Parse Document with Tensorlake API — tensorlake_parse

This function submits a document to Tensorlake for parsing using a file ID from tensorlake_upload_file(). Tensorlake offers high-accuracy document parsing (91.7% accuracy) with support for tables, forms, and structured data extraction.

Usage

tensorlake_parse_document(
  file_id,
  tensorlake_api_key,
  pages = NULL,
  figure_summarization = TRUE,
  chart_extraction = TRUE,
  figure_summarization_prompt = NULL,
  base_url = "https://api.tensorlake.ai"
)

Arguments

file_id: Character string. Tensorlake file ID from tensorlake_upload_file().
tensorlake_api_key: Character string. Tensorlake API key.
pages: Character string. Optional page range to parse (e.g., "1-5" or "1,3,5").
figure_summarization: Logical. Enable figure summarization enrichment. Default TRUE.
chart_extraction: Logical. Enable chart extraction enrichment. Default TRUE.
figure_summarization_prompt: Character string. Optional custom prompt to guide figure summarization. Default NULL uses Tensorlake's default prompt.
base_url: Character string. Base URL for Tensorlake API. Default is "https://api.tensorlake.ai".

Value

List containing the parse job details including:

parse_id: Unique ID for the parse job
status: Job status (processing, completed, failed)
result: Parsed document content (when completed)

Author

Nathan C. Layman