Preserves em and strong markers in the AST #373
Conversation
| break; | ||
| } | ||
| case 'em': { | ||
| yield '_'; | ||
| yield n.attributes.marker ?? '*'; |
There was a problem hiding this comment.
Why are we changing _ to *?
There was a problem hiding this comment.
Is there any harm in leaving it as _ (as a default convention)?
There was a problem hiding this comment.
Yeah, so the issue is that _ is treated like regular text and not an emphasis delimiter when it's used inside of a text node, e.g. foo_bar_baz is rendered literally whereas foo*bar*baz treats the *bar* part as emphasized. Because the underscore behaves differently in this specific case, it's not safe to convert foo*bar*baz to foo_bar_baz—it changes the meaning of the content. So using the asterisk is a safer universal default than the underscore. It's a little unfortunate, I'd prefer the underscore otherwise.
| }; | ||
|
|
||
| export const em: Schema = { | ||
| render: 'em', | ||
| children: ['strong', 's', 'link', 'code', 'text', 'tag'], | ||
| attributes: { | ||
| marker: { type: String, render: false }, | ||
| }, | ||
| }; | ||
|
|
||
| export const s: Schema = { |
There was a problem hiding this comment.
Should we add it for Strikethrough (s) too?
There was a problem hiding this comment.
It looks like strikethrough always uses the exact same marker syntax, right?
There was a problem hiding this comment.
I wonder if we can remove some of the character escape code I wrote now that we have captured the marker. Can you take a look at that briefly @rpaul-stripe (not a blocker for merging)?
This PR addresses issue #369. It makes the parser capture the actual symbols used for
emandstrongnodes and preserve that information in the AST so that it can be used in the formatter to reproduce the content accurately.emandstrongnodes and apply it to a node attribute calledmarkermarkerattributemarkerattribute to thestrongandemnode schemasemandstrongnodes using the original emphasis marker from the source content and falls back on*or**when not availableCloses #369