@azure/ai-form-recognizer package

A static-key-based credential that supports updating the underlying key value.

A client for interacting with the Form Recognizer service's analysis features.

Examples:

The Form Recognizer service and clients support two means of authentication:

Azure Active Directory

import { DefaultAzureCredential } from "@azure/identity";
import { DocumentAnalysisClient } from "@azure/ai-form-recognizer";

const credential = new DefaultAzureCredential();
const client = new DocumentAnalysisClient(
  "https://<resource name>.cognitiveservices.azure.com",
  credential,
);

API Key (Subscription Key)

import { AzureKeyCredential, DocumentAnalysisClient } from "@azure/ai-form-recognizer";

const credential = new AzureKeyCredential("<API key>");
const client = new DocumentAnalysisClient(
  "https://<resource name>.cognitiveservices.azure.com",
  credential,
);

A client for interacting with the Form Recognizer service's model management features, such as creating, reading, listing, deleting, and copying models.

Examples:

Azure Active Directory

import { DefaultAzureCredential } from "@azure/identity";
import { DocumentModelAdministrationClient } from "@azure/ai-form-recognizer";

const credential = new DefaultAzureCredential();
const client = new DocumentModelAdministrationClient(
  "https://<resource name>.cognitiveservices.azure.com",
  credential,
);

API Key (Subscription Key)

import { AzureKeyCredential, DocumentModelAdministrationClient } from "@azure/ai-form-recognizer";

const credential = new AzureKeyCredential("<API key>");
const client = new DocumentModelAdministrationClient(
  "https://<resource name>.cognitiveservices.azure.com",
  credential,
);

Address field value.

Options for the document analysis operation.

The result of an analysis operation. The type of the Document may be determined by the model used to perform the analysis.

The common fields of all AnalyzeResult-like types, such as LayoutResult, ReadResult, and GeneralDocumentResult.

An extracted document object.

An AnalyzedDocument is an instance of one of the document types within a model. Its fields correspond to the field schema of the document type.

A training data source defined by an Azure Blob Container and a JSONL file list within the container.

File list in Azure Blob Storage.

A training data source defined by an Azure Blob Container.

Azure Blob Storage content.

Options for the document classifier build operation.

Options for the model build operation.

Options for the model compose operation.

Options for the copy model operation.

Bounding polygon on a specific page of the input.

Classifier document type info.

Options for the document classification operation.

Options common to all operations that define new models, such as beginBuildDocumentModel, beginComposeDocumentModel, and getCopyAuthorization.

Authorization to copy a document model to the specified target resource and modelId.

Options for the model creation operation.

Currency field value.

Details regarding custom document models.

Options for model deletion.

A document field that describes a structured physical address.

Configurable options for DocumentAnalysisClient.

The state of an analysis operation, which will eventually produce the result type that corresponds to the model.

A visual annotation element in the document, such as a check mark or cross.

A DocumentField that consists of an array of nested fields. All fields in the array will have the same type.

An extracted barcode.

A DocumentField that has a boolean value.

An object representing the location and content of a table caption.

Get Operation response object.

Document classifier info.

A set of sources used to create a document classifier. This is a map of document type names to sources that will be used to train the model to classify documents of the corresponding source type.

The state of a model creation operation.

A DocumentField that has a value indicating a country or region, represented as a string.

A DocumentField that describes an amount of a certain currency.

A DocumentField that has a Date value.

Fields that are common to all DocumentField variants.

Description of the field semantic schema using a JSON Schema style syntax.

An object representing the location and content of a table footnote.

An extracted formula.

A DocumentField that has an integer value.

An object representing the field key or value in a key-value pair.

An object representing a form field with distinct field label (key) and field value (may be empty).

An object representing the detected language for a given text span.

A content line object consisting of an adjacent sequence of content elements, such as words and selection marks.

A well-known model specification that supports extracting structured documents.

See the beginAnalyzeDocument method of DocumentAnalysisClient, which supports consuming these DocumentModel objects instead of model ID strings to provide stronger result types.

Configurable options for DocumentModelAdministrationClient.

Get Operation response object.

Get Operation response object.

Get Operation response object.

Document model info.

The state of a model creation operation.

Document model summary.

A DocumentField that has a number value.

A DocumentField that consists of several named properties that have their own DocumentField values.

Content and layout elements extracted from a page from the input.

A paragraph object consisting with contiguous lines generally with common alignment and spacing.

A DocumentField that has a phone number value, represented as a string.

A selection mark object representing check boxes, radio buttons, and other elements indicating a selection.

A DocumentField that is has a value indicating a selection mark state (such as a checkbox or radio button), represented as a string.

A DocumentField that indicates the state of a signature, represented as a string.

Contiguous region of the concatenated content property, specified as an offset and length.

A DocumentField that has a string value.

An object representing observed text styles.

A table object consisting table cells arranged in a rectangular layout.

An object representing the location and content of a table cell.

A DocumentField that has a time value, represented as a string.

Document type info.

A simple field that has a primitive value, such as a string, number etc.

A word object consisting of a contiguous sequence of characters. For non-space delimited languages, such as Chinese, Japanese, and Korean, each character is represented as its own word.

Error info.

Options for the get copy authorization method.

Options for retrieving model information.

Options for retrieving an operation state.

Options for retrieving Form Recognizer resource information.

Simple document elements such as words, selection marks and lines are bounded by the polygon.

Detailed error.

Options for listing models.

Options for listing operations.

The set of fields common to all administration operations.

Get Operation response object.

Operation info.

Represents a point used to define bounding polygons. The unit is either 'pixel' or 'inch' (See LengthUnit).

Options for long-running operations (pollers) in the Form Recognizer clients.

Quota used, limit, and next reset date/time.

General information regarding the current resource.

A long-running operation (poller) that tracks the state of an analysis operation, eventually producing the result type that corresponds to the model.

Defines values for AnalyzeResultOperationStatus.

Defines values for DocumentBarcodeKind.
KnownDocumentBarcodeKind can be used interchangeably with DocumentBarcodeKind, this enum contains the known values that the service supports.

Known values supported by the service

QRCode: QR code, as defined in ISO/IEC 18004:2015.
PDF417: PDF417, as defined in ISO 15438.
UPCA: GS1 12-digit Universal Product Code.
UPCE: GS1 6-digit Universal Product Code.
Code39: Code 39 barcode, as defined in ISO/IEC 16388:2007.
Code128: Code 128 barcode, as defined in ISO/IEC 15417:2007.
EAN8: GS1 8-digit International Article Number (European Article Number).
EAN13: GS1 13-digit International Article Number (European Article Number).
DataBar: GS1 DataBar barcode.
Code93: Code 93 barcode, as defined in ANSI/AIM BC5-1995.
Codabar: Codabar barcode, as defined in ANSI/AIM BC3-1995.
DataBarExpanded: GS1 DataBar Expanded barcode.
ITF: Interleaved 2 of 5 barcode, as defined in ANSI/AIM BC2-1995.
MicroQRCode: Micro QR code, as defined in ISO/IEC 23941:2022.
Aztec: Aztec code, as defined in ISO/IEC 24778:2008.
DataMatrix: Data matrix code, as defined in ISO/IEC 16022:2006.
MaxiCode: MaxiCode, as defined in ISO/IEC 16023:2000.

Defines values for DocumentBuildMode.
KnownDocumentBuildMode can be used interchangeably with DocumentBuildMode, this enum contains the known values that the service supports.

Known values supported by the service

template: Target documents with similar visual templates.
neural: Support documents with diverse visual templates.

A long-running operation (poller) that tracks the state of a custom classifier creation operation, eventually producing a DocumentClassifierDetails.

A content source that may be used to build a document classifier.

One of:

• BlobSource
• BlobFileListSource

An extracted field. The kind property identifies (discriminates) the type of the DocumentField.

Defines values for DocumentFieldType.
KnownDocumentFieldType can be used interchangeably with DocumentFieldType, this enum contains the known values that the service supports.

Known values supported by the service

string: Plain text.
date: Date, normalized to ISO 8601 (YYYY-MM-DD) format.
time: Time, normalized to ISO 8601 (hh:mm:ss) format.
phoneNumber: Phone number, normalized to E.164 (+{CountryCode}{SubscriberNumber}) format.
number: Floating point number, normalized to double precision floating point.
integer: Integer number, normalized to 64-bit signed integer.
selectionMark: Is field selected?
countryRegion: Country/region, normalized to ISO 3166-1 alpha-3 format (ex. USA).
signature: Is signature present?
array: List of subfields of the same type.
object: Named list of subfields of potentially different types.
currency: Currency amount with optional currency symbol and unit.
address: Parsed address.
boolean: Boolean value, normalized to true or false.

Defines values for DocumentFormulaKind.
KnownDocumentFormulaKind can be used interchangeably with DocumentFormulaKind, this enum contains the known values that the service supports.

Known values supported by the service

inline: A formula embedded within the content of a paragraph.
display: A formula in display mode that takes up an entire line.

