3 Evan Martin <martine@danga.com>
9 Ninja is yet another build system. It takes as input the
10 interdependencies of files (typically source code and output
11 executables) and orchestrates building them, _quickly_.
13 Ninja joins a sea of other build systems. Its distinguishing goal is
14 to be fast. It is born from
15 http://neugierig.org/software/chromium/notes/2011/02/ninja.html[my
16 work on the Chromium browser project], which has over 30,000 source
17 files and whose other build systems (including one built from custom
18 non-recursive Makefiles) can take ten seconds to start building after
19 changing one file. Ninja is under a second.
21 Philosophical overview
22 ~~~~~~~~~~~~~~~~~~~~~~
24 Where other build systems are high-level languages, Ninja aims to be
27 Build systems get slow when they need to make decisions. When you are
28 in a edit-compile cycle you want it to be as fast as possible -- you
29 want the build system to do the minimum work necessary to figure out
30 what needs to be built immediately.
32 Ninja contains the barest functionality necessary to describe
33 arbitrary dependency graphs. Its lack of syntax makes it impossible
34 to express complex decisions.
36 Instead, Ninja is intended to be used with a separate program
37 generating its input files. The generator program (like the
38 `./configure` found in autotools projects) can analyze system
39 dependencies and make as many decisions as possible up front so that
40 incremental builds stay fast. Going beyond autotools, even build-time
41 decisions like "which compiler flags should I use?" or "should I
42 build a debug or release-mode binary?" belong in the `.ninja` file
48 Here are the design goals of Ninja:
50 * very fast (i.e., instant) incremental builds, even for very large
53 * very little policy about how code is built. Different projects and
54 higher-level build systems have different opinions about how code
55 should be built; for example, should built objects live alongside
56 the sources or should all build output go into a separate directory?
57 Is there an "package" rule that builds a distributable package of
58 the project? Sidestep these decisions by trying to allow either to
59 be implemented, rather than choosing, even if that results in
62 * get dependencies correct, and in particular situations that are
63 difficult to get right with Makefiles (e.g. outputs need an implicit
64 dependency on the command line used to generate them; to build C
65 source code you need to use gcc's `-M` flags for header
68 * when convenience and speed are in conflict, prefer speed.
70 Some explicit _non-goals_:
72 * convenient syntax for writing build files by hand. _You should
73 generate your ninja files using another program_. This is how we
74 can sidestep many policy decisions.
76 * built-in rules. _Out of the box, Ninja has no rules for
77 e.g. compiling C code._
79 * build-time customization of the build. _Options belong in
80 the program that generates the ninja files_.
82 * build-time decision-making ability such as conditionals or search
83 paths. _Making decisions is slow._
85 To restate, Ninja is faster than other build systems because it is
86 painfully simple. You must tell Ninja exactly what to do when you
87 create your project's `.ninja` files.
92 Ninja is closest in spirit and functionality to make, relying on
93 simple dependencies between file timestamps.
95 But fundamentally, make has a lot of _features_: suffix rules,
96 functions, built-in rules that e.g. search for RCS files when building
97 source. Make's language was designed to be written by humans. Many
98 projects find make alone adequate for their build problems.
100 In contrast, Ninja has almost no features; just those necessary to get
101 builds correct while punting most complexity to generation of the
102 ninja input files. Ninja by itself is unlikely to be useful for most
105 Here are some of the features Ninja adds to make. (These sorts of
106 features can often be implemented using more complicated Makefiles,
107 but they are not part of make itself.)
109 * Ninja has special support for discovering extra dependencies at build
110 time, making it easy to get <<ref_headers,header dependencies>>
111 correct for C/C++ code.
113 * A build edge may have multiple outputs.
115 * Outputs implicitly depend on the command line that was used to generate
116 them, which means that changing e.g. compilation flags will cause
117 the outputs to rebuild.
119 * Output directories are always implicitly created before running the
120 command that relies on them.
122 * Rules can provide shorter descriptions of the command being run, so
123 you can print e.g. `CC foo.o` instead of a long command line while
126 * Builds are always run in parallel, based by default on the number of
127 CPUs your system has. Underspecified build dependencies will result
130 * Command output is always buffered. This means commands running in
131 parallel don't interleave their output, and when a command fails we
132 can print its failure output next to the full command line that
133 produced the failure.
