MAINT: New LZW decoding implementation

[MartinThoma]

The basis for this implementation is https://github.com/empira/PDFsharp/blob/master/src/foundation/src/PDFsharp/src/PdfSharp/Pdf.Filters/LzwDecode.cs (MIT licensed)

As this removes the LZWDecode class from a public module, we have to do a major release when we release this change.

[codecov]

Codecov Report

Attention: Patch coverage is 97.14286% with 2 lines in your changes missing coverage. Please review.

Project coverage is 96.21%. Comparing base (8e1799e) to head (7c4df04).

Files with missing lines	Patch %	Lines
pypdf/_codecs/_codecs.py	98.50%	0 Missing and 1 partial ⚠️
pypdf/filters.py	66.66%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2887      +/-   ##
==========================================
- Coverage   96.27%   96.21%   -0.06%     
==========================================
  Files          52       52              
  Lines        8689     8703      +14     
  Branches     1733     1733              
==========================================
+ Hits         8365     8374       +9     
- Misses        187      192       +5     
  Partials      137      137              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[stefan6419846]

AFAIK our deprecation policy requires to keep this compatibility layer until we release version 7.0.

[MartinThoma]

I've added the class back, but the implementation now is the new _codecs.py one :-)

[stefan6419846]

Thanks. I have just removed some outdated comments. I guess we are going to keep the old class for now without any deprecation process?

[stefan6419846]

This indirectly exposes both the internal and the external implementation as a public module, including the encoder. Do we really want to do this? I tend to either make this import local or private by adding as _LzwCodec.