Open h-mathias opened 2 months ago
If we did this, it would be config only. We intentionally limit what capabilities we provide on the CLI as configuration like this is intended to be project configuration rather than run configuration.
Could you go into more detail on what problem you are running into with short words that you don't just need a short-word filter but you need it more efficient than what extend-ignore-words-re
provides?
It is mainly about words with 2 or 3 characters which are often abbreviations or acronyms. Some false positives findings ba, fo, seh, mis, ue, nd
.
cspell for examples has minWordLength
.
It would be a convenience option but as said it also works with extend-ignore-words-re
so feel free to close this issue.
From cspell's docs
minWordLength - defaults to 4 - the minimum length of a word before it is checked.
Not quickly finding when that was introduced to see all of the motivation. Unsure what it is about the code bases I work on that this hasn't really been a problem which biases me towards leaving this to users via regex.
random personal feeling: i generally find longer words have much more signal to noise
rather than a threshold, i have considered making a user interface that consumes the output of typos-cli to sort it long to short, and even make a TUI to approve a fix or "add to ignorelist". just dreaming though:)
It can currently be solved with adding
^[a-zA-Z]{1,3}$
toextend-ignore-words-re
but it is less efficient than just comparing the length of the word.