Beefy Boxes and Bandwidth Generously Provided by pair Networks
good chemistry is complicated,
and a little bit messy -LW
 
PerlMonks  

Re^3: Bug in XML::Parser

by graff (Chancellor)
on Oct 23, 2013 at 06:01 UTC ( #1059284=note: print w/ replies, xml ) Need Help??


in reply to Re^2: Bug in XML::Parser
in thread Bug in XML::Parser

If you are trying "to find if there are any mismatched tags", that sounds like you are looking for errors that would cause an XML parser to fail (and it appears that the sample xml data you posted has this kind of problem, so I understand your goal now).

But what that really means is that you can't really use an XML parser at all to solve this problem. As pointed out above, it's easy enough to check for xml errors using xmllint, although the error reports you get can sometimes be difficult to interpret, and the actual problem can still be hard to spot.

I would be inclined to use a regex-based diagnosis - something like this:

#!/usr/bin/perl use strict; use warnings; my $infile = shift; # get input file name from @ARGV open( my $fh, "<:utf8", $infile ) or die $!; local $/; # slurp the whole file in the next line $_ = <$fh>; s/^<\?.*>\s+//; # ditch the "<?xml...?>" line, if any my %open_tags; my %close_tags; for my $tkn (split/(?<=>)|(?=<)/) { # split on look-behind | look-ahe +ad for brackets if ( $tkn =~ m{^<(\/?)(\w+)} ) { if ( $1 eq '' ) { $open_tags{$2}++; } else { $close_tags{$2}++; } } } for my $tag ( sort keys %open_tags ) { if ( ! exists( $close_tags{$tag} )) { warn sprintf( "%s: open tag %s is never closed in %d occurrenc +e(s)\n", $infile, $tag, $open_tags{$tag} ); } else { if ( $close_tags{$tag} != $open_tags{$tag} ) { warn sprintf( "%s: element %s has %d open tags but %d clos +e tag(s)\n", $infile, $tag, $open_tags{$tag}, $close_tags +{$tag} ); } delete $close_tags{$tag}; } } for my $tag ( keys %close_tags ) { warn sprintf( "%s: close tag %s has no open tags in %d occurrence( +s)\n", $infile, $tag, $close_tags{$tag} ); }
That will at least give you a clear tally of imbalances (if any) in the open/close tag inventory for a given xml file. You should be able to use this information, together with the line numbers from the xmllint reports, to locate the problems.

So, when you find these mismatched tags, isn't the next step to look at the process that is creating the xml files, and fix that? (These xml files aren't being created by manual editing, are they??)

(Update: BTW, I forgot to mention... this new information in your reply makes your OP even more egregiously obtuse. If you had said at the beginning, "I have this xml file that has an error in the tags, and I need to figure out how to find the problem," then the discussion would have been more effective. I know, you already feel bad about the OP, and I shouldn't pile it on, but it needs to be said.)


Comment on Re^3: Bug in XML::Parser
Download Code
Re^4: Bug in XML::Parser
by manunamu (Initiate) on Oct 23, 2013 at 11:04 UTC
    Thanks graff. You have been great help. Yes, you are right about stressing on the point about OP and point is well taken. I understand your solution. But, what still confounds me completely is how the XML::Parser is able to detect the badly formed XML file even if the program is parsing one line at a time. This obviously could be because my understanding of the XML::Parser is limited. One of the unintended consequences that your reply had gave me an insight into a problem that I had earlier, but never got around asking the permonks site. Aren't unintended consequences great when they turn out to be good? Thanks macho again!
      If you happen to be passing one line at a time to the parser, and your input xml file happens to contain a close tag all by itself on one line (or a portion of a complicated open tag that has been broken up into multiple lines), I expect the parser would fail at that point. (At least, xmllint wouldn't accept such a file.)

      (Update: actually, if the data contains just a complete open tag on one line, and you pass that to an xml parser, that should fail as well.)

Log In?
Username:
Password:

What's my password?
Create A New User
Node Status?
node history
Node Type: note [id://1059284]
help
Chatterbox?
and the web crawler heard nothing...

How do I use this? | Other CB clients
Other Users?
Others meditating upon the Monastery: (7)
As of 2014-08-23 20:52 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    The best computer themed movie is:











    Results (178 votes), past polls