| blob | f47412042530a9324d0ce7313b8664484dd0dde1 |
1 Like other projects, we also have some guidelines for our code. For
2 Git in general, a few rough rules are:
4 - Most importantly, we never say "It's in POSIX; we'll happily
5 ignore your needs should your system not conform to it."
6 We live in the real world.
8 - However, we often say "Let's stay away from that construct,
9 it's not even in POSIX".
11 - In spite of the above two rules, we sometimes say "Although
12 this is not in POSIX, it (is so convenient | makes the code
13 much more readable | has other good characteristics) and
14 practically all the platforms we care about support it, so
15 let's use it".
17 Again, we live in the real world, and it is sometimes a
18 judgement call, the decision based more on real world
19 constraints people face than what the paper standard says.
21 - Fixing style violations while working on a real change as a
22 preparatory clean-up step is good, but otherwise avoid useless code
23 churn for the sake of conforming to the style.
25 "Once it _is_ in the tree, it's not really worth the patch noise to
26 go and fix it up."
27 Cf. https://lore.kernel.org/all/20100126160632.3bdbe172.akpm@linux-foundation.org/
29 - Log messages to explain your changes are as important as the
30 changes themselves. Clearly written code and in-code comments
31 explain how the code works and what is assumed from the surrounding
32 context. The log messages explain what the changes wanted to
33 achieve and why the changes were necessary (more on this in the
34 accompanying SubmittingPatches document).
36 Make your code readable and sensible, and don't try to be clever.
38 As for more concrete guidelines, just imitate the existing code
39 (this is a good guideline, no matter which project you are
40 contributing to). It is always preferable to match the _local_
41 convention. New code added to Git suite is expected to match
42 the overall style of existing code. Modifications to existing
43 code are expected to match the style the surrounding code already
44 uses (even if it doesn't match the overall style of existing code).
46 But if you must have a list of rules, here are some language
47 specific ones. Note that Documentation/ToolsForGit.adoc document
48 has a collection of tips to help you use some external tools
49 to conform to these guidelines.
51 For shell scripts specifically (not exhaustive):
53 - We use tabs for indentation.
55 - Case arms are indented at the same depth as case and esac lines,
56 like this:
58 case "$variable" in
59 pattern1)
60 do this
61 ;;
62 pattern2)
63 do that
64 ;;
65 esac
67 - Redirection operators should be written with space before, but no
68 space after them. In other words, write 'echo test >"$file"'
69 instead of 'echo test> $file' or 'echo test > $file'. Note that
70 even though it is not required by POSIX to double-quote the
71 redirection target in a variable (as shown above), our code does so
72 because some versions of bash issue a warning without the quotes.
74 (incorrect)
75 cat hello > world < universe
76 echo hello >$world
78 (correct)
79 cat hello >world <universe
80 echo hello >"$world"
82 - We prefer $( ... ) for command substitution; unlike ``, it
83 properly nests. It should have been the way Bourne spelled
84 it from day one, but unfortunately isn't.
86 - If you want to find out if a command is available on the user's
87 $PATH, you should use 'type <command>', instead of 'which <command>'.
88 The output of 'which' is not machine parsable and its exit code
89 is not reliable across platforms.
91 - We use POSIX compliant parameter substitutions and avoid bashisms;
92 namely:
94 - We use ${parameter-word} and its [-=?+] siblings, and their
95 colon'ed "unset or null" form.
97 - We use ${parameter#word} and its [#%] siblings, and their
98 doubled "longest matching" form.
100 - No "Substring Expansion" ${parameter:offset:length}.
102 - No shell arrays.
104 - No pattern replacement ${parameter/pattern/string}.
106 - We use Arithmetic Expansion $(( ... )).
108 - We do not use Process Substitution <(list) or >(list).
110 - Do not write control structures on a single line with semicolon.
111 "then" should be on the next line for if statements, and "do"
112 should be on the next line for "while" and "for".
