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-parservar 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-parserIf you run conventional-commits-parser without any arguments
$ conventional-commits-parserYou 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-parserAn 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