136 Using Ninja for your project
137 ----------------------------
139 Ninja currently works on Unix-like systems and Windows. It's seen the
140 most testing on Linux (and has the best performance there) but it runs
141 fine on Mac OS X and FreeBSD.
143 If your project is small, Ninja's speed impact is likely unnoticeable.
144 (However, even for small projects it sometimes turns out that Ninja's
145 limited syntax forces simpler build rules that result in faster
146 builds.) Another way to say this is that if you're happy with the
147 edit-compile cycle time of your project already then Ninja won't help.
149 There are many other build systems that are more user-friendly or
150 featureful than Ninja itself. For some recommendations: the Ninja
151 author found http://gittup.org/tup/[the tup build system] influential
152 in Ninja's design, and thinks https://github.com/apenwarr/redo[redo]'s
153 design is quite clever.
155 Ninja's benefit comes from using it in conjunction with a smarter
158 http://code.google.com/p/gyp/[gyp]:: The meta-build system used to
159 generate build files for Google Chrome and related projects (v8,
160 node.js). gyp can generate Ninja files for all platforms supported by
162 http://code.google.com/p/chromium/wiki/NinjaBuild[Chromium Ninja
163 documentation for more details].
165 http://www.cmake.org/[CMake]:: A widely used meta-build system that
166 can generate Ninja files on Linux as of CMake version 2.8.8. (There
167 is some Mac and Windows support -- http://www.reactos.org[ReactOS]
168 uses Ninja on Windows for their buildbots, but those platforms are not
169 yet officially supported by CMake as the full test suite doesn't
172 others:: Ninja ought to fit perfectly into other meta-build software
173 like http://industriousone.com/premake[premake]. If you do this work,
179 Run `ninja`. By default, it looks for a file named `build.ninja` in
180 the current directory and builds all out-of-date targets. You can
181 specify which targets (files) to build as command line arguments.
183 `ninja -h` prints help output. Many of Ninja's flags intentionally
184 match those of Make; e.g `ninja -C build -j 20` changes into the
185 `build` directory and runs 20 build commands in parallel. (Note that
186 Ninja defaults to running commands in parallel anyway, so typically
187 you don't need to pass `-j`.)
190 Environment variables
191 ~~~~~~~~~~~~~~~~~~~~~
193 Ninja supports one environment variable to control its behavior:
194 `NINJA_STATUS`, the progress status printed before the rule being run.
196 Several placeholders are available:
198 `%s`:: The number of started edges.
199 `%t`:: The total number of edges that must be run to complete the build.
200 `%p`:: The percentage of started edges.
201 `%r`:: The number of currently running edges.
202 `%u`:: The number of remaining edges to start.
203 `%f`:: The number of finished edges.
204 `%o`:: Overall rate of finished edges per second
205 `%c`:: Current rate of finished edges per second (average over builds specified by -j or its default)
206 `%e`:: Elapsed time in seconds. _(Available since Ninja 1.2.)_
207 `%%`:: A plain `%` character.
209 The default progress status is `"[%s/%t] "` (note the trailing space
210 to separate from the build rule). Another example of possible progress status
211 could be `"[%u/%r/%f] "`.
216 The `-t` flag on the Ninja command line runs some tools that we have
217 found useful during Ninja's development. The current tools are:
220 `query`:: dump the inputs and outputs of a given target.
222 `browse`:: browse the dependency graph in a web browser. Clicking a
223 file focuses the view on that file, showing inputs and outputs. This
224 feature requires a Python installation.
226 `graph`:: output a file in the syntax used by `graphviz`, a automatic
227 graph layout tool. Use it like:
230 ninja -t graph mytarget | dot -Tpng -ograph.png
233 In the Ninja source tree, `ninja graph.png`
234 generates an image for Ninja itself. If no target is given generate a
235 graph for all root targets.
237 `targets`:: output a list of targets either by rule or by depth. If used
238 like +ninja -t targets rule _name_+ it prints the list of targets
239 using the given rule to be built. If no rule is given, it prints the source
240 files (the leaves of the graph). If used like
241 +ninja -t targets depth _digit_+ it
242 prints the list of targets in a depth-first manner starting by the root
243 targets (the ones with no outputs). Indentation is used to mark dependencies.
