Tips and tricks
less
Why are you using editors to just look at a (large) file?
Under *nix or Cygwin, just use less. (There is a famous saying – "less is more, more or less" – because "less" replaced the earlier Unix command "more", with the addition that you could scroll back up.) Searching and navigating under less is very similar to Vim, but there is no swap file and little RAM used.
There is a Win32 port of GNU less. See the "less" section of the answer above.
Perl
Perl is good for quick scripts, and its .. (range flip-flop) operator makes for a nice selection mechanism to limit the crud you have to wade through.
For example:
$ perl -n -e 'print if ( 1000000 .. 2000000)' humongo.txt | less
This will extract everything from line 1 million to line 2 million, and allow you to sift the output manually in less.
Another example:
$ perl -n -e 'print if ( /regex one/ .. /regex two/)' humongo.txt | less
This starts printing when the "regular expression one" finds something, and stops when the "regular expression two" find the end of an interesting block. It may find multiple blocks. Sift the output...
logparser
This is another useful tool you can use. To quote the Wikipedia article:
logparser is a flexible command line utility that was initially written by Gabriele Giuseppini, a Microsoft employee, to automate tests for IIS logging. It was intended for use with the Windows operating system, and was included with the IIS 6.0 Resource Kit Tools. The default behavior of logparser works like a "data processing pipeline", by taking an SQL expression on the command line, and outputting the lines containing matches for the SQL expression.
Microsoft describes Logparser as a powerful, versatile tool that provides universal query access to text-based data such as log files, XML files and CSV files, as well as key data sources on the Windows operating system such as the Event Log, the Registry, the file system, and Active Directory. The results of the input query can be custom-formatted in text based output, or they can be persisted to more specialty targets like SQL, SYSLOG, or a chart.
Example usage:
C:\>logparser.exe -i:textline -o:tsv "select Index, Text from 'c:\path\to\file.log' where line > 1000 and line < 2000"
C:\>logparser.exe -i:textline -o:tsv "select Index, Text from 'c:\path\to\file.log' where line like '%pattern%'"
The relativity of sizes
100 MB isn't too big. 3 GB is getting kind of big. I used to work at a print & mail facility that created about 2% of U.S. first class mail. One of the systems for which I was the tech lead accounted for about 15+% of the pieces of mail. We had some big files to debug here and there.
And more...
Feel free to add more tools and information here. This answer is community wiki for a reason! We all need more advice on dealing with large amounts of data...
