[2.1][DomCrawler] Fix relative path handling in links #7244
Conversation
| $output = array(); | ||
|
|
||
| foreach (explode('/', $path) as $idx => $segment) { | ||
| if ('..' === $segment) { |
There was a problem hiding this comment.
&& $output to avoid a warning on invalid paths ?
There was a problem hiding this comment.
But why should it trigger a warning?
There was a problem hiding this comment.
Couldn't we have an invalid path that would trigger more pop than the array contains ?
In this case array_pop would generate a warning
There was a problem hiding this comment.
It wouldn't. AFAIK, it generates a warning only for non-arrays, but in that case there would be a warning on the last line of getUri().
I'll fix it and add another test. Thanks
There was a problem hiding this comment.
you're right, I should have checked the doc first !
There was a problem hiding this comment.
Would your code resolve an invalid path (too many "..") in the same manner of a real browser ? (I actually don't know what a browser would do)
There was a problem hiding this comment.
@vicb It does. A browser will simply ignore the extra .. when it reaches the root (at least FF does)
There was a problem hiding this comment.
I've just added the test for that case :)
There was a problem hiding this comment.
perfect then ! thanks @lazyhammer @stof
|
@fabpot seems like 2.1 as a deps issue if you look at travis logs |
Added relative path canonicalization according to RFC 3986, section 5.2.4
This PR was merged into the 2.1 branch. Commits ------- a4ec677 [DomCrawler] Fix relative path handling in links Discussion ---------- [2.1][DomCrawler] Fix relative path handling in links | Q | A | ------------- | --- | Bug fix? | yes | New feature? | no | BC breaks? | no | Deprecations? | no | Tests pass? | yes | Fixed tickets | #7219 | License | MIT | Doc PR | n/a Added relative path canonicalization according to RFC 3986, section 5.2.4 --------------------------------------------------------------------------- by vicb at 2013-03-02T14:48:46Z @fabpot seems like 2.1 as a deps issue if you look at travis logs
Added relative path canonicalization according to RFC 3986, section 5.2.4