@langchain/community
Version:
Third-party integrations for LangChain.js
50 lines (49 loc) • 1.62 kB
text/typescript
import { Document } from "@langchain/core/documents";
import { BaseDocumentLoader } from "@langchain/core/document_loaders/base";
import * as _$pdf_parse_lib_pdf_js_v1_10_100_build_pdf_js0 from "pdf-parse/lib/pdf.js/v1.10.100/build/pdf.js";
import * as _$pdf_parse0 from "pdf-parse";
//#region src/document_loaders/web/pdf.d.ts
type PDFLoaderV1Imports = {
isV2: false;
getDocument: typeof _$pdf_parse_lib_pdf_js_v1_10_100_build_pdf_js0.getDocument;
version: typeof _$pdf_parse_lib_pdf_js_v1_10_100_build_pdf_js0.version;
};
type PDFLoaderV2Imports = {
isV2: true;
PDFParse: typeof _$pdf_parse0.PDFParse;
};
type PDFLoaderImportsResult = PDFLoaderV1Imports | PDFLoaderV2Imports;
/**
* A document loader for loading data from PDFs.
* @example
* ```typescript
* const loader = new WebPDFLoader(new Blob());
* const docs = await loader.load();
* console.log({ docs });
* ```
*/
declare class WebPDFLoader extends BaseDocumentLoader {
protected blob: Blob;
protected splitPages: boolean;
private pdfjs;
protected parsedItemSeparator: string;
constructor(blob: Blob, {
splitPages,
pdfjs,
parsedItemSeparator
}?: {
parsedItemSeparator?: string | undefined;
pdfjs?: typeof PDFLoaderImports | undefined;
splitPages?: boolean | undefined;
});
/**
* Loads the contents of the PDF as documents.
* @returns An array of Documents representing the retrieved data.
*/
load(): Promise<Document[]>;
private parseWithV2;
}
declare function PDFLoaderImports(): Promise<PDFLoaderImportsResult>;
//#endregion
export { WebPDFLoader };
//# sourceMappingURL=pdf.d.cts.map