-
anarcat authored
This is pretty good, but really requires an extensive dictionary. Right now, we use a "vocabulary", but eventually we might want to make our own "style" with multiple plugins, in particular because this would allow us to split the dictionnaries between different things, like "usernames", "tech" and who knows. Right now it's all bundled up in the same file, but as it grows, it will become impractical. Right now, running vale over the entire wiki yields:
5742 errors, 0 warnings and 0 suggestions in 222 files. In comparison, codespell finds zero errors, which is suspicious, because there *are* errors, if only capitalisation inconsistencies, but also actual typos like `occured`. This is not yet in GitLab CI because i don't want to break the pipeline right off the bat. I also feel we're getting enough into linting that we should start looking at pre-commit so that those hooks are correctly run on the client side before push.anarcat authoredThis is pretty good, but really requires an extensive dictionary. Right now, we use a "vocabulary", but eventually we might want to make our own "style" with multiple plugins, in particular because this would allow us to split the dictionnaries between different things, like "usernames", "tech" and who knows. Right now it's all bundled up in the same file, but as it grows, it will become impractical. Right now, running vale over the entire wiki yields:
5742 errors, 0 warnings and 0 suggestions in 222 files. In comparison, codespell finds zero errors, which is suspicious, because there *are* errors, if only capitalisation inconsistencies, but also actual typos like `occured`. This is not yet in GitLab CI because i don't want to break the pipeline right off the bat. I also feel we're getting enough into linting that we should start looking at pre-commit so that those hooks are correctly run on the client side before push.