Cleaner file names

Despite Logseq is advertised as local-first, pages’ file names present several issues (non-standard alphanumeric characters, capitalisation…) that makes it really difficult to work with them outside logseq. The image below summarises most of the issues I have encountered so far:

  1. Using non alphanumeric characters: there’s plenty of literature on why this should be avoided (see references below, but in short: potential problem with certain OSes, file systems and syncronisation services)
    1. Using @ when inserting Zotero references
    2. Using %2Fwhen creating hierarchies
  2. Using whitespaces → difficult to work with command line
  3. Same capitalisation as page → lack of consistency, potential duplicates (only differing in capitalisation), file names may change easily, extra checks

Two possible workarounds could be:

  1. (Dendron inspired) Change default behaviour about automated filename creation so files follow the following criteria:
  • use - (or _ if - is used for hierarchies) instead of blank spaces
  1. Add the hability to customise the resulting filename at a note level, by creating a property (i.e. filename:: . If there’s such a property, filename should be created based on its content. If not, use the default behaviour.


Possibly related feature requuests:

+1 to this.

Having more easily readable and manageable files would make Logseq more coherent with the “local-first” approach. And it seems like it wouldn’t be hard to fix.

This PR, I believe addresses a lot of what’s mentioned in this.