3 ==============================
4 CommandLine 2.0 Library Manual
5 ==============================
10 This document describes the CommandLine argument processing library. It will
11 show you how to use it, and what it can do. The CommandLine library uses a
12 declarative approach to specifying the command line options that your program
13 takes. By default, these options declarations implicitly hold the value parsed
14 for the option declared (of course this `can be changed`_).
16 Although there are a **lot** of command line argument parsing libraries out
17 there in many different languages, none of them fit well with what I needed. By
18 looking at the features and problems of other libraries, I designed the
19 CommandLine library to have the following features:
21 #. Speed: The CommandLine library is very quick and uses little resources. The
22 parsing time of the library is directly proportional to the number of
23 arguments parsed, not the number of options recognized. Additionally,
24 command line argument values are captured transparently into user defined
25 global variables, which can be accessed like any other variable (and with the
28 #. Type Safe: As a user of CommandLine, you don't have to worry about
29 remembering the type of arguments that you want (is it an int? a string? a
30 bool? an enum?) and keep casting it around. Not only does this help prevent
31 error prone constructs, it also leads to dramatically cleaner source code.
33 #. No subclasses required: To use CommandLine, you instantiate variables that
34 correspond to the arguments that you would like to capture, you don't
35 subclass a parser. This means that you don't have to write **any**
38 #. Globally accessible: Libraries can specify command line arguments that are
39 automatically enabled in any tool that links to the library. This is
40 possible because the application doesn't have to keep a list of arguments to
41 pass to the parser. This also makes supporting `dynamically loaded options`_
44 #. Cleaner: CommandLine supports enum and other types directly, meaning that
45 there is less error and more security built into the library. You don't have
46 to worry about whether your integral command line argument accidentally got
47 assigned a value that is not valid for your enum type.
49 #. Powerful: The CommandLine library supports many different types of arguments,
50 from simple `boolean flags`_ to `scalars arguments`_ (`strings`_,
51 `integers`_, `enums`_, `doubles`_), to `lists of arguments`_. This is
52 possible because CommandLine is...
54 #. Extensible: It is very simple to add a new argument type to CommandLine.
55 Simply specify the parser that you want to use with the command line option
56 when you declare it. `Custom parsers`_ are no problem.
58 #. Labor Saving: The CommandLine library cuts down on the amount of grunt work
59 that you, the user, have to do. For example, it automatically provides a
60 ``-help`` option that shows the available command line options for your tool.
61 Additionally, it does most of the basic correctness checking for you.
63 #. Capable: The CommandLine library can handle lots of different forms of
64 options often found in real programs. For example, `positional`_ arguments,
65 ``ls`` style `grouping`_ options (to allow processing '``ls -lad``'
66 naturally), ``ld`` style `prefix`_ options (to parse '``-lmalloc
67 -L/usr/lib``'), and interpreter style options.
69 This document will hopefully let you jump in and start using CommandLine in your
70 utility quickly and painlessly. Additionally it should be a simple reference
71 manual to figure out how stuff works. If it is failing in some area (or you
72 want an extension to the library), nag the author, `Chris
73 Lattner <mailto:sabre@nondot.org>`_.
78 This section of the manual runs through a simple CommandLine'ification of a
79 basic compiler tool. This is intended to show you how to jump into using the
80 CommandLine library in your own program, and show you some of the cool things it
83 To start out, you need to include the CommandLine header file into your program:
87 #include "llvm/Support/CommandLine.h"
89 Additionally, you need to add this as the first line of your main program:
93 int main(int argc, char **argv) {
94 cl::ParseCommandLineOptions(argc, argv);
98 ... which actually parses the arguments and fills in the variable declarations.
100 Now that you are ready to support command line arguments, we need to tell the
101 system which ones we want, and what type of arguments they are. The CommandLine
102 library uses a declarative syntax to model command line arguments with the
103 global variable declarations that capture the parsed values. This means that
104 for every command line option that you would like to support, there should be a
105 global variable declaration to capture the result. For example, in a compiler,
106 we would like to support the Unix-standard '``-o <filename>``' option to specify
107 where to put the output. With the CommandLine library, this is represented like
110 .. _scalars arguments:
115 cl::opt<string> OutputFilename("o", cl::desc("Specify output filename"), cl::value_desc("filename"));
117 This declares a global variable "``OutputFilename``" that is used to capture the
118 result of the "``o``" argument (first parameter). We specify that this is a
119 simple scalar option by using the "``cl::opt``" template (as opposed to the
120 "``cl::list``" template), and tell the CommandLine library that the data
121 type that we are parsing is a string.
123 The second and third parameters (which are optional) are used to specify what to
124 output for the "``-help``" option. In this case, we get a line that looks like
129 USAGE: compiler [options]
132 -help - display available options (-help-hidden for more)
133 -o <filename> - Specify output filename
135 Because we specified that the command line option should parse using the
136 ``string`` data type, the variable declared is automatically usable as a real
137 string in all contexts that a normal C++ string object may be used. For
143 std::ofstream Output(OutputFilename.c_str());
144 if (Output.good()) ...
147 There are many different options that you can use to customize the command line
148 option handling library, but the above example shows the general interface to
149 these options. The options can be specified in any order, and are specified
150 with helper functions like `cl::desc(...)`_, so there are no positional
151 dependencies to remember. The available options are discussed in detail in the
154 Continuing the example, we would like to have our compiler take an input
155 filename as well as an output filename, but we do not want the input filename to
156 be specified with a hyphen (ie, not ``-filename.c``). To support this style of
157 argument, the CommandLine library allows for `positional`_ arguments to be
158 specified for the program. These positional arguments are filled with command
159 line parameters that are not in option form. We use this feature like this:
164 cl::opt<string> InputFilename(cl::Positional, cl::desc("<input file>"), cl::init("-"));
166 This declaration indicates that the first positional argument should be treated
167 as the input filename. Here we use the `cl::init`_ option to specify an initial
168 value for the command line option, which is used if the option is not specified
169 (if you do not specify a `cl::init`_ modifier for an option, then the default
170 constructor for the data type is used to initialize the value). Command line
171 options default to being optional, so if we would like to require that the user
172 always specify an input filename, we would add the `cl::Required`_ flag, and we
173 could eliminate the `cl::init`_ modifier, like this:
177 cl::opt<string> InputFilename(cl::Positional, cl::desc("<input file>"), cl::Required);
179 Again, the CommandLine library does not require the options to be specified in
180 any particular order, so the above declaration is equivalent to:
184 cl::opt<string> InputFilename(cl::Positional, cl::Required, cl::desc("<input file>"));
186 By simply adding the `cl::Required`_ flag, the CommandLine library will
187 automatically issue an error if the argument is not specified, which shifts all
188 of the command line option verification code out of your application into the
189 library. This is just one example of how using flags can alter the default
190 behaviour of the library, on a per-option basis. By adding one of the
191 declarations above, the ``-help`` option synopsis is now extended to:
195 USAGE: compiler [options] <input file>
198 -help - display available options (-help-hidden for more)
199 -o <filename> - Specify output filename
201 ... indicating that an input filename is expected.
206 In addition to input and output filenames, we would like the compiler example to
207 support three boolean flags: "``-f``" to force writing binary output to a
208 terminal, "``--quiet``" to enable quiet mode, and "``-q``" for backwards
209 compatibility with some of our users. We can support these by declaring options
210 of boolean type like this:
214 cl::opt<bool> Force ("f", cl::desc("Enable binary output on terminals"));
215 cl::opt<bool> Quiet ("quiet", cl::desc("Don't print informational messages"));
216 cl::opt<bool> Quiet2("q", cl::desc("Don't print informational messages"), cl::Hidden);
218 This does what you would expect: it declares three boolean variables
219 ("``Force``", "``Quiet``", and "``Quiet2``") to recognize these options. Note
220 that the "``-q``" option is specified with the "`cl::Hidden`_" flag. This
221 modifier prevents it from being shown by the standard "``-help``" output (note
222 that it is still shown in the "``-help-hidden``" output).
