 / Qwen-Image / Z-Image サブグラフ形式に対応済み",
helpMetadataTitle: "Metadata タブ",
helpMetadataDesc: "ComfyUI で生成した PNG / WebP または ワークフロー JSON をドロップすると、埋め込まれたモデル・プロンプト情報を即座に抽出・表示します。",
helpMetadata1: "左列: ドロップゾーン(PNG/WebP は画像プレビュー表示)+ファイル情報・対応フォーマット説明",
helpMetadata2: "中列: ワークフローから抽出した Checkpoint・VAE・Diffusion Model・Text Encoder の一覧",
helpMetadata3: "右列: LoRA 一覧(強度値付き)+プロンプト一覧(クリックで下部に全文表示)",
helpMetadata4: "プロンプトアクション — Copy: クリップボードにコピー; GenUI:P/N: GenerateUI のポジティブ/ネガティブプロンプトにセット; Prompt:P/N: Prompt タブ Presets のポジティブ/ネガティブにセット",
helpMetadata5: "対応フォーマット: ComfyUI(PNG/WebP/JSON)・SD WebUI・SD Forge・Fooocus。SD1.5/SDXL/Illustrious 以外のモデル形式への対応は今後予定。",
helpMetadata5: "対応フォーマット: ComfyUI PNG/WebP/JSON(標準形式 + Flux.2 / Qwen-Image / Z-Image サブグラフ形式)・SD WebUI・SD Forge・Fooocus。",
},
zh: {
@@ -1692,19 +1694,20 @@ const LANGUAGES = {
metaParseError: "元数据解析失败",
metaPromptPositive: "正向提示",
metaPromptNegative: "负向提示",
metaPromptText: "文本",
metaSelectPrompt: "选择上方提示词以查看全文...",
metaFormatNoteTitle: "支持格式",
metaFmtComfyui: "ComfyUI — PNG / WebP / JSON",
metaFmtSdwebui: "SD WebUI / SD Forge — PNG(parameters)",
metaFmtFooocus: "Fooocus — PNG(fooocus_scheme)",
metaFormatTodo: "更多格式规划中(SD1.5/SDXL/Illustrious 以外模型)",
metaFormatTodo: "已支持 Flux.2 / Qwen-Image / Z-Image 子图工作流",
helpMetadataTitle: "Metadata 标签页",
helpMetadataDesc: "拖放 ComfyUI 生成的 PNG/WebP 或工作流 JSON,即可提取并显示其中的模型和提示词信息。",
helpMetadata1: "左列:拖放区(PNG/WebP 显示图片预览)及文件信息与支持格式说明",
helpMetadata2: "中列:从工作流中提取的 Checkpoint、VAE、扩散模型、文本编码器列表",
helpMetadata3: "右列:LoRA 列表(含强度值)和提示词列表(点击在下方显示全文)",
helpMetadata4: "提示词操作 — Copy:复制到剪贴板;GenUI:P/N:设置 GenerateUI 正/负向提示;Prompt:P/N:设置 Prompt 标签页预设正/负向提示",
helpMetadata5: "支持格式:ComfyUI(PNG/WebP/JSON)、SD WebUI、SD Forge、Fooocus。SD1.5/SDXL/Illustrious 以外模型格式的支持正在规划中。",
helpMetadata5: "支持格式:ComfyUI PNG/WebP/JSON(标准格式 + Flux.2 / Qwen-Image / Z-Image 子图工作流)、SD WebUI、SD Forge、Fooocus。",
},
};
+167 -83
View File
@@ -112,12 +112,12 @@ const VAE_NONE = "None";
function extractCheckpoints(wf) {
if (!wf || typeof wf !== "object") return [];
if (Array.isArray(wf.nodes)) return collectUnique(wf.nodes.filter(n => n.type?.toLowerCase().includes("checkpoint") || META_NODE_TYPES.has(n.type)).map(n => n.widgets_values?.[0]));
if (Array.isArray(wf.nodes)) return collectUnique(collectAllNodes(wf).filter(n => n.type?.toLowerCase().includes("checkpoint") || META_NODE_TYPES.has(n.type)).map(n => n.widgets_values?.[0]));
return collectUnique(Object.values(wf).filter(n => n?.class_type?.toLowerCase().includes("checkpoint") || META_NODE_TYPES.has(n?.class_type)).map(n => n.inputs?.ckpt_name));
}
function extractVAEs(wf) {
if (!wf || typeof wf !== "object") return [];
if (Array.isArray(wf.nodes)) return collectUnique(wf.nodes.flatMap(n => { if (n.type === "VAELoader") return [n.widgets_values?.[0]]; if (META_NODE_TYPES.has(n.type ?? "")) { const v = n.widgets_values?.[1]; return v && v !== VAE_NONE ? [v] : []; } return []; }));
