Lists Home |
Date Index |
Thanks to Everyone, your insight is deeply appreciated.
I guess XML is harder to describe than to use. The best way to learn is by downloading the tools and prodding them to see how they react, but I've had my hands full with the dictionary...
My next step will be to write my own scripts in Perl to select information inside each page, stuff that can not be found in the index, tenses in the imperative, for example, for debugging the conjugations...
Perl on Linux is pretty easy to use. Only one surprise: I keep a copy of the dictionary in the Apache dictionary to test the search cgi function -- I need the server for that -- but I edit the originals in my home directory, and use the file:// access protocol...
I have written Perl scripts which apply global changes to the entire set of HTML pages, like when I added the <form>, or changed the navigation links, or correct some kind of silly boo-boo which got propogated through the whole set. The browser doesn't complain, but I like the tags to be properly balanced, you know, small stuff. That's another reason I'd like to use XML...
The perl scripts work fine in my home directory, so I thought I'd copy then over to the Apache directory and just work on the files there, but surprise, surprise!
The perl interpreter does not execute my scripts.
I have given myself all the proper permissions, there's no doubt it. I have no idea what could be wrong, none whatsoever. (I hate to always think "bug", because it could be my fault.)
I invoke a script from the current directory like this:
$ ./tripod *.html
The command line expands the arguments, starts the interpreter -- does not complain -- but does nothing with the input!
No error messages, no output, nothing. It just stops.
I can't figure it, so I just work in my home directory, where the same scripts work fine.
That's why I'm concerned about the Java RTE. I know, it should work, it probably does, but I don't have the time for surprises, so I'd prefer not using it, if I can.
It seems that Java is the preferred platform for XML tools. To me it's just C++ with garbage collection, but none of these tools are long-running, so C would be just fine.
As far as I know, Java has no built in pattern-matching, like Perl. I can write a quick and dirty parser in Perl without building a scanner, which is the real pain-in-the-ass with C, and must be dealt with in Java too.
In addition to the superbly written 3WC recommendations, it would be nice if they also posted also working models which excersized all the features they were writing about, no bells and whistles, just statically bound C programs which worked as canonical filters, no surprises.
Then I could read the litterature and experiment with the model at the same time.
The style of these documents is surely written for the gods, not mere-mortals. I'm sure they must be clear to the people writing them, but to me, they're simply exhausting.
That's why I appreciate your input.