114 (incorrect)
115 if test -f hello; then
116 do this
117 fi
119 (correct)
120 if test -f hello
121 then
122 do this
123 fi
125 - If a command sequence joined with && or || or | spans multiple
126 lines, put each command on a separate line and put && and || and |
127 operators at the end of each line, rather than the start. This
128 means you don't need to use \ to join lines, since the above
129 operators imply the sequence isn't finished.
131 (incorrect)
132 grep blob verify_pack_result \
133 | awk -f print_1.awk \
134 | sort >actual &&
135 ...
137 (correct)
138 grep blob verify_pack_result |
139 awk -f print_1.awk |
140 sort >actual &&
141 ...
143 - We prefer "test" over "[ ... ]".
145 - We do not write the noiseword "function" in front of shell
146 functions.
148 - We prefer a space between the function name and the parentheses,
149 and no space inside the parentheses. The opening "{" should also
150 be on the same line.
152 (incorrect)
153 my_function(){
154 ...
156 (correct)
157 my_function () {
158 ...
160 - As to use of grep, stick to a subset of BRE (namely, no \{m,n\},
161 [::], [==], or [..]) for portability.
163 - We do not use \{m,n\};
165 - We do not use ? or + (which are \{0,1\} and \{1,\}
166 respectively in BRE) but that goes without saying as these
167 are ERE elements not BRE (note that \? and \+ are not even part
168 of BRE -- making them accessible from BRE is a GNU extension).
170 - Use Git's gettext wrappers in git-sh-i18n to make the user
171 interface translatable. See "Marking strings for translation" in
172 po/README.
174 - We do not write our "test" command with "-a" and "-o" and use "&&"
175 or "||" to concatenate multiple "test" commands instead, because
176 the use of "-a/-o" is often error-prone. E.g.
178 test -n "$x" -a "$a" = "$b"
180 is buggy and breaks when $x is "=", but
182 test -n "$x" && test "$a" = "$b"
184 does not have such a problem.
186 - Even though "local" is not part of POSIX, we make heavy use of it
187 in our test suite. We do not use it in scripted Porcelains, and
188 hopefully nobody starts using "local" before all shells that matter
189 support it (notably, ksh from AT&T Research does not support it yet).
191 - Some versions of shell do not understand "export variable=value",
192 so we write "variable=value" and then "export variable" on two
193 separate lines.
195 - Some versions of dash have broken variable assignment when prefixed
196 with "local", "export", and "readonly", in that the value to be
197 assigned goes through field splitting at $IFS unless quoted.
199 (incorrect)
200 local variable=$value
201 local variable=$(command args)
203 (correct)
204 local variable="$value"
205 local variable="$(command args)"
207 - The common construct
209 VAR=VAL command args
211 to temporarily set and export environment variable VAR only while
212 "command args" is running is handy, but this triggers an
213 unspecified behaviour according to POSIX when used for a command
214 that is not an external command (like shell functions). Indeed,
215 dash 0.5.10.2-6 on Ubuntu 20.04, /bin/sh on FreeBSD 13, and AT&T
216 ksh all make a temporary assignment without exporting the variable,
217 in such a case. As it does not work portably across shells, do not
218 use this syntax for shell functions. A common workaround is to do
219 an explicit export in a subshell, like so:
221 (incorrect)
222 VAR=VAL func args
224 (correct)
225 (
226 VAR=VAL &&
227 export VAR &&
228 func args
229 )
231 but be careful that the effect "func" makes to the variables in the
232 current shell will be lost across the subshell boundary.
234 - Use octal escape sequences (e.g. "\302\242"), not hexadecimal (e.g.
235 "\xc2\xa2") in printf format strings, since hexadecimal escape
236 sequences are not portable.
239 For C programs:
241 - We use tabs to indent, and interpret tabs as taking up to
242 8 spaces.
244 - Nested C preprocessor directives are indented after the hash by one
245 space per nesting level.
247 #if FOO
248 # include <foo.h>
249 # if BAR
250 # include <bar.h>
251 # endif
252 #endif
254 - We try to keep to at most 80 characters per line.
256 - As a Git developer we assume you have a reasonably modern compiler
257 and we recommend you to enable the DEVELOPER makefile knob to
258 ensure your patch is clear of all compiler warnings we care about,
259 by e.g. "echo DEVELOPER=1 >>config.mak".