224 The CommandLine library uses a `different parser`_ for different data types.
225 For example, in the string case, the argument passed to the option is copied
226 literally into the content of the string variable... we obviously cannot do that
227 in the boolean case, however, so we must use a smarter parser. In the case of
228 the boolean parser, it allows no options (in which case it assigns the value of
229 true to the variable), or it allows the values "``true``" or "``false``" to be
230 specified, allowing any of the following inputs:
234 compiler -f # No value, 'Force' == true
235 compiler -f=true # Value specified, 'Force' == true
236 compiler -f=TRUE # Value specified, 'Force' == true
237 compiler -f=FALSE # Value specified, 'Force' == false
239 ... you get the idea. The `bool parser`_ just turns the string values into
240 boolean values, and rejects things like '``compiler -f=foo``'. Similarly, the
241 `float`_, `double`_, and `int`_ parsers work like you would expect, using the
242 '``strtol``' and '``strtod``' C library calls to parse the string value into the
245 With the declarations above, "``compiler -help``" emits this:
249 USAGE: compiler [options] <input file>
252 -f - Enable binary output on terminals
253 -o - Override output filename
254 -quiet - Don't print informational messages
255 -help - display available options (-help-hidden for more)
257 and "``compiler -help-hidden``" prints this:
261 USAGE: compiler [options] <input file>
264 -f - Enable binary output on terminals
265 -o - Override output filename
266 -q - Don't print informational messages
267 -quiet - Don't print informational messages
268 -help - display available options (-help-hidden for more)
270 This brief example has shown you how to use the '`cl::opt`_' class to parse
271 simple scalar command line arguments. In addition to simple scalar arguments,
272 the CommandLine library also provides primitives to support CommandLine option
273 `aliases`_, and `lists`_ of options.
280 So far, the example works well, except for the fact that we need to check the
281 quiet condition like this now:
286 if (!Quiet && !Quiet2) printInformationalMessage(...);
289 ... which is a real pain! Instead of defining two values for the same
290 condition, we can use the "`cl::alias`_" class to make the "``-q``" option an
291 **alias** for the "``-quiet``" option, instead of providing a value itself:
295 cl::opt<bool> Force ("f", cl::desc("Overwrite output files"));
296 cl::opt<bool> Quiet ("quiet", cl::desc("Don't print informational messages"));
297 cl::alias QuietA("q", cl::desc("Alias for -quiet"), cl::aliasopt(Quiet));
299 The third line (which is the only one we modified from above) defines a "``-q``"
300 alias that updates the "``Quiet``" variable (as specified by the `cl::aliasopt`_
301 modifier) whenever it is specified. Because aliases do not hold state, the only
302 thing the program has to query is the ``Quiet`` variable now. Another nice
303 feature of aliases is that they automatically hide themselves from the ``-help``
304 output (although, again, they are still visible in the ``-help-hidden output``).
306 Now the application code can simply use:
311 if (!Quiet) printInformationalMessage(...);
314 ... which is much nicer! The "`cl::alias`_" can be used to specify an
315 alternative name for any variable type, and has many uses.
317 .. _unnamed alternatives using the generic parser:
319 Selecting an alternative from a set of possibilities
320 ----------------------------------------------------
322 So far we have seen how the CommandLine library handles builtin types like
323 ``std::string``, ``bool`` and ``int``, but how does it handle things it doesn't
324 know about, like enums or '``int*``'s?
326 The answer is that it uses a table-driven generic parser (unless you specify
327 your own parser, as described in the `Extension Guide`_). This parser maps
328 literal strings to whatever type is required, and requires you to tell it what
329 this mapping should be.
331 Let's say that we would like to add four optimization levels to our optimizer,
332 using the standard flags "``-g``", "``-O0``", "``-O1``", and "``-O2``". We
333 could easily implement this with boolean options like above, but there are
334 several problems with this strategy:
336 #. A user could specify more than one of the options at a time, for example,
337 "``compiler -O3 -O2``". The CommandLine library would not be able to catch
338 this erroneous input for us.
340 #. We would have to test 4 different variables to see which ones are set.
342 #. This doesn't map to the numeric levels that we want... so we cannot easily
343 see if some level >= "``-O1``" is enabled.
