support releasing a custom version, including pre-releases #129

e-cloud · 2016-10-15T07:26:25Z

major changes

add a option --release-as to release a custom version like npm version
add a option --prerelease to name a pre-release or tag a release as prerelease without option value, it can be used with --release-as

this PR is similar to #83 , but this provide more version-control.

PS: some of the test code are borrowed from tests for #83

PPS: I'm bad at naming the option. Feel free to suggest new names for those options.

coveralls · 2016-10-15T07:45:22Z

Coverage remained the same at 100.0% when pulling f793bc5 on e-cloud:master into 8e4ec4a on conventional-changelog:master.

Tapppi

So this is related to #83 (Preid PR) #84 (releaseAs issue) and #122 (Preid Issue).

Thanks for contributing. A few concerns I have:

tag-id should IMO be the --prerelease [-p] that was discussed in Feature/add preid param for semver #122. This option should convert the bump to a pre(major|minor|patch) as in Feature/add preid param for semver #83. Otherwise you are forced to manually find out what version you want to prerelease and use it with the manual|release-as option, kind of going against the idea of standard-version. To set the prerelease tag you would -p alpha.
manual|release-as would only need to accept major|minor|patch as input, since the -p flag should set it as a prerelease when no manual|release-as is given and otherwise -p would set it as premajor|preminor|prepatch .

What this would entail is:

Either getting the recommended bump from conventional-recommended-bump or using the (major|minor|patch) from (manual|release-as) flag.
Prefixing the bump with pre if prerelase flag is used, and supplying either the given or default prerelease tag to semver.inc.

If the above doesn't resolve prereleases with breaking changes (e.g. patch prerelease into (major|minor) prerelease) correctly, it'll need a modified version of this function in #83.

/CC @bcoe @stevemao @nexdrew

Tapppi · 2016-10-16T11:43:01Z

.editorconfig

@@ -3,7 +3,25 @@
 root = true

 [*]


Could you create a separate PR for modifying the .editorconfig? Thanks 😄

Tapppi · 2016-10-16T11:45:37Z

cli.js

@@ -4,6 +4,21 @@ var defaults = require('./defaults')

 var argv = require('yargs')
  .usage('Usage: $0 [options]')
+  .option('manual', {


release-as [-r] would be a good name IMO (@stevemao opinion?).

Tapppi · 2016-10-16T11:46:10Z

cli.js

+    choices: ['major', 'premajor', 'minor', 'preminor', 'patch', 'prepatch', 'prerelease'],
+    global: true
+  })
+  .option('tag-id', {


prerelease would be my preference, see the summary comment on why.

Tapppi · 2016-10-16T11:47:53Z

test.js

+          version[type] += 1
+
+          var content = fs.readFileSync('CHANGELOG.md', 'utf-8')
+          content.should.match(new RegExp(version.major + '.' + version.minor + '.' + version.patch))


The dots in the regex should be escaped. E.g. version.major + '\.' + version.minor + '\.' + version.patch.

Tapppi · 2016-10-16T11:48:02Z

test.js

+          version[type.slice(3)] += 1
+
+          var content = fs.readFileSync('CHANGELOG.md', 'utf-8')
+          content.should.match(new RegExp(version.major + '.' + version.minor + '.' + version.patch + '-' + type + '.0'))


The dots in the regex should be escaped. E.g. version.major + '\.' + version.minor + '\.' + version.patch.

Tapppi · 2016-10-16T11:50:13Z

test.js

+      })
+    })
+
+    it('creates a prerelease with a new minor version after two prerelease patches', function () {


This works correctly when specifying the preminor manually, but how about prerelease? Would that bump the patch version? That would be very counterintuitive. This might not be worth fixing before deciding on the points in my summary though.

e-cloud · 2016-10-17T14:47:25Z

@Tapppi all your comment can be worked out except this:

Prefixing the bump with pre if prerelase flag is used, and supplying either the given or default prerelease tag to semver.inc.

