Document Editing

Fill forms and edit documents with natural language instructions.

What is Document Editing?

Document editing detects fillable fields and fills them:

• Form fields (text, checkboxes, dropdowns)
• Signature blocks
• Date fields
• Any editable area

Use cases:

• Fill out PDF forms programmatically
• Auto-populate contracts with client data
• Update template documents
• Batch form filling

Editing Documents

From the Dashboard

1. Go to Edit in the sidebar
2. Select a document (PDF or image)
3. DocLD detects fillable elements
4. Fill using:
   • Direct field mapping
   • Natural language instructions
   • Suggest — get AI-suggested values for fields
5. Download the filled document

You can create edit projects to group runs by document and config, and use saved edit configs (per organization) to apply the same settings across runs. For multiple documents, use batch edit to apply one config to many forms. See the Edit API for projects, batch, configs, and suggest endpoints.

Using Natural Language

Describe what you want to fill:

Fill the client name as "Acme Corporation",
the date as "January 15, 2024",
and the contract amount as "$25,000"

DocLD interprets your instructions and fills the appropriate fields.

Using Field Mappings

Map values directly to detected fields:

{
  "mappings": {
    "client_name": "Acme Corporation",
    "contract_date": "January 15, 2024",
    "amount": "$25,000",
    "signature_required": true
  }
}

Detected Elements

DocLD detects various field types:

Element Detection

Each detected element includes:

{
  "id": "element-uuid",
  "type": "text_field",
  "label": "Client Name",
  "page": 1,
  "bbox": {
    "x": 100,
    "y": 200,
    "width": 200,
    "height": 25
  },
  "required": true,
  "current_value": ""
}

Two-Step Process

For complex forms, use the two-step process: detect elements first, then fill using the returned edit job ID (edit_job_id).

Step 1: Detect Elements

curl -X POST "/api/edit/detect" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{"document_id": "form-uuid"}'

Review detected elements and their labels. The response includes a job_id (use this as edit_job_id in the next step).

Step 2: Fill Elements

Pass edit_job_id from the detect response so the server uses the same detected elements. You can send mappings as an array (with elementId and value) or as an object keyed by field label or element ID:

curl -X POST "/api/edit/run" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "edit_job_id": "detect-job-uuid",
    "mappings": {
      "client_name": "Acme Corporation"
    }
  }'

Configuration Options

Formatting

Control how filled values appear:

{
  "config": {
    "formatting_color": "#0000FF",
    "font_size": 12,
    "font_family": "Arial"
  }
}

Preservation

Keep original content:

{
  "config": {
    "preserve_original": true
  }
}

Supported Formats

Only PDF and image files are supported for detection and editing. See the Edit API for details.

Workflow Integration

Add editing to workflows:

{
  "type": "edit",
  "config": {
    "instructions": "Fill with extracted data",
    "mappings": {
      "name": "{{extracted.name}}",
      "date": "{{current_date}}"
    }
  }
}

This enables automated form filling as part of processing pipelines.

Combining with Extraction

Powerful workflow: Extract → Edit

1. Extract data from source document
2. Edit template document with extracted values

# Step 1: Extract from source
curl -X POST "/api/extract/run" \
  -d '{"document_id": "source-doc", "schema_id": "client-schema"}'
 
# Step 2: Fill template with extracted data
curl -X POST "/api/edit/run" \
  -d '{
    "document_id": "template-form",
    "mappings": {
      "client_name": "extracted_name_value",
      "address": "extracted_address_value"
    }
  }'

Best Practices

1. Preview first - Detect elements before filling
2. Verify mappings - Ensure fields match expected values
3. Use clear labels - Match field labels for accurate filling
4. Test with samples - Verify results before batch processing
5. Keep originals - Preserve original documents

CLI Support

The DocLD CLI supports editing:

# Fill a single form
docld edit form.pdf -i "Fill client name as 'Acme Corp'"
 
# Batch editing
docld edit ./contracts -i "Replace 'OLD COMPANY' with 'NEW COMPANY'"

API Reference

See the Edit API for programmatic access.