I'm getting an error with streams.
I'm working on adding istanbul to my existing mocha task. When I run this task I get the error below.
I'm using gulp-istanbul
(note: the config.test.src.bdd.features is set to the value 'test/bdd/features/**/*-spec.js')
var stream = gulp.src([config.test.src.bdd.features], { read: false });
gulp.task('mocha-bdd-features', function(cb) {
process.env.PORT = 8001;
return stream
.pipe(istanbul())
.pipe(istanbul.hookRequire())
.pipe(mocha({
compilers: {
js: babel
},
reporter: config.test.mocha.reporter,
ui: 'bdd'
}))
.on('finish', function () {
stream.pipe(istanbul.writeReports())
stream.pipe(istanbul.enforceThresholds({thresholds: {global: 90}}))
stream.on('end', cb);
});
});
the error I get is:
events.js:85
throw er; // Unhandled 'error' event
^
Error: streams not supported
and who knows I may not be setting up this task right when trying to incorporate gulp-istanbul but trying to at least get past this error first.
I was facing the exact same issue.
I believe the problem is in line:
var stream = gulp.src([config.test.src.bdd.features], { read: false });
Setting the read option to false causes the file.contents to be null and therefore istanbul is not able to cover the files. (check here)
so try the same thing but without the read option.
var stream = gulp.src([config.test.src.bdd.features]);
Hope this helps.
Related
I am trying to stream sass into browsersync without gulp. Here is my code
function streamSass() {
dev.sassTasks.forEach(task => {
const sass = new Sass(task.src, task.dest, task.opts)
sass.write().then(() => {
const stream = fs.createReadStream(sass.dest.segments.absolutePath)
stream.pipe(browserSync.stream())
})
})
}
The Sass class just process and creates css file. When the function executed nothing happens on the side of the browsersync. Why is that? Assume that browsersync is running and is active when function is executed.
If you are not seeing anything happen then I would wager that there is something wrong with your stream as it will not throw an error and without an event handler it will silently fail.
Try adding an error event handler:
stream.on('error', err => console.log(err));
I need help migrating my parse server and was directed here by the parse support page.
I have finished the database migration to an mLab database and have the parse server example running in heroku.
My project runs perfectly at retrieving my current data with the default parse server example main.js file, but I would like to still run ParseServer jobs from the client like in this iOS example so I would like to use my own main.js file.
[PFCloud callFunction:#"publishCard" withParameters:#{
#"cardID" : card.cardPF.objectId,
} error:&error];
When I try to run my project with the following file as main.js (I've renamed it newmainjs just for visibility), it doesn't let me login with parse at all anymore from my iOS Client--so it seems to be triggering errors and I have no idea how to debug it as it's not covered in the migration tutorial.
https://github.com/KetchupMoose/cardgameserver/blob/master/cloud/newmain.js
I am very amateur at backend/node so I would really appreciate some support, as I relied on Parse for a lot of things before.
take a look at: https://parse.com/migration
section 3 "cloud code".
You need to change any use of:
Parse.Cloud.useMasterKey()
which you use:
here:
https://github.com/KetchupMoose/cardgameserver/blob/master/cloud/newmain.js#L162
and here: https://github.com/KetchupMoose/cardgameserver/blob/master/cloud/newmain.js#L633
instead of using useMasterKey(), you need to pass 'useMasterKey: true' to queries and saves.
Here are a few examples from your code:
userQuery.find({
useMasterKey: true, // <-- note the addition here
success: function(results) {
console.log(results.count);
....
And:
}).then(function(saveObjects)
{
Parse.Object.saveAll(updatedUserObjects, {
useMasterKey: true, // <--- here
success: function(list) {
//assumes all are saved
response.success("user EloRatings Saved Successfully");
},
.....
And:
Parse.Cloud.define("giveSellerGold", function(request, response) {
Parse.Cloud.useMasterKey();
var user = new Parse.User();
var query = new Parse.Query(Parse.User);
query.equalTo("objectId", request.params.sellerID);
query.first({
useMasterKey: true, // <--- here
success: function(object) {
object.increment("gold", request.params.sellerGold);
object.save(null, { useMasterKey: true }); // <-- note how save is done.
response.success("Successfully saved gold");
},
error: function(error) {
response.error("update failed");
}
});
});
so just make sure that all the queries (that need it) are passing 'useMasterKey' in the options and that ought do it! Good luck.
I want to download a zip file and extract it with nodejs. This is what I have done so far:
var fs = require('fs');
var wget = require('wget-improved');
var filesizeHumanReadable = require('filesize');
var unzip = require('unzip');
var downloadCSS = function() {
var src = 'http://7-zip.org/a/7za920.zip';
var output = '/tmp/7z.zip';
var options = {};
var download = wget.download(src, output, options);
download.on('error', function(err) {
console.log(err);
});
download.on('start', function(fileSize) {
console.log(filesizeHumanReadable(fileSize));
});
download.on('end', function(outputMessage) {
console.log(outputMessage);
console.log(output);
fs.createReadStream(output).pipe(unzip.Extract({ path: '/tmp/' }));
});
download.on('progress', function(progress) {
// code to show progress bar
});
}
The error message I get when running it:
mles-MacBook-Pro:test-api mles$ node index.js
375.83 KB
Finished writing to disk
/tmp/7z.zip
events.js:85
throw er; // Unhandled 'error' event
^
Error: EPERM, unlink '/tmp'
at Error (native)
Now I'm a bit baffled how to handle the error event and what my actual error is?
Does the process have enough permission to write to /tmp? Does /tmp already have some files?
