Reasons for C: 1. The log files I am working are 60-100MB files. I *assumed* using C will reduce the parse time. 2. The records themselves are variable length and hence was concerned about the complexity for implementation in python. 3. Since I am not using a database, each request to refresh the graph (changing time duration for display) will have to re-parse the file all over again. And hence speed in parsing the log is important.
Hope that explains thanks, Ravi. On Thu, Dec 11, 2008 at 10:45 AM, Steve Willoughby <st...@alchemy.com>wrote: > Ravi Kondamuru wrote: > >> I am trying to read a binary log file to extract system counters. These >> counters will then be used to generate web-based graphs using the >> chart-director api in python. For extracting the counters I planning to >> write a program in C to read and give the output as lists for use by >> chart-director. If possible i would like to do the parsing of data only once >> in C on the log file and pass the processed output for direct use by python. >> >> > > Wait.. > Why not have Python read the binary log file itself, parse it and make the > API calls right there? Why do you even need the C program? > > > >> I have thought about having to populate a database first but will prefer >> to avoid having another intermediate datastore (apart from the log file). >> >> thanks, >> Ravi. >> >> On Thu, Dec 11, 2008 at 10:27 AM, Steve Willoughby <st...@alchemy.com<mailto: >> st...@alchemy.com>> wrote: >> >> Ravi Kondamuru wrote: >> >> Hi, >> I am writing a script to read list output from a C executable. >> How should c program be written so that python can read the >> output as a list? >> Any pointers to info on this appreciated. >> >> >> The possibilities are truly wide open on this. Python can read a >> variety of standard formats (and of course can have custom code to >> read anything. Depending on what the data involved actually are, >> you need to decide what format works best. >> >> A simple approach is to have the C program write simple CSV output, >> and use Python's csv module to read it. A more complex solution >> might be to use XML. There are approximately 52,495,102 other >> possibilities available too, so you have lots of room to work out >> what's best for your application. More specific information on >> what you're trying to accomplish would help narrow it down as well. >> >> >> >
_______________________________________________ Tutor maillist - Tutor@python.org http://mail.python.org/mailman/listinfo/tutor
