❗️ v0.3.11 is broken

[lsorber]

The new v0.3.11 release seems to be broken. Minimal reproducible example (e.g., in a Google Colab):

%pip install --quiet pdftext==0.3.11

from pathlib import Path
from pdftext.extraction import dictionary_output

pages = dictionary_output(Path("specrel.pdf"), sort=True, keep_chars=False)

Output:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
[<ipython-input-2-783804a94b7a>](https://localhost:8080/#) in <cell line: 4>()
      2 from pdftext.extraction import dictionary_output
      3 
----> 4 pages = dictionary_output(Path("specrel.pdf"), sort=True, keep_chars=False)

2 frames
[/usr/local/lib/python3.10/dist-packages/pdftext/extraction.py](https://localhost:8080/#) in _load_pdf(pdf, flatten_pdf)
     17 
     18     if not isinstance(pdf, pdfium.PdfDocument):
---> 19         raise TypeError("pdf must be a file path string or a PdfDocument object")
     20 
     21     # Must be called on the parent pdf, before the page was retrieved

TypeError: pdf must be a file path string or a PdfDocument object

[VikParuchuri]

Fix is out (v0.3.12), thanks for the catch