TMSU was born out of frustration with the hierarchical nature of filesystems - Home / Stackoverflow

see also

• I am endlessly fascinated with content tagging systems
• What’s a good solution for file-tagging in linux?
  • Modify ext4 file system to add custom attribute for files - There is no indexing of these values. You have to check all existing files - Extended attributes: the good, the not so good, the bad.
    • How I use the attr command with my Linux filesystem
    • Executing a bash script upon file creation

Principle

• Designing better file organization around tags, not hierarchies

Tsmu

• install using x86_64 binary release
  • fish autocompletion - [1]
  • zsh completion script

Works in user space, uses FUSE

Tagging

• support unicode tags (hopefully)

$ tmsu tag <file> --tags='score=★★★★☆'
$ tmsu tag --tags="iron-giant sci-fi" ./Pictures/iron-giant/*
  
# Mounting FUSE filesystem
tmsu mount mount-point

Tricks and Tips

• External Integration

  The versatility of TMSU means it isn’t very difficult to extract tags from text files (e.g. Markdown) and feed them into TMSU, for example the following could be used to extract #tags from markdown files in the current directory and assign tags to them.

Injecting my blog tags into tmsu, usig ripgrep and ruby one liner:

rg "tags:" | ruby -ne 'puts "tmsu tag \'#{$1}\' --tags=\'#{$2}\'" if $_ =~ /([^:]+):[^:]+:(.+)$/ ' | bash

• Tagging a Directory and its Contents

DEV

• [Proposal] Non-Sequential IDs in SQLite3 Database

Issue

FUSE error: Transport endpoint is not connected

• fusermount -u YOUR_MNT_DIR