This is a fork of Andy's orginal script, with a quick fix to try to make it to work with Roam backups - I use the page names as titles, and also add processing of #tags and #[[tag words]]. It's quick and dirty, and can probably be improved.
This script reads in a folder of Markdown files, notes all the [[wiki-style links]] between them, then adds a special "backlinks" section which lists passages which reference a given file.
For example, this text might get added to Sample note.md
:
## Backlinks
* [[Something that links here]]
* The block of text in the referencing note which contains the link to [[Sample note]].
* Another block in that same note which links to [[Sample note]].
* [[A different note that links here]]
* This is a paragraph from another note which links to [[Sample note]].
The script is idempotent; on subsequent runs, it will update that backlinks section in-place.
The backlinks section will be initially inserted at the end of the file. If there happens to be a HTML-style <!-- -->
block at the end of your note, the backlinks will be inserted before that block.
- Links are formatted
[[like this]]
. - Note titles are inferred from the first line of each note, which is assumed to be formatted as a heading, i.e.
# Note title
. - All
.md
files are siblings; the script does not currently recursively traverse subtrees (though that would be a simple modification if you need it; seelib/readAllNotes.ts
) - The backlinks "section" is defined as the AST span between
## Backlinks
and the next heading tag (or<!-- -->
tag). Any text you might add to this section will be clobbered. Don't append text after the backlinks list without a heading in between! (I like to leave my backlinks list at the end of the file)
This is FYI-style open source. I'm sharing it for interested parties, but without any stewardship commitment. Assume that my default response to issues and pull requests will be to ignore or close them without comment. If you do something interesting with this, though, please let me know.
To install a published release, run:
yarn global add @hannesdelbeke/note-link-janitor
Then to run it (note that it will modify your .md
files in-place; you may want to make a backup!):
note-link-janitor path/to/folder/containing/md/files
That will run it once; you'll need to create a cron job or a launch daemon to run it regularly.
If you store your notes in a Git repository and would like to run note-link-janitor on every push, see this Github Actions Workflow from Rajesh Shashi Kumar.
It's built to run against Node >=12, so you may need to upgrade or swap your runtime version.
yarn install
yarn run build
In the future, I intend to expand this project to monitor for broken links, orphans, and other interesting hypertext-y predicates.