As was sort-of mentioned, ebookmaker shares pandoc code to do its own rst -> html conversion. As a result, it doesn't support anything that's been added in pandoc in the past 10 years, while including some PG-specific stuff that isn't supported by modern pandoc (and not terribly useful, it appears).
The PG rst toolchain adds dependencies on xetex to make PDF. files.
I think it's generally a bad idea (i.e. not sustainable in the long term) to have post-submission hand-editing going on in more than one place. with rst, I see people using the rst->html conversion product to make new source files, which then get hand-editied. I've also seen the html files become zombie files which get edited but ignored by ebookmaker for one reason or another.
I don't have a clear sense of what our vision is for this use of RST. There were about 500 rst-sourced books not being updated (in danger of being zombies) when I started working on Ebookmaker and I revived the code that had stopped working. We could kill off the rst sources and coversion code and use exclusively the html; we could rip out our custom rst code and replace it with modern pandoc, or something else, but it would be good to not do all of the above.
Eric
> On Aug 29, 2021, at 1:43 PM, Greg Newby <gbnewby@pglaf.org> wrote:
>
> Thanks for confirming ebookmaker is working, James!
>
> Your RST production path looks good to me. As I mentioned in our correspondence, pandoc seems to do a good job.
>
> It's not required that the .txt include representation of italics or other typography. It seems that in either of the .txt generation steps, you might want to do a little further editing by hand to ensure the reading experience is as good as may be.
>
> Best,
> Greg
>
> On Sat, Aug 28, 2021 at 08:30:56PM -0500, James Simmons wrote:
>> Greg,
>>
>> Ebookmaker is working again. Thanks.
>>
>> Regarding using RST to generate PG formats, I found that if I avoid PG
>> extensions I can use the following command to generate my HTML from RST:
>>
>> pandoc Vishnupuranam.rst --shift-heading-level-by=1 -f rst -t html -s -o
>> Vishnupuranam.html
>>
>> The shift heading is necessary because if you don't use it the title and
>> the top level headings will both use H1. It sounds like Ebookmaker has a
>> problem with that.
>>
>> I use this command to generate plain text:
>>
>> pandoc Vishnupuranam.rst -f rst -t plain -s -o Vishnupuranam.txt
>>
>> followed by this command to fix the line endings (because I work on a Linux
>> box):
>>
>> unix2dos Vishnupuranam.txt
>>
>> This does a pretty good job, but it strips out asterisks that indicate
>> italicized text. I can live with that if PG can. There is a web utility
>> that handles the italics better:
>>
>> https://cloudconvert.com/rst-to-txt
>>
>> This produces output with some extra characters that need to be removed
>> (footnotes look like [^1] instead of [1], section headings begin with ###),
>> but that is simple enough.
>>
>> James Simmons
>>
>> On Sat, Aug 28, 2021 at 4:32 PM Greg Newby <gbnewby@pglaf.org> wrote:
>>
>>> Thanks again for this report, James.
>>>
>>> The issue was fixed by Eric, and the online ebookmaker is now updated.
>>> Please give it another try!
>>> Best,
>>> Greg
>>>
>>> On Fri, Aug 27, 2021 at 05:50:40PM -0500, James Simmons wrote:
>>>> I used Ebookmaker as recently as yesterday with no problems, but today
>>> I'm
>>>> getting no epub generated and output.txt shows the following:
>>>>
>>>> unzipping /htdocs/ebookmaker/cache/20210827154639/Vishnupuranam.zip
>>>> Input file: Vishnupuranam.html
>>>> EbookMaker 0.11.4
>>>> WARNING:root:Gutenberg Database is inactive because psycopg2 not
>>> installed
>>>> Traceback (most recent call last):
>>>> File
>>> "/var/www/.local/share/virtualenvs/ebookmaker-A6HCkzTJ/bin/ebookmaker",
>>>> line 18, in <module>
>>>> EbookMaker.main ()
>>>> File
>>> "/var/www/.local/share/virtualenvs/ebookmaker-A6HCkzTJ/lib/python3.6/site-packages/ebookmaker/EbookMaker.py",
>>>> line 534, in main
>>>> job.url = os.path.join(absoutputdir, output_files['epub.images'])
>>>> File
>>> "/var/www/.local/share/virtualenvs/ebookmaker-A6HCkzTJ/lib/python3.6/posixpath.py",
>>>> line 94, in join
>>>> genericpath._check_arg_types('join', a, *p)
>>>> File
>>> "/var/www/.local/share/virtualenvs/ebookmaker-A6HCkzTJ/lib/python3.6/genericpath.py",
>>>> line 149, in _check_arg_types
>>>> (funcname, s.__class__.__name__)) from None
>>>> TypeError: join() argument must be str or bytes, not 'NoneType'
>>>>
>>>> I tried it a couple of times to be certain.
>>>>
>>>> Other than correcting some spelling errors, my input today is no
>>>> different from yesterday.
>>>>
>>>> James Simmons
>>>
>>>> _______________________________________________
>>>> gutvol-d mailing list
>>>> gutvol-d@lists.pglaf.org
>>>> https://lists.pglaf.org/mailman/listinfo/gutvol-d
>>>> Unsubscribe: https://lists.pglaf.org/mailman/options/gutvol-d
>>>
>>> _______________________________________________
>>> gutvol-d mailing list
>>> gutvol-d@lists.pglaf.org
>>> https://lists.pglaf.org/mailman/listinfo/gutvol-d
>>> Unsubscribe: https://lists.pglaf.org/mailman/options/gutvol-d
>>>
>
>> _______________________________________________
>> gutvol-d mailing list
>> gutvol-d@lists.pglaf.org
>> https://lists.pglaf.org/mailman/listinfo/gutvol-d
>> Unsubscribe: https://lists.pglaf.org/mailman/options/gutvol-d
>
> _______________________________________________
> gutvol-d mailing list
> gutvol-d@lists.pglaf.org
> https://lists.pglaf.org/mailman/listinfo/gutvol-d
> Unsubscribe: https://lists.pglaf.org/mailman/options/gutvol-d
_______________________________________________
gutvol-d mailing list
gutvol-d@lists.pglaf.org
https://lists.pglaf.org/mailman/listinfo/gutvol-d
Unsubscribe: https://lists.pglaf.org/mailman/options/gutvol-d