if (Array.isArray(wf.nodes)) return collectUnique(collectAllNodes(wf).flatMap(n => { if (n.type === "VAELoader") return [n.widgets_values?.[0]]; if (META_NODE_TYPES.has(n.type ?? "")) { const v = n.widgets_values?.[1]; return v && v !== VAE_NONE ? [v] : []; } return []; }));
return collectUnique(Object.values(wf).flatMap(n => { if (!n || typeof n !== "object") return []; if (n.class_type === "VAELoader") return [n.inputs?.vae_name]; if (META_NODE_TYPES.has(n.class_type ?? "")) { const v = n.inputs?.vae_name; return v && v !== VAE_NONE ? [v] : []; } return []; }));
}
function extractDiffusionModels(wf) {
@@ -181,6 +181,113 @@ function extractPrompts(wf) {
return Array.isArray(wf.nodes) ? extractPromptsLiteGraph(wf) : extractPromptsAPI(wf);
}
// CLIPTextEncode + KSampler でプロンプト抽出(トップレベル・サブグラフ共用)
// サンプラーが見つかりテキストが取れた場合のみ非null を返す
function extractPromptsFromNodeSet(nodes, links) {
const nodeMap = new Map();
for (const n of nodes) nodeMap.set(n.id, n);
const linkOrigin = new Map(), linkSlot = new Map();
if (Array.isArray(links)) {
for (const lk of links) {
if (Array.isArray(lk)) {
linkOrigin.set(lk[0], lk[1]);
linkSlot.set(lk[0], lk[2] ?? 0);
} else if (lk && typeof lk === "object") {
const id = lk.id ?? lk[0], origin = lk.origin_id ?? lk[1], slot = lk.origin_slot ?? lk[2] ?? 0;
if (id != null && origin != null) { linkOrigin.set(id, origin); linkSlot.set(id, slot); }
}
}
}
const textMap = new Map();
for (const n of nodes) {
if (!isTextEncoderNode(n.type ?? "")) continue;
const text = n.widgets_values?.[0];
if (text && typeof text === "string") {
textMap.set(n.id, text);
} else if (Array.isArray(n.inputs)) {
const textInput = n.inputs.find(inp => inp.name === "text" || inp.name === "text_g" || inp.name === "prompt");
if (textInput?.link != null) {
const originId = linkOrigin.get(textInput.link);
const originSlot = linkSlot.get(textInput.link) ?? 0;
const srcNode = originId != null ? nodeMap.get(originId) : null;
if (srcNode) {
const srcType = srcNode.type ?? "";
if (isPromptStylerNode(srcType)) {
const v = srcNode.widgets_values?.[originSlot];
if (v && typeof v === "string") textMap.set(n.id, v);
} else {
// WFS_PromptText, PrimitiveStringMultiline, ComfySwitchNode 等、任意の STRING 出力ノード
const v = srcNode.widgets_values?.[originSlot] ?? srcNode.widgets_values?.[0];
if (v && typeof v === "string") textMap.set(n.id, v);
}
}
}
}
}
const pos = new Set(), neg = new Set();
let foundSampler = false;
for (const n of nodes) {
if (!isSamplerNode(n.type ?? "") || !Array.isArray(n.inputs)) continue;
foundSampler = true;
for (const inp of n.inputs) {
if (!inp || inp.link == null) continue;
const originId = linkOrigin.get(inp.link);
if (originId == null) continue;
const text = textMap.get(originId);
if (!text) continue;
const name = inp.name ?? "";
if (name === "positive" || name.startsWith("positive")) pos.add(text);
else if (name === "negative" || name.startsWith("negative")) neg.add(text);
}
}
if (!foundSampler) return null;
// SamplerCustomAdvanced などで positive/negative 直結がない場合、判別不能テキストとして返す
if (pos.size === 0 && neg.size === 0) {
