← Back to Documentation Home

PDF Batch Process API

Overview

The PDF Batch Process API handles batch processing of PDF documents for extraction and analysis.

Edge Function: pdf-batch-process
Base URL: https://bgbavxtjlbvgplozizxu.supabase.co/functions/v1/pdf-batch-process

Authentication

All requests require authentication via Supabase Auth:

Authorization: Bearer <supabase_access_token>

Endpoints

1. Create Batch Job

Create a new batch processing job for multiple documents.

Method: POST
Path: /

Request:

{
  documents: Array<{
    documentId: string,
    extractionType: 'markdown' | 'tables' | 'images' | 'all',
    priority?: 'low' | 'normal' | 'high'
  }>,
  workspaceId?: string,
  userId?: string,
  options?: {
    includeImages?: boolean,
    includeMetadata?: boolean,
    chunkSize?: number,
    overlapSize?: number,
    outputFormat?: 'json' | 'markdown',
    maxConcurrent?: number,
    notifyOnComplete?: boolean,
    webhookUrl?: string
  }
}

Response:

{
  success: true,
  data: {
    batchId: string,
    status: 'queued',
    totalDocuments: number,
    processedDocuments: 0,
    failedDocuments: 0,
    estimatedCompletionTime?: string,
    results: Array<{
      documentId: string,
      status: 'pending',
      extractionId?: string,
      error?: string,
      processingTime?: number
    }>
  }
}

Example:

const { data, error } = await supabase.functions.invoke('pdf-batch-process', {
  body: {
    documents: [
      {
        documentId: 'doc-123',
        extractionType: 'all',
        priority: 'high'
      },
      {
        documentId: 'doc-456',
        extractionType: 'markdown',
        priority: 'normal'
      }
    ],
    workspaceId: 'workspace-789',
    options: {
      includeImages: true,
      includeMetadata: true,
      chunkSize: 1000,
      overlapSize: 200,
      outputFormat: 'json',
      maxConcurrent: 3,
      notifyOnComplete: true
    }
  }
});

2. Get Batch Status

Get the status of a batch processing job.

Method: GET
Path: /?batchId={batchId}

Query Parameters:

Response:

{
  success: true,
  data: {
    batchId: string,
    status: 'queued' | 'processing' | 'completed' | 'failed' | 'partial',
    totalDocuments: number,
    processedDocuments: number,
    failedDocuments: number,
    createdAt: string,
    updatedAt: string,
    completedAt?: string,
    results: Array<{
      documentId: string,
      status: 'pending' | 'processing' | 'completed' | 'failed',
      extractionId?: string,
      error?: string,
      processingTime?: number
    }>
  }
}

Example:

const response = await fetch(
  `${API_BASE}/pdf-batch-process?batchId=batch-123`,
  {
    headers: {
      'Authorization': `Bearer ${session.access_token}`
    }
  }
);

3. Cancel Batch Job

Cancel a running or queued batch job.

Method: DELETE
Path: /?batchId={batchId}

Query Parameters:

Response:

{
  success: true,
  message: 'Batch job cancelled successfully',
  data: {
    batchId: string,
    status: 'cancelled',
    processedDocuments: number,
    cancelledDocuments: number
  }
}

Example:

const response = await fetch(
  `${API_BASE}/pdf-batch-process?batchId=batch-123`,
  {
    method: 'DELETE',
    headers: {
      'Authorization': `Bearer ${session.access_token}`
    }
  }
);

Extraction Types

Type Description
markdown Extract text content as markdown
tables Extract tables from PDF
images Extract images from PDF
all Extract everything (markdown + tables + images)

Priority Levels

Priority Description Processing Order
high Urgent processing Processed first
normal Standard processing Default queue
low Background processing Processed last

Batch Status Flow

queued → processing → completed
                   ↘ failed
                   ↘ partial (some docs failed)

Options

includeImages

Include image extraction in the processing (default: true)

includeMetadata

Include document metadata in results (default: true)

chunkSize

Size of text chunks for processing (default: 1000)

overlapSize

Overlap between chunks (default: 200)

outputFormat

Output format for results: json or markdown (default: json)

maxConcurrent

Maximum concurrent document processing (default: 5)

notifyOnComplete

Send notification when batch completes (default: false)

webhookUrl

Webhook URL to call when batch completes

Webhook Payload

If webhookUrl is provided, a POST request will be sent on completion:

{
  batchId: string,
  status: 'completed' | 'failed' | 'partial',
  totalDocuments: number,
  processedDocuments: number,
  failedDocuments: number,
  completedAt: string,
  results: Array<{
    documentId: string,
    status: string,
    extractionId?: string,
    error?: string
  }>
}

Error Handling

{
  success: false,
  error: string,
  statusCode?: number
}

Common Error Codes:

Related Documentation