Supported model build modes. The model build mode selects the engine that the service uses to train the model based on the labeled training data.

The options are:

• "neural", which yields the highest quality of model that is capable of extracting data from classes of documents that have the same structure of data, but different layouts (for example, W2 tax forms, which may vary from company to company, but always contain the same information).
• "template", which requires all documents to have the same fixed layout (template).

Please see the following link for more information: https://aka.ms/azsdk/formrecognizer/buildmode

A long-running operation (poller) that tracks the state of a model creation operation, eventually producing a DocumentModelDetails.

A content source that may be used to build a document model.

One of:

• BlobSource
• BlobFileListSource

Defines values for DocumentSignatureType.
KnownDocumentSignatureType can be used interchangeably with DocumentSignatureType, this enum contains the known values that the service supports.

Known values supported by the service

signed: A signature is detected.
unsigned: No signatures are detected.

Defines values for DocumentTableCellKind.
KnownDocumentTableCellKind can be used interchangeably with DocumentTableCellKind, this enum contains the known values that the service supports.

Known values supported by the service

content: Contains the main content/data.
rowHeader: Describes the content of the row.
columnHeader: Describes the content of the column.
stubHead: Describes the row headers, usually located at the top left corner of a table.
description: Describes the content in (parts of) the table.

Defines values for FontStyle.
KnownFontStyle can be used interchangeably with FontStyle, this enum contains the known values that the service supports.

Known values supported by the service

normal: Characters are represented normally.
italic: Characters are visually slanted to the right.

Defines values for FontWeight.
KnownFontWeight can be used interchangeably with FontWeight, this enum contains the known values that the service supports.

Known values supported by the service

normal: Characters are represented normally.
bold: Characters are represented with thicker strokes.

Add-on capabilities (features) that can be enabled for the request.

For more information about the features available in Form Recognizer, see the service documentation:

https://aka.ms/azsdk/formrecognizer/features

A request input that can be uploaded as binary data to the Form Recognizer service. Form Recognizer treats string inputs as URLs, so to send a string as a binary input, first convert the string to one of the following input types.

Defines values for LengthUnit.
KnownLengthUnit can be used interchangeably with LengthUnit, this enum contains the known values that the service supports.

Known values supported by the service

pixel: Length unit for image files.
inch: Length unit for PDF files.

Details about any of several different types of Form Recognizer management operations.

Defines values for OperationKind.
KnownOperationKind can be used interchangeably with OperationKind, this enum contains the known values that the service supports.

Known values supported by the service

documentModelBuild: Build a new custom document model.
documentModelCompose: Compose a new custom document model from existing models.
documentModelCopyTo: Copy an existing document model to potentially a different resource, region, or subscription.
documentClassifierBuild: Build a new custom classifier model.

Defines values for OperationStatus.

Defines values for ParagraphRole.
KnownParagraphRole can be used interchangeably with ParagraphRole, this enum contains the known values that the service supports.

Known values supported by the service

pageHeader: Text near the top edge of the page.
pageFooter: Text near the bottom edge of the page.
pageNumber: Page number.
title: Top-level title describing the entire document.
sectionHeading: Sub heading describing a section of the document.
footnote: A note usually placed after the main content on a page.
formulaBlock: A block of formulas, often with shared alignment.

Defines values for SelectionMarkState.
KnownSelectionMarkState can be used interchangeably with SelectionMarkState, this enum contains the known values that the service supports.

Known values supported by the service

selected: The selection mark is selected, often indicated by a check ✓ or cross X inside the selection mark.
unselected: The selection mark is not selected.

Valid string index types supported by the Form Recognizer service and SDK clients.

Known values of DocumentBarcodeKind that the service accepts.

Known values of DocumentBuildMode that the service accepts.

Known values of DocumentFieldType that the service accepts.

Known values of DocumentFormulaKind that the service accepts.

Known values of DocumentSignatureType that the service accepts.

Known values of DocumentTableCellKind that the service accepts.

Known values of FontStyle that the service accepts.

Known values of FontWeight that the service accepts.

Defines the known cloud audiences for Form Recognizer.

To authenticate with Entra Id (using a TokenCredential) in a Sovereign Cloud environment, provide the appropriate value below as the audience option when creating a DocumentAnalysisClient or DocumentModelAdministrationClient.

The default value is suitable for Form Recognizer resources created in the Azure Public Cloud, so this value is only required to use Form Recognizer in a different cloud environment.

Known values of LengthUnit that the service accepts.

Known values of OperationKind that the service accepts.

Known values of ParagraphRole that the service accepts.

Known values of SelectionMarkState that the service accepts.