Tom Harrington wrote:
> In article <1sime5-cj8.ln1@[EMAIL PROTECTED]
>,
> Lorenzopt <lorenzo@[EMAIL PROTECTED]
> wrote:
>
>> Tom Harrington wrote:
>>> In article <oster-21267B.20023929042008@[EMAIL PROTECTED]
>,
>>> David Phillip Oster <oster@[EMAIL PROTECTED]
> wrote:
>>>
>>>> I'm not. XML is supposed to be an all-or-nothing deal. If "tidy" says
>>>> there are errors, then a valid XML parser shouldn't parse it. A valid
>>>> parser SHOULD return an error in this case.
>>>>
>>>> Oh well, maybe NSXMLDo***ent is in some weird HTML compatibility
mode.
>>> NSXML is also surprisingly flexible when it comes to doing Xpath
lookups
>>> on do***ents with default namespaces. As in, you can leave the
>>> namespaces out of the lookup path and still get results.
>>>
>> It sure is. I've used it and it works quite nicely.
>
> In context though, what I meant is that this is not how the spec says
> Xpath is supposed to work. It's certainly convenient that it works this
> way, but it's a spec violation.
OK, I understand your point, but I'm not much of an XML person. What I'm
doing brings me into it for the first time, so I'm happy to find a
workaround. So, my question then is this:
What options does one have if parsing an XML do***ent fails due to
malformed content? If one does not control the XML source, what can be
done?
|
