Module spatial 

Spatial table reconstruction from hOCR bounding box coordinates

This module provides functions to detect and reconstruct tabular data from OCR’d text by analyzing the spatial positions of words using their bounding box (bbox) coordinates.

Structs

HocrWord

Represents a word extracted from hOCR with position and confidence information

Functions

detect_columns

Detect column positions from word positions

detect_rows

Detect row positions from word positions

extract_hocr_words

Extract hOCR words from a DOM tree

reconstruct_table

Reconstruct table structure from words

table_to_markdown

Convert table to markdown format