3 @page CodingSuggestions Coding suggestions
5 So you want to add code to Samba ...
7 One of the daunting tasks facing a programmer attempting to write code for
8 Samba is understanding the various coding conventions used by those most
9 active in the project. These conventions were mostly unwritten and helped
10 improve either the portability, stability or consistency of the code. This
11 document will attempt to document a few of the more important coding
12 practices used at this time on the Samba project. The coding practices are
13 expected to change slightly over time, and even to grow as more is learned
14 about obscure portability considerations. Two existing documents
15 samba/source/internals.doc and samba/source/architecture.doc provide
16 additional information.
18 The loosely related question of coding style is very personal and this
19 document does not attempt to address that subject, except to say that I
20 have observed that eight character tabs seem to be preferred in Samba
21 source. If you are interested in the topic of coding style, two oft-quoted
24 http://lxr.linux.no/source/Documentation/CodingStyle
25 http://www.fsf.org/prep/standards_toc.html
27 but note that coding style in Samba varies due to the many different
28 programmers who have contributed.
30 Following are some considerations you should use when adding new code to
31 Samba. First and foremost remember that:
33 Portability is a primary consideration in adding function, as is network
34 compatability with de facto, existing, real world CIFS/SMB implementations.
35 There are lots of platforms that Samba builds on so use caution when adding
36 a call to a library function that is not invoked in existing Samba code.
37 Also note that there are many quite different SMB/CIFS clients that Samba
38 tries to support, not all of which follow the SNIA CIFS Technical Reference
39 (or the earlier Microsoft reference documents or the X/Open book on the SMB
42 Here are some other suggestions:
44 1) use d_printf instead of printf for display text
45 reason: enable auto-substitution of translated language text
47 2) use SAFE_FREE instead of free
48 reason: reduce traps due to null pointers
50 3) don't use bzero use memset, or ZERO_STRUCT and ZERO_STRUCTP macros
53 4) don't use strcpy and strlen (use safe_* equivalents)
54 reason: to avoid traps due to buffer overruns
56 5) don't use getopt_long, use popt functions instead
59 6) explicitly add const qualifiers on parm passing in functions where parm
60 is input only (somewhat controversial but const can be #defined away)
62 7) when passing a va_list as an arg, or assigning one to another
63 please use the VA_COPY() macro
64 reason: on some platforms, va_list is a struct that must be
65 initialized in each function...can SEGV if you don't.
67 8) discourage use of threads
68 reason: portability (also see architecture.doc)
70 9) don't explicitly include new header files in C files - new h files
71 should be included by adding them once to includes.h
74 10) don't explicitly extern functions (they are autogenerated by
75 "make proto" into proto.h)
78 11) use endian safe macros when unpacking SMBs (see byteorder.h and
80 reason: not everyone uses Intel
82 12) Note Unicode implications of charset handling (see internals.doc). See
83 pull_* and push_* and convert_string functions.
84 reason: Internationalization
86 13) Don't assume English only
89 14) Try to avoid using in/out parameters (functions that return data which
90 overwrites input parameters)
91 reason: Can cause stability problems
93 15) Ensure copyright notices are correct, don't append Tridge's name to code
94 that he didn't write. If you did not write the code, make sure that it
95 can coexist with the rest of the Samba GPLed code.
97 16) Consider usage of DATA_BLOBs for length specified byte-data.
100 17) Take advantage of tdbs for database like function
103 18) Don't access the SAM_ACCOUNT structure directly, they should be accessed
104 via pdb_get...() and pdb_set...() functions.
105 reason: stability, consistency
107 19) Don't check a password directly against the passdb, always use the
108 check_password() interface.
109 reason: long term pluggability
111 20) Try to use asprintf rather than pstrings and fstrings where possible
113 21) Use normal C comments / * instead of C++ comments // like
114 this. Although the C++ comment format is part of the C99
115 standard, some older vendor C compilers do not accept it.
117 22) Try to write documentation for API functions and structures
118 explaining the point of the code, the way it should be used, and
119 any special conditions or results. Mark these with a double-star
120 comment start / ** so that they can be picked up by Doxygen, as in
123 23) Keep the scope narrow. This means making functions/variables
124 static whenever possible. We don't want our namespace
125 polluted. Each module should have a minimal number of externally
126 visible functions or variables.
128 24) Use function pointers to keep knowledge about particular pieces of
129 code isolated in one place. We don't want a particular piece of
130 functionality to be spread out across lots of places - that makes
131 for fragile, hand to maintain code. Instead, design an interface
132 and use tables containing function pointers to implement specific
133 functionality. This is particularly important for command
136 25) Think carefully about what it will be like for someone else to add
137 to and maintain your code. If it would be hard for someone else to
138 maintain then do it another way.
140 The suggestions above are simply that, suggestions, but the information may
141 help in reducing the routine rework done on new code. The preceeding list
142 is expected to change routinely as new support routines and macros are
145 Written by Steve French, with contributions from Simo Sorce, Andrew
146 Bartlett, Tim Potter and Martin Pool.