We have some (probably needlessly) complex gulp files and I can't find a tool that will show the dependencies in a some visual manner - even just a text output like npm does.
Eg given
gulp.task('something:else', ...
gulp.task('ts:compile', ['something:else'], function () {
gulp.task('copy:i18n', ...
gulp.task('test', ['ts:compile', 'copy:i18n'], ...
It will output something like
test
-> ts:compile
--> something:else
-> copy:i18n
or even better, some visual diagram/png/etc
I've found these:
https://www.npmjs.com/package/gulp-task-graph-visualizer - it presents a text representation like the output in the question.
Better was this one:
https://github.com/pavlov99/gulp-graph-tasks - it creates a visual graph of the task dependencies and is exactly what I was looking for.
Related
I have a typescript project in visual studio 2017 that has a gulpfile.js in it.
I Want to run my tasks when i click on run, build or clean buttons.
Is there anyway to do this?
I know i can do this by set tasks for before or after build, but i want to replace it, and add a task for run button.
A little bit later to answer this question, yet here is the summary of the answer combining comments on the original question and my experience so far
Visual Studio does contain Task runner, however, it does not recognize build in tasks defined in gulpfile.js.
In order to get this True integration with Visual Studio, we need to provide additional markup in the gulpfile.js by adding comment, that will populate task runner.
The one line we need is as follows:
/// <binding BeforeBuild='js,css' AfterBuild='move' Clean='cleanup' ProjectOpened='default' />
After adding the file your gulpgile.js should look like this:
/// <binding BeforeBuild='js,css' AfterBuild='move' Clean='cleanup' ProjectOpened='default' />
// Include gulp
var gulp = require('gulp');
If you add another line with additional definitions, it will not be recognized and only the top one is going to be used.
If you are a visual person you might want to to the Visual Studio. Nice guide for setting it up is populated here:
http://www.codedigest.com/quick-start/14/using-of-gulp-gulpfilejs-in-visual-studio-2017
UPDATED based on comments:
How to run this the task runner
After you can click on the task in question and it will execute.
if you have added the binding described above, it will be automatically bound to your bound actions.
IE:
for triggering compile, just build the solution.
I am new to gulp and i am getting object expected gulp error, but found solutions as rename file to gulfile.js
1) Need to understand every project has only one gulp file thats gulpfile.js? If i need to define more than one then how to and what will be the file name.
2)My requirement is to concatenate more than one less(convert to css) & js file into one and then apply it to index.html
3) I am using express to create gulp project structure. is this standard way? if no then how do i?
4) everytime i create gulp skeleton, do i need to install all packages again for every project?
Any references from where can learn gulp from basic.
I think one of the reasons you're getting downvoted is that on SO each question should be one question. This should really be four separate questions. Another reason is you haven't provided any of your code - add code (the { } icon) and include your gulpfile.js and your package.json.
1a) Yes, it has to be called gulpfile.js
1b) If you search SO and google for "multiple gulpfiles" you'll get a lot of solutions. If none of them work for you, let us know what you tried and what went wrong. But just so you know, it's better to start with just one gulpfile - it can be hard to get multiple gulpfiles working correctly, and using just one will help you learn gulp.
2) you'll need to use gulp-less and gulp-concat to turn multiple LESS files into one CSS file, and gulp-concat again to turn multiple js files into one.
3) You can use express, but you don't have to do. It depends what you're doing, and we have no idea what you're doing.
4) Not sure what you mean by "gulp skeleton". If you mean "Do I need to run npm install for every new project, yes you do.
5) Google "learn gulp"
6) If an image could just be text, it's better to just include the text.
If you need to, open new specific questions. For more on writing a great SO question, see https://stackoverflow.com/help/how-to-ask
I would like to use the ECMAScript 6 module system in a front-end project, so that the interdependencies of the code were more clear than simply loading "all that might be needed" up front, in the HTML.
However, having the following line in the main JavaScript file does not work:
import fuzLogin from 'fuzLogin'
The error in the browser's console is: can't find variable: require
The compiled code (created by Babel) is:
var _fuzLogin = require("fuzLogin");
var _fuzLogin2 = _interopRequireDefault(_fuzLogin);
function _interopRequireDefault(obj) { return obj && obj.__esModule ? obj : { default: obj }; }
Is ECMAScript 6 module system supposed to work, for compiled code, with WebStorm 10?
Should I maybe add some external dependency in my HTML, to provide the missing require?
