Match port when testing legacy browser #1511
Conversation
|
Are you sure this is safe? The test is titled I'm not sure why the original author thought this was an important test... but assuming that they were right, wouldn't this be a code bug? |
|
well this just adds the port number if the browser is IE 11, assuming this is right behavior for IE, since this implementation inside of |
|
I'll merge it cause it's testing a deprecated method, and it fixes our CI :/ ...but in general, we should be normalizing browser differences at as low a level as possible. jQuery was pretty much built on the idea... and consistency is the main motivation for the push for web standards at all. If the implementation is browser-dependent, then I'd consider it a buggy implementation. |
…built * 'master' of https://github.com/prebid/Prebid.js: (83 commits) feat(strAdapt): check if tagJS is already present (prebid#1500) Updated karma-mocha, and simplified the test framework for runs which dont include --watch. (prebid#1520) Revert "drop specific code for index adapter (prebid#1487)" (prebid#1529) Override default asset params when set on ad unit (prebid#1524) Adding new kv to xhb Adapter (prebid#1513) removing for...of loops because IE cannot handle them properly (prebid#1523) Increment pre version Prebid 0.27.0 Release Move unit test file to appropriate location (prebid#1516) Support 'cta' native asset (prebid#1505) Add adapter parameter types (prebid#1504) Register bid adapter (prebid#1514) Match port when testing legacy browser (prebid#1511) modify 'featureforward' adapter to be an indepandant adapter (prebid#1288) Feature/s2s client side fallback (prebid#1485) Make bid.vastUrl use the cache URL if the bid didnt already have one. (prebid#1506) OpenX Adapter Update: (prebid#1438) Add session id feature for roxot analytics adapter (prebid#1498) Update platformioBidAdapter (prebid#1493) Adding VAST Payload support for video bids (prebid#1407) ...
Type of change
Description of change
The
parsefunction in urls.js returns a port of 443 when tested in IE11, blank in other browsers. A hardcoded url that is matched against doesn't contain this port and causes the test mentioned in #1302 to fail. This PR adds that port to the hardcoded string when tested in IE11 and fixes #1302.