Huge amount of memory used with XML::DOM

[Sihal]

Fellow monks,

I'm using XML::DOM to parse a 250 kbytes XML document.
The parser works very fine indeed, but my perl process takes up to 1.85 Gigs of memory when parsing. Is that normal? I suspect no, but fail to have even the slightest hint on what might be causing this, bearing in mind that once the documents is parsed, the results are valid.

THanks in advance for any hints on what might be causing this.

[artist]

That's huge. If you post the code, we may be able to identify the problem.

[bear0053]

perhaps some of your code would help...post an example snippet of where the parsing is performed

[Sihal]

Ok

This is what I used to do:



my $xml = getXML($url); # retrieve it thru http
my $parser = new XML::DOM::Parser;
my $doc = $parser->parse($xml);

# This first snippet would take as much as 2Gigs of memory.

Now, I do this:



my $xml = getXML($url); # retrieve it thru http
my $file  = writeToFile($xml) or return; 
my $parser = new XML::DOM::Parser;
my $doc = $parser->parsefile($file);

# This isn't even noticeable



Any Ideas why the first form is so greedy in memory?
Never found anything about this in the doc.

Thanks a lot.

[Aristotle]

Maybe this is related to the leak described in (XML::Parser) Finding and fixing a bug? Although I'm rather puzzled by the disparity between a 250KB input size and a 1.9GB process size. Something has to have gone seriously awry there..