Get All Data

Retrieve all processed data for a model in a simplified, flattened JSON format. Data is organized by filename with normalized column names, making it ideal for bulk export, analytics, and integration with external systems.

Parameters

Request

const response = await fetch(
`https://app.limai.io/api/v1/extraction-schema/get-all-data?extractionSchemaId=${schemaId}`,
{
  headers: { "Authorization": `Bearer ${API_TOKEN}` }
}
)
const allData = await response.json()

for (const [filename, fileData] of Object.entries(allData)) {
const records = Array.isArray(fileData) ? fileData : [fileData]
for (const record of records) {
  console.log(filename, record)
}
}

Response

Data is organized by filename. Single-row documents return an object; multi-row documents return an array. Nested tables appear as direct properties.

Multi-Row Example

{
  "invoice_001.pdf": [
    {
      "invoice_number": "INV-001",
      "total_amount": "1000.00",
      "line_items": [
        {
          "item_description": "Software License",
          "quantity": "1",
          "unit_price": "800.00"
        }
      ]
    },
    {
      "invoice_number": "INV-002",
      "total_amount": "500.00"
    }
  ],
  "invoice_002.pdf": {
    "invoice_number": "INV-003",
    "total_amount": "750.00"
  }
}

Empty Response

Returns an empty object {} when no data exists for the model.

Column Name Normalization

All column and table names are automatically normalized in the response:

1. Convert to lowercase
2. Replace spaces and slashes with underscores
3. Remove special characters (keep letters, numbers, underscores)
4. Collapse multiple underscores into one
5. Remove leading/trailing underscores

If multiple columns normalize to the same name, the API returns a 400 error with details about the conflicting columns. Rename one column in the platform to resolve the conflict.