I know this is off topic (please forgive)
// loop logic to adjust and syncronize header table and
// data table
var ind;
var id1;
> > > One thing we could do, however, is implement char_traits
> > > such that
> > > user code could use basic_string >.
> > > That's something I've always wanted to do, but just haven't found
> > > the time yet.
> >
> > Yeah, I've wanted to do that too. I went so far as to look for an
> > existin
I could not possibly agree more!
I firmly believe that if we can just get vector and string working, the rest
of these problems will take care of themselves as newly-empowered
open-source coders have STL containers available to write code that deals
with the other weirder templates and their catc
[EMAIL PROTECTED] wrote:
I see it, I understand it, I disagree with it. I believe that
numpunct, moneypunct, time_get, time_put, and
messages are Not Our Problem (TM).
Our problem is getting the users their XML data. Shouldn't the other issues
be considered separately?
Exactly. Sometimes it re
I see it, I understand it, I disagree with it. I believe that
numpunct, moneypunct, time_get, time_put, and
messages are Not Our Problem (TM).
Our problem is getting the users their XML data. Shouldn't the other issues
be considered separately?
Regards,
Matt
-Original Message-
From: Axel
James Berry wrote:
> > One thing we could do, however, is implement char_traits
> > such that
> > user code could use basic_string >.
> > That's something I've always wanted to do, but just haven't found
> > the time yet.
>
> Yeah, I've wanted to do that too. I went so far as to look for an
> exis
On May 24, 2005, at 2:55 PM, [EMAIL PROTECTED] wrote:
From a quick grep of the code, I see calcRequiredSize() is used in the
deprecated DOM, and internally withing the transcoders to avoid
re-allocation memory with the local code page transcoders. Given
that,
perhaps it can deprecate it in 2
On May 24, 2005, at 3:02 PM, Axel Weiß wrote:
Am Dienstag, 24. Mai 2005 23:42 schrieb James Berry:
There's really little use for calcRequiredSize given that we also
support routines that allocate their own memory for the returned
string. Routines that need to do something fancier can do somet
On May 24, 2005, at 3:07 PM, [EMAIL PROTECTED] wrote:
It would be really nice if Xerces had a nice re-allocable data store
like a string. It's too bad the C++ standards folks didn't see fit to
give us a UTF-16 string (and UTF-8 too, for good measure). But they
didn't. And Xerces doesn't.
std
I actually was looking at what it would take to do this according to
http://www.sgi.com/tech/stl/char_traits.html . I think it would be a bit
difficult, but doable. And, if we did it, life in C++ for those using Xerces
and/or Unicode would be vastly improved. Here are the methods which we would
nee
> It would be really nice if Xerces had a nice re-allocable data store
> like a string. It's too bad the C++ standards folks didn't see fit to
> give us a UTF-16 string (and UTF-8 too, for good measure). But they
> didn't. And Xerces doesn't.
std::basic_string is completely compatible with UTF-
Am Mittwoch, 25. Mai 2005 00:00 schrieb
[EMAIL PROTECTED]:
> Well, on a mathematical level, it would not be possible for the output
> to take more than twice the space of the input if you are converting
> from UTF-16 to *-8, even accounting for a string composed entirely of
> surrogates.
>
> Surel
Well, on a mathematical level, it would not be possible for the output to
take more than twice the space of the input if you are converting from
UTF-16 to *-8, even accounting for a string composed entirely of surrogates.
Surely we could use a vector, where you reserved 1.5 times capacity (or
what
Am Dienstag, 24. Mai 2005 23:42 schrieb James Berry:
> There's really little use for calcRequiredSize given that we also
> support routines that allocate their own memory for the returned
> string. Routines that need to do something fancier can do something
> fancier.
James,
we do have a tr
> Wouldn't it be better if Xerces required users to use dynamic data
> structures like ?
I'd rather not go down the path of requiring something like that.
> Once you have a method like calcRequiredSize, people will start to
misuse
> it. This can be discouraged by not having such a method in the
[EMAIL PROTECTED] wrote:
> I am writing about this issue with calcRequiredSize. To me,
> calcRequiredSize seems like a rather bad method to have because you
> are converting an O(n) operation to transcode a string to at least a
> 2*O(n) operation, since you must make a pass across the data looking
Hi Matt,
On May 24, 2005, at 2:24 PM, [EMAIL PROTECTED] wrote:
I am writing about this issue with calcRequiredSize. To me,
calcRequiredSize
seems like a rather bad method to have because you are converting
an O(n)
operation to transcode a string to at least a 2*O(n) operation,
since you
mu
I am writing about this issue with calcRequiredSize. To me, calcRequiredSize
seems like a rather bad method to have because you are converting an O(n)
operation to transcode a string to at least a 2*O(n) operation, since you
must make a pass across the data looking for any surrogates, substitutable
On May 24, 2005, at 1:59 PM, Axel Weiß wrote:
Am Dienstag, 24. Mai 2005 02:24 schrieb James Berry:
- In general, about half of the transcoder code is taken up by
supporting the LCP functions, which are disjoint from the standard
virtual transcoder interface. As I proved recently for the
Am Dienstag, 24. Mai 2005 02:24 schrieb James Berry:
> - In general, about half of the transcoder code is taken up by
> supporting the LCP functions, which are disjoint from the standard
> virtual transcoder interface. As I proved recently for the Mac
> transcoder, the LCP transcoder can
[
http://issues.apache.org/jira/browse/XERCESC-1252?page=comments#action_66190 ]
Jason Ostermann commented on XERCESC-1252:
--
Same issue also exists with Xerces 2.6.0, IRIX 6.5.25m, MIPSPro 7.4.2m. Same
fix works.
