Error: `pelican-import --wpfile` interprets "/" on post titles as a filename/basename

[A-Infor]

• I have read the Filing Issues and subsequent “How to Get Help” sections of the documentation.
• I can reproduce this problem with stock/default settings file, theme, and sample content (as described in above “How to Get Help” sections of the documentation).
• I have searched the issues (including closed ones) and believe that this is not a duplicate.

• OS version and name: Debian 12
• Python version: 3.11.2
• Pelican version: 4.11.0.post0
• Link to theme: no theme set
• Links to plugins: no plugins installed
• Link to your site: still not published
• Link to your source: no source repository, just a local directory
• Link to a Gist with the contents of your settings file: https://gist.github.com/A-Infor/9c03b7f78dc32eac0f12868857c22c29

Issue

$ pelican-import --wpfile Posts.wordpress.xml 
content/🎞️-Alemanha-Socialista-∋-🕵️-Stasi-(Polícia-Secreta).rst
content/O-Perigo-da-Assexualidade.rst
content/Algumas-questões-que-afligem-os∕as-assexuais.rst
content/Sectarismo→hostilidade-∧-ideologias.rst
content/'Você-=-privilegiado!'.rst
content/Diversidade-sexual-≠-militância-pró-sexo.rst
content/Amor-e-sexo--diferenças-qualitativas.rst
content/Assexualidade-🆚-expectativas-alheias.rst
content/-Esquecer-&-Perdoar--≠--Vigiar-&-Punir-.rst
content/Assexualidade-∧-liberdade-de-não-buscar-parceiro(a)s.rst
content/Assexualidade--FAQ.rst
content/O-👑-poder-corrompe-∕-Lutzenberger.rst
content/⛪️-Violência-∈-Mal-→¬Mal-∌-Violência-∕-Tolstói.rst
content/🇧🇬-Bulgária--o-país-mais-corrupto-da-🇪🇺-União-Europeia.rst
content/Por-fora,-bela-viola-∧-por-dentro,-pão-bolorento.rst
content/🏴📚-Editora-Monstro-dos-Mares--conheça!.rst
Traceback (most recent call last):
  File "/home/analista/virtualenvs/pelican/bin/pelican-import", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/home/analista/virtualenvs/pelican/lib/python3.11/site-packages/pelican/tools/pelican_import.py", line 1259, in main
    fields2pelican(
  File "/home/analista/virtualenvs/pelican/lib/python3.11/site-packages/pelican/tools/pelican_import.py", line 973, in fields2pelican
    assert slug is None or filename == os.path.basename(
                           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AssertionError: filename is not a basename: Marxistas profissionais / 🏺 Platão

The post with title "Marxistas profissionais / 🏺 Platão" can't be imported due to the "/" on the title.
The other posts successfully imported before this one had the common slash "/" manually replaced by me for a division slash "∕", that does the trick of fooling Pelican.
But I have far too many posts with slashes on the title to keep doing that!

[A-Infor]

I noticed the problem wasn't on:
<title><![CDATA[Algumas questões que afligem os/as assexuais]]></title>
But instead on:
<wp:post_name><![CDATA[Algumas questões que afligem os∕as assexuais]]></wp:post_name>

So I used the following Regex on the ".xml" file:

• Searching by: <wp:post_name><!\[CDATA\[(.*?)/(.*?)\]\]></wp:post_name>
• Replacing by: <wp:post_name><![CDATA[\1∕\2]]></wp:post_name>

Now I need to do the same for every blog I have to import.

[A-Infor]

Now I need to do the same for every blog I have to import.

I did and it worked, but I don't think all this "manual" work should really be needed for such a simple matter (using / on post titles).

[justinmayer]

Hi Anderson. As I won't be able to allocate time to look into this, is this issue something you would be willing to investigate and potentially implement a solution for?

[A-Infor]

Hi @justinmayer .
I learnt Python for my Master's Degree studies, so I chose Pelican exactly for being a software for which I could contribute myself.
However, I still did not earn the degree, so I have a lot to do and little time for contributing to Pelican right now.
I'm only tinkering with Pelican in my spare time because the previous blog host that I used, https://blogs.sapo.pt/, is closing down this year and I need to make everything ready for https://neocities.org until June.

So:

1. I will probably find the cause of the error myself. It is probably the way Pelican handles special symbols;
2. I will probably fix it by hand, post by post;
3. I will probably not have time to study Pelican's code this semester and make a pull request to solve the issue.

[A-Infor]

@justinmayer In the case of this issue, simply replacing the common slash "/" by the division slash "∕" on the Wordpress XML before importing solved the problem.

I am not willing to implement this change on the Pelican Wordpress XML code myself, due to lack of time.
If you are not willing to implement yourself, feel free to close the issue.
I am too busy.