[Omake] Recursive invocation of omake

[Aleksey Nogin]

On 07.06.2006 07:12, Benjamin Pierce wrote:

> One specific question that I'm wondering about is when it is good to  
> invoke omake recursively.

Not only it's considered bad, but normally it's impossible. OMake will 
normally lock the project when it's running and will not allow other 
instances of OMake to interfere with the project. In general, many parts 
of OMake are optimized towards the notion that a single OMake process 
has a complete view of the whole project and the whole-project 
compilation (while still being able to pick and build a specific target 
from the big project) is one of the basic design goals of OMake.

> In my current tangle of (gnu) makefiles, when I do 'make X' in the  
> top-level directory (where X is 'all', 'clean', etc.), it does a  little 
> bit of work there and then does '$(MAKE) -C DIR X' for every  
> sub-directory DIR; the same process then repeats for sub-sub- 
> directories, and so on.
> 
> Is this a good way to set things up with omake too?  Or is there  
> another way that uses just a single omake process?

In OMake, there is a ".SUBDIRS: foo bar ..." syntax (in a for of a 
pseudo-rule without a body) that tells OMake to incorporate the 
subprojects from subdirectories foo and bar. Namely, it tells OMake to 
read the foo/OMakefile and interpret everything it finds there relative 
to directory foo, then read in bar/OMakefile and interpret it relative 
to bar.

BTW, note that normally OMake will attempt to read all the project's 
OMakefiles before executing anything.

P.S. ".SUBDIRS" also has a form with a body, but this more complicated 
form is only needed if you want to process a number of subdirectories in 
a similar manner and avoid having a separate OMakefile in each of them.

-- 
Aleksey Nogin

Home Page: http://nogin.org/
E-Mail: nogin at cs.caltech.edu (office), aleksey at nogin.org (personal)
Office: Moore 04, tel: (626) 395-2200