244 If the depth is zero it prints all targets. If no arguments are provided
245 +ninja -t targets depth 1+ is assumed. In this mode targets may be listed
246 several times. If used like this +ninja -t targets all+ it
247 prints all the targets available without indentation and it is faster
248 than the _depth_ mode.
250 `commands`:: given a list of targets, print a list of commands which, if
251 executed in order, may be used to rebuild those targets, assuming that all
252 output files are out of date.
254 `clean`:: remove built files. By default it removes all built files
255 except for those created by the generator. Adding the `-g` flag also
256 removes built files created by the generator (see <<ref_rule,the rule
257 reference for the +generator+ attribute>>). Additional arguments are
258 targets, which removes the given targets and recursively all files
261 If used like +ninja -t clean -r _rules_+ it removes all files built using
264 Files created but not referenced in the graph are not removed. This
265 tool takes in account the +-v+ and the +-n+ options (note that +-n+
268 `compdb`:: given a list of rules, each of which is expected to be a
269 C family language compiler rule whose first input is the name of the
270 source file, prints on standard output a compilation database in the
271 http://clang.llvm.org/docs/JSONCompilationDatabase.html[JSON format] expected
272 by the Clang tooling interface.
273 _Available since Ninja 1.2._
276 Writing your own Ninja files
277 ----------------------------
279 The remainder of this manual is only useful if you are constructing
280 Ninja files yourself: for example, if you're writing a meta-build
281 system or supporting a new language.
286 Ninja evaluates a graph of dependencies between files, and runs
287 whichever commands are necessary to make your build target up to date.
288 If you are familiar with Make, Ninja is very similar.
290 A build file (default name: `build.ninja`) provides a list of _rules_
291 -- short names for longer commands, like how to run the compiler --
292 along with a list of _build_ statements saying how to build files
293 using the rules -- which rule to apply to which inputs to produce
296 Conceptually, `build` statements describe the dependency graph of your
297 project, while `rule` statements describe how to generate the files
298 along a given edge of the graph.
303 Here's a basic `.ninja` file that demonstrates most of the syntax.
304 It will be used as an example for the following sections.
306 ---------------------------------
310 command = gcc $cflags -c $in -o $out
312 build foo.o: cc foo.c
313 ---------------------------------
317 Despite the non-goal of being convenient to write by hand, to keep
318 build files readable (debuggable), Ninja supports declaring shorter
319 reusable names for strings. A declaration like the following
325 can be used on the right side of an equals sign, dereferencing it with
326 a dollar sign, like this:
330 command = gcc $cflags -c $in -o $out
333 Variables can also be referenced using curly braces like `${in}`.
335 Variables might better be called "bindings", in that a given variable
336 cannot be changed, only shadowed. There is more on how shadowing works
337 later in this document.
342 Rules declare a short name for a command line. They begin with a line
343 consisting of the `rule` keyword and a name for the rule. Then
344 follows an indented set of `variable = value` lines.
346 The basic example above declares a new rule named `cc`, along with the
347 command to run. In the context of a rule, the `command` variable
348 defines the command to run, `$in` expands to the list of
349 input files (`foo.c`), and `$out` to the output files (`foo.o`) for the
350 command. A full list of special variables is provided in
351 <<ref_rule,the reference>>.
356 Build statements declare a relationship between input and output
357 files. They begin with the `build` keyword, and have the format
358 +build _outputs_: _rulename_ _inputs_+. Such a declaration says that
359 all of the output files are derived from the input files. When the
360 output files are missing or when the inputs change, Ninja will run the
361 rule to regenerate the outputs.
363 The basic example above describes how to build `foo.o`, using the `cc`
366 In the scope of a `build` block (including in the evaluation of its
367 associated `rule`), the variable `$in` is the list of inputs and the
368 variable `$out` is the list of outputs.
370 A build statement may be followed by an indented set of `key = value`
371 pairs, much like a rule. These variables will shadow any variables
372 when evaluating the variables in the command. For example:
375 cflags = -Wall -Werror
377 command = gcc $cflags -c $in -o $out
379 # If left unspecified, builds get the outer $cflags.
