Build warnings triggered when run using Sphinx 2.2

[johnthagen]

Description of problem

Several warnings are triggered when building PDF output the latest master branch using Sphinx 2.2 and Python 3.7.

$ sphinx-build -a -n -W -b pdf . _build
...
.tox/docs/lib/python3.7/importlib/__init__.py:127: RemovedInSphinx30Warning: sphinx.environment.NoUri is deprecated. Check CHANGES for Sphinx API modifications.
  return _bootstrap._gcd_import(name[level:], package, level)

.tox/docs/lib/python3.7/site-packages/rst2pdf/rson.py:140: FutureWarning: Possible nested set at position 2
  splitter = re.compile(pattern).split

.tox/docs/lib/python3.7/site-packages/rst2pdf/pdfbuilder.py:602: RemovedInSphinx30Warning: sphinx.util.osutil.ustrtime is deprecated for removal
  date=ustrftime(self.config.today_fmt or _('%B %d, %Y'))

At least the bottom two are generated from this package, the first one I'm not sure on.

🖥 Versions

Python 3.7

Sphinx 2.2.0

rst2pdf master branch 2571340

sphinx-rtd-theme==0.4.3

[johnthagen]

See #788

[akrabat]

Thanks for the report. Out of interest, have you just started to use rst2pdf and how did you hear about it?

[johnthagen]

have you just started to use rst2pdf

I have just started using rst2pdf, yes.

how did you hear about it

I believe I googled "Create PDF using Sphinx" and found this SO post: https://stackoverflow.com/questions/39534718/how-to-create-a-pdf-out-of-sphinx-documentation-tool/47720278#47720278

So far it's been great! All we need is probably some doc updates (#802) and to get a new release on PyPI. Given Python 2 is EOL in just a couple months, all of our projects are Python 3-only (#796) so we have to point to the master branch on GitHub in our requirements.

[akrabat]

Thanks @johnthagen. I think #785 is the only blocker for a PyPI release with Python 3 support.

[akrabat]

I believe that this is fixed as #788 is now merged.

[johnthagen]

@akrabat Can confirm the warnings are now gone as of the latest master.

[johnthagen]

@akrabat We are still getting this warning using rst2pdf==0.96:

docs installed: alabaster==0.7.12,Babel==2.8.0,certifi==2019.11.28,chardet==3.0.4,Click==7.0,docutils==0.16,filelock==3.0.12,idna==2.8,imagesize==1.2.0,importlib-metadata==1.4.0,Jinja2==2.10.3,MarkupSafe==1.1.1,more-itertools==8.1.0,packaging==20.1,pdfrw==0.4,Pillow==7.0.0,pip-tools==4.4.0,pluggy==0.13.1,py==1.8.1,Pygments==2.5.2,pyparsing==2.4.6,pytz==2019.3,reportlab==3.5.34,requests==2.22.0,rst2pdf==0.96,six==1.14.0,smartypants==2.0.1,snowballstemmer==2.0.0,Sphinx==2.3.1,sphinx-rtd-theme==0.4.3,sphinxcontrib-applehelp==1.0.1,sphinxcontrib-devhelp==1.0.1,sphinxcontrib-htmlhelp==1.0.2,sphinxcontrib-jsmath==1.0.1,sphinxcontrib-qthelp==1.0.2,sphinxcontrib-serializinghtml==1.1.3,toml==0.10.0,tox==3.14.3,urllib3==1.25.8,virtualenv==16.7.9,zipp==2.0.1
docs run-test-pre: PYTHONHASHSEED='2706317774'
docs run-test: commands[0] | sphinx-build -a -n -W . _build
Running Sphinx v2.3.1
/.tox/docs/lib/python3.7/site-packages/rst2pdf/rson.py:140: FutureWarning: Possible nested set at position 2
  splitter = re.compile(pattern).split

I think one of the warnings still needs to be fixed.

[dave-lawrence-dev]

Hi, still getting this on 0.96.dev0 running on Ubuntu focal:

/usr/lib/python3/dist-packages/rst2pdf/rson.py:140: FutureWarning: Possible nested set at position 2 splitter = re.compile(pattern).split

[samuelharmer]

Windows with Miniconda installed Python, pip installed rst2pdf. Doesn't happen with all my input files though so assume there is a particular problem with some of them. Let me know if I can help identify the cause, or whether this wants its own bug report?

${HOME}\.conda\envs\general\lib\site-packages\rst2pdf\rson.py:119: FutureWarning: Possible nested set at position 2
  splitter = re.compile(pattern).split

python --version
Python 3.8.5
rst2pdf --version
0.98

OK, I was going to attempt to provide a code sample but this is especially strange. Given the exact same input, it doesn't happen every single time.

[akrabat]

@samuelharmer This has been fixed in #880 and will be in the next release. It is harmless on Python 3.8 as it's a warning for a change to Python expected in the future.

[samuelharmer]

@akrabat thanks for the pointer, #880 didn't come up in my Google for the error message so hopefully this will help anyone else too 😄