Goal:
Build /src/assets/(theme)/*.scss by passing the parameter in the command line using flag --theme. I solved step one using gulp-utils
function sass() {
// return gulp.src('/src/assets/scss/**/*.scss')
return gulp.src(['src/assets/scss/' + (util.env.theme ? util.env.theme : 'app') + '.scss'])
.pipe($.sourcemaps.init())
.pipe($.sass({
includePaths: PATHS.sass
})
.on('error', $.sass.logError))
.pipe($.autoprefixer({
browsers: COMPATIBILITY
}))
// Comment in the pipe below to run UnCSS in production
//.pipe($.if(PRODUCTION, $.uncss(UNCSS_OPTIONS)))
.pipe($.if(PRODUCTION, $.cleanCss({ compatibility: 'ie9' })))
.pipe($.if(!PRODUCTION, $.sourcemaps.write()))
.pipe(gulp.dest(PATHS.dist + '/assets/css'))
.pipe(browser.reload({ stream: true }));
}
Remaining:
gulp build --production --theme folderName1, folderName2, folderName*
Passing multiple theme values --theme folderName1 folderName2 to the same sass() function/task.
This would send folderName to function sass()
function sass() {
// return gulp.src('/src/assets/scss/**/*.scss')
return gulp.src([
'src/assets/scss/folderName1/*.scss',
'src/assets/scss/folderName2/*.scss',
'src/assets/scss/folderName3/*.scss'
])
...
...
}
Would creating an argsList like this SitePoint article discusses work?
Please review your idea. This package looks like solve similar topic
https://www.npmjs.com/package/gulp-sass-themes
You need only group your themes in sub-folder.
From other side gulp.src accepts string and array so you can pass this information from command line.
https://github.com/gulpjs/gulp/blob/master/docs/API.md
One of my colleagues was able to help resolve this issue. Instead of using gulp.utils he was able to use yargs.argv.theme with slipt(,). This allows for a user to pass the --theme param and pass as many comma-separated items as desired.
function sass() {
if (yargs.argv.theme) {
let collection = yargs.argv.theme.split(','),
results = collection.map(item => `./src/assets/scss/${item}/*.scss`);
return gulp.src(results)
.pipe($.sourcemaps.init())
.pipe($.sass({
includePaths: PATHS.sass
})
.on('error', $.sass.logError))
.pipe($.autoprefixer({
browsers: COMPATIBILITY
}))
// Comment in the pipe below to run UnCSS in production
//.pipe($.if(PRODUCTION, $.uncss(UNCSS_OPTIONS)))
.pipe($.if(PRODUCTION, $.cleanCss({ compatibility: 'ie9' })))
.pipe($.if(!PRODUCTION, $.sourcemaps.write()))
.pipe(gulp.dest(PATHS.dist + '/assets/css'))
.pipe(browser.reload({ stream: true }));
}
}
Related
I'm using NeoVim with autocomplete using nvim-lspconfig and nvim-cmp. I would like to know if there's a way of filtering out text feeds from the autocompletion menu, so that they don't appear in the contextual menu:
In your setup you can exclude any kind if suggestions thanks to this merged PR.
What is happening is the function "entry_filter" is getting called whenever a suggestion for nvim_lsp is being made. in it we return false if the entry is of the kind "text".
local cmp = require "cmp"
cmp.setup {
...
sources = cmp.config.sources({
-- Dont suggest Text from nvm_lsp
{ name = "nvim_lsp",
entry_filter = function(entry, ctx)
return require("cmp").lsp.CompletionItemKind.Text ~= entry:get_kind()
end },
})
}
Check out nvim-cmp sources list and remove whatever source you don't want to use. Text is quite probably coming from buffer:
cmp.setup({
...
sources = cmp.config.sources({
{ name = 'buffer' }, -- <- remove
{ name = 'nvim_lsp' },
...
})
})
I am trying to replace dynamically "import" statements.
Here is an example that checks if the import ends with a Plus.
module.exports = function(babel) {
return {
visitor: {
ImportDeclaration: function(path, state) {
// import abc from "./logic/+"
if( ! path.node.source.value.endsWith("/+"))
return;
path.replaceWithSourceString('import all from "./logic/all"')
}
}
}
}
This gives an error of
SyntaxError: src/boom.js: Unexpected token (1:1) - make sure this is an expression.
> 1 | (import all from "./logic/all")
The problem is that replaceWithSourceString is wrapping the string in rounded braces.
