[Omake] Case-insensitivity and cons-hashing.

[Jason Hickey]

This seems like a good solution.  Note that, when an explicit rule is  
created, it is saved in the rule database, and the rule refers to the  
saved nodes (with the case at the time the rule was created).  This  
may be an issue in particular with active rules.

    %.a:
        section rule
           %.a: %.b
              touch $@

    touch foo.b

    target-exists(Foo.a)

    .DEFAULT: foo.a

My guess is that this will build Foo.a, but I'm not sure.

Some stylistic questions:
    - Why is it called MakeHashMarshalEq.WeakCompare? -- it uses the  
strong_compare function...
    - Is it required that strong_compare be finer than compare?  If  
they are unrelated, then
      perhaps internal_compare or hash_compare is a better name than  
strong_compare.

Jason

On May 3, 2007, at 11:43 AM, Aleksey Nogin wrote:

> I tried addressing this by introducing the notion of "two levels of  
> equality" to the cons-hashing of nodes. Basically, for the purposes  
> of NodeSet, NodeTable, etc, the two nodes are equal if the  
> canonical forms are equal. However, for the purposes of the cons- 
> hashing, one node can only be substituted for another when they are  
> identical...
>
> What do you think?
>
> Aleksey
>
> -- 
> Aleksey Nogin, Research Staff Member
> Advanced Technologies Department, Information & System Sciences Lab
> HRL Laboratories, LLC, Malibu, CA
> _______________________________________________
> OMake-Devel mailing list
> OMake-Devel at metaprl.org
> https://lists.metaprl.org/mailman/listinfo/omake-devel

--
Jason Hickey                  http://www.cs.caltech.edu/~jyh
Caltech Computer Science      Tel: 626-395-6568 FAX: 626-792-4257