345 To cope with these problems, we can use an enum value, and have the CommandLine
346 library fill it in with the appropriate level directly, which is used like this:
354 cl::opt<OptLevel> OptimizationLevel(cl::desc("Choose optimization level:"),
356 clEnumVal(g , "No optimizations, enable debugging"),
357 clEnumVal(O1, "Enable trivial optimizations"),
358 clEnumVal(O2, "Enable default optimizations"),
359 clEnumVal(O3, "Enable expensive optimizations"),
363 if (OptimizationLevel >= O2) doPartialRedundancyElimination(...);
366 This declaration defines a variable "``OptimizationLevel``" of the
367 "``OptLevel``" enum type. This variable can be assigned any of the values that
368 are listed in the declaration (Note that the declaration list must be terminated
369 with the "``clEnumValEnd``" argument!). The CommandLine library enforces that
370 the user can only specify one of the options, and it ensure that only valid enum
371 values can be specified. The "``clEnumVal``" macros ensure that the command
372 line arguments matched the enum values. With this option added, our help output
377 USAGE: compiler [options] <input file>
380 Choose optimization level:
381 -g - No optimizations, enable debugging
382 -O1 - Enable trivial optimizations
383 -O2 - Enable default optimizations
384 -O3 - Enable expensive optimizations
385 -f - Enable binary output on terminals
386 -help - display available options (-help-hidden for more)
387 -o <filename> - Specify output filename
388 -quiet - Don't print informational messages
390 In this case, it is sort of awkward that flag names correspond directly to enum
391 names, because we probably don't want a enum definition named "``g``" in our
392 program. Because of this, we can alternatively write this example like this:
400 cl::opt<OptLevel> OptimizationLevel(cl::desc("Choose optimization level:"),
402 clEnumValN(Debug, "g", "No optimizations, enable debugging"),
403 clEnumVal(O1 , "Enable trivial optimizations"),
404 clEnumVal(O2 , "Enable default optimizations"),
405 clEnumVal(O3 , "Enable expensive optimizations"),
409 if (OptimizationLevel == Debug) outputDebugInfo(...);
412 By using the "``clEnumValN``" macro instead of "``clEnumVal``", we can directly
413 specify the name that the flag should get. In general a direct mapping is nice,
414 but sometimes you can't or don't want to preserve the mapping, which is when you
420 Another useful argument form is a named alternative style. We shall use this
421 style in our compiler to specify different debug levels that can be used.
422 Instead of each debug level being its own switch, we want to support the
423 following options, of which only one can be specified at a time:
424 "``--debug-level=none``", "``--debug-level=quick``",
425 "``--debug-level=detailed``". To do this, we use the exact same format as our
426 optimization level flags, but we also specify an option name. For this case,
427 the code looks like this:
432 nodebuginfo, quick, detailed
435 // Enable Debug Options to be specified on the command line
436 cl::opt<DebugLev> DebugLevel("debug_level", cl::desc("Set the debugging level:"),
438 clEnumValN(nodebuginfo, "none", "disable debug information"),
439 clEnumVal(quick, "enable quick debug information"),
440 clEnumVal(detailed, "enable detailed debug information"),
443 This definition defines an enumerated command line variable of type "``enum
444 DebugLev``", which works exactly the same way as before. The difference here is
445 just the interface exposed to the user of your program and the help output by
446 the "``-help``" option:
450 USAGE: compiler [options] <input file>
453 Choose optimization level:
454 -g - No optimizations, enable debugging
455 -O1 - Enable trivial optimizations
456 -O2 - Enable default optimizations
457 -O3 - Enable expensive optimizations
458 -debug_level - Set the debugging level:
459 =none - disable debug information
460 =quick - enable quick debug information
461 =detailed - enable detailed debug information
462 -f - Enable binary output on terminals
463 -help - display available options (-help-hidden for more)
464 -o <filename> - Specify output filename
465 -quiet - Don't print informational messages
467 Again, the only structural difference between the debug level declaration and
468 the optimization level declaration is that the debug level declaration includes
469 an option name (``"debug_level"``), which automatically changes how the library
470 processes the argument. The CommandLine library supports both forms so that you
471 can choose the form most appropriate for your application.
