An XML realization of Twtxt

[sjehuda]

Greetings.

I am working on a document uniformication platform which enables a uniform support to any syndicated document, to any software which parses The Atom Syndication Format.

• I have created a script to convert Twtxt into XML; and

• I have created an XSLT stylesheet to transform that XML into Atom.

• https://git.xmpp-it.net/sch/Focus/src/branch/main/utilities/twtxt-to-xml.py

• https://git.xmpp-it.net/sch/Focus/src/branch/main/templates/twtxt-1.0.xslt

python utilities/twtxt-to-xml.py twtxt.txt
xsltproc templates/twtxt-1.0.xslt twtxt.txt.xml > result.atom

I desire to have a defined XML form of Twtxt, in order to provide a proper XML format, and a subsequent stylesheet.

This may be considered as related to #193

[prologic]

Does your script implement Twtxt Extension here: https://twtxt.dev ?

[sjehuda]

Yes, only to some extensions.

Extension "subject" is realized to "twtxt:context", because I did not find any equivalent of Atom; though, RFC 4685: Atom Threading Extensions might be a good choice.

Extension "prev" is realized as "atom:link" with attribute rel="previous" as specified in RFC 5005: Feed Paging and Archiving.

Extension "link" is realized as "atom:link" with attribute rel="related" with a respective title.

Extension "follow" is realized as "atom:link" with attribute rel="follow" with a respective title.

Extension "avatar" is realized as "atom:logo".

Please. Explore the attached documents.

twtxt.atom.xml
twtxt_fine.xml
twtxt.txt.xml
twtxt.txt

Process order.

python utilities/twtxt-to-xml.py twtxt.txt 
xsltproc templates/twtxt-1.0.xslt twtxt.txt.xml > twtxt.atom

For convenience.

xmllint --format twtxt.txt.xml > twtxt_fine.xml

---

Post script

@prologic

I have tried with your Twtxt documents, and the script failed due to this line which has two "tabs".

'2025-04-08T14:39:13Z\t(#2znenta) @<movq https://www.uninformativ.de/twtxt.txt> Not according to the output of `./yarnc debug <your feed url>`:\u2028\u2028```\u2028znf6csa 2025-04-07T19:59:51+00:00\tI wonder if my current Linux installation will actually make it to 20 years:\u2028\u2028    $ head -n 1 /var/log/pacman.log\u2028    [2011-07-07 11:19] installed filesystem (2011.04-1)\u2028\u2028It’s not toooo far into the future.\u2028\u2028It would be crazy … 20 years without reinstalling once … phew. 🥴\u2028```'

Error

  File "/The_Focus/utilities/twtxt-to-xml.py", line 42, in twtxt_entries
    date, text = line.split("   ")
    ^^^^^^^^^^
ValueError: too many values to unpack (expected 2)

[sjehuda]

@prologic

Please. Provide a Twtxt document with all extensions, and I will act upon the new ones which are not yet realized.

[prologic]

@prologic

Please. Provide a Twtxt document with all extensions, and I will act upon the new ones which are not yet realized.

That's the twtxt.dev site my friend 👌

[prologic]

I think your parser might have a bug? 🧐 Does it recognize the miltiline extension?

[sjehuda]

That's the twtxt.dev site my friend.

Yes. I realize it. It be easier when all are arrayed in the same document.

Nevertheless, please advise changes.

[sjehuda]

I think your parser might have a bug? 🧐 Does it recognize the miltiline extension?

No. It does not.

[sjehuda]

I added support for multiple lines, by appending a line to previously created object. https://git.xmpp-it.net/sch/Focus/commit/6dcbbbcb9458462dbff0bf87f51c0f6a44bcee00 However, line breaks are not realized; and I do not know whether it is a fault of the module "markdown" which converts the data to HTML or my fault. Generally, it works better; yet, I would be glad to receive help.