mirror of
https://github.com/conventional-commits/conventionalcommits.org.git
synced 2026-03-14 12:55:12 -05:00
accepted characters for type and scope #87
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @hdevalke on GitHub (Nov 27, 2019).
This is the definition of type and scope:
It is unclear to me what are the allowed characters of a noun. Is it limited to ascii word characters, can it have numbers, underscores, dashes?
@tunnckoCore commented on GitHub (Nov 28, 2019):
Good point. I guess everything is allowed. It can be seen that some add the package name in the scope like
fix(@scope/pkg-name): foo bar, so..In javascript land, it's probably easy to assume that what's supported as
pkg.namethat's supported as scope.But yea, good question, I'm wondering almost every time, haha.
@bcoe commented on GitHub (Nov 29, 2019):
@hdevalke @tunnckoCore what if we use the term
identifierrather thannoun, and then explicitly define this?@hdevalke commented on GitHub (Dec 1, 2019):
@bcoe That would be better indeed. Maybe a formal grammar of conventional commits could be helpful as well, see semver.org.
@mainrs commented on GitHub (Jul 27, 2020):
If we're already there: What about any valid unicode character that is not ')'? So for example, Japanese, Korean etc. would work as well. Numbers are included, emoji. It just has to terminate with a
)and does not contain newline characters 🤔@damianopetrungaro commented on GitHub (Jul 27, 2020):
Good point @SirWindfield
@mainrs commented on GitHub (Jul 27, 2020):
The same could be applied to the type, however, that does bring up the question: What about
fixandfeat? Should they be translated into other languages as well? For example, Japanese does normally not use our alphabet letters. How would those languages get handled?Or would
fix,featbe "locked in" into English as they are right now? :)@tunnckoCore commented on GitHub (Jul 28, 2020):
@SirWindfield yea, I think that they definitely should be locked. I see CC as conventional standard, not a framework. The difference is that in "conventional standard", there're some locked things and terms, and rules, a convention. If you look at it as "framework" it just draws some frame and suggestions and allows interpretation (in this case allowing type/scope/desc to be in a different language). It's... thin difference and I may not be able to explain it well but it's there.
The beauty in the software industry is that English is THE language and everyone can understand it, or at least should at a basic level, not to mention the Google Translate which is quite good these days. And... in this global world, almost everyone knows at least 2 or 3 languages.
Another strong point against such transliteration is that if we allow it, it will be harder for other non-japanese (or Korean, or... other) people to understand what's happening. If we don't, at least, we will know is it a fix or what.
@mainrs commented on GitHub (Jul 28, 2020):
Good points @tunnckoCore!
I do agree on the lock, I just wanted to gather other opinions :) That being said, is there some RFC process involved for making changes to the specification (or maybe something like "extensions"). The "unicode" extension for example 🤷 :D
@tunnckoCore commented on GitHub (Jul 28, 2020):
I think there's no RFC process, but we are commenting/voting/sharing our thoughts on the issues and PRs, plus the other members have a higher view on the community users even outside the JavaScript.
@vlad0337187 commented on GitHub (Sep 23, 2020):
Hello.
Are points allowed as scope identifiers right now ?
like:
feat(users.seeds): add adminsCan word
featurebe used instead offeat?(according to clause 14 of specification, types other than feat and fix MAY be used in your commit messages, e.g., docs: updated ref docs. Also I see issue about feature, but I'd like to know, is
featureallowed right now, or current standard discourages such word)Thanks.
@damianopetrungaro commented on GitHub (Sep 24, 2020):
Il mer 23 set 2020, 21:42 Vladislav notifications@github.com ha scritto:
@bcoe commented on GitHub (Dec 24, 2020):
👋 we've been pulling together a formal grammar for Conventional Commits:
https://github.com/conventional-commits/parser#the-grammar
This grammar defines tokens explicitly, as unicode characters with some exceptional tokens depending on context (e.g.,
:and(in the header).Let's move conversation to the parser repo.