50 lines
1.3 KiB
Markdown
50 lines
1.3 KiB
Markdown
|
# PDF Metadata Post-Processor
|
||
|
|
|
||
|
|
Register a post-processor that extracts and enriches extraction results with PDF metadata.
|
||
|
|
|
||
|
|
```typescript title="WASM"
|
||
|
|
import init, { registerPostProcessor, extractBytes } from "kreuzberg-wasm";
|
||
|
|
|
||
|
|
await init();
|
||
|
|
|
||
|
|
// Define a PDF metadata extractor post-processor
|
||
|
|
const pdfMetadataProcessor = {
|
||
|
|
processingStage: () => "post-extraction",
|
||
|
|
process: (extractionResult) => {
|
||
|
|
// Enrich extraction with metadata
|
||
|
|
const enriched = {
|
||
|
|
...extractionResult,
|
||
|
|
metadata: {
|
||
|
|
...extractionResult.metadata,
|
||
|
|
processorName: "pdf-metadata",
|
||
|
|
processedAt: new Date().toISOString(),
|
||
|
|
wordCount: (extractionResult.text || "").split(/\s+/).length,
|
||
|
|
},
|
||
|
|
};
|
||
|
|
|
||
|
|
return enriched;
|
||
|
|
},
|
||
|
|
};
|
||
|
|
|
||
|
|
try {
|
||
|
|
registerPostProcessor(pdfMetadataProcessor);
|
||
|
|
console.log("PDF metadata post-processor registered");
|
||
|
|
} catch (error) {
|
||
|
|
console.error("Failed to register post-processor:", error);
|
||
|
|
}
|
||
|
|
|
||
|
|
// Extract with post-processing
|
||
|
|
const pdfBytes = new Uint8Array([
|
||
|
|
/* PDF content */
|
||
|
|
]);
|
||
|
|
const config = {
|
||
|
|
ocr: null,
|
||
|
|
chunking: null,
|
||
|
|
};
|
||
|
|
|
||
|
|
const result = await extractBytes(pdfBytes, "application/pdf", config);
|
||
|
|
console.log("Enriched metadata:", result.metadata);
|
||
|
|
```
|
||
|
|
|
||
|
|
The post-processor runs after extraction to enrich or transform the results.
|