261 - When using DEVELOPER=1 mode, you may see warnings from the compiler
262 like "error: unused parameter 'foo' [-Werror=unused-parameter]",
263 which indicates that a function ignores its argument. If the unused
264 parameter can't be removed (e.g., because the function is used as a
265 callback and has to match a certain interface), you can annotate
266 the individual parameters with the UNUSED (or MAYBE_UNUSED)
267 keyword, like "int foo UNUSED".
269 - We try to support a wide range of C compilers to compile Git with,
270 including old ones. As of Git v2.35.0 Git requires C99 (we check
271 "__STDC_VERSION__"). You should not use features from a newer C
272 standard, even if your compiler groks them.
274 New C99 features have been phased in gradually, if something's new
275 in C99 but not used yet don't assume that it's safe to use, some
276 compilers we target have only partial support for it. These are
277 considered safe to use:
279 . since around 2007 with 2b6854c863a, we have been using
280 initializer elements which are not computable at load time. E.g.:
282 const char *args[] = { "constant", variable, NULL };
284 . since early 2012 with e1327023ea, we have been using an enum
285 definition whose last element is followed by a comma. This, like
286 an array initializer that ends with a trailing comma, can be used
287 to reduce the patch noise when adding a new identifier at the end.
289 . since mid 2017 with cbc0f81d, we have been using designated
290 initializers for struct (e.g. "struct t v = { .val = 'a' };").
292 . since mid 2017 with 512f41cf, we have been using designated
293 initializers for array (e.g. "int array[10] = { [5] = 2 }").
295 . since early 2021 with 765dc168882, we have been using variadic
296 macros, mostly for printf-like trace and debug macros.
298 . since late 2021 with 44ba10d6, we have had variables declared in
299 the for loop "for (int i = 0; i < 10; i++)".
301 . since late 2023 with 8277dbe987 we have been using the bool type
302 from <stdbool.h>.
304 C99 features we have test balloons for:
306 . since late 2024 with v2.48.0-rc0~20, we have test balloons for
307 compound literal syntax, e.g., (struct foo){ .member = value };
308 our hope is that no platforms we care about have trouble using
309 them, and officially adopt its wider use in mid 2026. Do not add
310 more use of the syntax until that happens.
312 New C99 features that we cannot use yet:
314 . %z and %zu as a printf() argument for a size_t (the %z being for
315 the POSIX-specific ssize_t). Instead you should use
316 printf("%"PRIuMAX, (uintmax_t)v). These days the MSVC version we
317 rely on supports %z, but the C library used by MinGW does not.
319 . Shorthand like ".a.b = *c" in struct initializations is known to
320 trip up an older IBM XLC version, use ".a = { .b = *c }" instead.
321 See the 33665d98 (reftable: make assignments portable to AIX xlc
322 v12.01, 2022-03-28).
324 - Variables have to be declared at the beginning of the block, before
325 the first statement (i.e. -Wdeclaration-after-statement). It is
326 encouraged to have a blank line between the end of the declarations
327 and the first statement in the block.
329 - Do not explicitly initialize global variables to 0 or NULL;
330 instead, let BSS take care of the zero initialization.
332 - NULL pointers shall be written as NULL, not as 0.
334 - When declaring pointers, the star sides with the variable
335 name, i.e. "char *string", not "char* string" or
336 "char * string". This makes it easier to understand code
337 like "char *string, c;".
339 - Use whitespace around operators and keywords, but not inside
340 parentheses and not around functions. So:
342 while (condition)
343 func(bar + 1);
345 and not:
347 while( condition )
348 func (bar+1);
350 - A binary operator (other than ",") and ternary conditional "?:"
351 have a space on each side of the operator to separate it from its
352 operands. E.g. "A + 1", not "A+1".
354 - A unary operator (other than "." and "->") have no space between it
355 and its operand. E.g. "(char *)ptr", not "(char *) ptr".