const allTexts = [...textMap.values()].filter(t => t.trim());
if (allTexts.length > 0) return { positives: [], negatives: [], texts: allTexts };
return null;
}
return { positives: [...pos], negatives: [...neg] };
}
// MarkdownNote の **section** → - [name](url) パターンからモデルを抽出
// flux/qwen/z-image などのサブグラフ形式ワークフロー向け補完用
function extractMarkdownNoteModels(wf) {
const allNodes = [];
if (Array.isArray(wf.nodes)) allNodes.push(...wf.nodes);
for (const sg of wf.definitions?.subgraphs ?? []) if (Array.isArray(sg.nodes)) allNodes.push(...sg.nodes);
const result = { checkpoints: [], vaes: [], diffusionModels: [], textEncoders: [], loras: [] };
const seen = { checkpoints: new Set(), vaes: new Set(), diffusionModels: new Set(), textEncoders: new Set(), loras: new Set() };
function addU(arr, set, name) { if (name && typeof name === "string" && !set.has(name)) { set.add(name); arr.push(name); } }
for (const n of allNodes) {
if (n.type !== "MarkdownNote") continue;
const raw = n.widgets_values;
const text = Array.isArray(raw) ? raw[0] : (typeof raw === "string" ? raw : null);
if (!text) continue;
const sRe = /\*\*([^*\n]+)\*\*/g;
let sm;
while ((sm = sRe.exec(text)) !== null) {
const sec = sm[1].trim().toLowerCase().replace(/\s+/g, "_");
if (!["text_encoders", "diffusion_models", "vae", "checkpoints", "loras"].includes(sec)) continue;
const rest = text.slice(sm.index + sm[0].length);
const end = rest.search(/\n\*\*|\n##/);
const content = end >= 0 ? rest.slice(0, end) : rest;
const lRe = /^- \[([^\]]+)\]/gm;
let lm;
while ((lm = lRe.exec(content)) !== null) {
const name = lm[1].trim();
if (sec === "text_encoders") addU(result.textEncoders, seen.textEncoders, name);
else if (sec === "diffusion_models") addU(result.diffusionModels, seen.diffusionModels, name);
else if (sec === "vae") addU(result.vaes, seen.vaes, name);
else if (sec === "checkpoints") addU(result.checkpoints, seen.checkpoints, name);
else if (sec === "loras") addU(result.loras, seen.loras, name);
}
}
}
const hasAny = result.checkpoints.length || result.vaes.length || result.diffusionModels.length || result.textEncoders.length || result.loras.length;
return hasAny ? result : null;
}
// API形式: リンク参照 [srcNodeId, slot] からテキストを解決
function resolveLinkedText(wf, srcId, slot) {
const src = wf[String(srcId)];
@@ -234,19 +341,23 @@ function extractPromptsAPI(wf) {
else if (key === "negative" || key.startsWith("negative")) neg.add(text);
}
}
if (!foundSampler || (pos.size === 0 && neg.size === 0)) { const all = [...textMap.values()]; return { positives: all, negatives: all }; }
if (!foundSampler || (pos.size === 0 && neg.size === 0)) { const all = [...textMap.values()].filter(t => t && t.trim()); return { positives: [], negatives: [], texts: all }; }
return { positives: [...pos], negatives: [...neg] };
}
function extractPromptsLiteGraph(wf) {
const { nodes, links } = wf;
if (!Array.isArray(nodes)) return { positives: [], negatives: [] };
// 1. ImageMetadataPromptLoader (WFS専用)
const metaNodes = nodes.filter(n => n.type === "ImageMetadataPromptLoader");
