Parse html using $.parseHTML.

[ssafejava]

See #351. In version 0.12.1, cheerio supports $.parseHTML. More on $.parseHTML from jQuery's tracker.

@jugglinmike, what do you think about wrapping the return value of $.parseHTML in cheerio inside $()? While all tests pass in the browser with an unwrapped $.parseHTML call, they fail in cheerio - I get the following output using console.log in the tests:

actual:
<div class="listContainer">List Container<div class="list">[object Object][object Object]</div></div>
expected:
<div class="listContainer">List Container<div class="list"><div class="item"><span class="title highlight">Item 1</span></div><div class="item"><span class="title highlight">Item 2</span></div></div></div>

[SBoudrias]

Is that really necessary? Can a string end up here without that being a bug?

[ssafejava]

Yeah, strings would reach it through applyTemplate, if manager.noel was false. I fixed that in the next commit and removed $.parseHTML from options.html.

[SBoudrias]

I like the simplified version!

Although, I think the $.parseHTML bug should be adressed in Cheerio and eventually remove the fix here. (Although like you, I wonder if Mike have any insight on the behaviour - maybe this was a conscient choice for Cheerio folks)

[jugglinmike]

@ssafejava I wrote the same code yesterday after Cheerio 0.12.1 was published, but I felt that it's only exchanging one sub-optimal shim for another. @SBoudrias is correct--this is a bug in Cheerio. We might as well stick with what works for now until we have a complete solution.

As I mentioned earlier, I'm taking responsibility for #351. It's just going to take a little more time:

1. Yesterday, I took steps to update the jQuery documentation..
2. Once that lands, I'll fix the bug in Cheerio.
3. Finally, we'll be able to simply use $.parseHTML consistently in both environments.

[ssafejava]

Thanks for handling it @jugglinmike! I can leave the PR open and we can just patch it to remove the $() wrapper around $.parseHTML when Cheerio is ready.
On Aug 1, 2013, at 8:25 PM, jugglinmike notifications@github.com wrote:

@ssafejava I wrote the same code yesterday after Cheerio 0.12.1 was published, but I felt that it's only exchanging one sub-optimal shim for another. @SBoudrias is correct--this is a bug in Cheerio. We might as well stick with what works for now until we have a complete solution.

As I mentioned earlier, I'm taking responsibility for #351. It's just going to take a little more time:

Yesterday, I took steps to update the jQuery documentation..
Once that lands, I'll fix the bug in Cheerio.
Finally, we'll be able to simply use $.parseHTML consistently in both environments.
—
Reply to this email directly or view it on GitHub.

[tbranyen]

@ssafejava yup, that's what we're gonna do. We got the jQuery API docs updated (live now) and are progressing with the next step in keeping the Cheerio API matching closer.

[tbranyen]

I removed the wrapper and updated the test's intention to not be focused on trimming, but instead the retaining source via parseHTML.

We're going to merge this as soon as Cheerio releases a version bump with @jugglinmike's fixes.

[tbranyen]

Closed in #380