Re: Seafox HTML -> Seaside render methods translator

2010/8/2 Nick Ager <nick.ager <at> gmail.com>:
> Hi Philippe,
>>
>> The algorithm you use seems to work well for body tags but have
>> problems with header tags. For example look at the style tag:
>> - there's no such tag, simply the link tag which explains the selector
>> - it says the rel attribute is set with #beAlternateStylesheet
>> - it says the media attribute is set with #addHandheld
>> - it misses the attributes from the superclass
>> - it says the type attribute is set with #beRss
>>
>> Apart from that, very nice work, thank you.
>>
>
> Thanks for the feedback. You're right I didn't spend as long on the header
> mappings, though I added a special case in the parser for the link tag so
> that it picks up some of the common link to stylesheet mappings. See
> SFLibrary>>#browserJs function createLinkHTMLElementBrush (aNode). The
> problem with putting special cases in the parser is that it leads to a
> discrepancy between what the web app shows the mappings to be and the
> results the parser generates. Though hopefully the parser output is more
> interesting to people.

As long as it picks up commons CSS, and JavaScript it doesn't matter
when it's not complete.

> Perhaps we can come up with a better approach at ESUG?

I won't be able to make to to ESUG, sorry.

Cheers
Philippe