Closed akinsho closed 2 years ago
kdheepak
commented
2 years ago Try it like this:
<!-- panvimdoc-ignore-start -->
<h1 align="center">
toggleterm.nvim
</h1>
<p align="center">A <i>neovim</i> plug
<!-- panvimdoc-ignore-end -->
### Multiple orientations
…
Whatever is in between the start and end comment “tag” is ignored. Does that make sense?
akinsho
commented
2 years ago Will give that a shot. So do I essentially have to wrap each html tag individually? I wonder why the the tags placed as they are in my example don't work i.e. both start and end are present it just doesn't seem to catch the end tag in my example
kdheepak
commented
2 years ago So do I essentially have to wrap each html tag individually?
If you want the html tag to appear in the markdown, and do not want them to appear in the vimdoc, you'll have to do that.
kdheepak
commented
2 years ago it just doesn't seem to catch the end tag in my example
What do you mean it doesn't catch the end tag?
kdheepak
commented
2 years ago This script is what replaces everything between start and end tags with empty pandoc objects: https://github.com/kdheepak/panvimdoc/blob/main/scripts/skip-blocks.lua
kdheepak
commented
2 years ago @akinsho is this working for you? If yes, I'll close this issue.
akinsho
commented
2 years ago Sorry @kdheepak for leaving this open for this long, I've actually been completely sideswided by work and life and haven't had the opportunity to try it yet. I'll give it a go this evening and report back. If you don't hear back from me, just close the issue and I'll comment on here something if comes up
kdheepak
commented
2 years ago No worries, I’ve been swamped with work too. I can leave it open until you get back. I just tested it too, so I’m pretty sure it works as a GitHub action with the pandoc version in the docker image. So I’m curious what’s going on on your end.
akinsho
commented
2 years ago @kdheepak I just tried again wrapping each tag I wanted to exclude manually and it essentially removes all the contents from my help file.
<!-- panvimdoc-ignore-start -->
<h1 align="center">
toggleterm.nvim
</h1>
<!-- panvimdoc-ignore-end -->
<!-- panvimdoc-ignore-start -->
<p align="center">A <i>neovim</i> plugin to persist and toggle multiple terminals during an editing session</p>
<!-- panvimdoc-ignore-end -->
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
### Multiple orientations
- **Float**
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
- **Vertical**
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
- **Tab**
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
### Send commands to different terminals
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
This results in
*toggleterm.txt*plugin to persist and toggle multiple terminals during an editing session
==============================================================================
Table of Contents *toggleterm-table-of-contents*
Generated by panvimdoc <https://github.com/kdheepak/panvimdoc>
vim:tw=78:ts=8:noet:ft=help:norl:I'm able to replicate this! I'll look into this now.
kdheepak
commented
2 years ago I know what is going on but I don't know how to solve it. Basically, pandoc runs the filter for all RawInline elements before it runs the filter for RawBlock elements. And this means that if you have:
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->It gets parsed as
{
"pandoc-api-version": [
1,
22,
1
],
"meta": {
"project": {
"t": "MetaString",
"c": "panvimdoc"
}
},
"blocks": [
{
"t": "RawBlock",
"c": [
"html",
"<!-- panvimdoc-ignore-start -->"
]
},
{
"t": "Para",
"c": [
{
"t": "Image",
"c": [
[
"",
[],
[]
],
[
{
"t": "Str",
"c": "screenshot"
}
],
[
"./github/toggleterm.gif",
"Toggleterm in action"
]
]
},
{
"t": "SoftBreak"
},
{
"t": "RawInline",
"c": [
"html",
"<!-- panvimdoc-ignore-end -->"
]
}
]
}
]
}And because it parses the "start" and "end" as different types, the checks for all the "end" tags happen before the "start" tag. And the "start" tag is the one that sets a variable that tracks whether the filter should ignore the content, and because the "end"s have already been processed, it is never unset.
The easiest solution to this is to add new lines before and after all "comment" tags, e.g.:
<!-- panvimdoc-ignore-start -->
<h1 align="center">
toggleterm.nvim
</h1>
<!-- panvimdoc-ignore-end -->
<!-- panvimdoc-ignore-start -->
<p align="center">A <i>neovim</i> plugin to persist and toggle multiple terminals during an editing session</p>
<!-- panvimdoc-ignore-end -->
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
### Multiple orientations
- **Float**
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
- **Vertical**
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
- **Tab**
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
### Send commands to different terminals
<!-- panvimdoc-ignore-start -->
<!-- panvimdoc-ignore-end -->
This gets rendered as the following:
*panvimdoc.txt* For NVIM v0.7.0 Last change: 2022 January 26
==============================================================================
Table of Contents *panvimdoc-table-of-contents*
MULTIPLE ORIENTATIONS ~
- **Float**
- **Vertical**
- **Tab**
SEND COMMANDS TO DIFFERENT TERMINALS ~
Generated by panvimdoc <https://github.com/kdheepak/panvimdoc>
vim:tw=78:ts=8:noet:ft=help:norl:I added a section to the specification documentation: https://github.com/kdheepak/panvimdoc/blob/c50c5d5b898dbf1f850f0144f5b7f863bb4f1d8c/doc/panvimdoc.md?plain=1#L319-L335.
Short of using a different markdown parser, this appears to be the only way to resolve this issue.
akinsho
commented
2 years ago Awesome, that works, thanks 👍🏿
Hi @kdheepak,
Thanks for your work on this project, it's super helpful 🙏🏿. I just added it to my https://github.com/akinsho/toggleterm.nvim plugin, but I'm seeing that when I use the ignore comments to try to omit the HTML at the start of the README the rest of the help doc is not converted correctly ie.
When adjusting the README to the above, this does not work as intended and the rest of the README is omitted during the conversion