A few weeks ago I was happily parsing xml files that contained named entities ( , ’, é, etc.). The xml files contain a dtd reference, that on disk is a link to a dtd document which in turn references an entities file. This all worked well on a pretty standard RH6 machine, Perl 5.005, XML::DOM 1.30, XML::Parser 2.30.
Now, following an upgrade to RH8, Perl 5.80, XML::DOM 1.43, XML::Parser 2.31, I am not having as much luck. The entities are unrecognized and now die a horrible death. There are a number of scripts and programs based on these modules, so I'm looking for something that doesn't require a complete rewrite. My suspicion is that the entities file never was found before, but the default behavior was to use LWP or some other source for identifying these common named entities. Perhaps that's my biggest question, how did it work before? I can't isolate the difference between XML::Parser 2.30 and 2.31 to determine what has changed, or what it might be calling that has changed. One alternative I've found is to use ParseParamEnt as a parameter that forces the dtd and entities files to be read, but this has other side-effects such as loading all the definitions into the current xml file. Not what I have in mind plus it doesn't explain why it worked before without the entities file. Transforming the entities into other notations is not an option either. I feel like I've just read the wrong documents or am failing to see something obvious. What are others doing to parse their documents when named html entities are involved? Thanks for any help. -Lance _______________________________________________ Boston-pm mailing list [EMAIL PROTECTED] http://mail.pm.org/mailman/listinfo/boston-pm
