Error on .find_tables()

[JorjMcKie]

Discussed in #3190

^{Originally posted by bjmvercelli February 20, 2024}
Hello, hope you guys are doing great.

I'm getting an error in version 1.23.24 (latest) using find_tables() method, more specific on extract_text() call.

The following code was extracted from table.py (lines 606 and 607). The error happens when extract_words(chars) returns an empty array.

words = extractor.extract_words(chars)
rotation = words[0]["rotation"]  # rotation cannot change within a cell

I do not believe that there's a problem in extract_words(), but i do believe that's an edge case from my PDF and, if thats the case, we could fix it by validating the length of words:

words = extractor.extract_words(chars)
if len(words) == 0:
  return ""
rotation = words[0]["rotation"]  # rotation cannot change within a cell

You can reproduce here

[julian-smith-artifex-com]

Fixed in 1.23.25.