If I change the replaceWithSourceString to
path.replaceWithSourceString('console.log("Hi")')
and this works.. ¯_(ツ)_/¯
Any and all help you be great
replaceWithSourceString should really be avoided, because it is just not a very good API, as you are seeing. The recommended approach for creating ASTs to insert into the script is to use template. Assuming this is for Babel 7.x, you can do
const importNode = babel.template.statement.ast`import all from "./logic/all"`;
path.replaceWith(importNode);
One of my test expects an error message text to be one of multiple values. Since getText() returns a promise I cannot use toContain() jasmine matcher. The following would not work since protractor (jasminewd under-the-hood) would not resolve a promise in the second part of the matcher, toContain() in this case:
expect(["Unknown Error", "Connection Error"]).toContain(page.errorMessage.getText());
Question: Is there a way to check if an element is in an array with jasmine+protractor where an element is a promise?
In other words, I'm looking for inverse of toContain() so that the expect() would implicitly resolve the promise passed in.
As a workaround, I can explicitly resolve the promise with then():
page.errorMessage.getText().then(function (text) {
expect(["Unknown Error", "Connection Error"]).toContain(text);
});
I'm not sure if this is the best option. I would also be okay with a solution based on third-parties like jasmine-matchers.
As an example, this kind of assertion exists in Python:
self.assertIn(1, [1, 2, 3, 4])
Looks like you need a custom matcher. Depending on the version of Jasmine you are using:
With Jasmine 1:
this.addMatchers({
toBeIn: function(expected) {
var possibilities = Array.isArray(expected) ? expected : [expected];
return possibilities.indexOf(this.actual) > -1;
}
});
With Jasmine 2:
this.addMatchers({
toBeIn: function(util, customEqualityTesters) {
return {
compare: function(actual, expected) {
var possibilities = Array.isArray(expected) ? expected : [expected];
var passed = possibilities.indexOf(actual) > -1;
return {
pass: passed,
message: 'Expected [' + possibilities.join(', ') + ']' + (passed ? ' not' : '') + ' to contain ' + actual
};
}
};
}
});
You'll have to execute this in the beforeEach section on each of your describe blocks it's going to be used in.
Your expect would look like:
expect(page.errorMessage.getText()).toBeIn(["Unknown Error", "Connection Error"]);
The alternative solution is to use .toMatch() matcher with Regular Expressions and specifically a special character | (called "or"), which allows to match only one entry to succeed:
expect(page.errorMessage.getText()).toMatch(/Unknown Error|Connection Error/);
To me, the work-around that you identified is the best solution. However, we should not forget that this is an asynchronous execution and you might want to consider Jasmine's asynchronous support.
Then, your test will look like the following one:
it('should check item is in array', function(done){ //Note the argument for callback
// do your stuff/prerequisites for the spec here
page.errorMessage.getText().then(function (text) {
expect(["Unknown Error", "Connection Error"]).toContain(text);
done(); // Spec is done!
});
});
Note: If you don't pass this done argument to the spec callback, it is going to run to completion without failures, but no assertions are going to be reported in the execution results for that spec (in other words, that spec will have 0 assertions) and it might lead to confusions.
I am using gulp and gulp-shell packages for a php Laravel application, I just need to know if this is possible to pass argument from cmd to gulpfile.js ? this is my file:
gulp.task('default', shell.task([
'echo user',
]));
Question:
Is it possible to pass an argument from command-line when running gulp and then inside the gulpfile print it out instead of user?
var command_line_args = require('yargs').argv
Not sure if this is of any use to you or others but I did this manually by passing the arguments explicitly through a custom function. It's not super elegant but it gets the job done.
var appendWithCommandLineArguments = function(cmd, arguments) {
var to_append = _.chain(command_line_args)
.pick(arguments)
.reduce(function(string, val, prop){
return string+"--"+prop+"="+val+" ";
}, " ")
.value();
return cmd + to_append
}
gulp.task('taskmailer', shell.task([
appendWithCommandLineArguments('node automate/build/mail/taskMailer.js', ["email", "template"])
]))
I'm using the config from my project but would like to run karma just for one specific test script one time. I don't want to have to create a whole new config file just for this case and would prefer just passing in the script I want run (so basically telling karma to use files: ['myTest.js'].
But there don't seem to be any options for that AFAICT in the docs. Why would this be missing? It seems like a fundamental feature IMO.
in karma.conf something like that:
function mergeFilesWithArgv(staticFiles) {
var source = staticFiles, argv = process.argv;
argv.forEach(function (arg) {
var index = arg.indexOf('--check=');
if (index !== -1) {
source.push(arg.substring(8));
}
});
return source;
}
config.set({
...
files: mergeFilesWithArgv([
'js_src/tests/*.test.js'
]),
...
});
use: karma start --check='./path/to/file.js'
or for multiple files: karma start --check='./path/to/file.js' --check='/another/path/to/another/file.js'