$ npm install @marionebl/conventional-commits-parser
Parse raw conventional commits
A minimum input should contain a raw message.
Each commit message consists of a merge header, a header (mandatory), a body and a footer. Mention (optional) someone using the @
notation.
<merge>
<header>
<body>
<footer>
The merge header may optionally have a special format that includes other parts, such as branch, issueId or source.
Merge branch <branch>
Merge pull request <issue-id> from <source>
The header may optionally have a special format that includes other parts, such as type, scope and subject. You could reference (optional) issues here.
<type>(<scope>): <subject>
The footer should contain any information about Important Notes (optional) and is also the place to reference (optional) issues.
<important note>
<references>
This module will only parse the message body. However, it is possible to include other fields such as hash, committer or date.
My commit message
-sideNotes-
It should warn the correct unfound file names.
Also it should continue if one file cannot be found.
Tests are added for these
Then sideNotes
will be It should warn the correct unfound file names.\nAlso it should continue if one file cannot be found.\nTests are added for these
. You can customize the fieldPattern
.
$ npm install --save conventional-commits-parser
var conventionalCommitsParser = require('conventional-commits-parser');
conventionalCommitsParser(options);
It returns a transform stream and expects an upstream that looks something like this:
'feat(scope): broadcast $destroy event on scope destruction\nCloses #1'
'feat(ng-list): Allow custom separator\nbla bla bla\n\nBREAKING CHANGE: some breaking change.\nThanks @stevemao\n'
Each chunk should be a commit. The downstream will look something like this:
{ type: 'feat',
scope: 'scope',
subject: 'broadcast $destroy event on scope destruction',
merge: null,
header: 'feat(scope): broadcast $destroy event on scope destruction',
body: null,
footer: 'Closes #1',
notes: [],
references:
[ { action: 'Closes',
owner: null,
repository: null,
issue: '1',
raw: '#1',
prefix: '#' } ],
mentions: [],
revert: null }
{ type: 'feat',
scope: 'ng-list',
subject: 'Allow custom separator',
merge: null,
header: 'feat(ng-list): Allow custom separator',
body: 'bla bla bla',
footer: 'BREAKING CHANGE: some breaking change.\nThanks @stevemao',
notes:
[ { title: 'BREAKING CHANGE',
text: 'some breaking change.\nThanks @stevemao' } ],
references: [],
mentions: [ 'stevemao' ],
revert: null }
Returns an transform stream. If there is any malformed commits it will be gracefully ignored (an empty data will be emitted so down stream can notice).
Type: object
Type: regex
or string
Default: null
Pattern to match merge headers. EG: branch merge, GitHub or GitLab like pull requests headers. When a merge header is parsed, the next line is used for conventionnal header parsing.
For example, if we have a commit
Merge pull request #1 from user/feature/feature-name
feat(scope): broadcast $destroy event on scope destruction
We can parse it with these options and the default headerPattern:
{
mergePattern: /^Merge pull request #(\d+) from (.*)$/,
mergeCorrespondence: ['id', 'source']
}
Type: array
of string
or string
Default: null
Used to define what capturing group of mergePattern
.
If it's a string
it will be converted to an array
separated by a comma.
Type: regex
or string
Default: /^(\w*)(?:\(([\w\$\.\-\* ]*)\))?\: (.*)$/
Used to match header pattern.
Type: array
of string
or string
Default ['type', 'scope', 'subject']
Used to define what capturing group of headerPattern
captures what header part. The order of the array should correspond to the order of headerPattern
's capturing group. If the part is not captured it is null
. If it's a string
it will be converted to an array
separated by a comma.
Type: array
of string
or string
Default:
[ 'close', 'closes', 'closed', 'fix', 'fixes', 'fixed', 'resolve', 'resolves', 'resolved' ]
Keywords to reference an issue. This value is case insensitive. If it's a string
it will be converted to an array
separated by a comma.
Set it to null
to reference an issue without any action.
Type: array
of string
or string
Default: ['#']
The prefixes of an issue. EG: In gh-123
gh-
is the prefix.
Type: array
of string
or string
Default: ['BREAKING CHANGE']
Keywords for important notes. This value is case insensitive. If it's a string
it will be converted to an array
separated by a comma.
