write-good: Naive linter for English prose for developers who can't write good and wanna learn to do other stuff good too.
This repository contains a Vale-compatible implementation of the guidelines enforced by the write-good (LICENSE) linter.
To get started, add the package to your configuration file (as shown below) and then run vale sync.
StylesPath = styles
MinAlertLevel = suggestion
Packages = write-good
[*]
BasedOnStyles = write-goodThe rules are generated from the checks that write-good 1.0.8 ships, one
rule per check, with the same word lists and the same patterns.
| Rule | write-good check |
|---|---|
Weasel |
weasel |
Illusions |
illusion |
So |
so |
ThereIs |
thereIs |
Passive |
passive |
Adverbs |
adverb |
TooWordy |
tooWordy |
Cliches |
cliches |
E-Prime |
eprime |
write-good leaves eprime off unless asked. To do the same:
write-good.E-Prime = NOOn plain text the style reports what write-good reports, at the same line
and column, except:
write-goodreads every file as plain text. In Markdown and the other formats Vale understands, the style skips code, URLs, and front matter, and reads through inline markup such as**bold**.write-goodfolds two findings on the same words into one line. Vale reports one per rule.- When more than one space leads a sentence,
write-goodplacesSoandThereIsone character into the gap. Vale reports where the word is.
$ cd script
$ npm install
$ npm run generate # rewrite write-good/*.yml and fixtures/phrases.md
$ cd ..
$ ./test.sh # compare with the golden file, then with write-good itselfTo move to a newer write-good, change its version in script/package.json
and run the same steps. script/compare.mjs takes any files and prints each
finding the two tools disagree on.