@heripo/pdf-parser 0.1.39 → 0.1.41

package/README.ko.md

@@ -65,7 +65,7 @@
 brew install node
 ```
 
-#### 2. pnpm >= 11.0.0
+#### 2. pnpm >= 11
 
 ```bash
 npm install -g pnpm
@@ -653,9 +653,13 @@ type PDFConvertOptions = {
 interface PDFCorrectionOptions {
   models: {
     textCorrection: LanguageModel; // 필수: 페이지 텍스트와 표 셀 OCR 보정
+    textCorrectionFallback?: LanguageModel; // 선택: textCorrection 실패 시 fallback
     pageGate: LanguageModel; // 필수: 구조 Review Assistance page gate
+    pageGateFallback?: LanguageModel; // 선택: pageGate 실패 시 fallback
     reviewAssistance: LanguageModel; // 필수: 기본 구조 review 모델
+    reviewAssistanceFallback?: LanguageModel; // 선택: reviewAssistance 공통 fallback
     tableCorrection?: LanguageModel; // 선택: 표 전용 보정 override
+    tableCorrectionFallback?: LanguageModel; // 선택: tableCorrection 실패 시 fallback
     reviewAssistanceTasks?: Partial<
       Record<
         | 'text_ocr_hanja'
@@ -667,6 +671,17 @@ interface PDFCorrectionOptions {
         LanguageModel
       >
     >;
+    reviewAssistanceTasksFallback?: Partial<
+      Record<
+        | 'text_ocr_hanja'
+        | 'text_integrity'
+        | 'text_role_footnote'
+        | 'tables'
+        | 'pictures_captions'
+        | 'layout_bbox_order',
+        LanguageModel
+      >
+    >;
   };
   concurrency?: {
     pages?: number; // 페이지 단위 text correction/page gate 동시성
@@ -687,6 +702,9 @@ interface PDFCorrectionOptions {
   };
   autoApplyThreshold?: number; // 자동 적용 최소 confidence (기본값: 0.85)
   proposalThreshold?: number; // sidecar proposal 최소 confidence (기본값: 0.5)
+  forceAutoApply?: boolean; // true일 경우, confidence 임계값에 관계없이 review-assistance 명령을 자동 적용 (기본값: false)
+  reviewAssistanceEnabled?: boolean; // false일 경우, review-assistance 단계를 완전히 건너뜀 (기본값: true)
+  tableCorrectionEnabled?: boolean; // false일 경우, table-correction 작업을 건너뜀 (기본값: true)
   temperature?: number; // VLM 생성 temperature (기본값: 0)
 }
 ```

package/README.md

@@ -65,7 +65,7 @@
 brew install node
 ```
 
-#### 2. pnpm >= 11.0.0
+#### 2. pnpm >= 11
 
 ```bash
 npm install -g pnpm
@@ -653,9 +653,13 @@ type PDFConvertOptions = {
 interface PDFCorrectionOptions {
   models: {
     textCorrection: LanguageModel; // Required: page text and table-cell OCR correction
+    textCorrectionFallback?: LanguageModel; // Optional: fallback for text correction
     pageGate: LanguageModel; // Required: structural Review Assistance page gate
+    pageGateFallback?: LanguageModel; // Optional: fallback for page gate
     reviewAssistance: LanguageModel; // Required: default structural review model
+    reviewAssistanceFallback?: LanguageModel; // Optional: common fallback for review assistance
     tableCorrection?: LanguageModel; // Optional: table-specific correction override
+    tableCorrectionFallback?: LanguageModel; // Optional: fallback for table correction
     reviewAssistanceTasks?: Partial<
       Record<
         | 'text_ocr_hanja'
@@ -667,6 +671,17 @@ interface PDFCorrectionOptions {
         LanguageModel
       >
     >;
+    reviewAssistanceTasksFallback?: Partial<
+      Record<
+        | 'text_ocr_hanja'
+        | 'text_integrity'
+        | 'text_role_footnote'
+        | 'tables'
+        | 'pictures_captions'
+        | 'layout_bbox_order',
+        LanguageModel
+      >
+    >;
   };
   concurrency?: {
     pages?: number; // Page-level text correction and page gate concurrency
@@ -687,6 +702,9 @@ interface PDFCorrectionOptions {
   };
   autoApplyThreshold?: number; // Minimum confidence for direct mutation (default: 0.85)
   proposalThreshold?: number; // Minimum confidence for sidecar proposal (default: 0.5)
+  forceAutoApply?: boolean; // When true, review-assistance commands auto-apply regardless of threshold (default: false)
+  reviewAssistanceEnabled?: boolean; // When false, review-assistance is skipped completely (default: true)
+  tableCorrectionEnabled?: boolean; // When false, table-correction is skipped (default: true)
   temperature?: number; // VLM generation temperature (default: 0)
 }
 ```

package/dist/index.js

@@ -64,7 +64,7 @@ var IMAGE_PDF_CONVERTER = {
   QUALITY: 100
 };
 
-// ../shared/dist/index.mjs
+// ../shared/dist/index.js
 import { spawn } from "child_process";
 import {
   NoObjectGeneratedError,
@@ -309,7 +309,7 @@ var LLMCaller = class {
     return this.executeWithFallback(
       config,
       (model) => this.generateStructuredOutput(model, config.schema, {
-        system: config.systemPrompt,
+        instructions: config.systemPrompt,
         prompt: config.userPrompt,
         temperature: config.temperature,
         maxRetries: config.maxRetries,
@@ -1328,8 +1328,8 @@ var PdfLanguageDetector = class {
               text: KOREAN_DOCUMENT_DETECTION_PROMPT
             },
             {
-              type: "image",
-              image: imageData,
+              type: "file",
+              data: imageData,
               mediaType: "image/png"
             }
           ]
@@ -3309,8 +3309,8 @@ var ReviewAssistancePageGate = class {
               text: this.buildPrompt(context, options.outputLanguage)
             },
             {
-              type: "image",
-              image,
+              type: "file",
+              data: image,
               mediaType: "image/png"
             }
           ]
@@ -3582,8 +3582,8 @@ var PostDoclingPageProcessor = class {
                 text: fullPrompt
               },
               {
-                type: "image",
-                image,
+                type: "file",
+                data: image,
                 mediaType: "image/png"
               }
             ]
@@ -9207,8 +9207,8 @@ var ReviewAssistanceRunner = class {
                 content: [
                   { type: "text", text: prompt },
                   {
-                    type: "image",
-                    image,
+                    type: "file",
+                    data: image,
                     mediaType: "image/png"
                   }
                 ]
@@ -10033,8 +10033,8 @@ var ReviewAssistanceRunner = class {
             content: [
               { type: "text", text: prompt },
               {
-                type: "image",
-                image,
+                type: "file",
+                data: image,
                 mediaType: "image/png"
               }
             ]