Highlighting API

API to generate highlight locations for a search term in the document which has been extracted using LLMWhisperer. This can be used to highlight the search terms in your frontend application.

Note: The parameter store_metadata_for_highlighting should be set to true while starting the whisper process to use this API.

Endpoint	/highlight-data
URL	https://llmwhisperer-api.unstract.com/v1/highlight-data
Method	POST
Headers	unstract-key: <YOUR API KEY>

Parameters

Parameter	Type	Default	Required	Description
whisper-hash	string		Yes	The whisper hash returned while starting the whisper process.

Request Body

The request body should contain the search term. The Content-Type should be text/plain.

Example Curl Request

curl -X POST --location 'https://llmwhisperer-api.unstract.com/v1/highlight-data?whisper-hash=XXXXXXXXXXXXXXXXXXX' \
-H 'unstract-key: <Your API Key>' \
-H 'Content-Type: text/plain' \
--data 'search term'

info

To include the headers in the response use curl -i in the request.

Response

HTTP Status	Content-Type	Headers	Description
200	application/json		Highlight information
400	application/json		Error while retrieveing. Refer below for JSON format

Highlight Data

The API will return the highlight data in the response body. The JSON response will contain an array of JSON objects. Each object will contain the following fields:

Field	Type	Description
text	string	The search term (or part of)
page	integer	The page number where the search term is found. 0 indexed
pos	integer	The position of the search term in the text. 0 indexed
left	float	The left position of the search term in the page.
top	float	The top position of the search term in the page.
width	float	The width of the search term in the page.
height	float	The height of the search term in the page.
distance	integer	The Levenshtein distance of the search term from the match.
location_type	string	The unit of the location values. Can be inches or points based on how the original document was created and processed

Example 200 Response

{
    "highlight_data": [
        {
            "distance": 4,
            "height": 0.11460000000000026,
            "left": 6.2699,
            "location_type": "inches",
            "page": 0,
            "pos": 960,
            "text": "U.S.",
            "top": 2.9505,
            "width": 0.20060000000000056
        },
        {
            "distance": 4,
            "height": 0.11940000000000017,
            "left": 6.4944,
            "location_type": "inches",
            "page": 0,
            "pos": 965,
            "text": "Citizen",
            "top": 2.9505,
            "width": 0.34860000000000024
        }
    ]
}

Example 400 Response

{
    "message": "No Metadata found"
}