import question: only if xpath true syntax on output attribute

[naglists]

Hi,
My import map (FrameMaker MIF to XML) is emitting:
morerows="0" namest="1" nameend="1"

on every <entry> (namest/end values increase with column position).

I'm trying to modify the "Only Output Attribute if XPath is True:" dialog to
only output something if namest != nameend and morerows if !=0.

Anyone know the right way to present the XPath syntax in that box?
@namest != @nameend

Does not work.

Neither do:
@morerows !=0
@morerows !="0"

--
Paul Nagai

[GarethOakes]

Hi Paul,

Your Xpath syntax seems correct, and I think XPath won't care about your
string or non-string zero for the comparison you're doing.

What might be tripping you up is that you're testing in the wrong spot.
It's hard to tell without seeing your Import map file. How are the
<entry> attributes being generated? Are they already in the ppXML once
the first-stage import has done it's work on your MIF? If so, you might
need to look more closely at how the tables are being handled - if it's
a simple passthru type of thing that is handling the "table" element,
then your rules for "entry" may never in fact be triggered (as the table
will just be copied through as-is to the output XML).

If you end up spinning your wheels on this, another method to consider
is a second pass over the file to strip bad attributes. You can chain up
(multiple) post-processes for one import job, I think it's under one of
the option dialogs. You could add a simple XSLT post-process which
whizzes through the file pulling out any stupid attributes. Just another
idea in case you end up getting stuck under pressure (I know what that's
like..).

Cheers,
Gareth

[naglists]

Thanks, Gareth.

Am a little further along, now. Turns out I was, uh, "unplugged" and
reviewing a *copy* of my XML output that wasn't being refreshed by
reprocessing the map template. Duh. The namest, nameend, and morerows are
all working now. Funny how stuff like that helps 😉
The correct (well, one set of correct) settings are:

@namest != @nameend
and
@morerows != 0

Oh, and FWIW, 1=0 on colname. Don't need that attr ever, I don't think.

[ClayHelberg]

Heh, I hate it when I do that, it's always a real head smacker.



"Why aren't my changes showing up in the output?!!!"



"D'oh! I'm looking at the wrong output file!"



BTW, XPath has a false() function for the situation you describe with
colname. It's just a function that always returns false, since there's
no literal "false" value in XPath. Your way (1 = 0) works just as well,
but might be a little opaque to someone trying to debug the map template
at some later date. (There's also a corresponding true() function for
cases where you always want something to happen.)



--Clay

[naglists]

It took a dinner break for the light to go on.

Subbed false() for 1=0. Guess this also protects me from quantum
instabilities ...