Add lines of code statistics

I need this hack. Thanks.

Plan to work on this if more people want it.

Replying to khundeen:

Plan to work on this if more people want it.

Trac definately needs this. :)

+ vote. will be usefull for all the dev. team member to evaluate their performance. :)

a must-have feature for trac

+1

A must-have feature also. But must also work with other VCS systems not just SubVersion?! --JamesMills

That would be a really great addition to Trac.

I see this feature like this: When you "browse source" in trac there's an additional button "LOC stats" displayed next to the directory or file.

• If you press it next to the file, you'll see a plot of LOC over time for that file.
• If you press it next to some directory, you'll see a plot of LOC over time for the whole directory (sum of LOC of all files).
• It might be useful to jump there from the changeset entries in the timeline

me too :)

I vote for this

Needs it we do, yeees...

Why does Lines of code == coder performance?

byte[] buffer;
buffer = byte[]{48, 65, 6C, 6C, 6F, 20, 57, 6F, 72, 6C, 64};
String s = new String(buffer);
OutputStream out = System.out;
out.print(s);
out.print("\n");

vs

System.out.println("Hello World");

Top version = 6x as performing than the bottom one, and thats not even with comments.

would also love this

LOC alone is a pretty useless stat. I wouldn't mind some "code statistics" not just LOC, but %Lines comments, largest file. file with largest "scoped object" i.e. function, method, whatever your language offers. complexity...all that.

anyway, if my idea of the plugin is the idea, then the plugin should just read/display, and define a format for the data to adhere too, and let external tools, or other plugins (or sub parts) do the work. (i.e., I don't want the # of classes in Java crap, it has bazillions, and my c code has none.).

Well, that's my take, maybe it's a different plugin.

+1

I agree that LOC in no very meaningful, but at least it gives some insight if a developer changed 1 or 100 lines of code over a certain time period.

LOC stats would be very very useful. All coders should know how efficient code they write so the "Hello World" example above is pretty irrelevant. Even if you're a team leader you should have guidelines and control over your team. And if not you would be able to monitor the lines of code are written by each member of your team over time, so would be able to pick up on coders with LOC above the average (either they work too hard (yeah right) or they write inefficient code).

When can we have this hack then? :)

Have a look at statSVN, this does most of this already, but it would be good to have it integrated into trac.

I would be careful with LOC metrics as a productiviy measure, as others have highlighted. I would also suggest that for simplicity a simple physical lines of code count is done (i.e. none comment, none blank lines) as that can be made to work for a large number of languages. More detailed lines of code, such as a logical LOC are difficult to develop and are language specific. My company has Ada, Java, C, C++, C.NET, ASP and assembler in use as a mimimum!