475 Parsing a list of options
476 -------------------------
478 Now that we have the standard run-of-the-mill argument types out of the way,
479 lets get a little wild and crazy. Lets say that we want our optimizer to accept
480 a **list** of optimizations to perform, allowing duplicates. For example, we
481 might want to run: "``compiler -dce -constprop -inline -dce -strip``". In this
482 case, the order of the arguments and the number of appearances is very
483 important. This is what the "``cl::list``" template is for. First, start by
484 defining an enum of the optimizations that you would like to perform:
489 // 'inline' is a C++ keyword, so name it 'inlining'
490 dce, constprop, inlining, strip
493 Then define your "``cl::list``" variable:
497 cl::list<Opts> OptimizationList(cl::desc("Available Optimizations:"),
499 clEnumVal(dce , "Dead Code Elimination"),
500 clEnumVal(constprop , "Constant Propagation"),
501 clEnumValN(inlining, "inline", "Procedure Integration"),
502 clEnumVal(strip , "Strip Symbols"),
505 This defines a variable that is conceptually of the type
506 "``std::vector<enum Opts>``". Thus, you can access it with standard vector
511 for (unsigned i = 0; i != OptimizationList.size(); ++i)
512 switch (OptimizationList[i])
515 ... to iterate through the list of options specified.
517 Note that the "``cl::list``" template is completely general and may be used with
518 any data types or other arguments that you can use with the "``cl::opt``"
519 template. One especially useful way to use a list is to capture all of the
520 positional arguments together if there may be more than one specified. In the
521 case of a linker, for example, the linker takes several '``.o``' files, and
522 needs to capture them into a list. This is naturally specified as:
527 cl::list<std::string> InputFilenames(cl::Positional, cl::desc("<Input files>"), cl::OneOrMore);
530 This variable works just like a "``vector<string>``" object. As such, accessing
531 the list is simple, just like above. In this example, we used the
532 `cl::OneOrMore`_ modifier to inform the CommandLine library that it is an error
533 if the user does not specify any ``.o`` files on our command line. Again, this
534 just reduces the amount of checking we have to do.
536 Collecting options as a set of flags
537 ------------------------------------
539 Instead of collecting sets of options in a list, it is also possible to gather
540 information for enum values in a **bit vector**. The representation used by the
541 `cl::bits`_ class is an ``unsigned`` integer. An enum value is represented by a
542 0/1 in the enum's ordinal value bit position. 1 indicating that the enum was
543 specified, 0 otherwise. As each specified value is parsed, the resulting enum's
544 bit is set in the option's bit vector:
548 bits |= 1 << (unsigned)enum;
550 Options that are specified multiple times are redundant. Any instances after
551 the first are discarded.
553 Reworking the above list example, we could replace `cl::list`_ with `cl::bits`_:
557 cl::bits<Opts> OptimizationBits(cl::desc("Available Optimizations:"),
559 clEnumVal(dce , "Dead Code Elimination"),
560 clEnumVal(constprop , "Constant Propagation"),
561 clEnumValN(inlining, "inline", "Procedure Integration"),
562 clEnumVal(strip , "Strip Symbols"),
565 To test to see if ``constprop`` was specified, we can use the ``cl:bits::isSet``
570 if (OptimizationBits.isSet(constprop)) {
574 It's also possible to get the raw bit vector using the ``cl::bits::getBits``
579 unsigned bits = OptimizationBits.getBits();
581 Finally, if external storage is used, then the location specified must be of
582 **type** ``unsigned``. In all other ways a `cl::bits`_ option is equivalent to a
585 .. _additional extra text:
587 Adding freeform text to help output
588 -----------------------------------
590 As our program grows and becomes more mature, we may decide to put summary
591 information about what it does into the help output. The help output is styled
592 to look similar to a Unix ``man`` page, providing concise information about a
593 program. Unix ``man`` pages, however often have a description about what the
594 program does. To add this to your CommandLine program, simply pass a third
595 argument to the `cl::ParseCommandLineOptions`_ call in main. This additional
596 argument is then printed as the overview information for your program, allowing
597 you to include any additional information that you want. For example:
601 int main(int argc, char **argv) {
602 cl::ParseCommandLineOptions(argc, argv, " CommandLine compiler example\n\n"
603 " This program blah blah blah...\n");
607 would yield the help output:
611 **OVERVIEW: CommandLine compiler example
613 This program blah blah blah...**
615 USAGE: compiler [options] <input file>
619 -help - display available options (-help-hidden for more)
620 -o <filename> - Specify output filename
627 Now that you know the basics of how to use the CommandLine library, this section
628 will give you the detailed information you need to tune how command line options
629 work, as well as information on more "advanced" command line option processing
633 .. _positional argument:
634 .. _Positional Arguments:
635 .. _Positional arguments section:
636 .. _positional options:
641 Positional arguments are those arguments that are not named, and are not
642 specified with a hyphen. Positional arguments should be used when an option is
643 specified by its position alone. For example, the standard Unix ``grep`` tool
644 takes a regular expression argument, and an optional filename to search through
645 (which defaults to standard input if a filename is not specified). Using the
646 CommandLine library, this would be specified as:
650 cl::opt<string> Regex (cl::Positional, cl::desc("<regular expression>"), cl::Required);
651 cl::opt<string> Filename(cl::Positional, cl::desc("<input file>"), cl::init("-"));
653 Given these two option declarations, the ``-help`` output for our grep
654 replacement would look like this:
658 USAGE: spiffygrep [options] <regular expression> <input file>
661 -help - display available options (-help-hidden for more)
663 ... and the resultant program could be used just like the standard ``grep``
666 Positional arguments are sorted by their order of construction. This means that
667 command line options will be ordered according to how they are listed in a .cpp
668 file, but will not have an ordering defined if the positional arguments are
669 defined in multiple .cpp files. The fix for this problem is simply to define
670 all of your positional arguments in one .cpp file.
