You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I try to use a custom Markdown parser by setting parser = ./custom-markdown-parser in my .prezrc file.
However, I keep getting the following error:
/usr/lib/node_modules/prez/lib/convert.js:32
throw new Error("Error while loading parser '" + parser + "' (" + e1 + ")");
^
Error: Error while loading parser './custom-markdown-parser' (Error: Cannot find module './parsers/./custom-markdown-parser')
The problem seems to be that the script never looks for my custom parser in my local project dir instead of the global prez install path.
In convert.js:20, it seems to never enter the // No module: try to load user implementation conditional branch.
If I modify the line to omit the && e1.message.indexOf("'" + parser + "'") !== -1) condition, it works. (But I guess that condition has a reason to be there?)
The text was updated successfully, but these errors were encountered:
Honestly I feel quite uncomfortable with this condition ;) so I guess you nailed it and there is a bug lying in there.
In the first place it was here in case of missing optional dependency. For example, to use "marky" you have to install a prez's peer dependency "marky". But:
This is not well documented
This is hard to detect
I'm not sure how to handle those "default but not enabled parsers" properly, in the meantime your fix is OK and I'll push a version modifying this.
I try to use a custom Markdown parser by setting
parser = ./custom-markdown-parser
in my.prezrc
file.However, I keep getting the following error:
The problem seems to be that the script never looks for my custom parser in my local project dir instead of the global prez install path.
In convert.js:20, it seems to never enter the
// No module: try to load user implementation
conditional branch.If I modify the line to omit the
&& e1.message.indexOf("'" + parser + "'") !== -1)
condition, it works. (But I guess that condition has a reason to be there?)The text was updated successfully, but these errors were encountered: