2 ** Source tree organization
4 ******************************************************
6 There is 4 projects in the tree:
8 - GROS: no fancy name yet (low-level toolbox: logging, datatypes).
9 - GRAS: Grid Reality And Simulation (message passing API with two
10 implementations allowing to compile programs on top of the
11 simulator or for the real life without code modification)
12 - SURF: Server for the Use of Resource Fictions (the simulator used
14 - AMOK: Advanced Metacomputing Overlay Kit (high level toolbox; Grid
15 application elements such as distributed database, topology
16 discovery service, and so on)
18 They are all in the same tree because GRAS depends on SURF which depends on
19 GRAS (that's the only cycle here, we're not *that* vicious).
21 The tree is not splited on projects, but on file finality:
22 include/ -> all *public* headers
23 include/gros/*.h -> one file per module
24 include/gros.h -> file including all modules headers
25 (same for gras, surf and amok instead of gros)
27 src/Makefile.am -> main makefile. All projects should fit in only one
28 library (I mean 2, RL+SG), which is compiled here.
30 Since all object.o files are placed here, you should
31 choose the name of c files carfully to avoid
34 src/gras/DataDesc -> typical project module
35 src/gras/DataDesc/datadesc_interface.h -> visible to any GRAS modules;
36 masked to the user and GROS/AMOK/SURF
37 src/gras/DataDesc/datadesc_private.h -> visible only from this module
39 So, the modules have 3 levels of publicity for their interface.
40 Private, internal to GRAS, public. Of course, I try to keep as much stuff
43 testsuite/ -> The more test the better.
44 Same organization than src/ and include/
45 Tests are allowed to load some headers of the module they test.
46 All tests should be listed in run_test.in so that they get
48 They are not listed directly in the check_PROGRAMS part of
49 the makefile because run_test knows about the gras logging
50 features and relaunch with full details the failed tests.
52 examples/ -> Supposed to be copy/pastable by the user, so keep it clear and
53 avoid any kind of trick. In particular, do only include the
57 ** Type naming standard
59 *****************************************************
61 It may sound strange, but the type naming convention was source of intense
62 discution between da GRAS posse members. The convention we came to may not
63 be the best solution, but it has the merit to exist and leave everyone work.
64 So please stick to it.
66 - ???_t is a valid type (builded with typedef)
67 - s_toto_t is a structure (access to fields with .)
68 - s_toto is a structure needing 'struct' keyword to be used
70 - u_toto_t is an union
71 - u_toto is an union needing 'union' keyword to be used
72 - toto_t is an 'object' (struct*)
74 Please to not call toto_t something else than an 'object' (ie, something you
75 have to call _new and _free on it).
78 typedef struct s_toto {} s_toto_t, *toto_t;
79 typedef enum {} e_toto_t;
81 Moreover, only toto_t (and e_toto_t) are public. The rest (mainly s_toto_t)
84 If you see any part of the code not following this convention, this is a
85 bug. Please report it (or fix it yourself if you can).
88 ** Random bits about coding standards and portability
90 *****************************************************
93 You must cast the result of malloc on AIX.
94 It's even better to use gras_new when possible.
97 If possible, avoid size_t and use unsigned long instead. If not,
99 #include <sys/types.h> in all files manipulating size_t
100 do cast it to unsigned long before printing (and use %lu)
102 PRINTF pointer difference
103 printf ("diff = %ld\n", (long) (pointer2 - pointer1));