672 Specifying positional options with hyphens
673 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
675 Sometimes you may want to specify a value to your positional argument that
676 starts with a hyphen (for example, searching for '``-foo``' in a file). At
677 first, you will have trouble doing this, because it will try to find an argument
678 named '``-foo``', and will fail (and single quotes will not save you). Note
679 that the system ``grep`` has the same problem:
683 $ spiffygrep '-foo' test.txt
684 Unknown command line argument '-foo'. Try: spiffygrep -help'
686 $ grep '-foo' test.txt
687 grep: illegal option -- f
688 grep: illegal option -- o
689 grep: illegal option -- o
690 Usage: grep -hblcnsviw pattern file . . .
692 The solution for this problem is the same for both your tool and the system
693 version: use the '``--``' marker. When the user specifies '``--``' on the
694 command line, it is telling the program that all options after the '``--``'
695 should be treated as positional arguments, not options. Thus, we can use it
700 $ spiffygrep -- -foo test.txt
703 Determining absolute position with getPosition()
704 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
706 Sometimes an option can affect or modify the meaning of another option. For
707 example, consider ``gcc``'s ``-x LANG`` option. This tells ``gcc`` to ignore the
708 suffix of subsequent positional arguments and force the file to be interpreted
709 as if it contained source code in language ``LANG``. In order to handle this
710 properly, you need to know the absolute position of each argument, especially
711 those in lists, so their interaction(s) can be applied correctly. This is also
712 useful for options like ``-llibname`` which is actually a positional argument
713 that starts with a dash.
715 So, generally, the problem is that you have two ``cl::list`` variables that
716 interact in some way. To ensure the correct interaction, you can use the
717 ``cl::list::getPosition(optnum)`` method. This method returns the absolute
718 position (as found on the command line) of the ``optnum`` item in the
721 The idiom for usage is like this:
725 static cl::list<std::string> Files(cl::Positional, cl::OneOrMore);
726 static cl::list<std::string> Libraries("l", cl::ZeroOrMore);
728 int main(int argc, char**argv) {
730 std::vector<std::string>::iterator fileIt = Files.begin();
731 std::vector<std::string>::iterator libIt = Libraries.begin();
732 unsigned libPos = 0, filePos = 0;
734 if ( libIt != Libraries.end() )
735 libPos = Libraries.getPosition( libIt - Libraries.begin() );
738 if ( fileIt != Files.end() )
739 filePos = Files.getPosition( fileIt - Files.begin() );
743 if ( filePos != 0 && (libPos == 0 || filePos < libPos) ) {
744 // Source File Is next
747 else if ( libPos != 0 && (filePos == 0 || libPos < filePos) ) {
752 break; // we're done with the list
756 Note that, for compatibility reasons, the ``cl::opt`` also supports an
757 ``unsigned getPosition()`` option that will provide the absolute position of
758 that option. You can apply the same approach as above with a ``cl::opt`` and a
759 ``cl::list`` option as you can with two lists.