if (metaNodes.length > 0) {
const pos = new Set(), neg = new Set();
for (const n of metaNodes) { const p = n.widgets_values?.[2], ng = n.widgets_values?.[3]; if (p) pos.add(p); if (ng) neg.add(ng); }
if (pos.size > 0 || neg.size > 0) return { positives: [...pos], negatives: [...neg] };
}
// 2. WFS_PromptText (WFS専用)
const wfsNodes = nodes.filter(n => n.type === "WFS_PromptText");
if (wfsNodes.length > 0) {
const pos = new Set(), neg = new Set();
@@ -254,89 +365,47 @@ function extractPromptsLiteGraph(wf) {
if (pos.size > 0 || neg.size > 0) return { positives: [...pos], negatives: [...neg] };
}
const nodeMap = new Map(); for (const n of nodes) nodeMap.set(n.id, n);
// 3. トップレベルの CLIPTextEncode + KSampler
const topResult = extractPromptsFromNodeSet(nodes, links ?? []);
if (topResult) return topResult;
// links[]: [id, origin_id, origin_slot, dest_id, dest_slot, type]
const linkOrigin = new Map(); // linkId → originNodeId
const linkSlot = new Map(); // linkId → originSlot
if (Array.isArray(links)) {
for (const lk of links) {
if (Array.isArray(lk)) {
linkOrigin.set(lk[0], lk[1]);
linkSlot.set(lk[0], lk[2] ?? 0);
} else if (lk && typeof lk === "object") {
const id = lk.id ?? lk[0], origin = lk.origin_id ?? lk[1], slot = lk.origin_slot ?? lk[2] ?? 0;
if (id != null && origin != null) { linkOrigin.set(id, origin); linkSlot.set(id, slot); }
}
}
// 4. PrimitiveStringMultiline(flux2-klein / ernie など)- サブグラフ内も探す
const primTexts = [];
for (const n of collectAllNodes(wf)) {
if (n.type !== "PrimitiveStringMultiline") continue;
const t = Array.isArray(n.widgets_values) ? n.widgets_values[0] : n.widgets_values;
if (t && typeof t === "string" && t.trim()) primTexts.push(t.trim());
}
if (primTexts.length > 0) return { positives: [], negatives: [], texts: primTexts };
// 5. サブグラフ内の CLIPTextEncode + KSampler(z-image / qwen / flux など)
for (const sg of wf.definitions?.subgraphs ?? []) {
if (!Array.isArray(sg.nodes)) continue;
const sgResult = extractPromptsFromNodeSet(sg.nodes, sg.links ?? []);
if (sgResult) return sgResult;
}
// CLIPTextEncode系ノードからテキストを収集
// ・widgets_values[0] に直接入っている場合
// ・inputs にリンクがある場合(PromptStyler → CLIPTextEncode)
const textMap = new Map();
// 6. PromptStyler フォールバック
const stylerPos = new Set(), stylerNeg = new Set();
for (const n of nodes) {
if (!isPromptStylerNode(n.type ?? "")) continue;
const vals = n.widgets_values ?? [];
for (let i = 0; i < vals.length; i++) {
if (typeof vals[i] !== "string" || !vals[i].trim()) continue;
if (i % 2 === 0) stylerPos.add(vals[i]);
else stylerNeg.add(vals[i]);
}
}
if (stylerPos.size > 0 || stylerNeg.size > 0) return { positives: [...stylerPos], negatives: [...stylerNeg] };
// 7. テキストエンコーダー全テキスト(最終フォールバック)- サブグラフも含む・判別不能
const all = [];
for (const n of collectAllNodes(wf)) {
if (!isTextEncoderNode(n.type ?? "")) continue;
const text = n.widgets_values?.[0];
if (text && typeof text === "string") {
textMap.set(n.id, text);
} else if (Array.isArray(n.inputs)) {
// テキスト入力がリンクの場合 → リンク先ノードから解決
