pyppd 1.1.1

Bug fix release, to make compressing PPD files vastly faster, avoid weird "*Product" entries in PPD files, and to not use wrong model names for Brother BR-Script printers.

• Replace bytes with bytearray for massive speed up
  Bytes type is very inefficient for append/concatenation and involves copying the entire data. This one-line fix speeds up foomatic-db compression from ~20 minutes to 10 seconds (Pull request #5).
• Correction for Brother BR-Script PPD files
  Use the "*Product" entry for the model name.
• When generating PPD index, avoid using single "*Product" entries in PPD
  When creating the PPD index and there is a PPD with only one, single "*Product" line, use the model name of this line only if the PPD has no "*1284DeviceID" and no "*ModelName", as a single "*Product" entry usually either repeats "*1284DeviceID" and/or "*ModelName" or contains something weird.

pyppd 1.1.0 release

Vastly reduced memory consumption when extracting a PPD file and incorporated the patches of the Debian/Ubuntu package

• When extracting a PPD file from the archive streaming decompression is used now instead of decompressing the whole archive into memory at once (Issue #2), Thanks to Sambhav Dusad (@dsam82).
• Use JSON dumps instead of Pickle dumps, from Ubuntu/Debian to make archive reproducible, but also faster than Pickle dumps. Thanks to Didier Raboud (@OdyX).
• Sort PPD list before archiving, from Ubuntu/Debian to make archive reproducable. Thanks to Didier Raboud (@OdyX)
• Use "python3" in shebangs, from Debian/Ubuntu