Using ​XDocument.​Load() to load an .aspx file

I'm trying to manipulate an ASPX file via Linq to XML, but when I try to load the ASPX file using XDocument.Load(), it throws up the "Name cannot begin with the '%' character" exception, obviously because of the <%@ Page %> directive.

Is there anyway to have Linq ignore that line or something? I'm noticing that ScottGu's tutorial also uses XDocument.Load() to load an aspx file, but I don't see anything in his code that might prevent the % exception.

Where is he loading a .aspx file? I don't see it.

Fact is that raw aspx files are not valid XML, (not just the lines at the top, but any <% %> block including databinding expressions etc.) so you can't do this without modifying the aspx file first.

Sven Groot said:

Where is he loading a .aspx file? I don't see it.

Fact is that raw aspx files are not valid XML, (not just the lines at the top, but any <% %> block including databinding expressions etc.) so you can't do this without modifying the aspx file first.

My bad, I thought he did, but on closer inspection he's loading it via HTTP, so it won't be ASPX by the time it's on his side.

Anyway, that's annoying. I hoped Linq would be more forgiving about invalid XML.

Could perhaps configure/adapt an xml reader so that it can understand / ignore code blocks? not sure how much work that would be..

Bas said:

Sven Groot said:

*snip*

My bad, I thought he did, but on closer inspection he's loading it via HTTP, so it won't be ASPX by the time it's on his side.

Anyway, that's annoying. I hoped Linq would be more forgiving about invalid XML.

Its not even HTML on his side, the example is using REST and he is loading an XML file from someplace.rss.aspx RSS is always XML and not HTML or XHTML which is what ASPX is. I'd be interested fo find out what you are tying to do?

Going the _stevo route would require writing a parser (non trivial) so you'd probably use Sharp Develop or Actipro

vesuvius said:

Bas said:

*snip*

Its not even HTML on his side, the example is using REST and he is loading an XML file from someplace.rss.aspx RSS is always XML and not HTML or XHTML which is what ASPX is. I'd be interested fo find out what you are tying to do?

Going the _stevo route would require writing a parser (non trivial) so you'd probably use Sharp Develop or Actipro

 

I'm automating some of our application-building, which involves adding several  pages to a project and then modifying them based on certain stuff. Changing the title, adding controls, that sort of stuff. The EnvDTE namespace allows for this sort of stuff with classes, but not ASPX pages, it seems. So instead I figured I'd manipulate the ASPX file through Linq to XML because that's pretty convenient with XML, but now I have to figure out something more unwieldy.

I might read the aspx into memory, remove the page directive, feed it to an XDocument, manipulate it, write it back and then manually add the page directive back, but that somehow seems error-prone.

Bas said:

vesuvius said:

*snip*

I'm automating some of our application-building, which involves adding several  pages to a project and then modifying them based on certain stuff. Changing the title, adding controls, that sort of stuff. The EnvDTE namespace allows for this sort of stuff with classes, but not ASPX pages, it seems. So instead I figured I'd manipulate the ASPX file through Linq to XML because that's pretty convenient with XML, but now I have to figure out something more unwieldy.

I might read the aspx into memory, remove the page directive, feed it to an XDocument, manipulate it, write it back and then manually add the page directive back, but that somehow seems error-prone.

I wouldn't recommend that, because the <% %> tags won't parse with XDocument.

joechung said:

Bas said:

*snip*

I wouldn't recommend that, because the <% %> tags won't parse with XDocument.

How about reading the ASPX file into memory, replacing the '<%' and '%>' with XML-valid markers ('<_percentgoesghere_' and '_percentgoeshere_>' ?), process as XML, then put back the '<%' and '%>'.

Herbie

Ion Todirel said:

Dr Herbie said:

*snip*

aspx files can contain anything (CSS, JavaScript...), and HTML != XHTML, it would require a bit more effort than that

Hmm.  Sounding more and more like a customer parser is required

Herbie

This is a total PITA. I thought the same thing Dr. Herbie suggested, and it worked, until I indeed came across some scripts like Ion Todirel warned against.

I'm already using custom project templates, but they only go so far. I need to modify the ASPX code (as in, add a bunch of controls) for each page generated based on different criteria. Also, each page needs to have a different title, etcetera.

Any other ideas? I'm almost ready to just read the whole thing into a massive string, find the opening and closing tags of the form element, and insert a string of code between it. Which sounds incredibly error-prone and unwieldy.