const textInput = n.inputs.find(inp => inp.name === "text" || inp.name === "text_g");
if (textInput?.link != null) {
const originId = linkOrigin.get(textInput.link);
const originSlot = linkSlot.get(textInput.link) ?? 0;
const srcNode = originId != null ? nodeMap.get(originId) : null;
if (srcNode) {
const srcType = srcNode.type ?? "";
if (isPromptStylerNode(srcType) || srcType === "WFS_PromptText") {
const v = srcNode.widgets_values?.[originSlot];
if (v && typeof v === "string") textMap.set(n.id, v);
}
}
}
}
const t = n.widgets_values?.[0];
if (t && typeof t === "string" && t.trim()) all.push(t);
}
const pos = new Set(), neg = new Set();
let foundSampler = false;
for (const n of nodes) {
if (!isSamplerNode(n.type ?? "") || !Array.isArray(n.inputs)) continue;
foundSampler = true;
for (const inp of n.inputs) {
if (!inp || inp.link == null) continue;
const originId = linkOrigin.get(inp.link);
if (originId == null) continue;
const text = textMap.get(originId);
if (!text) continue;
const name = inp.name ?? "";
if (name === "positive" || name.startsWith("positive")) pos.add(text);
else if (name === "negative" || name.startsWith("negative")) neg.add(text);
}
}
// フォールバック: サンプラーが見つからないか両方空
if (!foundSampler || (pos.size === 0 && neg.size === 0)) {
// PromptStylerノードを直接スキャン
const stylerPos = new Set(), stylerNeg = new Set();
for (const n of nodes) {
if (!isPromptStylerNode(n.type ?? "")) continue;
// widgets_values からポジティブ/ネガティブを探す
// 多くのPromptStylerはindex 0=positive, 1=negativeまたはスタイル文字列
const vals = n.widgets_values ?? [];
for (let i = 0; i < vals.length; i++) {
if (typeof vals[i] !== "string" || !vals[i].trim()) continue;
if (i % 2 === 0) stylerPos.add(vals[i]);
else stylerNeg.add(vals[i]);
}
}
if (stylerPos.size > 0 || stylerNeg.size > 0) {
return { positives: [...stylerPos], negatives: [...stylerNeg] };
}
const all = [...textMap.values()];
return { positives: all, negatives: all };
}
return { positives: [...pos], negatives: [...neg] };
return { positives: [], negatives: [], texts: all };
}
// ── SD/Fooocus prompt extraction ──────────────────────────────
@@ -372,7 +441,17 @@ async function extractAllMetadata(file) {
const isWebP = file.type === "image/webp" || name.endsWith(".webp");
function fromWorkflow(wf, source) {
return { source, checkpoints: extractCheckpoints(wf), vaes: extractVAEs(wf), diffusionModels: extractDiffusionModels(wf), textEncoders: extractTextEncoders(wf), loras: extractLoRAs(wf), ...extractPrompts(wf) };
const base = { source, checkpoints: extractCheckpoints(wf), vaes: extractVAEs(wf), diffusionModels: extractDiffusionModels(wf), textEncoders: extractTextEncoders(wf), loras: extractLoRAs(wf), ...extractPrompts(wf) };
// MarkdownNote からモデル情報を補完(subgraph形式の flux/qwen/z-image など)
const mdm = extractMarkdownNoteModels(wf);
if (mdm) {
if (!base.checkpoints.length) base.checkpoints = mdm.checkpoints;
if (!base.vaes.length) base.vaes = mdm.vaes;
if (!base.diffusionModels.length) base.diffusionModels = mdm.diffusionModels;