357 - Do not explicitly compare an integral value with constant 0 or '\0',
358 or a pointer value with constant NULL. For instance, to validate that
359 counted array <ptr, cnt> is initialized but has no elements, write:
361 if (!ptr || cnt)
362 BUG("empty array expected");
364 and not:
366 if (ptr == NULL || cnt != 0);
367 BUG("empty array expected");
369 - We avoid using braces unnecessarily. I.e.
371 if (bla) {
372 x = 1;
373 }
375 is frowned upon. But there are a few exceptions:
377 - When the statement extends over a few lines (e.g., a while loop
378 with an embedded conditional, or a comment). E.g.:
380 while (foo) {
381 if (x)
382 one();
383 else
384 two();
385 }
387 if (foo) {
388 /*
389 * This one requires some explanation,
390 * so we're better off with braces to make
391 * it obvious that the indentation is correct.
392 */
393 doit();
394 }
396 - When there are multiple arms to a conditional and some of them
397 require braces, enclose even a single line block in braces for
398 consistency. E.g.:
400 if (foo) {
401 doit();
402 } else {
403 one();
404 two();
405 three();
406 }
408 - We try to avoid assignments in the condition of an "if" statement.
410 - Try to make your code understandable. You may put comments
411 in, but comments invariably tend to stale out when the code
412 they were describing changes. Often splitting a function
413 into two makes the intention of the code much clearer.
415 - Multi-line comments include their delimiters on separate lines from
416 the text. E.g.
418 /*
419 * A very long
420 * multi-line comment.
421 */
423 Note however that a comment that explains a translatable string to
424 translators uses a convention of starting with a magic token
425 "TRANSLATORS: ", e.g.
427 /*
428 * TRANSLATORS: here is a comment that explains the string to
429 * be translated, that follows immediately after it.
430 */
431 _("Here is a translatable string explained by the above.");
433 - Double negation is often harder to understand than no negation
434 at all.
436 - There are two schools of thought when it comes to comparison,
437 especially inside a loop. Some people prefer to have the less stable
438 value on the left hand side and the more stable value on the right hand
439 side, e.g. if you have a loop that counts variable i down to the
440 lower bound,
442 while (i > lower_bound) {
443 do something;
444 i--;
445 }
447 Other people prefer to have the textual order of values match the
448 actual order of values in their comparison, so that they can
449 mentally draw a number line from left to right and place these
450 values in order, i.e.
452 while (lower_bound < i) {
453 do something;
454 i--;
455 }
457 Both are valid, and we use both. However, the more "stable" the
458 stable side becomes, the more we tend to prefer the former
459 (comparison with a constant, "i > 0", is an extreme example).
460 Just do not mix styles in the same part of the code and mimic
461 existing styles in the neighbourhood.
463 - There are two schools of thought when it comes to splitting a long
464 logical line into multiple lines. Some people push the second and
465 subsequent lines far enough to the right with tabs and align them:
467 if (the_beginning_of_a_very_long_expression_that_has_to ||
468 span_more_than_a_single_line_of ||
469 the_source_text) {
470 ...
472 while other people prefer to align the second and the subsequent
473 lines with the column immediately inside the opening parenthesis,
474 with tabs and spaces, following our "tabstop is always a multiple
475 of 8" convention:
477 if (the_beginning_of_a_very_long_expression_that_has_to ||
478 span_more_than_a_single_line_of ||
479 the_source_text) {
480 ...
482 Both are valid, and we use both. Again, just do not mix styles in
483 the same part of the code and mimic existing styles in the
484 neighbourhood.
486 - When splitting a long logical line, some people change line before
487 a binary operator, so that the result looks like a parse tree when
488 you turn your head 90-degrees counterclockwise:
490 if (the_beginning_of_a_very_long_expression_that_has_to
491 || span_more_than_a_single_line_of_the_source_text) {
493 while other people prefer to leave the operator at the end of the
494 line:
496 if (the_beginning_of_a_very_long_expression_that_has_to ||
497 span_more_than_a_single_line_of_the_source_text) {
499 Both are valid, but we tend to use the latter more, unless the
500 expression gets fairly complex, in which case the former tends to
501 be easier to read. Again, just do not mix styles in the same part
502 of the code and mimic existing styles in the neighbourhood.
