A fast, helpful, and open-source document parser
perf: disable image extraction when --no-ocr
Logan Markewicha703c0e9fc3a
diff --git a/src/core/parser.ts b/src/core/parser.tsindex a7b7439..2557c0f 100644--- a/src/core/parser.ts+++ b/src/core/parser.ts@@ -162,7 +162,8 @@ export class LiteParse {const pages = await this.pdfEngine.extractAllPages(doc,this.config.maxPages,- this.config.targetPages+ this.config.targetPages,+ { extractImages: this.config.ocrEnabled });// run BEFORE grid projection@@ -310,7 +311,7 @@ export class LiteParse {log(`Rendering page ${pageNum}...`);const imageBuffer = await renderer.renderPageToBuffer(pdfInput, pageNum, this.config.dpi);- const pageData = await this.pdfEngine.extractPage(doc, pageNum);+ const pageData = await this.pdfEngine.extractPage(doc, pageNum, { extractImages: false });results.push({pageNum,diff --git a/src/engines/pdf/interface.ts b/src/engines/pdf/interface.tsindex 36e6844..94bd0ab 100644--- a/src/engines/pdf/interface.ts+++ b/src/engines/pdf/interface.ts@@ -1,10 +1,21 @@import { TextItem } from "../../core/types.js";+/** Options for page extraction */+export interface ExtractOptions {+ /** Whether to extract embedded image bounds (needed for OCR). Default: true */+ extractImages?: boolean;+}+export interface PdfEngine {name: string;loadDocument(input: string | Uint8Array, password?: string): Promise<PdfDocument>;- extractPage(doc: PdfDocument, pageNum: number): Promise<PageData>;- extractAllPages(doc: PdfDocument, maxPages?: number, targetPages?: string): Promise<PageData[]>;+ extractPage(doc: PdfDocument, pageNum: number, options?: ExtractOptions): Promise<PageData>;+ extractAllPages(+ doc: PdfDocument,+ maxPages?: number,+ targetPages?: string,+ options?: ExtractOptions+ ): Promise<PageData[]>;renderPageImage(doc: PdfDocument,pageNum: number,diff --git a/src/engines/pdf/pdfjs.ts b/src/engines/pdf/pdfjs.tsindex 210b12a..287084b 100644--- a/src/engines/pdf/pdfjs.ts+++ b/src/engines/pdf/pdfjs.ts@@ -1,5 +1,13 @@import fs from "node:fs/promises";-import { PdfEngine, PdfDocument, PageData, Image, Annotation, BoundingBox } from "./interface.js";+import {+ PdfEngine,+ PdfDocument,+ PageData,+ Image,+ Annotation,+ BoundingBox,+ ExtractOptions,+} from "./interface.js";import { TextItem } from "../../core/types.js";import { PdfiumRenderer } from "./pdfium-renderer.js";import { importPdfJs } from "./pdfjsImporter.js";@@ -795,7 +803,7 @@ export class PdfJsEngine implements PdfEngine {} as PdfJsExtendedDocument;}- async extractPage(doc: PdfDocument, pageNum: number): Promise<PageData> {+ async extractPage(doc: PdfDocument, pageNum: number, options?: ExtractOptions): Promise<PageData> {const pdfDocument = (doc as PdfJsExtendedDocument)._pdfDocument;const page = await pdfDocument.getPage(pageNum);@@ -942,21 +950,23 @@ export class PdfJsEngine implements PdfEngine {}let images: Image[] = [];- try {- const pdfInput = this.currentPdfPath || this.currentPdfData || doc.data;- if (!this.pdfiumRenderer) {- this.pdfiumRenderer = new PdfiumRenderer();- await this.pdfiumRenderer.loadDocument(pdfInput);+ if (options?.extractImages !== false) {+ try {+ const pdfInput = this.currentPdfPath || this.currentPdfData || doc.data;+ if (!this.pdfiumRenderer) {+ this.pdfiumRenderer = new PdfiumRenderer();+ await this.pdfiumRenderer.loadDocument(pdfInput);+ }+ const imageBounds = await this.pdfiumRenderer.extractImageBounds(pdfInput, pageNum);+ images = imageBounds.map((bounds) => ({+ x: bounds.x,+ y: bounds.y,+ width: bounds.width,+ height: bounds.height,+ }));+ } catch {+ // Image extraction is best-effort}- const imageBounds = await this.pdfiumRenderer.extractImageBounds(pdfInput, pageNum);- images = imageBounds.map((bounds) => ({- x: bounds.x,- y: bounds.y,- width: bounds.width,- height: bounds.height,- }));- } catch {- // Image extraction is best-effort}// Skip annotation extraction - not currently used in processing pipeline@@ -1027,7 +1037,8 @@ export class PdfJsEngine implements PdfEngine {async extractAllPages(doc: PdfDocument,maxPages?: number,- targetPages?: string+ targetPages?: string,+ options?: ExtractOptions): Promise<PageData[]> {const numPages = Math.min(doc.numPages, maxPages || doc.numPages);@@ -1048,7 +1059,7 @@ export class PdfJsEngine implements PdfEngine {if (maxPages && pages.length >= maxPages) {break;}- const pageData = await this.extractPage(doc, pageNum);+ const pageData = await this.extractPage(doc, pageNum, options);pages.push(pageData);}