Are there other ways I could reach a modular front-end orchestration of my JavaScript side?
I think that you're babel configuration is set up to use commonjs that transpiles with require (requirejs)... so, in order to work with that configuration you need to include requirejs: http://requirejs.org/
I found two ways that fulfil what I was looking for, in slightly different ways:
jspm
Rollup
JSPM allows on-the-fly loading of ES2015 modules, so that the transpiling happens in the browser. This is pretty awesome, really, and something I wasn't expecting.
In addition, JSPM also provides traditional build tools for doing the bundling for production.
But I actually chose to go with Rollup.
Rollup gathers all kinds of build systems together, and is based on ES2015 packaging, providing what I was after. Most important for me were the brilliant blog posts by Jason Lengstorf (just 1 and 2 weeks old, btw) that walk one through the whole practical setup.
References:
jspm-trial (GitHub) repo that I did, experimenting these things
Smaller, More Efficient JavaScript Bundling Using Rollup (blog, Aug 2016)
I'm trying to make my tasks run tests in a certain directory. I was looking at sourceSets, however I inferred that they are useful if you are running outside the test/groovy folder. All of my tests are within the test/groovy folder.
I've got a set of Geb tests as well as a set of service tests. I would like to run them both together and independently. Essentially my tree structure would look like this, being able to run all tests.
Test
--gebTest
----firefoxTest
----chromeTest
----ieTest
--servicesTest
----service1Test
----service2Test
----service3Test
----etc.
My file structure is as follows:
project
-src
--test
---groovy
----com
-----acme
------functional <---where my geb tests sit
------services <---umbrella for services
-------service1 <---each unique service
-------service2
-------service3
-------etc
Can anyone lend me a hand. For the life of me I don't know how Gradle picks what tests to execute.
Thank you in advanced.
SourceSets are indeed a solution to your problem, but I notice you only differentiate your tests by their package names. I'm not sure but that may prove problematic with source sets.
Personally I would prefer a directory structure like this anyway
src
-test
--groovy
---functional
----com etc
---services
----com etc
However, if you are attached to your current structure then take a look at Gradle's test filtering support, which will allow you to filter by package name.
http://www.gradle.org/docs/current/userguide/java_plugin.html#sec:java_test
I was trying to add some automated Unit Tests to my project with Bamboo and have been facing some problems. The Unit Tests themselves are done with googletest, which creates an XML file which should be compatible with the JUnit parser.
However, I'm getting the following error when executing Bamboo's JUnit Parser:
02-Apr-2013 12:11:22 Starting task ''Parse UnitTest output' of type 'com.atlassian.bamboo.plugins.testresultparser:task.testresultparser.junit'
02-Apr-2013 12:11:22 Parsing test results...
02-Apr-2013 12:11:22 Failing task since test cases were expected but none were found.
02-Apr-2013 12:11:22 Finished task 'Parse UnitTest output'
This doesn't seem to have anything to do with the .xml file itself, as I've tried a few. This included my own output.xml, generated by googletest and the sample outputs from https://confluence.atlassian.com/display/BAMBOO/JUnit+parsing+in+Bamboo.
I also adapted said files against the two proposed .xsd files, which should match the output that the JUnit Parser expects, but all to no effect.
Update:
Up until now I told the JUnit Parser to look for ${bamboo.build.working.directory}/output.xml
When I tried **/*.xml it worked.
As I understand it now, after very carefully reading the task description, I have to give it a folder. But I can also give it the files, if I do it in ant-style (with a glob?). This is at the very least very confusing and still doesn't fully answer the inital question. So if anyone could enlighten me, please do.
This is a super-old question, I figured I'd add an answer for posterity. As a few people have commented, the configuration value for test output files requires a relative path. The question is, relative to what?
I think the answer to that depends on how you have your source repositories configured, but in general it will be relative to the root of your project. If all else fails, look at where bamboo is putting your source code when it gets checked out; that'll be the directory to which bamboo appends the test output search path.
For the configuration syntax, you're correct that ant-style patterns can be used (Learning Ant path style for reference).
Just as an example, if you have a project which on your local machine lives at C:\git\MyProject, and your test results end up at C:\git\MyProject\Output\Tests\output.xml, then you'd specify Output/Tests/output.xml in the 'Specify custom results directories' field of the appropriate task configuration. You could also use Output/**/*.xml to search for all .xml files in the Output directory.