380 build foo.o: cc foo.c
382 # But you can shadow variables like cflags for a particular build.
383 build special.o: cc special.c
386 # The variable was only shadowed for the scope of special.o;
387 # Subsequent build lines get the outer (original) cflags.
388 build bar.o: cc bar.c
392 For more discussion of how scoping works, consult <<ref_scope,the
395 If you need more complicated information passed from the build
396 statement to the rule (for example, if the rule needs "the file
397 extension of the first input"), pass that through as an extra
398 variable, like how `cflags` is passed above.
400 If the top-level Ninja file is specified as an output of any build
401 statement and it is out of date, Ninja will rebuild and reload it
402 before building the targets requested by the user.
407 _Available since Ninja 1.1._
409 Pools allow you to allocate one or more rules or edges a finite number
410 of concurrent jobs which is more tightly restricted than the default
413 This can be useful, for example, to restrict a particular expensive rule
414 (like link steps for huge executables), or to restrict particular build
415 statements which you know perform poorly when run concurrently.
417 Each pool has a `depth` variable which is specified in the build file.
418 The pool is then referred to with the `pool` variable on either a rule
419 or a build statement.
421 No matter what pools you specify, ninja will never run more concurrent jobs
422 than the default parallelism, or the number of jobs specified on the command
426 # No more than 4 links at a time.
430 # No more than 1 heavy object at a time.
431 pool heavy_object_pool
441 # The link_pool is used here. Only 4 links will run concurrently.
442 build foo.exe: link input.obj
444 # A build statement can be exempted from its rule's pool by setting an
445 # empty pool. This effectively puts the build statement back into the default
446 # pool, which has infinite depth.
447 build other.exe: link input.obj
450 # A build statement can specify a pool directly.
451 # Only one of these builds will run at a time.
452 build heavy_object1.obj: cc heavy_obj1.cc
453 pool = heavy_object_pool
454 build heavy_object2.obj: cc heavy_obj2.cc
455 pool = heavy_object_pool
460 Generating Ninja files from code
461 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
463 `misc/ninja_syntax.py` in the Ninja distribution is a tiny Python
464 module to facilitate generating Ninja files. It allows you to make
465 Python calls like `ninja.rule(name='foo', command='bar',
466 depfile='$out.d')` and it will generate the appropriate syntax. Feel
467 free to just inline it into your project's build system if it's
477 The special rule name `phony` can be used to create aliases for other
478 targets. For example:
481 build foo: phony some/file/in/a/faraway/subdir/foo
484 This makes `ninja foo` build the longer path. Semantically, the
485 `phony` rule is equivalent to a plain rule where the `command` does
486 nothing, but phony rules are handled specially in that they aren't
487 printed when run, logged (see below), nor do they contribute to the
488 command count printed as part of the build process.
490 `phony` can also be used to create dummy targets for files which
491 may not exist at build time. If a phony build statement is written
492 without any dependencies, the target will be considered out of date if
493 it does not exist. Without a phony build statement, Ninja will report
494 an error if the file does not exist and is required by the build.
497 Default target statements
498 ~~~~~~~~~~~~~~~~~~~~~~~~~
500 By default, if no targets are specified on the command line, Ninja
501 will build every output that is not named as an input elsewhere.
502 You can override this behavior using a default target statement.
503 A default target statement causes Ninja to build only a given subset
504 of output files if none are specified on the command line.
506 Default target statements begin with the `default` keyword, and have
507 the format +default _targets_+. A default target statement must appear
508 after the build statement that declares the target as an output file.
509 They are cumulative, so multiple statements may be used to extend
510 the list of default targets. For example:
517 This causes Ninja to build the `foo`, `bar` and `baz` targets by
525 For each built file, Ninja keeps a log of the command used to build
526 it. Using this log Ninja can know when an existing output was built
527 with a different command line than the build files specify (i.e., the
528 command line changed) and knows to rebuild the file.
530 The log file is kept in the build root in a file called `.ninja_log`.
531 If you provide a variable named `builddir` in the outermost scope,
532 `.ninja_log` will be kept in that directory instead.