Because unlink is a node.js function to delete directories. Apparently, unzip.Extract calls it. So, unlink fails if the folder isn't empty (in your case /tmp).
Setting the unzip location to a specific directory fixes it
fs.createReadStream(output).pipe(unzip.Extract({ path: '/tmp/7zip' }));
I'm marking mostruash answer as correct since he brought me on the right track.
I'm having a very weird EPIPE write error when I'm trying to generate a PDF from HTML with this module:
https://www.npmjs.com/package/html-pdf
The exact error:
events.js:72
throw er; // Unhandled 'error' event
^
Error: write EPIPE
at errnoException (net.js:904:11)
at Object.afterWrite (net.js:720:19)
The line where I call the PDF gen:
var pdfgen = require("html-pdf");
pdfgen.create(html, options).toFile('whappbook.pdf', function(err, res) {
var errorCount = 0;
if (err) {
console.log(err);
errorCount++;
if(errorCount > 2) {
return console.log(err);
} else {
return create();
}
}
console.log(res); // { filename: '/app/businesscard.pdf' }
});
I've tried just using <h1>Testing</h1> as the HTML to see if the file was too big, but that wasn't the case.
Note: local everything works perfectly fine, but when I run this code on Appfog it doesn't work, so I suspect it has something to do with missing dependencies or file writing rights.
I'm writing directly to the home directory (in Appfog's case thats /app).
HTML-PDF is using PhantomJS and I've also directly npm'd that in my app because some people were reporting issues when it wasn't directly installed, but that didn't solve my problem either.
If there is any information that I can additionally provide please let me know!
watch stops when error messages occur.
stream.js:94
throw er; // Unhandled stream error in pipe.
^
source string:51: error: invalid property name
How I can keep watch running and just to tell me where is the error located.
grunt could deal with errors and doesn't need to stop,
styleSheet.scss:41: error: invalid property name
otherwise, I need to keep typing "gulp" in the command-line when an error occurs.
This answer has been appended to reflect recent changes to Gulp. I've retained the original response, for relevance to the OPs question. If you are using Gulp 2.x, skip to the second section
Original response, Gulp 1.x
You may change this default behavior by passing errLogToConsole: true as an option to the sass() method.
Your task might look something like this, right now:
gulp.task('sass', function () {
gulp.src('./*.scss')
.pipe(sass())
.pipe(gulp.dest('./'));
});
Change the .pipe(sass()) line to include the errLogToConsole: true option:
.pipe(sass({errLogToConsole: true}))
This is what the task, with error logging, should look like:
gulp.task('sass', function () {
gulp.src('./*.scss')
.pipe(sass({errLogToConsole: true}))
.pipe(gulp.dest('./'));
});
Errors output will now be inline, like so:
[gulp] [gulp-sass] source string:1: error: invalid top-level expression
You can read more about gulp-sass options and configuration, on nmpjs.org
Gulp 2.x
In Gulp 2.x errLogToConsole may no longer be used. Fortunately, gulp-sass has a method for handling errors. Use on('error', sass.logError):
gulp.task('sass', function () {
gulp.src('./sass/**/*.scss')
.pipe(sass().on('error', sass.logError))
.pipe(gulp.dest('./css'));
});
If you need more fine-grained control, feel free to provide a callback function:
gulp.task('sass', function () {
gulp.src('./sass/**/*.scss')
.pipe(sass()
.on('error', function (err) {
sass.logError(err);
this.emit('end');
})
)
.pipe(gulp.dest('./css'));
});
This is a good thread to read if you need more information on process-control: https://github.com/gulpjs/gulp/issues/259#issuecomment-55098512
Actually above anwsers doesn't work for me (Im using gulp-sass 3.XX). What really worked:
gulp.task('sass', function () {
return gulp.src(config.scssPath + '/styles.scss')
.pipe(sourcemaps.init())
.pipe(sass({ outputStyle: 'compressed' })
.on('error', sass.logError)
)
.pipe(sourcemaps.write('./'))
.pipe(gulp.dest(config.cssPath))
});
In gulp-sass 3.x.x when I was using "sass.logError(err);" I constantly recive error that "this.emit('end'); is not a function". Now when I'm using:
.pipe(sass({ outputStyle: 'compressed' })
.on('error', sass.logError)
)
everything is working like a charm
In gulp "^2.0.0" the option errLogToConsole will no longer work. Instead gulp-sass has a built in error logging callback that uses gulp-util under the hood. Also, because gulp has some problems with killing the process on errors, if you are using with watch you will have to call this.emit('end')
https://github.com/gulpjs/gulp/issues/259#issuecomment-55098512
var sass = require('gulp-sass');
//dev
sass(config.sassDev)
.on('error', function(err) {
sass.logError(err);
this.emit('end'); //continue the process in dev
})
)
//prod
sass(config.sassProd).on('error', sass.logError)
A heads up for Gulp 3 users:
I liked #dtotheftp solution above, regarding gulp 2.x. Interestingly, it doesn't work unter Gulp3, at least not under #3.9.1:
.on('error', function(err){
sass.logError(err);
this.emit('end'); //continue the process in dev
})
gets me
TypeError: this.emit is not a function
at Function.logError (/depot/myproject/node_modules/gulp-sass/index.js:181:8)
Note, that the complaint is not coming from his this.emit() in the gulpfile but rather from the sass node-module, hence from the prior line.
This works for me:
.on('error', function(err){
gutil.log(err);
this.emit('end');
})
I do get all errors², and the watch never ends ;) (I am also using plumber() right after gulp.src(), which might help with that).
(Yes, the fix might be highly illogical, since sass.logError is said to be based on gutil...)
—
²also on undefined macros which went silent before on my setup for whatever reason.