504 - When splitting a long logical line, with everything else being
505 equal, it is preferable to split after the operator at higher
506 level in the parse tree. That is, this is more preferable:
508 if (a_very_long_variable * that_is_used_in +
509 a_very_long_expression) {
510 ...
512 than
514 if (a_very_long_variable *
515 that_is_used_in + a_very_long_expression) {
516 ...
518 - Some clever tricks, like using the !! operator with arithmetic
519 constructs, can be extremely confusing to others. Avoid them,
520 unless there is a compelling reason to use them.
522 - Use the API. No, really. We have a strbuf (variable length
523 string), several arrays with the ALLOC_GROW() macro, a
524 string_list for sorted string lists, a hash map (mapping struct
525 objects) named "struct decorate", amongst other things.
527 - When you come up with an API, document its functions and structures
528 in the header file that exposes the API to its callers. Use what is
529 in "strbuf.h" as a model for the appropriate tone and level of
530 detail.
532 - The first #include in C files, except in platform specific compat/
533 implementations and sha1dc/, must be <git-compat-util.h>. This
534 header file insulates other header files and source files from
535 platform differences, like which system header files must be
536 included in what order, and what C preprocessor feature macros must
537 be defined to trigger certain features we expect out of the system.
538 A collorary to this is that C files should not directly include
539 system header files themselves.
541 There are some exceptions, because certain group of files that
542 implement an API all have to include the same header file that
543 defines the API and it is convenient to include <git-compat-util.h>
544 there. Namely:
546 - the implementation of the built-in commands in the "builtin/"
547 directory that include "builtin.h" for the cmd_foo() prototype
548 definition,
550 - the test helper programs in the "t/helper/" directory that include
551 "t/helper/test-tool.h" for the cmd__foo() prototype definition,
553 - the xdiff implementation in the "xdiff/" directory that includes
554 "xdiff/xinclude.h" for the xdiff machinery internals,
556 - the unit test programs in "t/unit-tests/" directory that include
557 "t/unit-tests/test-lib.h" that gives them the unit-tests
558 framework, and
560 - the source files that implement reftable in the "reftable/"
561 directory that include "reftable/system.h" for the reftable
562 internals,
564 are allowed to assume that they do not have to include
565 <git-compat-util.h> themselves, as it is included as the first
566 '#include' in these header files. These headers must be the first
567 header file to be "#include"d in them, though.
569 - A C file must directly include the header files that declare the
570 functions and the types it uses, except for the functions and types
571 that are made available to it by including one of the header files
572 it must include by the previous rule.
574 - If you are planning a new command, consider writing it in shell
575 or perl first, so that changes in semantics can be easily
576 changed and discussed. Many Git commands started out like
577 that, and a few are still scripts.
579 - Avoid introducing a new dependency into Git. This means you
580 usually should stay away from scripting languages not already
581 used in the Git core command set (unless your command is clearly
582 separate from it, such as an importer to convert random-scm-X
583 repositories to Git).
585 - When we pass <string, length> pair to functions, we should try to
586 pass them in that order.
588 - Use Git's gettext wrappers to make the user interface
589 translatable. See "Marking strings for translation" in po/README.
591 - Variables and functions local to a given source file should be marked
592 with "static". Variables that are visible to other source files
593 must be declared with "extern" in header files. However, function
594 declarations should not use "extern", as that is already the default.
596 - You can launch gdb around your program using the shorthand GIT_DEBUGGER.
597 Run `GIT_DEBUGGER=1 ./bin-wrappers/git foo` to simply use gdb as is, or
598 run `GIT_DEBUGGER="<debugger> <debugger-args>" ./bin-wrappers/git foo` to
599 use your own debugger and arguments. Example: `GIT_DEBUGGER="ddd --gdb"
600 ./bin-wrappers/git log` (See `bin-wrappers/wrap-for-bin.sh`.)
602 - The primary data structure that a subsystem 'S' deals with is called
603 `struct S`. Functions that operate on `struct S` are named
604 `S_<verb>()` and should generally receive a pointer to `struct S` as
605 first parameter. E.g.
