Adding support for "role" attributes for the DocBook reader (second try!) #10932
Conversation
This commits adds a getRoleAttr function to get the value from the role attribute of a DocBook element. This value is then added to the returned Pandoc object by using addPandocAttributes.
Sections have to be considered in a different way than regular blocks and inlines. In this case, we wrap the content in a Div element with a `section` class and a `level` attribute in addition to the `role` attribute. See this comment: jgm@d097c64
Elements parsed with withOptionalTitle do not automatically get the role attributes transfered to them. This covers this problem.
|
Hello! A month has passed, and in the meantime I have been working on my IDML reader... It is now giving pretty nice results :), and is now easy to install and test out: https://outdesign.deborderbollore.fr/en/md/2_installation.html. If there is anything I can do at this point to help with the PR, don't hesitate. I'd love to see that be integrated in the main branch, as it would definitely help in making IDML files really parsable by Pandoc! (ping @lifeunleaded who might be interested as well.) |
|
Hi! As you can see, the Windows build has a warning that should be fixed. (Actually all the others should have given this too, but the CI was misconfigured; this has now been fixed.) Fix that up, rebase, and hopefully we can get it merged soon. |
|
Sorry - this warning is not from your code, it comes from a bad commit in main that you rebased on top of. So if you just rebase your changes on top of current HEAD, everything should pass. |
| return $ case attrValue "role" e of | ||
| "" -> content | ||
| _ -> divWith ("", ["section"], | ||
| ("level", T.pack $ show n') : attrs) content |
There was a problem hiding this comment.
We shouldn't be mixing the use of section Divs and bare headings in the same document the way you do here. Why not add the role attribute to the Header? When the resulting AST is passed through makeSections, it will become a section div.
There was a problem hiding this comment.
Hello, thanks for taking the time to have a look at it again. I proposed this modification following what I understood from your recommendation in #10665 (comment).
The problem that led to this modification was that the role attributes were applied recursively to all child sections, because of the way addPandocAttributes is designed. From my understanding, we arrived to this "acceptable" solution to avoid this recursion.
Should we figure out something else then?
Sections have to be considered in a different way than regular blocks and inlines. In this case, we wrap the content in a Div element with a `section` class and a `level` attribute in addition to the `role` attribute.
Elements parsed with withOptionalTitle do not automatically get the role attributes transfered to them. This covers this problem.
Functions that expect UTF-8-encoded filenames should make it easier to write platform-independent scripts, as the encoding of the actual filename depends on the system. Additionally, this also adds a generalized method to run commands, and functions to retrieve XDG directory names. The new functions are `command`, `copy`, `read_file`, `remove`, `rename`, `times`, `write_file`, `xdg`.
for every PDF engine, not just LaTeX/ConTeXt. This is part of the fix for jgm#10911.
Export `copyrightMessage` from the unexported module Text.Pandoc.App.CommandLineOptions and reexport from Text.Pandoc.App [API change]. This avoids the need for a duplicated version in pandoc-cli, which can now depend on the library's exported version.
This reverts commit 2cf9b55.
Text.Pandoc.App.CommandLineOptions and pandoc-cli/src/pandoc.hs had similar code for generating version information. To avoid duplication, we now export `versionInfo` from Text.Pandoc.App [API change]. (The function is reexported from the non-public module Text.Pandoc.App.CommandLineOptions.) This function has three parameters that can be filled in when it is called by pandoc-cli. This change will make it simpler to revise version information.
This function performs a normalization of Pandoc documents. E.g., multiple successive spaces are collapsed, and tables are normalized such that all rows and columns contain the same number of cells. Closes: jgm#10356
from the latest chicago-author-date.csl. (Note that this goes from the 17th to the 18th edition.) Update tests.
When we updated to the latest chicago-author-date.csl, this test no longer tested what it was supposed to; so we use a different csl.
It was meant to test subsequent author substitution, but the new chicago-author-date doesn't do this. So we use a different CSL.
[API change] New Avif constructor on ImageType. Closes jgm#10979.
We were only getting the return status for the tests, apparently, from `cabal test`. So now we run `cabal build` separately.
The functions allows to check the existence of file-system objects.
Closes jgm#10983 by allowing `nocase` spans to be used to suppress capitalization of initial word in a footnote.
Previously we set `--ghc-options` in Makefile and CI; but this overrides the ghc-options set in the pandoc.cabal file. Better to add options one-by-one using `--ghc-option`. We no longer use GHC_OPTIONS and just put these extra options in CABAL_OPTIONS.
For now I want to avoid having to put in lots of CPP.
This implements the changes suggested in jgm#9956, with the exception of the filecolor/urlcolor one. These would require adding some regex to guess the link types. This is theoretically possible to do, but it wasn't clear to me that this is a good thing to put in a default template. Happy to adjust if you have thoughts on this. Closes jgm#9956. Some things to note: I'm converting colors by passing them as content, as I was seeing pandoc escape # if that was included. I set the default fonts for math and code ("raw") to fonts that are bundled with Typst. These need not be those fonts if there are more familiar pandoc preferences.
This solves the problem of unwanted capitalization of names at the beginning of citations in footnotes. Closes jgm#10983.
These are the development versions of the LaTeX binaries; installable, e.g., with `tlmgr install latex-base-dev`. Closes: jgm#10991
Each supported engine is now printed on a line of its own.
Sections have to be considered in a different way than regular blocks and inlines. In this case, we wrap the content in a Div element with a `section` class and a `level` attribute in addition to the `role` attribute.
Elements parsed with withOptionalTitle do not automatically get the role attributes transfered to them. This covers this problem.
|
... I am so bad at rebasing. I am sorry. The only thing I tried to do was to change the commit-msg-length, which apparently did not work, and instead it added all those commits that basically are all the commits that were the changes on top of current HEAD since I made my PR. I can make yet another PR if this is cleaner, sorry for this. |
|
You need to use Try |
This PR is the direct result of what has been discussed on #10665. This proposition is basically a cleaner version based on the previous contribution attempt.
More details on what this PR cover can be found in this comment.