I took a break from doing useful things, and made this:

http://imgur.com/Bro0DYT

Infer what you will from the frequencies of certain keywords and identifiers.

On Saturday, 17 August 2013 at 05:57:08 UTC, Brian Schott wrote:
> I took a break from doing useful things, and made this:
>
> http://imgur.com/Bro0DYT
>
> Infer what you will from the frequencies of certain keywords and identifiers.

Cool. What is the input data? Phobos?

I find it *awesome* that assert is one of the most used keywords.

Am 17.08.2013 07:57, schrieb Brian Schott:
> I took a break from doing useful things, and made this:
>
> http://imgur.com/Bro0DYT
>
> Infer what you will from the frequencies of certain keywords and
> identifiers.

Cool idea! But even considering that std.datetime is a monster module, can it really be true that "Date" is as frequent as "assert" throughout the code base? It somehow looks like _only_ std.datetime made it into the final statistic.

On Saturday, August 17, 2013 09:25:41 Sönke Ludwig wrote:
> Am 17.08.2013 07:57, schrieb Brian Schott:
> > I took a break from doing useful things, and made this:
> > 
> > http://imgur.com/Bro0DYT
> > 
> > Infer what you will from the frequencies of certain keywords and identifiers.
> 
> Cool idea! But even considering that std.datetime is a monster module, can it really be true that "Date" is as frequent as "assert" throughout the code base? It somehow looks like _only_ std.datetime made it into the final statistic.

It's due to how many unit tests std.datetime has. There are a _lot_ of assertions in there. I reworked some of them to loop more rather than just be a straight set of tests, but I still have a long way to go, so assert and assertPred are used a _lot_ in that module (and even after that refactoring is done, there will still be plenty of assertions - just not quite so many). If the unit tests weren't counted, std.datetime would probably be a much, much smaller portion of the count (something like 3/4 of that module is unit tests).

- Jonathan M Davis

On Saturday, 17 August 2013 at 07:25:53 UTC, Sönke Ludwig wrote:
> Am 17.08.2013 07:57, schrieb Brian Schott:
>> I took a break from doing useful things, and made this:
>>
>> http://imgur.com/Bro0DYT
>>
>> Infer what you will from the frequencies of certain keywords and
>> identifiers.
>
> Cool idea! But even considering that std.datetime is a monster module, can it really be true that "Date" is as frequent as "assert" throughout the code base? It somehow looks like _only_ std.datetime made it into the final statistic.

I ran through my code again, increasing the number of words that made it to the final list by quite a bit. (For a minute or two I thought you were right and that I had a bug)

This is actually all of Phobos. "EmailStatusCode" and CodePointSet" aren't in datetime.