CamelcasedClassNameResolver now can resolve FQCN page and element #81
Conversation
|
If a totally new and separated one resolver is required (it was my first idea, but reading your suggestion drove me to this implementation) let me know. |
|
@jakzal now build is passing, let me know if there is something else I need to do or if this need changes. Peace! |
| } | ||
| } | ||
|
|
||
| class Foo extends Page |
There was a problem hiding this comment.
These two classes (Foo and Bar) are not used. Please remove.
There was a problem hiding this comment.
@jakzal I'm using them in spec examples to validate if it's a valid FQCN
There was a problem hiding this comment.
Do you mean arguments passed to the class under test? These are only namespace prefixes. In your case these would be ignored, as a FQCN is passed to the resolve method.
There was a problem hiding this comment.
I mean this
private function isPageObjectNameFQCN($pageObjectName)
{
return class_exists($pageObjectName);
}
If I remove those classes, spec examples I add will fail
There was a problem hiding this comment.
Ah right, can you use spec\\SensioLabs\\Behat\\PageObjectExtension\\PageObject\\Factory\\Fixtures\\NamespacedArticleList and spec\\SensioLabs\\Behat\\PageObjectExtension\\PageObject\\Factory\\Fixtures\\Element\\NamespacedSearchBox instead of the new ones? They're used in other examples.
There was a problem hiding this comment.
Umh, yes, but in that case I could not tell the difference in the examples, I mean, they're going to work even if FQCN part is broken. Maybe I can create brand new ones that not respects Page/Element structures in Fixtures folder?
There was a problem hiding this comment.
There's still a difference, as in the existing examples only part of the namespace is passed (and a whole namespace is returned with prefixes). In your new examples, you'd be passing a full namespace in, and expecting it back.
|
|
||
| throw new \InvalidArgumentException($message); | ||
| if ($this->isPageObjectNameFQCN($pageObjectName)) { |
There was a problem hiding this comment.
Would it be possible to move this condition to the beginning of the method? We wouldn't have to evaluate other parts of this method if the passed argument is already a class name.
There was a problem hiding this comment.
Yes, I would have to but I was not sure about that. Indeed is far better.
|
We will also need to document this new behaviour :) |
Yes, you're right! :) |
|
I've made some edits. If everything is ok I'll edit the documentation also. Let me know :) |
|
Looks great! A scenario (if easy to add), and docs would be appreciated :) |
|
Don't know if I understand correctly all behat scenarios you already have (going to take a look later). |
|
I've updated docs, please review them @jakzal Thanks :) |
- working with elements - working with page objects Introduction to FQCN resolver for page/element factory
|
I've also introduced a scenario adapted from existing one |
|
@jakzal ping :) |
|
Thank you @DonCallisto. Excellent work! 🍺 |
|
It was a pleasure: you're welcome! :) |
| } | ||
|
|
||
| If you choose FQCN strategy, you can organize your page directories freely as you are not bounded to page namespace | ||
| (see `configuration <http://behat-page-object-extension.readthedocs.io/en/latest/guide/configuration.html>`_) |
There was a problem hiding this comment.
I converted this link to an internal reference to make docs more portable: #83
Issue ref.: #76
Don't know if specs are good ones, could you check @jakzal?
Thanks :)