536 Version compatibility
537 ~~~~~~~~~~~~~~~~~~~~~
539 _Available since Ninja 1.2._
541 Ninja version labels follow the standard major.minor.patch format,
542 where the major version is increased on backwards-incompatible
543 syntax/behavioral changes and the minor version is increased on new
544 behaviors. Your `build.ninja` may declare a variable named
545 `ninja_required_version` that asserts the minimum Ninja version
546 required to use the generated file. For example,
549 ninja_required_version = 1.1
552 declares that the build file relies on some feature that was
553 introduced in Ninja 1.1 (perhaps the `pool` syntax), and that
554 Ninja 1.1 or greater must be used to build. Unlike other Ninja
555 variables, this version requirement is checked immediately when
556 the variable is encountered in parsing, so it's best to put it
557 at the top of the build file.
559 Ninja always warns if the major versions of Ninja and the
560 `ninja_required_version` don't match; a major version change hasn't
561 come up yet so it's difficult to predict what behavior might be
565 C/C++ header dependencies
566 ~~~~~~~~~~~~~~~~~~~~~~~~~
568 To get C/C++ header dependencies (or any other build dependency that
569 works in a similar way) correct Ninja has some extra functionality.
571 The problem with headers is that the full list of files that a given
572 source file depends on can only be discovered by the compiler:
573 different preprocessor defines and include paths cause different files
574 to be used. Some compilers can emit this information while building,
575 and Ninja can use that to get its dependencies perfect.
577 Consider: if the file has never been compiled, it must be built anyway,
578 generating the header dependencies as a side effect. If any file is
579 later modified (even in a way that changes which headers it depends
580 on) the modification will cause a rebuild as well, keeping the
581 dependencies up to date.
583 When loading these special dependencies, Ninja implicitly adds extra
584 build edges such that it is not an error if the listed dependency is
585 missing. This allows you to delete a header file and rebuild without
586 the build aborting due to a missing input.
591 `gcc` (and other compilers like `clang`) support emitting dependency
592 information in the syntax of a Makefile. (Any command that can write
593 dependencies in this form can be used, not just `gcc`.)
595 To bring this information into Ninja requires cooperation. On the
596 Ninja side, the `depfile` attribute on the `build` must point to a
597 path where this data is written. (Ninja only supports the limited
598 subset of the Makefile syntax emitted by compilers.) Then the command
599 must know to write dependencies into the `depfile` path.
600 Use it like in the following example:
605 command = gcc -MMD -MF $out.d [other gcc flags here]
608 The `-MMD` flag to `gcc` tells it to output header dependencies, and
609 the `-MF` flag tells it where to write them.
614 _(Available since Ninja 1.3.)_
616 It turns out that for large projects (and particularly on Windows,
617 where the file system is slow) loading these dependency files on
620 Ninja 1.3 can instead process dependencies just after they're generated
621 and save a compacted form of the same information in a Ninja-internal
624 Ninja supports this processing in two forms.
626 1. `deps = gcc` specifies that the tool outputs `gcc`-style dependencies
627 in the form of Makefiles. Adding this to the above example will
628 cause Ninja to process the `depfile` immediately after the
629 compilation finishes, then delete the `.d` file (which is only used
632 2. `deps = msvc` specifies that the tool outputs header dependencies
633 in the form produced by Visual Studio's compiler's
634 http://msdn.microsoft.com/en-us/library/hdkef6tk(v=vs.90).aspx[`/showIncludes`
635 flag]. Briefly, this means the tool outputs specially-formatted lines
636 to its stdout. Ninja then filters these lines from the displayed
637 output. No `depfile` attribute is necessary.
642 command = cl /showIncludes -c $in /Fo$out
649 A file is a series of declarations. A declaration can be one of:
651 1. A rule declaration, which begins with +rule _rulename_+, and
652 then has a series of indented lines defining variables.
654 2. A build edge, which looks like +build _output1_ _output2_:
655 _rulename_ _input1_ _input2_+. +
656 Implicit dependencies may be tacked on the end with +|
657 _dependency1_ _dependency2_+. +
658 Order-only dependencies may be tacked on the end with +||
659 _dependency1_ _dependency2_+. (See <<ref_dependencies,the reference on
662 3. Variable declarations, which look like +_variable_ = _value_+.