That involves some problems:

yargs can specify a default value for an option, but can't not set as undefined when not using that option in cli. Then we aren't able to see if user specify a prerelease flag with simple way
when someone wants no tag id, with default tag-id, it would also be hard to implement the requirement

so, the default value for prerelease is undefined and when a use specify it with option value, its value would be ''(empty string). If a user need a custom tag id, he/she should specify it explicitly for now.

coveralls · 2016-10-17T15:01:42Z

Coverage remained the same at 100.0% when pulling e3a1892 on e-cloud:master into 8e4ec4a on conventional-changelog:master.

stevemao · 2016-10-17T23:15:50Z

package.json

@@ -2,12 +2,12 @@
  "name": "standard-version",
  "version": "3.0.0",
  "description": "replacement for `npm version` with automatic CHANGELOG generation",
-  "bin": "cli.js",
+  "bin": "bin/standard-version.js",


the naming needs to be changed. cli should be the bin. So you'd need to rename the files to whatever you think makes more sense :)

cli by convention should be the bin

got it. what if i rename bin/standard-version.js to bin/cli.js and cli.js to command.js?

coveralls · 2016-10-26T02:27:59Z

Coverage remained the same at 100.0% when pulling a111a4b on e-cloud:master into 8e4ec4a on conventional-changelog:master.

e-cloud · 2016-11-08T06:40:47Z

@stevemao do you have time to review this again?

stevemao · 2016-11-13T04:24:45Z

test.js

+    describe('release-types', function () {
+      var regularTypes = ['major', 'minor', 'patch']
+
+      regularTypes.forEach(function (type) {


I wouldn't auto generate tests like this. I'd keep it DAMP.

But in this case we dont add a bunch of fixtures to avoid duplication. I think its readability is acceptable, just opinion. If you insist, i can expand it. I would be much longer.

I'd probably just test one for each assertion. EG: say test 'major' for "creates a major release" and 'patch' for "creates a pre-patch release". Not a big deal :)

stevemao · 2016-11-13T04:27:02Z

it looks awesome In general. Thanks, @e-cloud.
We also need to add docs before we can merge :)
I'd also like to hear other maintainers opinions :)

this is for better modularity, and, for more convenient test. before, you can't breakpoint-debug the source mostly when testing for the usage of shelljs. now, some of the test cases can be done potentially with this commit

Here uses promise for some of the test cases, works nicely, thanks bluebird.

coveralls · 2016-11-22T11:19:02Z

Coverage remained the same at 100.0% when pulling d784d46 on e-cloud:master into 80004ec on conventional-changelog:master.

e-cloud · 2016-11-25T02:17:30Z

I just add docs in Readme.md. Anybody could review this?

bcoe · 2016-11-26T21:09:58Z

@e-cloud really appreciate this work; sorry for the slow turn around on this repo lately, going to work on seeing this over the finish line 👍

coveralls · 2016-11-26T21:12:12Z

Coverage remained the same at 100.0% when pulling d784d46 on e-cloud:master into 80004ec on conventional-changelog:master.

bcoe · 2016-11-26T21:17:04Z

@stevemao @Tapppi sorry that I've been so AWOL the past couple months -- haven't been able to carve as much OSS time out as I would like to.

Please feel free to land any pull requests that meet your guidelines; I've added you both to the npm module so that I don't block you on publishing, and I very much trust your instincts with regards to features.

bcoe · 2016-11-26T22:31:50Z

@e-cloud @Tapppi give this a spin when you have a moment:

npm cache clear; npm i standard-version@next

I've landed these changes in a pre-release! how meta is that.

Tapppi · 2016-12-01T22:28:20Z

@bcoe @e-cloud sorry to bump this, but this is fricking awesome, I just released my first pre-release of an internal company package automagically from our CI server after versioning with standard-version@4.0.0-0 🎉 😍 ! All the changes look pretty good, I don't think there are any blocking issues for v4?!

bcoe · 2016-12-02T00:14:54Z

@Tapppi if you're comfortable signing off on everything, I say release it! my two cents:

I think your point about rolling back the operation if an error is a good thing for us to aspire to; but perhaps we can take this on as we start to abstract the checkpoint logic... maybe model after database migrations? have a rollback for each checkpoint.
I love the pre-release functionality, my only beef is changelog management gets slightly weird if you release several pre-releases in a row -- but I think this is something we can work on in the future too.

want to cut a release @Tapppi you should have access to the module on npm now too.

Tapppi · 2016-12-02T11:18:10Z

@bcoe I think both of those points are crucial but big enough to warrant thinking a bit more about them instead of trying to rush them in. I'll give releasing v4.0.0 a shot right now, thanks! 👍

e-cloud force-pushed the master branch from a70117f to f793bc5 Compare October 15, 2016 07:42

Tapppi reviewed Oct 16, 2016

View reviewed changes

e-cloud force-pushed the master branch from f793bc5 to d6bb3d4 Compare October 17, 2016 14:47

e-cloud force-pushed the master branch from d6bb3d4 to e3a1892 Compare October 17, 2016 14:58

stevemao requested changes Oct 17, 2016

View reviewed changes

e-cloud force-pushed the master branch 2 times, most recently from 06d413c to a111a4b Compare October 26, 2016 02:25