607 struct strbuf;
609 void strbuf_add(struct strbuf *buf, ...);
611 void strbuf_reset(struct strbuf *buf);
613 is preferred over:
615 struct strbuf;
617 void add_string(struct strbuf *buf, ...);
619 void reset_strbuf(struct strbuf *buf);
621 - There are several common idiomatic names for functions performing
622 specific tasks on a structure `S`:
624 - `S_init()` initializes a structure without allocating the
625 structure itself.
627 - `S_release()` releases a structure's contents without freeing the
628 structure.
630 - `S_clear()` is equivalent to `S_release()` followed by `S_init()`
631 such that the structure is directly usable after clearing it. When
632 `S_clear()` is provided, `S_init()` shall not allocate resources
633 that need to be released again.
635 - `S_free()` releases a structure's contents and frees the
636 structure.
638 - Function names should be clear and descriptive, accurately reflecting
639 their purpose or behavior. Arbitrary suffixes that do not add meaningful
640 context can lead to confusion, particularly for newcomers to the codebase.
642 Historically, the '_1' suffix has been used in situations where:
644 - A function handles one element among a group that requires similar
645 processing.
646 - A recursive function has been separated from its setup phase.
648 The '_1' suffix can be used as a concise way to indicate these specific
649 cases. However, it is recommended to find a more descriptive name wherever
650 possible to improve the readability and maintainability of the code.
652 For Perl programs:
654 - Most of the C guidelines above apply.
656 - We try to support Perl 5.8.1 and later ("use Perl 5.008001").
658 - use strict and use warnings are strongly preferred.
660 - Don't overuse statement modifiers unless using them makes the
661 result easier to follow.
663 ... do something ...
664 do_this() unless (condition);
665 ... do something else ...
667 is more readable than:
669 ... do something ...
670 unless (condition) {
671 do_this();
672 }
673 ... do something else ...
675 *only* when the condition is so rare that do_this() will be almost
676 always called.
678 - We try to avoid assignments inside "if ()" conditions.
680 - Learn and use Git.pm if you need that functionality.
682 For Python scripts:
684 - We follow PEP-8 (https://peps.python.org/pep-0008/).
686 - As a minimum, we aim to be compatible with Python 2.7.
688 - Where required libraries do not restrict us to Python 2, we try to
689 also be compatible with Python 3.1 and later.
692 Program Output
694 We make a distinction between a Git command's primary output and
695 output which is merely chatty feedback (for instance, status
696 messages, running transcript, or progress display), as well as error
697 messages. Roughly speaking, a Git command's primary output is that
698 which one might want to capture to a file or send down a pipe; its
699 chatty output should not interfere with these use-cases.
701 As such, primary output should be sent to the standard output stream
702 (stdout), and chatty output should be sent to the standard error
703 stream (stderr). Examples of commands which produce primary output
704 include `git log`, `git show`, and `git branch --list` which generate
705 output on the stdout stream.
707 Not all Git commands have primary output; this is often true of
708 commands whose main function is to perform an action. Some action
709 commands are silent, whereas others are chatty. An example of a
710 chatty action commands is `git clone` with its "Cloning into
711 '<path>'..." and "Checking connectivity..." status messages which it
712 sends to the stderr stream.
714 Error messages from Git commands should always be sent to the stderr
715 stream.
718 Error Messages
720 - Do not end a single-sentence error message with a full stop.
722 - Do not capitalize the first word, only because it is the first word
723 in the message ("unable to open '%s'", not "Unable to open '%s'"). But
724 "SHA-3 not supported" is fine, because the reason the first word is
725 capitalized is not because it is at the beginning of the sentence,
726 but because the word would be spelled in capital letters even when
727 it appeared in the middle of the sentence.
729 - Say what the error is first ("cannot open '%s'", not "%s: cannot open").
731 - Enclose the subject of an error inside a pair of single quotes,
732 e.g. `die(_("unable to open '%s'"), path)`.
734 - Unless there is a compelling reason not to, error messages from
735 porcelain commands should be marked for translation, e.g.
736 `die(_("bad revision %s"), revision)`.
