Pypdf3

-----------------------

- Fix a bug where merging pages did not work in "no-rename" cases when the
second page has an array of content streams.

- Remove some debugging output that should not have been present.

-----------------------

- Add capability to merge pages from multiple PDF files into a single page
using the PageObject.mergePage function. See example code (README or web
site) for more information.

- Add ability to modify a page's MediaBox, CropBox, BleedBox, TrimBox, and
ArtBox properties through PageObject. See example code (README or web site)
for more information.

- Refactor pdf.py into multiple files: generic.py (contains objects like
NameObject, DictionaryObject), filters.py (contains filter code),
utils.py (various). This does not affect importing PdfFileReader
or PdfFileWriter.

- Add new decoding functions for standard PDF filters ASCIIHexDecode and
ASCII85Decode.

- Change url and download_url to refer to new pybrary.net web site.

-----------------------

- Fix new bug introduced in 1.2 where PDF files with \r line endings did not
work properly anymore. A new test suite developed with various PDF files
should prevent regression bugs from now on.

- Fix a bug where inheriting attributes from page nodes did not work.

-----------------------

- Improved support for files with CRLF-based line endings, fixing a common
reported problem stating "assertion error: assert line == "%%EOF"".

- Software author/maintainer is now officially a proud married person, which
is sure to result in better software... somehow.

-----------------------

- Add capability to rotate pages.

- Improved PDF reading support to properly manage inherited attributes from
/Type=/Pages nodes. This means that page groups that are rotated or have
different media boxes or whatever will now work properly.

- Added PDF 1.5 support. Namely cross-reference streams and object streams.
This release can mangle Adobe's PDFReference16.pdf successfully.

--------------------------
- Fixed an issue where embedded font names would not be written when reading and saving old Japanese PDF files with font names specified in Shift-JIS