The purpose of the auto-documentation procedure we are using with the LAL is to adhere the common wisdom: ``keep the documentation close to the code''. The system is built around a a code parser laldoc that allows LAL programmers automatically extract fragments of code or comments from the source files and include them in their documentation. This is done in such a way that if the fragment in the source code is modified, then the change is automatically incorporated into the documentation the next time the document is built.
This extraction is accomplished by having the programmer surround the
fragments of code or comments he or she wishes to incorporate in the
document with key-words. Currently, we only have three pairs of
key-words, so the learning curve is flat and short! The source code
(the .c and .h files) is then parsed and the fragments written to
storage files. (The key-word also includes a space for user-specified file name
where the fragment will be stored.) When you write your documentation simply
use the LATEX command \input{} to put the contents of the
storage file in the document where you need it.
When you install the LAL on your machine, the parser is automatically run on all .c and .h files. The extracted files are stuffed into the directory lal/doc/autodoc/. If you have installed the LAL software package, take a look at the contents of lal/doc/autodoc/. Their are dozens of extracted files their.