738 - Error messages from the plumbing commands are sometimes meant for
739 machine consumption and should not be marked for translation,
740 e.g., `die("bad revision %s", revision)`.
742 - BUG("message") are for communicating the specific error to developers,
743 thus should not be translated.
746 Externally Visible Names
748 - For configuration variable names, follow the existing convention:
750 . The section name indicates the affected subsystem.
752 . The subsection name, if any, indicates which of an unbounded set
753 of things to set the value for.
755 . The variable name describes the effect of tweaking this knob.
757 The section and variable names that consist of multiple words are
758 formed by concatenating the words without punctuation marks (e.g. `-`),
759 and are broken using bumpyCaps in documentation as a hint to the
760 reader.
762 When choosing the variable namespace, do not use variable name for
763 specifying possibly unbounded set of things, most notably anything
764 an end user can freely come up with (e.g. branch names). Instead,
765 use subsection names or variable values, like the existing variable
766 branch.<name>.description does.
769 Writing Documentation:
771 Most (if not all) of the documentation pages are written in the
772 AsciiDoc format in *.adoc files (e.g. Documentation/git.adoc), and
773 processed into HTML and manpages (e.g. git.html and git.1 in the
774 same directory).
776 The documentation liberally mixes US and UK English (en_US/UK)
777 norms for spelling and grammar, which is somewhat unfortunate.
778 In an ideal world, it would have been better if it consistently
779 used only one and not the other, and we would have picked en_US
780 (if you wish to correct the English of some of the existing
781 documentation, please see the documentation-related advice in the
782 Documentation/SubmittingPatches file).
784 In order to ensure the documentation is inclusive, avoid assuming
785 that an unspecified example person is male or female, and think
786 twice before using "he", "him", "she", or "her". Here are some
787 tips to avoid use of gendered pronouns:
789 - Prefer succinctness and matter-of-factly describing functionality
790 in the abstract. E.g.
792 `--short`:: Emit output in the short-format.
794 and avoid something like these overly verbose alternatives:
796 `--short`:: Use this to emit output in the short-format.
797 `--short`:: You can use this to get output in the short-format.
798 `--short`:: A user who prefers shorter output could....
799 `--short`:: Should a person and/or program want shorter output, he
800 she/they/it can...
802 This practice often eliminates the need to involve human actors in
803 your description, but it is a good practice regardless of the
804 avoidance of gendered pronouns.
806 - When it becomes awkward to stick to this style, prefer "you" when
807 addressing the hypothetical user, and possibly "we" when
808 discussing how the program might react to the user. E.g.
810 You can use this option instead of `--xyz`, but we might remove
811 support for it in future versions.
813 while keeping in mind that you can probably be less verbose, e.g.
815 Use this instead of `--xyz`. This option might be removed in future
816 versions.
818 - If you still need to refer to an example person that is
819 third-person singular, you may resort to "singular they" to avoid
820 "he/she/him/her", e.g.
822 A contributor asks their upstream to pull from them.
824 Note that this sounds ungrammatical and unnatural to those who
825 learned that "they" is only used for third-person plural, e.g.
826 those who learn English as a second language in some parts of the
827 world.
829 Every user-visible change should be reflected in the documentation.
830 The same general rule as for code applies -- imitate the existing
831 conventions.
834 Markup:
836 Literal parts (e.g. use of command-line options, command names,
837 branch names, URLs, pathnames (files and directories), configuration and
838 environment variables) must be typeset as verbatim (i.e. wrapped with
839 backticks):
840 `--pretty=oneline`
841 `git rev-list`
842 `remote.pushDefault`
843 `http://git.example.com`
844 `.git/config`
845 `GIT_DIR`
846 `HEAD`
847 `umask`(2)
849 An environment variable must be prefixed with "$" only when referring to its
850 value and not when referring to the variable itself, in this case there is
851 nothing to add except the backticks:
852 `GIT_DIR` is specified
853 `$GIT_DIR/hooks/pre-receive`
855 Word phrases enclosed in `backtick characters` are rendered literally
856 and will not be further expanded. The use of `backticks` to achieve the
857 previous rule means that literal examples should not use AsciiDoc
858 escapes.