if (!base.textEncoders.length) base.textEncoders = mdm.textEncoders;
if (!base.loras.length) base.loras = mdm.loras;
}
return base;
}
if (isJSON) {
@@ -439,13 +518,17 @@ function buildPromptItem(label, type, full, fullArea, fullLabel, listEl) {
const el = document.createElement("div");
el.className = "wfm-meta-item wfm-meta-item-clickable";
const snippet = label.length > 60 ? label.slice(0, 60) + "…" : label;
const typeBadge = type === "positive" ? `<span class="wfm-meta-badge-pos">POS</span>` : `<span class="wfm-meta-badge-neg">NEG</span>`;
const typeBadge = type === "positive" ? `<span class="wfm-meta-badge-pos">POS</span>`
: type === "negative" ? `<span class="wfm-meta-badge-neg">NEG</span>`
: ``;
el.innerHTML = `${typeBadge}<span class="wfm-meta-item-name">${escapeHtml(snippet)}</span>`;
el.addEventListener("click", () => {
listEl.querySelectorAll(".wfm-meta-item-clickable").forEach(e => e.classList.remove("selected"));
el.classList.add("selected");
fullArea.value = full;
fullLabel.textContent = type === "positive" ? t("metaPromptPositive") : t("metaPromptNegative");
fullLabel.textContent = type === "positive" ? t("metaPromptPositive")
: type === "negative" ? t("metaPromptNegative")
: t("metaPromptText") || "Text";
});
return el;
}
@@ -589,6 +672,7 @@ export function initMetadataTab() {
const allPrompts = [
...meta.positives.map(p => ({ type: "positive", text: p })),
...meta.negatives.map(p => ({ type: "negative", text: p })),
...(meta.texts ?? []).map(p => ({ type: "text", text: p })),
];
if (allPrompts.length > 0) {
promptSection.classList.remove("wfm-meta-section-empty");
+2 -2
View File
@@ -723,7 +723,7 @@
<li id="wfm-meta-fmt-sdwebui">SD WebUI / SD Forge — PNG (parameters)</li>
<li id="wfm-meta-fmt-fooocus">Fooocus — PNG (fooocus_scheme)</li>
</ul>
<div class="wfm-meta-format-todo" id="wfm-meta-format-todo">More formats planned (non-SD1.5/SDXL/Illustrious models)</div>
<div class="wfm-meta-format-todo" id="wfm-meta-format-todo">Flux.2 / Qwen-Image / Z-Image subgraph workflows supported</div>
</div>
<input type="file" id="wfm-meta-file-input" accept="image/png,image/webp,.json,application/json" style="display:none;">
</div>
@@ -864,7 +864,7 @@
<li id="wfm-help-metadata-2">Center column: Checkpoint, VAE, Diffusion Model, and Text Encoder lists extracted from the workflow</li>
<li id="wfm-help-metadata-3">Right column: LoRA list (with strength values) and Prompt list (click to view full text below)</li>
<li id="wfm-help-metadata-4">Prompt actions — Copy: copy to clipboard; GenUI:P/N: set GenerateUI positive/negative prompt; Prompt:P/N: set Prompt tab preset positive/negative</li>
<li id="wfm-help-metadata-5">Supported formats: ComfyUI (PNG/WebP/JSON), SD WebUI, SD Forge, Fooocus. Support for additional model types (beyond SD1.5/SDXL/Illustrious) is planned.</li>
<li id="wfm-help-metadata-5">Supported: ComfyUI PNG/WebP/JSON (standard + Flux.2 / Qwen-Image / Z-Image subgraph workflows), SD WebUI, SD Forge, Fooocus.</li>
</ul>
</div>