664 4. Default target statements, which look like +default _target1_ _target2_+.
666 5. References to more files, which look like +subninja _path_+ or
667 +include _path_+. The difference between these is explained below
668 <<ref_scope,in the discussion about scoping>>.
673 Ninja is mostly encoding agnostic, as long as the bytes Ninja cares
674 about (like slashes in paths) are ASCII. This means e.g. UTF-8 or
675 ISO-8859-1 input files ought to work. (To simplify some code, tabs
676 and carriage returns are currently disallowed; this could be fixed if
677 it really mattered to you.)
679 Comments begin with `#` and extend to the end of the line.
681 Newlines are significant. Statements like `build foo bar` are a set
682 of space-separated tokens that end at the newline. Newlines and
683 spaces within a token must be escaped.
685 There is only one escape character, `$`, and it has the following
689 `$` followed by a newline:: escape the newline (continue the current line
690 across a line break).
692 `$` followed by text:: a variable reference.
694 `${varname}`:: alternate syntax for `$varname`.
696 `$` followed by space:: a space. (This is only necessary in lists of
697 paths, where a space would otherwise separate filenames. See below.)
699 `$:` :: a colon. (This is only necessary in `build` lines, where a colon
700 would otherwise terminate the list of outputs.)
702 `$$`:: a literal `$`.
704 A `build` or `default` statement is first parsed as a space-separated
705 list of filenames and then each name is expanded. This means that
706 spaces within a variable will result in spaces in the expanded
711 build $spaced/baz other$ file: ...
712 # The above build line has two outputs: "foo bar/baz" and "other file".
715 In a `name = value` statement, whitespace at the beginning of a value
716 is always stripped. Whitespace at the beginning of a line after a
717 line continuation is also stripped.
720 two_words_with_one_space = foo $
722 one_word_with_no_space = foo$
726 Other whitespace is only significant if it's at the beginning of a
727 line. If a line is indented more than the previous one, it's
728 considered part of its parent's scope; if it is indented less than the
729 previous one, it closes the previous scope.
735 Two variables are significant when declared in the outermost file scope.
737 `builddir`:: a directory for some Ninja output files. See <<ref_log,the
738 discussion of the build log>>. (You can also store other build output
741 `ninja_required_version`:: the minimum version of Ninja required to process
742 the build correctly. See <<ref_versioning,the discussion of versioning>>.
749 A `rule` block contains a list of `key = value` declarations that
750 affect the processing of the rule. Here is a full list of special
753 `command` (_required_):: the command line to run. This string (after
754 $variables are expanded) is passed directly to `sh -c` without
755 interpretation by Ninja. Each `rule` may have only one `command`
756 declaration. To specify multiple commands use `&&` (or similar) to
757 concatenate operations.
759 `depfile`:: path to an optional `Makefile` that contains extra
760 _implicit dependencies_ (see <<ref_dependencies,the reference on
761 dependency types>>). This is explicitly to support C/C++ header
762 dependencies; see <<ref_headers,the full discussion>>.
764 `deps`:: _(Available since Ninja 1.3.)_ if present, must be one of
765 `gcc` or `msvc` to specify special dependency processing. See
766 <<ref_headers,the full discussion>>. The generated database is
767 stored as `.ninja_deps` in the `builddir`, see <<ref_toplevel,the
768 discussion of `builddir`>>.
770 `description`:: a short description of the command, used to pretty-print
771 the command as it's running. The `-v` flag controls whether to print
772 the full command or its description; if a command fails, the full command
773 line will always be printed before the command's output.
775 `generator`:: if present, specifies that this rule is used to
776 re-invoke the generator program. Files built using `generator`
777 rules are treated specially in two ways: firstly, they will not be
778 rebuilt if the command line changes; and secondly, they are not
781 `in`:: the shell-quoted space-separated list of files provided as
782 inputs to the build line referencing this `rule`. (`$in` is provided
783 solely for convenience; if you need some subset or variant of this
784 list of files, just construct a new variable with that list and use
787 `in_newline`:: the same as `$in` except that multiple inputs are
788 separated by newlines rather than spaces. (For use with
789 `$rspfile_content`; this works around a bug in the MSVC linker where
790 it uses a fixed-size buffer for processing input.)