859 Correct:
860 `--pretty=oneline`
861 Incorrect:
862 `\--pretty=oneline`
864 Placeholders are spelled in lowercase and enclosed in
865 angle brackets surrounded by underscores:
866 _<file>_
867 _<commit>_
869 If a placeholder has multiple words, they are separated by dashes:
870 _<new-branch-name>_
871 _<template-directory>_
873 When needed, use a distinctive identifier for placeholders, usually
874 made of a qualification and a type:
875 _<git-dir>_
876 _<key-id>_
878 Characters are also surrounded by underscores:
879 _LF_, _CR_, _CR_/_LF_, _NUL_, _EOF_
881 Git's Asciidoc processor has been tailored to treat backticked text
882 as complex synopsis. When literal and placeholders are mixed, you can
883 use the backtick notation which will take care of correctly typesetting
884 the content.
885 `--jobs <n>`
886 `--sort=<key>`
887 `<directory>/.git`
888 `remote.<name>.mirror`
889 `ssh://[<user>@]<host>[:<port>]/<path-to-git-repo>`
891 As a side effect, backquoted placeholders are correctly typeset, but
892 this style is not recommended.
894 When documenting multiple related `git config` variables, place them on
895 a separate line instead of separating them by commas. For example, do
896 not write this:
897 `core.var1`, `core.var2`::
898 Description common to `core.var1` and `core.var2`.
900 Instead write this:
901 `core.var1`::
902 `core.var2`::
903 Description common to `core.var1` and `core.var2`.
905 Synopsis Syntax
907 The synopsis (a paragraph with [synopsis] attribute) is automatically
908 formatted by the toolchain and does not need typesetting.
910 A few commented examples follow to provide reference when writing or
911 modifying command usage strings and synopsis sections in the manual
912 pages:
914 Possibility of multiple occurrences is indicated by three dots:
915 <file>...
916 (One or more of <file>.)
918 Optional parts are enclosed in square brackets:
919 [<file>...]
920 (Zero or more of <file>.)
922 An optional parameter needs to be typeset with unconstrained pairs
923 [<repository>]
925 --exec-path[=<path>]
926 (Option with an optional argument. Note that the "=" is inside the
927 brackets.)
929 [<patch>...]
930 (Zero or more of <patch>. Note that the dots are inside, not
931 outside the brackets.)
933 Multiple alternatives are indicated with vertical bars:
934 [-q | --quiet]
935 [--utf8 | --no-utf8]
937 Use spacing around "|" token(s), but not immediately after opening or
938 before closing a [] or () pair:
939 Do: [-q | --quiet]
940 Don't: [-q|--quiet]
942 Don't use spacing around "|" tokens when they're used to separate the
943 alternate arguments of an option:
944 Do: --track[=(direct|inherit)]
945 Don't: --track[=(direct | inherit)]
947 Parentheses are used for grouping:
948 [(<rev>|<range>)...]
949 (Any number of either <rev> or <range>. Parens are needed to make
950 it clear that "..." pertains to both <rev> and <range>.)
952 [(-p <parent>)...]
953 (Any number of option -p, each with one <parent> argument.)
955 git remote set-head <name> (-a|-d|<branch>)
956 (One and only one of "-a", "-d" or "<branch>" _must_ (no square
957 brackets) be provided.)
959 And a somewhat more contrived example:
960 --diff-filter=[(A|C|D|M|R|T|U|X|B)...[*]]
961 Here "=" is outside the brackets, because "--diff-filter=" is a
962 valid usage. "*" has its own pair of brackets, because it can
963 (optionally) be specified only when one or more of the letters is
964 also provided.
966 A note on notation:
967 Use 'git' (all lowercase) when talking about commands i.e. something
968 the user would type into a shell and use 'Git' (uppercase first letter)
969 when talking about the version control system and its properties.
971 If some place in the documentation needs to typeset a command usage
972 example with inline substitutions, it is fine to use +monospaced and
973 inline substituted text+ instead of `monospaced literal text`, and with
974 the former, the part that should not get substituted must be
975 quoted/escaped.