Skip to content

unicode normalization #7

@pwinckles

Description

@pwinckles

I think rocfl needs to unicode normalize object id and path comparisons by default because it is difficult/impossible to control the normalization from the cli. If there is a collision, warn and attempt to resolve base on the un-normalized form.

It is unclear if writes should also be normalized. File names are typically based on their on disk representation. My inclination is not to change this. The main problem with not normalizing is that it could result in files/objects that only differ by their normal form.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions