JS: Are ignoreUrls regexps tested against param values too?

SDKs
1

I know the documentation reads “will ignore errors from whole urls matching a regex pattern”, but I wanted to clarify. I’m hoping to get the Raven client to ignore errors from urls that indicate a brower addon is present, typically something like some.domain.com?addonname=DNSUnlocker&someotherstuff. Will adding /DNSUnlocker/ to the ignoreUrls config value get the job done?

2

If those URLs appear in the stack trace file paths, then yes. (They should.)

Will adding /DNSUnlocker/ to the ignoreUrls config value get the job done?

It should.

3

@benvinegar Thanks for the reply. We’ve added regex-formatted strings to ignoreUrls, but we’re still getting errors that we expected to be ignored. To be concrete, we added this to our Raven.config ignoreUrls attribute:

'akamaihd.net/loaders/\\d+/l\\.js\\?(?:.+=.*&)*ext=(?:DNSUnlocker|SurfBuyer)'

But yesterday an error was reported in a script from https://cdncache-a.akamaihd.net/loaders/2484/l.js?pid=2484&ext=SurfBuyer&nocache=1&zoneid=88602545.

The url reported in the error is from our domain, which we can call https:my-website.com, but the first line of the stacktrace cites the script from cdncache-a.akamaihd.net. I’m attaching a screenshot of the issue page in case it helps.

loaders-error-redacted.png1113×832 112 KB

I confirmed that our regex string should match the url of the script by entering the following into the Chrome console:

>> url = 'https://cdncache-a.akamaihd.net/loaders/2484/l.js?pid=2484&ext=SurfBuyer&nocache=1&zoneid=88602545';
>> regexString = 'akamaihd.net/loaders/\\d+/l\\.js\\?(?:.+=.*&)*ext=(?:DNSUnlocker|SurfBuyer)';
>> new RegExp(regexString).test(url)
<- true

Am I misunderstanding how ignoreUrls works? Or is it possible this is a bug?

4

@benvinegar: Just to be clear, we are assigning an array to ignoreUrls. It looks like

ignoreUrls:   [
    'akamaihd.net/loaders/\\d+/l\\.js\\?(?:.+=.*&)*ext=(?:DNSUnlocker|SurfBuyer)',
    'other string'
]
5

OK OK. The problem is that although I thought I was providing strings to go directly to RegExp(), in fact there is an intermediate step in raven.js. The function joinRegExp transforms strings by escaping some characters: https://github.com/getsentry/raven-js/blob/master/src/raven.js#L1822

I’ve spent a little time trying to massage the pattern strings in our config, but after failing to find strings that will produce the desired RegExp, I’m going to replace them with RegExp literals.