> Compilation problem on IRIX
Yes, there actually is a theoretically better way. The only problem is that
it did not work for me when I tried it before, and I do not like
recommending methods that I have not had success with.
Since schemas are, in and of themselves, XML instance documents, you can
actually validate them with X
Matt,
My question is what's the way to validate JUST an XSD schema file, not
how to validate an XML file that has a schema. For the latter case your
example is helpful; however, the need I have is to validate only the XSD
file. Your method would imply that one needs to create a dummy XML file
as
Xerces will check out the schema before you parse an XML file with it.
If there are problems, it prints detailed errors, then attempts to process
the XML file if the errors are not fatal.
http://ultrasparcy.mhcomputing.net/~mhall/validator.zip might help you.
HTH!
Matt
-Original Message-
Neil,
Thanks for the reply.
Your curiosity is well-founded in my case c).
I re-tested this and I DO NOT get validation errors (i.e. parse is
successful) with:
- use cached grammars == true, and
- PSVIHandler NOT registered.
I have tried calling lockPool after the ::loadGrammar, but before th
Hi,
Sorry infrastructure I can't do it. I cannot commit to xerces and I
cannot check out infrastructure to find out how to add people (as per
the FAQ). The command I use to check out infrastructure is
svn co https://svn.apache.org/repos/asf/infrastructure/trunk infrastructure
I check out x
> I was wondering if anyone has suggestions for fixing this link error.
> The string "XercesMessages2_6_0_dat" seems to only appear in
> ICUMsgLoader.cpp.
That symbol will be built into the message library if you built it
properly. You can use the Perl script packageBinaries.pl in the scripts
d
Hi James,
I agree - we should keep the HEAD as the main development line, and
make a Xerces-2.7 branch as suggested.
I'll try to explain the tagging thing a bit more clearly.
James Berry <[EMAIL PROTECTED]> writes:
> The reason for a branch-point tag is that it gives you an easy way to
> get di
Hey,
Sorry, I have not been following the thread Jason. I am supposed to
have the ability to do this. Let me look into it and get back to you in
an hour. If I can't get it to work it will be back to infrastructure :)
Cheers,
Gareth
Jason E. Stewart wrote:
Hi Infrastructure team,
One of t
Hi Infrastructure team,
One of the xerces/c committers tried using the instructions for
accessing SVN using:
http://www.apache.org/dev/version-control.html
But hasn't had any success.
Q: Is this something that committers can take care of themselves, or
do we need intervention from infrast
Hi Jeff,
Not sure why this isn't working for you. I am a little curious about why
you're getting validation errors in case c); is this something you expect?
i.e., are you using an invalid instance as your test case?
Have you tried calling lockPool() on your XMLGrammarPool implementation
befo
Hey Jason,
I don't think there is any access at all set up for the xerces-c
repository. The file that controls svn access on svn.apache.org
contains no entries for xerces-c:
grep xerces /x1/svn/asf-authorization
xerces-p=jasons
[/xml/xerces-p]
@xerces-p = rw
What are you in
On May 23, 2005, at 9:55 PM, Jason E. Stewart wrote:
James Berry <[EMAIL PROTECTED]> writes:
After following the directions here (http://www.apache.org/dev/
version-control.html) for running svnpasswd, I tried to do a commit,
which failed with an authentication error. So either I messed up
s
On May 23, 2005, at 9:46 PM, Jason E. Stewart wrote:
[EMAIL PROTECTED] (Jason E. Stewart) writes:
- Tag the branch point: tags/Xerces-C-2.7-bp
I'm pretty sure this is not needed in SVN. You can if you want, but
all the information that is needed can be gotten in other ways in
SVN.
Hi,
The problem is that I am getting a lot of multiply
defined symbol errors while linking.
A few errors are reported below.
ld: fatal: symbol
`xercesc_2_6::chLatin_f' is
multiply-defined: (file
SOLARIS/ASCIIRangeFactory.o and file SOLARIS/MemoryManagerArrayImpl.o);ld:
fatal: sym
35 matches
Mail list logo