Type: regex
or string
Default: /^-(.*?)-$/
Pattern to match other fields.
Type: regex
or string
Default: /^Revert\s"([\s\S]*)"\s*This reverts commit (\w*)\./
Pattern to match what this commit reverts.
Type: array
of string
or string
Default: ['header', 'hash']
Used to define what capturing group of revertPattern
captures what reverted commit fields. The order of the array should correspond to the order of revertPattern
's capturing group.
For example, if we had commit
Revert "throw an error if a callback is passed"
This reverts commit 9bb4d6c.
If configured correctly, the parsed result would be
{
revert: {
header: 'throw an error if a callback is passed',
hash: '9bb4d6c'
}
}
It implies that this commit reverts a commit with header 'throw an error if a callback is passed'
and hash '9bb4d6c'
.
If it's a string
it will be converted to an array
separated by a comma.
Type: function
or boolean
Default: function() {}
What warn function to use. For example, console.warn.bind(console)
or grunt.log.writeln
. By default, it's a noop. If it is true
, it will error if commit cannot be parsed (strict).
The sync version. Useful when parsing a single commit. Returns the result.
A single commit to be parsed.
Same as the options
of conventionalCommitsParser
.
You can use cli to practice writing commit messages or parse messages from files. Note: the sample output might be different. It's just for demonstration purposes.
$ npm install --global conventional-commits-parser
If you run conventional-commits-parser
without any arguments
$ conventional-commits-parser
You will enter an interactive shell. To show your parsed output enter "return" three times (or enter your specified separator).
> fix(title): a title is fixed
{"type":"fix","scope":"title","subject":"a title is fixed","header":"fix(title): a title is fixed","body":null,"footer":null,"notes":[],"references":[],"revert":null}
You can also use cli to parse messages from files.
If you have log.txt
feat(ngMessages): provide support for dynamic message resolution
Prior to this fix it was impossible to apply a binding to a the ngMessage directive to represent the name of the error.
BREAKING CHANGE: The `ngMessagesInclude` attribute is now its own directive and that must be placed as a **child** element within the element with the ngMessages directive.
Closes #10036
Closes #9338
And you run
$ conventional-commits-parser log.txt
# or
$ cat log.txt | conventional-commits-parser
An array of json will be printed to stdout.
[
{"type":"feat","scope":"ngMessages","subject":"provide support for dynamic message resolution","header":"feat(ngMessages): provide support for dynamic message resolution","body":"Prior to this fix it was impossible to apply a binding to a the ngMessage directive to represent the name of the error.","footer":"BREAKING CHANGE: The `ngMessagesInclude` attribute is now its own directive and that must be placed as a **child** element within the element with the ngMessages directive.\nCloses #10036\nCloses #9338","notes":[{"title":"BREAKING CHANGE","text":"The `ngMessagesInclude` attribute is now its own directive and that must be placed as a **child** element within the element with the ngMessages directive."}],"references":[{"action":"Closes","owner":null,"repository":null,"issue":"10036","raw":"#10036"},{"action":"Closes","owner":null,"repository":null,"issue":"9338","raw":"#9338"}],"revert":null}
]
Commits should be split by at least three newlines (\n\n\n
) or you can specify a separator as the second argument.
Eg: in log2.txt
docs(ngMessageExp): split ngMessage docs up to show its alias more clearly
===
fix($animate): applyStyles from options on leave
Closes #10068
And you run
$ conventional-commits-parser log2.txt '==='
[
{"type":"docs","scope":"ngMessageExp","subject":"split ngMessage docs up to show its alias more clearly","header":"docs(ngMessageExp): split ngMessage docs up to show its alias more clearly","body":null,"footer":null,"notes":[],"references":[],"revert":null}
,
{"type":"fix","scope":"$animate","subject":"applyStyles from options on leave","header":"fix($animate): applyStyles from options on leave","body":null,"footer":"Closes #10068","notes":[],"references":[{"action":"Closes","owner":null,"repository":null,"issue":"10068","raw":"#10068"}],"revert":null}
]
Will be printed out.
You can specify one or more files. The output array will be in order of the input file paths. If you specify more than one separator, the last one will be used.
MIT © Steve Mao
© 2010 - cnpmjs.org x YWFE | Home | YWFE