Reorganize ext/uri tests - withers #19970
Conversation
ext/uri/tests/whatwg/modification/host_error_forbidden_host_codepoint_opaque2.phpt
Outdated
Show resolved
Hide resolved
There was a problem hiding this comment.
The ^ character is in the percent-encode set of the path (https://url.spec.whatwg.org/#path-percent-encode-set), but somehow it's not percent-encoded.
kocsismate
force-pushed
the
ext-uri-tests1
branch
from
00e380e to
dd5e549
Compare
kocsismate
force-pushed
the
ext-uri-tests1
branch
from
dd5e549 to
059b1a1
Compare
There was a problem hiding this comment.
I went through all RFC 3986 tests. Looking pretty good overall, some nits. I also think it would be helpful to print the entire resulting URL in addition to printing the output of the getters, this makes it easier to double-check how exactly the resulting URL will look, particularly with regard to the difference between null and "" (e.g. whether there is a ? for the query or not).
ext/uri/tests/rfc3986/modification/fragment_success_encoded.phpt
Outdated
Show resolved
Hide resolved
| <?php | ||
|
|
||
| $uri1 = Uri\Rfc3986\Uri::parse("https://example.com"); | ||
| $uri2 = $uri1->withFragment("foo%3Dbar"); // foo=bar |
There was a problem hiding this comment.
Should this also have a normalizable percent-encoded character similarly to ext/uri/tests/rfc3986/modification/host_success_encoded.phpt?
There was a problem hiding this comment.
This seems to be unresolved still.
ext/uri/tests/rfc3986/modification/userinfo_success_existing.phpt
Outdated
Show resolved
Hide resolved
Yes, I was also thinking about something similar (at least in some edge cases), but I didn't want to make the output too "crowded". My other concern is that most states can be tested via regular parsing 🤔 (we don't necessary have to use withers for them) |
|
@TimWolla What do you think about verifying URI recomposition separately? Or do you think we should always test the recomposed URI after URI modification? |
Always, because this also helps the human reader understand the test better. |
If the tests follow a common structure / order, that's easy to check and scan. You could also insert an additional newline between the various output “sections”. |
|
Unless I'm really needed here I'm removing my review request. |
| <?php | ||
|
|
||
| $uri1 = Uri\Rfc3986\Uri::parse("https://example.com"); | ||
| $uri2 = $uri1->withHost("t%65st.com"); // test.com |
There was a problem hiding this comment.
Can we please use t%65st.example as the replacement to avoid using “legit” domain names?
| <?php | ||
|
|
||
| $uri1 = Uri\Rfc3986\Uri::parse("https://example.com"); | ||
| $uri2 = $uri1->withFragment("foo%3Dbar"); // foo=bar |
There was a problem hiding this comment.
This seems to be unresolved still.
| <?php | ||
|
|
||
| $uri1 = Uri\Rfc3986\Uri::parse("https://example.com"); | ||
| $uri2 = $uri1->withFragment("foo%3dbar"); // foo=bar |
There was a problem hiding this comment.
Can we have a second encoding that can be normalized? e.g. %61 for the a.
| <?php | ||
|
|
||
| $uri1 = Uri\Rfc3986\Uri::parse("https://example.com"); | ||
| $uri2 = $uri1->withPath("/foo%2Fbar"); // /foo/bar |
There was a problem hiding this comment.
Can we have a second encoding that can be normalized? e.g. %61 for the a.
| <?php | ||
|
|
||
| $uri1 = Uri\Rfc3986\Uri::parse("https://example.com"); | ||
| $uri2 = $uri1->withQuery("foo%3dbar"); // foo=bar |
There was a problem hiding this comment.
Can we have a second encoding that can be normalized? e.g. %61 for the a.
| <?php | ||
|
|
||
| $uri1 = Uri\Rfc3986\Uri::parse("https://example.com"); | ||
| $uri2 = $uri1->withScheme("http"); |
There was a problem hiding this comment.
| $uri2 = $uri1->withScheme("http"); | |
| $uri2 = $uri1->withScheme("HTTP"); |
To allow for normalizing to happen.
No description provided.