Hi,

I'm trying to extract information from text logs, the task is a bit non-trivial though.
First of all, I'd like to do it a portable way. Personally I'm in the linux world, but I'd like to share the script with a community which I know to largely consist of Windows users.

So here's an idea of what the raw logs look like:
The parsing itself can be broken down into the following points:

1. getting rid of some useless information, eg deleting all the lines matching certain regex
2. counting all the class2 items with similar names and writing one line that includes the name and number of occurrences
3. sorting according to classes and names
4. this is where it gets a bit challenging: getting rid of 'static' data in class1

The last point requires some further explanation:
I want to match entries from class1 with some .lua configuration files. There are like 10 files and the every class1 entry has a small configuration matching its name. Now I need to compare the data of Name and keep only the lines containing numeric values in the range specified in the configuration. Here's a snippet of a .lua config:

Unfortunately foo and bar are not found in the config, i just have foobar there, so the matching has to be done according to the value ranges, these are always under the form '[0-9]\{2,3\}-[0-9]\{2,3\}'.

The output of the log above would then look like this:
So, how do I get started? As I mentioned earlier I'd prefer not having to rely on typical linux tools like sed. Maybe Python could do the job? If so, please give me some hints what to do/where to look.

thanks for reading and any help!

There is, of course, Perl for Linux and there is Strawberry Perl for Windows. The beauty of the latter Perl is that it also comes as a self-sufficient tree which can be put into any directory and run from there, i.e. one doesn't even need to have Administrator privileges under Windows to be able to run Strawberry Perl.

Look for Portable Strawberry Perl in order to get the version not requiring Administrator privileges/installation.

yes of course. Its portable as well.

first you should learn about it if you have not already. see my sig


As for your questions.
Python has very good string manipulation capabilities, so normally, regex is not needed. for example to delete all line matching certain pattern, in the simplest case,

in the above, you can use fileinput module to do "inplace" editing. (much like sed's -i option). change to "1" to take effect.

there are data structures you can use , eg lists, dictionaries to enable you to count stuff. for example, once you put your stuffs into lists, you can use len() function to count. eg
Python has sort(), sorted() functions to do this

these are just tip of the iceberg, to get the real deal, head down to the Python doc site and learn about it.