Yes, it would be quite "challenging" to deduce it from C/C++ code - and likely highly error-prone. That's why I asked would it be worth the trouble. :)
At first I was going to say it would be nuts, but I've been looking at our API code and, for my company at least, it may be possible... for us. But we followed a (fairly) consistent pattern. In fact, so consistent a pattern I'm now debating whether we should have not written much C/C++ code to begin with, and instead done it as XML and generated the code through an XSLT script.
But anyway... yes having the doc languages be the same would be nice, but it's not that hard for folks to just write a script to convert their particular format into
the Komeki LDT one. (Assuming they wrote their comments in some consistent format to begin with, like Doxygen format for example)
-hadriel
From: Fabien Fleutot <fleutot+koneki@xxxxxxxxx>
To: Hadriel Kaplan <hadrielk@xxxxxxxxx>
Cc: General development dicussions for the Koneki project <koneki-dev@xxxxxxxxxxx>
Sent: Thursday, March 1, 2012 12:44 PM
Subject: Re: [koneki-dev] The Record-block/Type-block
On Thu, Mar 1, 2012 at 5:53 PM, Hadriel Kaplan
<hadrielk@xxxxxxxxx> wrote:
As for the inferring of type information from code - assuming you mean from C/C++ code, is it really worth it?
I was talking about Lua code. Parsing C++ code to guess what it does to an embedded scripting language would be... challenging. However, the doc languages in C++ and Lua comments sure should be the same.
