-
-
Notifications
You must be signed in to change notification settings - Fork 14.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Migrate modules docs to markdown #175586
Comments
It would be nice but there was opposition to including pandoc in the NixOS system closure. So we would either need to decouple the manual from the system, choose some lighter parser (maybe lowdown + HTML→DocBook via XSL), or cache pre-generated DocBook like we currently do with the NixOS manual. |
Sure! If we don't want pandoc in the build closure for NixOS, we can choose to implement step 5. NixOS' dependency on pandoc will be a temporary affair.
We'll need a tool that can convert tons of snippets at pace. Pandoc can do this (as linked before). Shelling out will not be an option, so I suspect that this would involve a custom tool. Could be worth it.
This is terrible developer UX. If we want people to write less docs, this is one way to do it. |
probably, yeah. there's not that much literal docbook in nixos modules that's properly labelled as such, and among the great many docbook-using descriptions that are not properly labelled by far the most use for docbook uses only a few tags (literal, option, filename, programlisting, ...). parsing descriptions as xml fragments and running a custom transform over them actually looks quite feasible! |
Hmm yeah, that makes me think that perhaps my migration path is overly correct. We could just do it and fix the occurrences of the small set of tags used after switching.
I would avoid writing any and all parsing logic, and only go from semi-structured data to {markdown + pandoc raw attribute} in our own tooling. |
I bet it's possible to introduce and get rid of pandoc until the next release. It would be a shame to block the refactor for a temporary dependency, as long as it doesn't stay around forever. |
We do not have a time frame, so it might take longer than the next release. Especially when we have not even settled on a toolchain yet (currently the candidates are #105036 and #108063). And people who would be annoyed by the closure do often use unstable. So, while deciding that having them choose between having pandoc in their system closure and disabling local docs temporarily can be a reasonable trade off for the long term goal of improving the docs infrastructure, it might still turn out few |
I would keep it simple and just automatically convert all P.S. NixOS/nix#6583 already assumes that option descriptions are in markdown. Supporting other markup languages would just make life more complicated. |
we've made an attempt at producing a script to migrate docs. perl -pi -e 's,<literal>([^`]*?)</literal>,`$1`,g' nixos/modules/**/*.nix
perl -pi -e 's,<replaceable>([^»]*?)</replaceable>,«$1»,g' nixos/modules/**/*.nix
perl -pi -e 's,<filename>([^`]*?)</filename>,`$1`,g' nixos/modules/**/*.nix
perl -pi -e 's,<code>([^`]*?)</code>,`$1`,g' nixos/modules/**/*.nix
perl -pi -e 's,<option>([^`]*?)</option>,`$1`,g' nixos/modules/**/*.nix
perl -pi -e 's,<command>([^`]*?)</command>,`$1`,g' nixos/modules/**/*.nix
perl -pi -e 's,<link xlink:href="(.+?)" ?/>,$1,g' nixos/modules/**/*.nix
perl -pi -e 's,<link xlink:href="(.+?)">(.*?)</link>,[$2]($1),g' nixos/modules/**/*.nix
perl -pi -e 's,<package>([^`]*?)</package>,`$1`,g' nixos/modules/**/*.nix
perl -pi -e 's,<emphasis>([^*]*?)</emphasis>,*$1*,g' nixos/modules/**/*.nix
perl -pi -e 'BEGIN{undef $/;} s,<citerefentry>\s*<refentrytitle>\s*(.*?)\s*</refentrytitle>\s*<manvolnum>\s*(.*?)\s*</manvolnum>\s*</citerefentry>,`$1 ($2)`,smg' nixos/modules/**/*.nix does a reasonable job of replacing most docbook tag with markdown constructs. around 2000 matches that look like xml tags remain after that, but a lot of that is either actual xml than can be safely ignored (eg fontconfig) or extended lists that have to converted with more thought anyway. |
having spent a couple hours trying to hack together something to process the docbook xml snippets, it doesn't quite seem to be worth the trouble over using regexes. descriptions are not well-formed xml to begin with, interpolations split them up even further (and are a mess to track), some descriptions aren't even valid xml fragments—and the main benefit of automatic translation (being able to translate eg varlists) requires all of that to not be like it is. and we'll need sapient inspection of the results anyway. :/ |
This is why I proposed doing it as part of the existing pipeline, where we do have quality data. Another option is to write a script that takes writes out a long markdown document with markers to delineate the individual I guess though, docs quality during the transition is not a priority, so then I suppose leaving some descriptions in a broken state is more acceptable than having pandoc in the closure? Let me also mention |
maybe we should do something that's a mix of all of the above: temporarily extend/augment the doc merging script we already have to render markdown to docbook, add (as we've learned from the split docs changes, a flag day event really doesn't work very well. the split docs builds at least broke to call attention to breakages, docbook somehow still making it into the MD future might not be as loud) |
while trying some things out we've noticed we can't change some docbook tags without changing eg the manpages. one we noticed in particular was |
interest seems low, but that might be a discoverability issue. perhaps a round through discourse would be useful, what do you think? |
Forgot to share my thoughts: this is great. There are currently at least three places where the "NixOS module system option documentation rendering logic" is defined: nixpkgs, nmd (used by Home Manager, ping @rycee), and nixos-search. It would be nice to unify the three. Merging this now would break some options at nixos-search. Maybe I'll find the energy to port the changes there soon. |
No problems here.
Home Manager doesn't have to use
Flakes are experimental anyway, so I don't think this is a blocker?
