getElementsByTagName, tag does not exist

H

hymie!

Greetings. I'm just starting to dabble in XML, and I've run across a
problem.

I'm going through my XML document using this construct:

use XML::DOM;
my $parser = new XML::DOM::parser
or bail ("Unable to create XML parser");
my $story = $parser->parse($data);
$out{"SOURCE"} = $story->getElementsByTagName("Source")->
item(0)-> getFirstChild->getData;
$out{"DATE"} = $story->getElementsByTagName("Publication_Date")->
item(0)-> getFirstChild->getData;
$out{"TEXT"} = $story->getElementsByTagName("Body_Text")->
item(0)-> getFirstChild->getData or die "$!";

Everything works fine, until I get to a $story where Body_Text doesn't
exist. I've looked through all of the XML::DOM docs that I can find,
but I can't find either a way to test if Body_Text exists, or what happens
when Body_Text doesn't exist. The script stops with no obvious
diagnostic output -- it appears that the "die" never happens.

Can somebody show me the light?

hymie! http://www.smart.net/~hymowitz hymie_@_lactose.homelinux.net
===============================================================================
I've got an answer. I'm going to fly away. What have I got to lose?
--Crosby, Stills, and Nash
===============================================================================
 
M

Matt Garrish

hymie! said:
Greetings. I'm just starting to dabble in XML, and I've run across a
problem.

I'm going through my XML document using this construct:

use XML::DOM;
my $parser = new XML::DOM::parser
or bail ("Unable to create XML parser");
my $story = $parser->parse($data);
$out{"SOURCE"} = $story->getElementsByTagName("Source")->
item(0)-> getFirstChild->getData;
$out{"DATE"} = $story->getElementsByTagName("Publication_Date")->
item(0)-> getFirstChild->getData;
$out{"TEXT"} = $story->getElementsByTagName("Body_Text")->
item(0)-> getFirstChild->getData or die "$!";

Everything works fine, until I get to a $story where Body_Text doesn't
exist. I've looked through all of the XML::DOM docs that I can find,
but I can't find either a way to test if Body_Text exists, or what happens
when Body_Text doesn't exist. The script stops with no obvious
diagnostic output -- it appears that the "die" never happens.

It's pretty clear from the docs that getElementsByTagName returns an array
containing all the nodes found, so why don't you just check whether there
are any nodes? As per the documentation:

[untested]

my $nodes = $story->getElementsByTagName('Body_Text');
unless ($nodes->getLength > 0) {
print "Sorry, no nodes!\n"
}

I would recommend XML-Libxml as a better alternative to XML-DOM, though.

Matt
 
R

robic0

Greetings. I'm just starting to dabble in XML, and I've run across a
problem.

I'm going through my XML document using this construct:

use XML::DOM;
my $parser = new XML::DOM::parser
or bail ("Unable to create XML parser");
my $story = $parser->parse($data);
$out{"SOURCE"} = $story->getElementsByTagName("Source")->
item(0)-> getFirstChild->getData;
$out{"DATE"} = $story->getElementsByTagName("Publication_Date")->
item(0)-> getFirstChild->getData;
$out{"TEXT"} = $story->getElementsByTagName("Body_Text")->
item(0)-> getFirstChild->getData or die "$!";

Everything works fine, until I get to a $story where Body_Text doesn't
exist. I've looked through all of the XML::DOM docs that I can find,
but I can't find either a way to test if Body_Text exists, or what happens
when Body_Text doesn't exist. The script stops with no obvious
diagnostic output -- it appears that the "die" never happens.

Can somebody show me the light?

hymie! http://www.smart.net/~hymowitz hymie_@_lactose.homelinux.net
===============================================================================
I've got an answer. I'm going to fly away. What have I got to lose?
--Crosby, Stills, and Nash
===============================================================================
The alternative to DOM is SAX, widely used in modern code.
Its basically a simple event driven model, calling handlers
when the basic structured xml components are encountered. This
allows you to control going from xml to internal data structures
and/or back out to xml. Expat provides hooking handlers to most
of the current W3c constructs. These are just the basic ones.
Its up to you to extract the data into internal structures.
For that XML:Simple is a good tool. With Expat you can accumulate
nested data in a single string. Then Simple will create nested
Perl structures using tag names. Then you can Dumper it.
But, nobody uses Xml that doesen't know ahead of time what those
structures are both out and in. This is a way to control/validate/
populate them. SAX gives you a much simpler model and allows
much better control of the data. If you need more information
let me know. Getting Xerces working is a chore (you could do
without it for now, its only being used for schema checking here).
This code chunk sample is from 7,000 line code I wrote that was
converted
to a binary with Perl2Exe (including Xerces). I've chopped it up,
you can't see or know what it does so it will look nasty but
all the clues are there for you to investigate SAX and thats enough.
-gluck


---
This code is chopped out of a large practical xml code base and is NOT
cut & paste workable. Its just for instructional purposes
for the poster to give a flavor of SAX: Simple Api Xml.

