Skip to main content

Update a Record

To update a new record, use the update_record method. This method requires two primary parameters:

  • collection_id: Required. The identifier of the collection where the record will be stored.
  • record_id: Required. The identifier of the record to be updated.
  • content: Optional. The textual content of the record.
  • type: Optional. The type of record. Supported types are text, web, and file.
  • url: Optional. The URL of the webpage to be scraped.
  • file_id: Optional. The file ID of the file to be record.
  • text_splitter: Optional. A dictionary containing the text splitter configuration or a TextSplitter object.
  • metadata: Optional. A dictionary containing the metadata of the record.

If the content of the record is to be updated. Please make sure the combination of content, type, url, and file_id are valid. You should specify the type and set value for one of the content, url, or file_id parameters that matches with the type.

note

Records of file type cannot be updated. Records of web or text type can be updated with new content, or switch to a different type.

Here is an example of updating a record with new content:

import taskingai

record = taskingai.retrieval.update_record(
    collection_id="YOUR_COLLECTION_ID",
    record_id="YOUR_RECORD_ID",
    type="text",
    content="Machine learning is a subfield of artificial intelligence...",
    text_splitter={"type": "token", "chunk_size": 200, "chunk_overlap": 20},
)
note

After content update, all the old text chunks associated with the record will be deleted and replaced with the new ones.

Another example of updating a record with new metadata:

import taskingai

record = taskingai.retrieval.update_record(
    collection_id="YOUR_COLLECTION_ID",
    record_id="YOUR_RECORD_ID",
    metadata={"file_name":"machine_learning.pdf", "author": "James Brown"}
)