I don't think the PR needs more review. If anything pops up, we can fix it as we go. It'd be a nicer experience for contributors if their changes are actually mergeable. I'd say merge and then announce on discourse. |
A tracking issue for the conversion project would also be helpful for discoverability. |
I think this situation is not optimal: since Home Manager inherits e.g.
The confusingly-named |
FWIW, nmd has been moving towards asciidoc for some time and much of the HM manual is written in asciidoc. So it is likely that option descriptions in HM will go that way as well. I've been meaning to switch over but have never found the time to do so. I'll have more spare time sometime in July so might make some moves then. |
that's a point, actually. nixos does have an asciidoc exporter for module docs, but it's kind of not useful. same goes for the MD exporter at this point, but that will become more useful in the future at least. the asciidoc exporter will likely have to go away though at some point, unless we want to write multiple renderers. |
Arion's docs are also asciidoc based and it's the first consumer of the Nixpkgs asciidoc renderer. I'm not aware of any others. Asciidoc was a good contender before markdown was chosen for the manuals, but I'm avoiding it for new projects. Migrating arion's docs to markdown wouldn't be bad. |
Do I read correctly that we don't have any blockers for #176146? |
don't think there are any, especially since the scope is still rather limited and changes to the rendering tool are easy enough should problems pop up. |
@pennae, do you think the script you wrote in #175586 (comment) is worth promoting to a It may create a better first iteration than I would, because I still forget to add |
I am not going to waste my time prefixing every documentation string in my modules with |
@ehmry the markers were added to facilitate a smooth transition, taking into account 3rd party repos. The "extra processing" is very minimal, because conversion happens at build time, not eval time. The markers will be removed in due time. |
Except for stuff like I guess nmd could just use |
|
AFAICT that leaves |
what's wrong with |
It uses |
ah, well. that we probably can't fix without either adding special cases to the "no docbook descriptions in nixpkgs" rule, or dropping the rule altogether. neither of those sound like a good path forward. 🙁 maybe this (very special) option could be hidden for now and moved to the manual instead? |
Yes, we could make it invisible/internal (can't remember the distinction) for now and add a |
|
FWIW I've put together a preliminary PR to support AsciiDoc descriptions in nmd. See https://gitlab.com/rycee/nmd/-/merge_requests/7/diffs Since CommonMark to some extent is similar I also use this for the |
Just to justify the use of AsciiDoc a bit. The primary reason that HM uses that instead of Markdown is that it simplifies the build pipeline a great deal and reduces the build closure size quite a lot. It also integrates well with the existing DocBook setup. For example, you can see that it was reasonably simple to support AsciiDoc module descriptions in the nmd merge request above without noticeably expanding the closure or build time. The build closure and build time is more important for HM documentation since it will almost never be fetched from a central binary cache while the Nixpkgs/NixOS documentation will nearly always be available from That said, nmd uses the legacy Python version of the asciidoc tooling, which does add a dependency on Python and does not support recent AsciiDoc features. The more up-to-date Asciidoctor pulls in Ruby, which is even heavier so I would be very reluctant to move to that. With the recent resurgence of AsciiDoc development (see https://asciidoc.org/) perhaps there will be some more light-weight tool that can do AsciiDoc → DocBook in the future, perhaps based on libasciidoc. Of course, if somebody knows of a light-weight Markdown → DocBook converter (that supports description lists) then we could equally well use that in nmd. It's just that I haven't found anything like that in my, admittedly shallow, investigations. |
I would highly recommend to follow the module system's migration to markdown, so that all markers can be removed and the contributor experience is consistent. |
@rycee did you have a look at the script we use in nixpkgs? It depends on Python + mistune, would that be small enough for your purposes? I also think we should keep in mind the goal to make consumers able to use +1 on strongly recommending not adding a third |
@ncfavier Ah yeah, the |
correct. we tried to add those, but the mistune plugin API seems to not be powerful enough to support the syntax nixpkgs already uses. since these lists were rather rare in nixos options we decided to convert them to regular lists instead, but if something pops up that lets us support them: all for it! :) |
unfortunately we can't unconditionally make this text markdown without impacting downstream users of docs generation (as noted in #175586). hide it entirely until the transition is complete.
for the record: option docs now support all features of nixpkgs-flavored markdown, including definition lists. |
can this be considered done, or only once docbook in option docs is no longer allowed? |
It's done when "mdDoc" is the default behavior of the
|
technically we can do it, but that would go against the deprecation process for docbook descriptions that we agreed on elsewhere. so let's keep this open until after 23.05, when we can disable docbook support. :) |
unfortunately we can't unconditionally make this text markdown without impacting downstream users of docs generation (as noted in NixOS#175586). hide it entirely until the transition is complete.
#239636 removed the last vestiges of the docbook toolchain in our manual builds, and mdDoc behavior has been the default since #237557. removing mdDoc should probably wait for another release to keep backporting effort reasonable both within and without the nixpkgs repo, but the migration is now finally done. |
Project description
Migrate module docs to markdown.
Pandoc should be able to mix and match 🤞
Suggested pipeline
Suggested migration path
description = lib.docMD "..."
, etcdescription
(and other such attrs)lib.docMD
calls.From the broader ecosystem perspective, it would be good to keep the flexible pipeline, as we can use it to dismantle the cottage industry of custom module docs generators. It's a matter of wrapping it with the right pandoc syntax.
Technical details
(just to get a feel for it)
->
->
Some XML
->
Markdown:
Alternatives
Metadata
lib/
,nixos/
,pkgs/
The text was updated successfully, but these errors were encountered: