Skip to content

Latest commit

 

History

History
284 lines (208 loc) · 28.3 KB

README.md

File metadata and controls

284 lines (208 loc) · 28.3 KB
Raw

ExtractStructuredData

(files.extractStructuredData)

Overview

Extract structured data from files

Available Operations

  • sync - Extract Structured Data - Sync
  • start - Extract Structured Data - Async Start
  • result - Extract Structured Data - Async Result

sync

Extracts structured data from a file. Supported file types are image, pdf (more coming soon!). It accepts the file and requested schema for the data to be extracted.

Example Usage

import { IntunedClient } from "@intuned/client";

const intunedClient = new IntunedClient({
  apiKey: "<YOUR_API_KEY_HERE>",
  workspaceId: "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
});

async function run() {
  const result = await intunedClient.files.extractStructuredData.sync({
      type: "pdf",
    source:     {
          type: "url",
          data: "http://unconscious-margin.name",
        },
    }, {
    "key": "<value>",
  });

  // Handle the result
  console.log(result)
}

run();

Standalone function

The standalone function version of this method:

import { IntunedClientCore } from "@intuned/client/core.js";
import { filesExtractStructuredDataSync } from "@intuned/client/funcs/filesExtractStructuredDataSync.js";

// Use `IntunedClientCore` for best tree-shaking performance.
// You can create one instance of it to use across an application.
const intunedClient = new IntunedClientCore({
  apiKey: "<YOUR_API_KEY_HERE>",
  workspaceId: "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
});

async function run() {
  const res = await filesExtractStructuredDataSync(intunedClient, {
      type: "image",
    source:     {
          type: "base64",
          data: "<value>",
        },
    }, {
    "key": "<value>",
  });

  if (!res.ok) {
    throw res.error;
  }

  const { value: result } = res;

  // Handle the result
  console.log(result)
}

run();

Parameters

Parameter Type Required Description
file components.FileT ✔️ N/A
dataSchema Record<string, any> ✔️ The schema (in JSONSchema) to which the extracted data should conform to. More about JSONSchema

JSONSchema
strategy components.ExtractStructuredDataStrategy ➖ The strategy to extract structured data from a file. Includes the model and type of extraction.
prompt string ➖ Prompt given to the AI model to help with extraction.
options RequestOptions ➖ Used to set various options for making HTTP requests.
options.fetchOptions RequestInit ➖ Options that are passed to the underlying HTTP request. This can be used to inject extra headers for examples. All Request options, except method and body, are allowed.
options.retries RetryConfig ➖ Enables retrying HTTP requests under certain failure conditions.

Response

Promise<components.StructuredDataExtractionSyncResponse>

Errors

Error Object Status Code Content Type
errors.ApiErrorInvalidInput 400 application/json
errors.ApiErrorUnauthorized 401 application/json
errors.SDKError 4xx-5xx /

start

Starts an asynchronous operation to extract structured data from a file. Supported file types are image, pdf (more coming soon!). It accepts the file and requested schema for the data to be extracted. It responds with an ID to track the operation status and retrieve the result.

Example Usage

import { IntunedClient } from "@intuned/client";

const intunedClient = new IntunedClient({
  apiKey: "<YOUR_API_KEY_HERE>",
  workspaceId: "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
});

async function run() {
  const result = await intunedClient.files.extractStructuredData.start({
      type: "image",
    source:     {
          type: "base64",
          data: "<value>",
        },
    }, {
    "key": "<value>",
  });

  // Handle the result
  console.log(result)
}

run();

Standalone function

The standalone function version of this method:

import { IntunedClientCore } from "@intuned/client/core.js";
import { filesExtractStructuredDataStart } from "@intuned/client/funcs/filesExtractStructuredDataStart.js";

// Use `IntunedClientCore` for best tree-shaking performance.
// You can create one instance of it to use across an application.
const intunedClient = new IntunedClientCore({
  apiKey: "<YOUR_API_KEY_HERE>",
  workspaceId: "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
});

async function run() {
  const res = await filesExtractStructuredDataStart(intunedClient, {
      type: "pdf",
    source:     {
          type: "base64",
          data: "<value>",
        },
    }, {
    "key": "<value>",
  });

  if (!res.ok) {
    throw res.error;
  }

  const { value: result } = res;

  // Handle the result
  console.log(result)
}

run();

Parameters

Parameter Type Required Description
file components.FileT ✔️ N/A
dataSchema Record<string, any> ✔️ The schema (in JSONSchema) to which the extracted data should conform to. More about JSONSchema

JSONSchema
strategy components.ExtractStructuredDataStrategy ➖ The strategy to extract structured data from a file. Includes the model and type of extraction.
prompt string ➖ Prompt given to the AI model to help with extraction.
options RequestOptions ➖ Used to set various options for making HTTP requests.
options.fetchOptions RequestInit ➖ Options that are passed to the underlying HTTP request. This can be used to inject extra headers for examples. All Request options, except method and body, are allowed.
options.retries RetryConfig ➖ Enables retrying HTTP requests under certain failure conditions.

Response

Promise<components.AsyncFilePendingResponse>

Errors

Error Object Status Code Content Type
errors.ApiErrorInvalidInput 400 application/json
errors.ApiErrorUnauthorized 401 application/json
errors.SDKError 4xx-5xx /

result

Gets the result of the structured data extraction operation using the operation ID.

Example Usage

import { IntunedClient } from "@intuned/client";

const intunedClient = new IntunedClient({
  apiKey: "<YOUR_API_KEY_HERE>",
  workspaceId: "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
});

async function run() {
  const result = await intunedClient.files.extractStructuredData.result("aaaabbbCCCCdddd");

  // Handle the result
  console.log(result)
}

run();

Standalone function

The standalone function version of this method:

import { IntunedClientCore } from "@intuned/client/core.js";
import { filesExtractStructuredDataResult } from "@intuned/client/funcs/filesExtractStructuredDataResult.js";

// Use `IntunedClientCore` for best tree-shaking performance.
// You can create one instance of it to use across an application.
const intunedClient = new IntunedClientCore({
  apiKey: "<YOUR_API_KEY_HERE>",
  workspaceId: "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
});

async function run() {
  const res = await filesExtractStructuredDataResult(intunedClient, "aaaabbbCCCCdddd");

  if (!res.ok) {
    throw res.error;
  }

  const { value: result } = res;

  // Handle the result
  console.log(result)
}

run();

Parameters

Parameter Type Required Description Example
operationId string ✔️ The ID for the requested file operation. This is obtained from the start request. [object Object]
options RequestOptions ➖ Used to set various options for making HTTP requests.
options.fetchOptions RequestInit ➖ Options that are passed to the underlying HTTP request. This can be used to inject extra headers for examples. All Request options, except method and body, are allowed.
options.retries RetryConfig ➖ Enables retrying HTTP requests under certain failure conditions.

Response

Promise<components.StructuredDataExtractionAsyncResponse>

Errors

Error Object Status Code Content Type
errors.ApiErrorInvalidInput 400 application/json
errors.ApiErrorUnauthorized 401 application/json
errors.SDKError 4xx-5xx /