792 `out`:: the shell-quoted space-separated list of files provided as
793 outputs to the build line referencing this `rule`.
795 `restat`:: if present, causes Ninja to re-stat the command's outputs
796 after execution of the command. Each output whose modification time
797 the command did not change will be treated as though it had never
798 needed to be built. This may cause the output's reverse
799 dependencies to be removed from the list of pending build actions.
801 `rspfile`, `rspfile_content`:: if present (both), Ninja will use a
802 response file for the given command, i.e. write the selected string
803 (`rspfile_content`) to the given file (`rspfile`) before calling the
804 command and delete the file after successful execution of the
807 This is particularly useful on Windows OS, where the maximal length of
808 a command line is limited and response files must be used instead.
810 Use it like in the following example:
814 command = link.exe /OUT$out [usual link flags here] @$out.rsp
816 rspfile_content = $in
818 build myapp.exe: link a.obj b.obj [possibly many other .obj files]
825 There are three types of build dependencies which are subtly different.
827 1. _Explicit dependencies_, as listed in a build line. These are
828 available as the `$in` variable in the rule. Changes in these files
829 cause the output to be rebuilt; if these file are missing and
830 Ninja doesn't know how to build them, the build is aborted.
832 This is the standard form of dependency to be used for e.g. the
833 source file of a compile command.
835 2. _Implicit dependencies_, either as picked up from
836 a `depfile` attribute on a rule or from the syntax +| _dep1_
837 _dep2_+ on the end of a build line. The semantics are identical to
838 explicit dependencies, the only difference is that implicit dependencies
839 don't show up in the `$in` variable.
841 This is for expressing dependencies that don't show up on the
842 command line of the command; for example, for a rule that runs a
843 script, the script itself should be an implicit dependency, as
844 changes to the script should cause the output to rebuild.
846 Note that dependencies as loaded through depfiles have slightly different
847 semantics, as described in the <<ref_rule,rule reference>>.
849 3. _Order-only dependencies_, expressed with the syntax +|| _dep1_
850 _dep2_+ on the end of a build line. When these are out of date, the
851 output is not rebuilt until they are built, but changes in order-only
852 dependencies alone do not cause the output to be rebuilt.
854 Order-only dependencies can be useful for bootstrapping dependencies
855 that are only discovered during build time: for example, to generate a
856 header file before starting a subsequent compilation step. (Once the
857 header is used in compilation, a generated dependency file will then
858 express the implicit dependency.)
863 Variables are expanded in paths (in a `build` or `default` statement)
864 and on the right side of a `name = value` statement.
866 When a `name = value` statement is evaluated, its right-hand side is
867 expanded immediately (according to the below scoping rules), and
868 from then on `$name` expands to the static string as the result of the
869 expansion. It is never the case that you'll need to "double-escape" a
870 value to prevent it from getting expanded twice.
872 All variables are expanded immediately as they're encountered in parsing,
873 with one important exception: variables in `rule` blocks are expanded
874 when the rule is _used_, not when it is declared. In the following
875 example, the `demo` rule prints "this is a demo of bar".
879 command = echo "this is a demo of $foo'
886 Evaluation and scoping
887 ~~~~~~~~~~~~~~~~~~~~~~
889 Top-level variable declarations are scoped to the file they occur in.
891 The `subninja` keyword, used to include another `.ninja` file,
892 introduces a new scope. The included `subninja` file may use the
893 variables from the parent file, and shadow their values for the file's
894 scope, but it won't affect values of the variables in the parent.
896 To include another `.ninja` file in the current scope, much like a C
897 `#include` statement, use `include` instead of `subninja`.
899 Variable declarations indented in a `build` block are scoped to the
900 `build` block. The full lookup order for a variable expanded in a
901 `build` block (or the `rule` is uses) is:
903 1. Special built-in variables (`$in`, `$out`).
905 2. Build-level variables from the `build` block.
907 3. Rule-level variables from the `rule` block (i.e. `$command`).
908 (Note from the above discussion on expansion that these are
909 expanded "late", and may make use of in-scope bindings like `$in`.)
911 4. File-level variables from the file that the `build` line was in.
913 5. Variables from the file that included that file using the