Subversion committer statistics

I am sure there are simpler ways to do it. If you know them.

Here is a Perl one-liner that will count the number of commits in a Subversion repository and print out the statistics.

svn log | perl -MData::Dumper -n -e'if (/^r\d+\s*\|\s*(\w+)\s*\|/) {$c{$1}++} END {print Dumper \%c}'

and the result for the Padre Subversion repository.

$VAR1 = { 'tsee' => 19, 'adamk' => 86, 'patspam' => 6, 'bricas' => 2, 'szabgab' => 450 };

To have a quick explanation:

svn log

Is just to get all the logs from Subversion then we pipe it (using |) to the perl one-liner.

-MData::Dumper tells perl to load the Data::Dumper module which provides us the Dumper function used later.

-n Basically means to split up the input stream into lines and feed one by one to the perl code.

-e'' tells perl that instead of running a script in a file, expect the Perl code between the quotation marks.

if (/^r\d+\s*\|\s*(\w+)\s*\|/) { $c{$1}++ }

The above means: Check the current line (that is hidden away in a variable called $_ but we don't really care). We use a regular expression to match those lines that start with and r (^r), followed by some digits (\d+), followed by optional spaces (\s*), followed by a vertical bar (\|), followed by more optional spaces (\s*), followed by some letters, numbers (\w+), followed by more optional spaces (\s*) and followed by another vertical bar (\|).

This might be a bit long explanation, but this is how we can describe the beginning of those lines that look like the following:

r638 | szabgab | 2008-11-01 18:06:45 +0200 (Sat, 01 Nov 2008) | 1 line

If I was writing this in a file I probably would write it a bit more clearer by using the /x modifier like this:

if (/^r\d+ \s* \| \s* (\w+) \s* \|/x) { $c{$1}++ };

As you can see the \w* is within parentheses which means the string it captures will magically end up in the $1 variable. So for the above input line $1 will hold "szabgab". I use that as a key to the %c hash incrementing the value of $c{"szabgab"} by one.

The remaining part, the END block is a special statement that will only execute after all the rest of the perl script ran, so it will only execute once after all the lines were processed and it simply generates the out you can see above inserting an arbitrary $VAR1 at the beginning.

END {print Dumper \%c}

If this is the kind of things you might need to do, you don't need to learn the full Perl language so Minimal Perl by Tim Maher might be a good book for you.