Skip to content

Detection and Tiny sequences #391

@dpatryas-rtbh

Description

@dpatryas-rtbh

Describe the bug
The issue pertains to the charset_normalizer.detect() method, which fails to perform valid encoding detection for expression sequences. Specifically, the method incorrectly recognizes the sequence as Big5 and returns it as a Chinese character. This behavior can be observed with the provided code snippet:

To Reproduce
Execute the code snippet, where the charset_normalizer.detect() method misidentifies the encoding of the sequence.

import charset_normalizer
text = b"What Actions Will Keep Us at 1.5-2\xbaC?"
text.decode(charset_normalizer.detect(text)["encoding"])

Expected behavior
The expected behavior can be demonstrated using the chardet library, which accurately recognizes the encoding as ISO 8859-1. The correct degree character is then returned from the sequence:

import chardet
text = b"What Actions Will Keep Us at 1.5-2\xbaC?"
text.decode(chardet.detect(text)["encoding"])

Logs

charset-normalizer:
What Actions Will Keep Us at 1.5-2慢?

chardet:
What Actions Will Keep Us at 1.5-2ºC?

{'encoding': 'ISO-8859-1', 'confidence': 0.73, 'language': ''}  # chardet
{'encoding': 'Big5', 'language': 'Chinese', 'confidence': 1.0}  # charset-normalizer

Desktop (please complete the following information):

  • OS: Linux
  • Python version 3.10
  • Package version 3.3.2 / 2.1.1

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions