[PATCH] docs-rst: automatically convert Graphviz and SVG images

[Markus Heiser]

> Am 02.03.2017 um 20:34 schrieb Mauro Carvalho Chehab <mchehab at s-opensource.com>:
> 
> Em Thu, 2 Mar 2017 20:06:39 +0100
> Markus Heiser <markus.heiser at darmarit.de> escreveu:
> 
>> Hi Mauro,
>> 
>>> Tested here with the enclosed patch.  
>> 
>> great, big step forward making /media/Makefile smaller ...  thanks a lot!!!!
>> 
>>> It crashed:
>>> Exception occurred:
>>> File "/devel/v4l/patchwork/Documentation/sphinx/kfigure.py", line 222, in dot2format
>>>   sys.stderr.write(err)
>>> TypeError: write() argument must be str, not bytes
>>> The full traceback has been saved in /tmp/sphinx-err-_1vahbmg.log, if you want to report the issue to the developers.
>>> Please also report this if it was a user error, so that a better error message can be provided next time.
>>> A bug report can be filed in the tracker at <https://github.com/sphinx-doc/sphinx/issues>. Thanks!
>>> Documentation/Makefile.sphinx:69: recipe for target 'htmldocs' failed
>>> make[1]: *** [htmldocs] Error 1
>>> Makefile:1450: recipe for target 'htmldocs' failed
>>> make: *** [htmldocs] Error 2
>>> 
>>> Weird enough, it produced a Documentation/output/media/uapi/v4l/pipeline.svg file.  
>> 
>> I guess that the dot command writes something to stderr. This is captured 
>> by the extension and printed to stderr ...
>> 
>> +def dot2format(dot_fname, out_fname):
>> ...
>> +    exit_code = 42
>> +    with open(out_fname, "w") as out:
>> +        p = subprocess.Popen(
>> +            cmd, stdout = out, stderr = subprocess.PIPE )
>> +        nil, err = p.communicate()
>> +
>> +        sys.stderr.write(err)
>> +
>> +        exit_code = p.returncode
>> +        out.flush()
>> +    return bool(exit_code == 0)
>> 
>>> File "/devel/v4l/patchwork/Documentation/sphinx/kfigure.py", line 222, in dot2format
>>>   sys.stderr.write(err)
>>> TypeError: write() argument must be str, not bytes  
>> 
>> Do we need this stderr output? For a first test, uncomment the 
>> "sys.stderr.write(err)“ in line 222. Or, if we really need the
>> stderr, try:
>> 
>> -        sys.stderr.write(err)
>> +        sys.stderr.write(str(err))
> 
> Yes, this fixed. I actually did:
> 
> -        sys.stderr.write(err)
> +        sys.stderr.write(str(err))
> +        sys.stderr.write("\n")
> 
> It is now printing:
> 	b''
> 
> I added the \n print to avoid it to be mixed with the "writing output"
> prints.
> No idea how to make sense from it - but clearly, the error report
> logic require some care ;-)


Aargh, I’am a idiot ... I guess 'sys.stderr.write(err)‘ is a artefact
of my development, simply drop it and the subprocess.PIPE of stderr
also.

+    with open(out_fname, "w") as out:
+        p = subprocess.Popen(
-            cmd, stdout = out, stderr = subprocess.PIPE )
+            cmd, stdout = out)
+        nil, err = p.communicate()
-
-        sys.stderr.write(err)
-
+        exit_code = p.returncode
+        out.flush()
+    return bool(exit_code == 0)

I can’t test it ATM, but without redirect stderr, the stderr
of the parent process is inherited.

  https://docs.python.org/3.6/library/subprocess.html#popen-constructor

The Popen.communicate() always returns a tuple (stdout_data, stderr_data)
with above the tuple is always (None, None).

  https://docs.python.org/3.6/library/subprocess.html#subprocess.Popen.communicate

  """to get anything other than None in the result tuple,
     you need to give stdout=PIPE and/or stderr=PIPE too."""

Sorry, that I made all this mistakes, but „here“ I have only mail
and web, no dev-env and I miss my emacs  ;) 

If the suggestion above does not work, I have to investigate
more time next weekend.

-- Markus --


> 
> Thanks,
> Mauro