use XML::Xerces;

use XML::parser::Expat;
use XML::Simple;

## main
{
## Initialize program / build list of xml files (ie: glob)
for (@XmlFiles)
{
/.+$dlimsep(.+)$/; (defined elsewhere for win/unix os)
$XML_File = $1;
Log ($XML_File);

## Validate Schema with Xerces
## note: Xerces is being used for schema validation
and
## as backup xml integrity (done elsewhere)
next if (!ValidateSchema ($_));

if (!open(SAMP, $_)) {
Log (...);
next;
}

## Parse xml and integrity check (Expat-SAX)
my $parser = new XML::parser::Expat;
$parser->setHandlers('Start' => \&stag_h,
'End' => \&etag_h,
'Char' => \&cdata_h);
$parser->setHandlers('Comment' => \&comment_h) if
($hVars{'CommentLogging'});

eval {$parser->parse(*SAMP)};
if ($@) {
## xml integrity failed -log this error
$@ =~ s/^[\x20\n\t]+//; $@ =~
s/[\x20\n\t]+$//;
# attempt strip off program line,col info at
end
$@ =~ s/(at line [0-9]+,.+)?at .+ line
[0-9]+$/$1/;
Log (...error...);
}
close(SAMP);
$parser->release;
}
}

########################################################
# EXPAT Event Handlers - start/end/content (defaults)
########################################################
##
sub stag_h # -- Start Tag --
{
my ($p, $element, %atts) = @_;
$element = uc($element);
$last_content = '';
$last_syntax_content = '';

## -- construct & Print start tag --
my $tag = "\<$element\>";
if ($XML_PRINT) {
printf ("%3d", $p->current_line);
print get_indent();
print "$tag";
print " Attr" if (keys %atts);
foreach my $key (keys %atts) {
print ", $key=".$atts{$key};
}
print "\n";
}
$tab_lev++;

## -- set Detached special content handler --
if (exists ($Content_hash{$element}) &&
$Content_hash{$element}->[1]) {
$p->setHandlers('Char' =>
$Content_hash{$element}->[0]);
}

## do something with attributes
## start keying (populating) your data structures
## set flags, etc ...
}

##
##
sub etag_h # -- End Tag --
{
my ($p, $element) = @_;
$element = uc($element);

## -- Construct & Print end tag --
my $tag = "\</$element\>";
$tab_lev--;
if ($XML_PRINT) {
printf ("%3d", $p->current_line);
print get_indent();
print "$tag\n";
}
## -- store last Content in hash (do more stuff)

## then:
$last_content = '';

## -- Restore default content handler --
if (exists ($Content_hash{$element}) &&
$Content_hash{$element}->[1]) {
$p->setHandlers('Char' => \&cdata_h);
my $last = (@Action) - 1;
my $aref = $Action[$last];
$Content_hash{$element}->[2]($last_syntax_content,
$aref, $element);
}
}

##
##
sub cdata_h # -- Default Content Data --
{
my ($p, $str) = @_;
# use original for entities, incase reparse
$str = $p->original_string;
# remove leading/trailing space, newline, tab
$str =~ s/^[\x20\n\t]+//; $str =~ s/[\x20\n\t]+$//;
if (length ($str) > 0)
{
if ($XML_PRINT) {
printf ("%3d", $p->current_line);
print get_indent();
print "$str (".length($str).")\n";
}
$last_content .= $str;
}
}

##
##
sub comment_h # -- Default Comment Data --
{
my ($p, $str) = @_;
# use original for entities, incase reparse
$str = $p->original_string;
# remove leading/trailing space, newline, tab
$str =~ s/^[\x20\n\t]+//; $str =~ s/[\x20\n\t]+$//;
if (length ($str) > 0)
{
printf (" %d,%d\n",
$p->current_line,$p->current_column);
}
}

##
##
sub cdata_x_h # -- Special Content Data --
{
my ($p, $str) = @_;
cdata_h ($p, $str);
# remove leading/trailing space, newline, tab
$str =~ s/^[\x20\n\t]+//; $str =~ s/[\x20\n\t]+$//;
$last_syntax_content .= $str if (length ($str) > 0);
}

########################################################
# Xerces - too much to explain
########################################################
#
sub ValidateSchema {
my ($xfile) = @_;
#my $valerr = 0;

# Docs:
http://xml.apache.org/xerces-c/apiDocs/classAbstractDOMParser.html#z869_9
my $Xparser = XML::Xerces::XercesDOMParser->new();
$Xparser->setValidationScheme(1);
$Xparser->setDoNamespaces(1);
$Xparser->setDoSchema(1);
#$Xparser->setValidationSchemaFullChecking(1); # full
constraint (if enabled, may be time-consuming)


$Xparser->setExternalNoNamespaceSchemaLocation($hVdef{'Schema'});

my $ERROR_HANDLER = XLoggingErrorHandler->new(\&LogX_warn,
\&LogX_error, \&LogX_ferror, );
#my $ERROR_HANDLER = XML::Xerces::perlErrorHandler->new();
$Xparser->setErrorHandler($ERROR_HANDLER);

# no need for eval on parse with handlers.. just insurance on
die
eval {$Xparser->parse
(XML::Xerces::LocalFileInputSource->new($xfile));};
if ($@) {
}
return 1;
}

## handlers (alot more not shown)
 
H

hymie!

In our last episode, the evil Dr. Lacto had captured our hero,
$out{"TEXT"} = $story->getElementsByTagName("Body_Text")->
item(0)-> getFirstChild->getData or die "$!";

Everything works fine, until I get to a $story where Body_Text doesn't
exist. [...]
The script stops with no obvious
diagnostic output -- it appears that the "die" never happens.
It's pretty clear from the docs that getElementsByTagName returns an array
containing all the nodes found, so why don't you just check whether there
are any nodes? As per the documentation:

Thanks for pointing out my mistake. My program doesn't die at the
getElementsByTagName , it apparantly fails at the ->item(0) .
my $nodes = $story->getElementsByTagName('Body_Text');
unless ($nodes->getLength > 0) {
print "Sorry, no nodes!\n"
}

All is happy now.

hymie! http://www.smart.net/~hymowitz (e-mail address removed)
===============================================================================
My brothers and sisters all hated me, cuz I was an only child.
--'Weird Al' Yankovic
===============================================================================
 
M

Matt Garrish

[ TOFU corrected ]
You know, I'm gonna go one step further here and say:
If you use nodes your some kind of a dumb ass. Not the
modern thinking at all !!

Your logic has me convinced. Oh wait, there is no logic. You do realize that
you're just trying to create your own DOM tree by jumping through a bunch of
hoops, right? SAX will let you do that, but I don't see how what you're
suggesting would be an improvement over a regular DOM tree.

SAX has its uses (especially for large documents you don't want to read into
memory), but it's hardly a reason not to use DOM. If I need the whole
document in memory before beginning to process, SAX is a needlessly complex
way of doing that. If I just want to fire off events as I come across
elements, DOM is a needlessly complex way of doing that.

Picking one method over the other requires analyzing what your needs are.
Simplistic statements like sax for everything aren't helpful.

Matt
 
R

robic0

On Thu, 20 Oct 2005 08:04:45 -0400, "Matt Garrish"

Actually, your right! Analysis always does the trick.
I personally would not do DOM, I want intimate controll.
I feel I can controll the entire event driven model,
on data large and small. Dom will fade imho, but
you have a point..
[ TOFU corrected ]
You know, I'm gonna go one step further here and say:
If you use nodes your some kind of a dumb ass. Not the
modern thinking at all !!

Your logic has me convinced. Oh wait, there is no logic. You do realize that
you're just trying to create your own DOM tree by jumping through a bunch of
hoops, right? SAX will let you do that, but I don't see how what you're
suggesting would be an improvement over a regular DOM tree.

SAX has its uses (especially for large documents you don't want to read into
memory), but it's hardly a reason not to use DOM. If I need the whole
document in memory before beginning to process, SAX is a needlessly complex
way of doing that. If I just want to fire off events as I come across
elements, DOM is a needlessly complex way of doing that.

Picking one method over the other requires analyzing what your needs are.
Simplistic statements like sax for everything aren't helpful.

Matt
 
R

robic0

On Thu, 20 Oct 2005 08:04:45 -0400, "Matt Garrish"

I didn't getcha on this statement..
 
M

Matt Garrish

I didn't getcha on this statement..

You only talk about DOM and SAX in your original post (and being the two
most common and supported parsers, that's not surprising). By following up
and saying you'll never use nodes again you're implying that SAX is the only
way to go. That's all that was meant.

Matt
 
R

robic0

On Sat, 22 Oct 2005 09:57:21 -0400, "Matt Garrish"
Your right Matt. I sometimes make those comments on
Friday nights after the pub. Dissregard those,
and sorry for them. But I am usually coherent on here
and am interrested in technical development.
I mean no harm. - thanks!
 
R

robic0

On Sun, 23 Oct 2005 15:04:54 -0700, (e-mail address removed) wrote:
Actually, DOM is exponentionaly irrelavent in modern xml parsing.
Its use is fadinding like an old 59 Buick in 1965.
 
M

Matt Garrish

Actually, DOM is exponentionaly irrelavent in modern xml parsing.
Its use is fadinding like an old 59 Buick in 1965.

I get the impression you just like to hear yourself talk, so I'm not even
going to bother arguing this. I would suggest in the future you not get
carried away by the latest fad technology you just googled.

Matt
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

Forum statistics

Threads
473,995
Messages
2,570,233
Members
46,820
Latest member
GilbertoA5

Latest Threads

Top