This module is API-compatible with
node-phantom but doesn't rely on
WebSockets
/ socket.io
. In essence the communication between Node and
Phantom / Slimer has been simplified significantly. It has the following advantages
over node-phantom
:
- Fewer dependencies/layers.
- Doesn't use the unreliable and huge socket.io.
- Works under
cluster
(node-phantom does not, due to how it works)server.listen(0)
works in cluster. - Supports SlimerJS.
Your software should work without changes, but can show deprecation warning about outdated signatures. You need to update:
options.phantomPath
->options.path
- in
.create()
.evaluate()
&.waitForSelector()
-> movecallback
to last position of arguments list.
That's all!
npm install node-phantom-simple
# Also need phantomjs OR slimerjs:
npm install phantomjs
# OR
npm install slimerjs
Note. SlimerJS is not headless and requires a windowing environment.
Under Linux/FreeBSD/OSX xvfb can be used to run headlessly.. For example, if you wish
to run SlimerJS on Travis-CI, add those lines to your .travis.yml
config:
before_script:
- export DISPLAY=:99.0
- "sh -e /etc/init.d/xvfb start"
You should manualy install slimerjs
to run npm test
:
npm install slimerjs
It's excluded from devDeps, because slimerjs binary download is banned on Tvavice-CI network by authors.
You can use it exactly like node-phantom, and the entire API of PhantomJS should work, with the exception that every method call takes a callback (always as the last parameter), instead of returning values.
For example, this is an adaptation of a web scraping example:
var driver = require('node-phantom-simple');
driver.create({ path: require('phantomjs').path }, function (err, browser) {
return browser.createPage(function (err, page) {
return page.open("http://tilomitra.com/repository/screenscrape/ajax.html", function (err,status) {
console.log("opened site? ", status);
page.includeJs('http://ajax.googleapis.com/ajax/libs/jquery/1.7.2/jquery.min.js', function (err) {
// jQuery Loaded.
// Wait for a bit for AJAX content to load on the page. Here, we are waiting 5 seconds.
setTimeout(function () {
return page.evaluate(function () {
//Get what you want from the page using jQuery. A good way is to populate an object with all the jQuery commands that you need and then return the object.
var h2Arr = [],
pArr = [];
$('h2').each(function () { h2Arr.push($(this).html()); });
$('p').each(function () { pArr.push($(this).html()); });
return {
h2: h2Arr,
p: pArr
};
}, function (err,result) {
console.log(result);
browser.exit();
});
}, 5000);
});
});
});
});
options (not mandatory):
- path (String) - path to phantomjs/slimerjs, if not set - will search in $PATH
- parameters (Array) - CLI params for executed engine, [ { nave: value } ]. You can also pass in an array to use verbatim names and values.
- ignoreErrorPattern (RegExp) - a regular expression that can be used to
silence spurious warnings in console, generated by Qt and PhantomJS.
On Mavericks, you can use
/CoreText/
to suppress some common annoying font-related warnings.
For example
driver.create({ parameters: { 'ignore-ssl-errors': 'yes' } }, callback)
driver.create({ parameters: ['-jsconsole', '-P', 'myVal']} }, callback)
will start phantom as:
phantomjs --ignore-ssl-errors=yes
You can rely on globally installed engines, but we recommend to pass path explicit:
driver.create({ path: require('phantomjs').path }, callback)
// or for slimer
driver.create({ path: require('slimerjs').path }, callback)
You can also have a look at the test directory to see some examples of using the API, however the de-facto reference is the PhantomJS documentation. Just mentally substitute all return values for callbacks.
All of the WebPage
callbacks have been implemented including onCallback
,
and are set the same way as with the core phantomjs library:
page.onResourceReceived = function(response) {
console.log('Response (#' + response.id + ', stage "' + response.stage + '"): ' + JSON.stringify(response));
};
This includes the onPageCreated
callback which receives a new page
object.
Properties on the WebPage
and Phantom
objects are accessed via the get()
/set()
method calls:
page.get('content', function (err, html) {
console.log("Page HTML is: " + html);
});
page.set('zoomfactor', 0.25, function () {
page.render('capture.png');
});
// You can get/set nested values easy!
page.set('settings.userAgent', 'PhAnToSlImEr', callback);
Engines are buggy. Here are some cases you should know.
.evaluate
can return corrupted result:- SlimerJS: undefined -> null.
- PhantomJS:
- undefined -> null
- null -> '' (empty string)
- [ 1, undefined, 2 ] -> null
page.onConfirm()
handler can not return value due async driver nature. Use.setFn()
instead:page.setFn('onConfirm', function () { return true; })
.
Made by Matt Sergeant for Hubdoc Inc.