Tapppi mentioned this pull request Oct 26, 2016

support pre-patch #133

Closed

stevemao reviewed Nov 13, 2016

View reviewed changes

e-cloud added 7 commits November 22, 2016 18:31

feat: support specifying release type manually with identifier option

1e1c50c

test: full-test for manaul-release and tag-id

ea2e94f

refactor(cli): split out cli module and executable module

15ab90a

this is for better modularity, and, for more convenient test. before, you can't breakpoint-debug the source mostly when testing for the usage of shelljs. now, some of the test cases can be done potentially with this commit

test(pre): adjust tests for the new options and cli

2936aea

Here uses promise for some of the test cases, works nicely, thanks bluebird.

chore: add rule to ignore coverage folder

17ae0e4

refactor: rename modules name for convention

a1840dc

docs: add guidance for the new features

d784d46

e-cloud force-pushed the master branch from a111a4b to d784d46 Compare November 22, 2016 11:14

stevemao approved these changes Nov 25, 2016

View reviewed changes

bcoe mentioned this pull request Nov 26, 2016

Feature/add preid param for semver #83

Closed

bcoe closed this Nov 26, 2016

bcoe reopened this Nov 26, 2016

bcoe merged commit 068008d into conventional-changelog:master Nov 26, 2016

bcoe mentioned this pull request Nov 26, 2016

pre-release improvements #146

Open

Tapppi mentioned this pull request Dec 23, 2016

add a releaseAs option? #84

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

support releasing a custom version, including pre-releases #129

support releasing a custom version, including pre-releases #129

e-cloud commented Oct 15, 2016 •

edited

Loading

coveralls commented Oct 15, 2016 •

edited

Loading

Tapppi left a comment •

edited

Loading

Tapppi Oct 16, 2016

Tapppi Oct 16, 2016

Tapppi Oct 16, 2016

Tapppi Oct 16, 2016

Tapppi Oct 16, 2016

Tapppi Oct 16, 2016

e-cloud commented Oct 17, 2016

coveralls commented Oct 17, 2016 •

edited

Loading

stevemao Oct 17, 2016

e-cloud Oct 18, 2016

stevemao Oct 18, 2016

e-cloud Oct 18, 2016

coveralls commented Oct 26, 2016 •

edited

Loading

e-cloud commented Nov 8, 2016

stevemao Nov 13, 2016

e-cloud Nov 22, 2016 •

edited

Loading

stevemao Nov 25, 2016

stevemao commented Nov 13, 2016 •

edited

Loading

coveralls commented Nov 22, 2016 •

edited

Loading

e-cloud commented Nov 25, 2016

bcoe commented Nov 26, 2016

coveralls commented Nov 26, 2016 •

edited

Loading

bcoe commented Nov 26, 2016

bcoe commented Nov 26, 2016

Tapppi commented Dec 1, 2016

bcoe commented Dec 2, 2016

Tapppi commented Dec 2, 2016

support releasing a custom version, including pre-releases #129

support releasing a custom version, including pre-releases #129

Conversation

e-cloud commented Oct 15, 2016 • edited Loading

major changes

coveralls commented Oct 15, 2016 • edited Loading

Tapppi left a comment • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

e-cloud commented Oct 17, 2016

coveralls commented Oct 17, 2016 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

coveralls commented Oct 26, 2016 • edited Loading

e-cloud commented Nov 8, 2016

Choose a reason for hiding this comment

e-cloud Nov 22, 2016 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

stevemao commented Nov 13, 2016 • edited Loading

coveralls commented Nov 22, 2016 • edited Loading

e-cloud commented Nov 25, 2016

bcoe commented Nov 26, 2016

coveralls commented Nov 26, 2016 • edited Loading

bcoe commented Nov 26, 2016

bcoe commented Nov 26, 2016

Tapppi commented Dec 1, 2016

bcoe commented Dec 2, 2016

Tapppi commented Dec 2, 2016

e-cloud commented Oct 15, 2016 •

edited

Loading

coveralls commented Oct 15, 2016 •

edited

Loading

Tapppi left a comment •

edited

Loading

coveralls commented Oct 17, 2016 •

edited

Loading

coveralls commented Oct 26, 2016 •

edited

Loading

e-cloud Nov 22, 2016 •

edited

Loading

stevemao commented Nov 13, 2016 •

edited

Loading

coveralls commented Nov 22, 2016 •

edited

Loading

coveralls commented Nov 26, 2016 •

edited

Loading