Until now we're not using t.Run to create a sub test for each fixture - for the
current bug I want to only run a single file as I'm a print debugger and don't
care for the noise of logs from other tests. While at it, it made sense to
merge the implementations.
org mode supports [1] setting the value attribute [2] of ordered list items to
change the numbering for the current and following items. Let's do the same. As
the attribute has no meaning for other types of lists [2] we'll just not
support it for those cases [3].
[1] https://orgmode.org/manual/Plain-Lists.html#Plain-Lists
[2] https://developer.mozilla.org/en-US/docs/Web/HTML/Element/li#attributes
[3]
Org mode seems to instead set the id attribute for e.g. unordered lists
starting with `[@\d+]\s` - but I don't really see the value in that and will
skip that for now.
The actual parameter parsing logic in emacs org mode is quite complex [1]. All
we want for now is to handle parameter values containing spaces [2]. Splitting
on ` :` gets us close enough for now. As I'm very opposed to copying 100 lines
of parameter parsing logic just to get exports right let's wait for use cases -
no hurt in gathering requirements as we go.
[1] https://github.com/bzg/org-mode/blob/master/lisp/ob-core.el#L1481
[2] I never ran into such parameters before and wrongly assumed that splitting
on spaces would be enough. boy was i wrong - just look at that massive
function [1]! that's why we can't have nice things!
This reverts commit bf7f957af2.
We still have to manually install it for now but we don't have to fuck around
with env vars anymore. github actions still defaults to go 1.15 [1]. Whatever.
[1]
```
ls -lisah $(which go)
108022 0 lrwxrwxrwx 1 root root 42 Mar 18 22:15 /usr/bin/go ->
/opt/hostedtoolcache/go/1.15.10/x64/bin/go
```
org mode allows rendering the toc anywhere in the html document using the `TOC`
keyword [1]. There's more options but `#+TOC: headlines $n` should be enough
for starters. Note that org mode still requires setting `#+OPTIONS: toc:nil` to
disable the default toc
[1] https://orgmode.org/manual/Table-of-Contents.html
Hugo defaults to serving files with pretty urls [1] - this means
`/posts/foo.org` is served at `/posts/foo/`. This works because servers
default to serving index.html when a directory is specified and hugo renders
the post to `/posts/foo/index.html` instead of `/posts/foo.html`. To make
relative links work we need to (1) remove the fake `foo/` subdirectory from
unrooted links and (2) replace any `.org` suffix with `/`.
[1] https://gohugo.io/content-management/urls/#pretty-urls
As headlines are always lvl (indent) 0 I thought it would be clever to abuse
the lvl field to store the headline lvl. Well, here we are - it wasn't clever.
List items only end when their parent ends or they run into something that's
not indented enough - everything else becomes part of the list item. Abusing
the token.lvl field (indent) for the headline lvl means headlines look indented
to the list item parsing logic - i.e. they become part of the list item if the
headline has a high enough lvl. That should never happen - so let's get rid of
the hack and (re-)calculate the headline lvl when we need it.
Turns out Org mode supports image links natively and we don't have to go out of
spec!
From https://orgmode.org/manual/Images-in-HTML-export.html:
[...] if the description part of the Org link is itself another link, such as
‘file:’ or ‘http:’ URL pointing to an image, the HTML export back-end in-lines
this image and links to [...]
html does not support table separator rows as Org mode does. Emacs org export
simulates rows as defined by separators by wrapping all the rows between 2
separators into a separate tbody. The html spec is fine with that [0] so we
follow.
[0] https://developer.mozilla.org/en-US/docs/Web/HTML/Element/tbody
All tags are put on a line by themselves to help with visual
diffing. Apparently this extra cosmetic whitespace causes problems inside p
tags for ppl who want to use `white-space: pre`. Not much hurt for visual
diffing in removing cosmetic whitespace for just p tags and can't think of
anything that would break because of this right now. So let's do it and wait
for things to break.
if we define a custom LINK for file we run into index problems bc it's trimmed
before already - this fixes that. Shouldn't ever happen but whatever, fuzzing
found it.
To support code block directives like :exports none we need context - i.e. we
need to have the block and it's results at once and can't just render them
independently.
inside src example blocks lines starting with `\s*,` are escaped - i.e. org
mode will not try to parse them as e.g. a headline. We don't want to render the
escape commata to html so let's take them out - and put them back in before
rendering to org. Doing it this way allows us to render them correctly even
when the input did not include them.
see https://orgmode.org/manual/Literal-Examples.html#Literal-Examples
While example blocks do not render inline markup and are thus parsed raw in
some way, their contents are not literal html and thus still need to be html
escaped.
delimiters are not always 2 in length - $ being the exception. Also we have to
make sure to handle both $ and $$ delimiters.
Not proud of this solution but can't think of anything simpler right now - will
come back to this later hopefully.
The org mode toc OPTION does not just support true/false - it also allows
specifying the max headline level [1] to be included in the toc.
[1] headline level as seen in org mode - not the html tag level
We want original whitespace to be rendered in some cases (e.g. verse
blocks). This requires information about the original whitespace to be
preserved during paragraph parsing. As html ignores (collapses) whitespace by
default we don't have to adapt the html writer and can just selectively enable
rendering of the preseverved whitespace wherever we want it using
css (white-space: pre).
To differentiate meaningful whitespace from document structure based
indentation (i.e. list item base indentation) we need to introduce
document.baseLvl. A paragraph by itself does not have enough information to
differentiate both kinds of whitespace and needs this information as context
[0].
As we're already touching list indentation i went along and improved (fixed?)
descriptive list item indentation rendering in the org writer (it should match
emacs tab behavior - i.e. indent subsequent lines up to the `:: `).
[0] e.g. list items can contain blank lines - a paragraph starting with a blank
line would not know that it is part of a list item / has a base indentation -
the blank line would suggest a baseLvl of 0.
WriteNodesAsString is simple enough to implement but exposing it is helpful in
the implementation of extending writers and we don't aim to keep writer a small
interface so let's expose it.
Extension of the org & html writers is made possible by creating circular
references between the extending and extended writer - that way the extending
writer can forward all methods it doesn't implement to the extended writer and the
extended writer can use the extending writer as the root for method calls to
make sure methods overridden in the extending writer are used even for nested
method calls.
This circular reference leads to problems when cloning writers - cloning the
extended writer merely copies the pointer to the extending writer - i.e. the
extending writer does not get cloned with an updated reference to the extended
writer. Thus method calls to the extending writer act as if no cloning took
place and things break.
The easiest solution is to just get rid of cloning. We could also clone the
ExtendingWriter and replace it's reference to the extended writer with the just
cloned one but that's harder so we just remove it.
As there are a lot of "extending writer" and "extended writer" in the above
paragraphs and I'm too lazy to write up something better here's another attempt
at a TLDR:
Cloning is broken as ExtendingWriter is a reference to a writer that has
a reference to the writer we are cloning - that writer would have to have it's
reference updated but that's hard. So we solve it it by not cloning at all.
