| [reply] [d/l] |
| [reply] |
There may be opportunities to "cheat", if you know something about the processes generating the XML. For instance, are the sections always in the same order? The message id's on the same lines in each section? etc...
If you have to handle arbitrary XML, or you don't have control of the sources so you can't guarantee that any cheats will remain valid, I'd say find the fastest XML parser you can and go with that...
| [reply] |
380MB, really, is not "large" for most computers these days, which could easily handle both data structures side-by-side in memory without serious swapping. Therefore you could simply suck the two files into memory and use something like Data::Compare. If you are looking for specific comparisons, XSLT might be useful to "drill down" to exactly what you are looking for. | [reply] |