I came across this idea of better search in Obsidian notes: https://simplyexplained.com/blog/how-i-use-alfred-to-search-obsidian-notes/

This should work nicely, but:

• it would require installing an additional tool Alfred,
• would bind even more to MacOS ecosystem - what if I want to use the same workflow on Linux,
• would not be too VCS friendly - hard to version those UI wizards

So I’ve figured I can do better, with just CLI tools. That resulted in the following short script:

RG_PREFIX="rg --column --line-number --no-heading --color=always --smart-case "
INITIAL_QUERY=$1
export FZF_DEFAULT_COMMAND="$RG_PREFIX '$INITIAL_QUERY'"

selected_file=`fzf --bind "change:reload:$RG_PREFIX {q} || true" \
      --ansi --disabled --query "$INITIAL_QUERY" | awk -F : '{ print $1 }'`

set -x
encoded_url=`python3 -c "import urllib.parse; import sys; print(urllib.parse.quote(sys.argv[1]))" "$selected_file"`
open "obsidian://open?vault=obsidian-notes&file=$encoded_url"

Steps done here:

• make ripgrep preserve colors when piping its output to fzf
• prepare ripgrep output in a specific way, for fzf to digest
• encode file name, so that MacOS open is comfortable with

What would make this script more bullet-proof:

• check if tools are installed - fzf and ripgrep
• some parameter checking
• use some other version of open on non-Mac systems.