[ccan] [PATCH 4/4] agar: Re-entrant Abstract Graph Algorithms

[Gianluca Della Vedova]

-- 
Gianluca Della Vedova
http://gianluca.dellavedova.org
On 11/07/2014 at 15:31, David Gibson wrote:

>On Fri, Jul 11, 2014 at 12:18:04PM +0200, Gianluca Della Vedova wrote:
>> I have two newbie questions:
>> 1. what are the pros of the aga(r) module vs igraph (http://igraph.org/c/)?
>
>So, there are an awful lot of common problems which can be regarded as
>instances of a BFS, DFS or Dijkstra's, if you think in terms of the
>right graph.  But, usually that graph is never explicitly constructed
>in your program - it's an implicit set of possible states and
>transitions which can be derived from the data model you do have.
>
>Although it's certainly possible to explicitly construct a graph data
>structure from whatever you have internally, using igraph or even the
>existing CCAN dgraph module, it's often clumsy and inconvenient to do
>so.
>
>The idea of the aga(r) modules is that you need minimal graph related
>data structures beyond what you already have in  your code.  The graph
>structure is derived on the fly via your callbacks and used to drive
>the graph algorithms.
>
>In short, aga is suppoed to make it easy to use simple graph
>algorithms on real world problems (I don't, rather than to be a tool
>to explore graph theory itself.

That's great, thanks.