761 .. _interpreter style options:
762 .. _cl::ConsumeAfter:
763 .. _this section for more information:
765 The ``cl::ConsumeAfter`` modifier
766 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
768 The ``cl::ConsumeAfter`` `formatting option`_ is used to construct programs that
769 use "interpreter style" option processing. With this style of option
770 processing, all arguments specified after the last positional argument are
771 treated as special interpreter arguments that are not interpreted by the command
774 As a concrete example, lets say we are developing a replacement for the standard
775 Unix Bourne shell (``/bin/sh``). To run ``/bin/sh``, first you specify options
776 to the shell itself (like ``-x`` which turns on trace output), then you specify
777 the name of the script to run, then you specify arguments to the script. These
778 arguments to the script are parsed by the Bourne shell command line option
779 processor, but are not interpreted as options to the shell itself. Using the
780 CommandLine library, we would specify this as:
784 cl::opt<string> Script(cl::Positional, cl::desc("<input script>"), cl::init("-"));
785 cl::list<string> Argv(cl::ConsumeAfter, cl::desc("<program arguments>..."));
786 cl::opt<bool> Trace("x", cl::desc("Enable trace output"));
788 which automatically provides the help output:
792 USAGE: spiffysh [options] <input script> <program arguments>...
795 -help - display available options (-help-hidden for more)
796 -x - Enable trace output
798 At runtime, if we run our new shell replacement as ```spiffysh -x test.sh -a -x
799 -y bar``', the ``Trace`` variable will be set to true, the ``Script`` variable
800 will be set to "``test.sh``", and the ``Argv`` list will contain ``["-a", "-x",
801 "-y", "bar"]``, because they were specified after the last positional argument
802 (which is the script name).
804 There are several limitations to when ``cl::ConsumeAfter`` options can be
805 specified. For example, only one ``cl::ConsumeAfter`` can be specified per
806 program, there must be at least one `positional argument`_ specified, there must
807 not be any `cl::list`_ positional arguments, and the ``cl::ConsumeAfter`` option
808 should be a `cl::list`_ option.
811 .. _Internal vs External Storage:
813 Internal vs External Storage
814 ----------------------------
816 By default, all command line options automatically hold the value that they
817 parse from the command line. This is very convenient in the common case,
818 especially when combined with the ability to define command line options in the
819 files that use them. This is called the internal storage model.
821 Sometimes, however, it is nice to separate the command line option processing
822 code from the storage of the value parsed. For example, lets say that we have a
823 '``-debug``' option that we would like to use to enable debug information across
824 the entire body of our program. In this case, the boolean value controlling the
825 debug code should be globally accessible (in a header file, for example) yet the
826 command line option processing code should not be exposed to all of these
827 clients (requiring lots of .cpp files to ``#include CommandLine.h``).
829 To do this, set up your .h file with your option, like this for example:
833 // DebugFlag.h - Get access to the '-debug' command line option
836 // DebugFlag - This boolean is set to true if the '-debug' command line option
837 // is specified. This should probably not be referenced directly, instead, use
838 // the DEBUG macro below.
840 extern bool DebugFlag;
842 // DEBUG macro - This macro should be used by code to emit debug information.
843 // In the '-debug' option is specified on the command line, and if this is a
844 // debug build, then the code specified as the option to the macro will be
845 // executed. Otherwise it will not be.
849 #define DEBUG(X) do { if (DebugFlag) { X; } } while (0)
852 This allows clients to blissfully use the ``DEBUG()`` macro, or the
853 ``DebugFlag`` explicitly if they want to. Now we just need to be able to set
854 the ``DebugFlag`` boolean when the option is set. To do this, we pass an
855 additional argument to our command line argument processor, and we specify where
856 to fill in with the `cl::location`_ attribute:
860 bool DebugFlag; // the actual value
861 static cl::opt<bool, true> // The parser
862 Debug("debug", cl::desc("Enable debug output"), cl::Hidden, cl::location(DebugFlag));
864 In the above example, we specify "``true``" as the second argument to the
865 `cl::opt`_ template, indicating that the template should not maintain a copy of
866 the value itself. In addition to this, we specify the `cl::location`_
867 attribute, so that ``DebugFlag`` is automatically set.
872 This section describes the basic attributes that you can specify on options.
874 * The option name attribute (which is required for all options, except
875 `positional options`_) specifies what the option name is. This option is
876 specified in simple double quotes:
880 cl::opt<**bool**> Quiet("quiet");
884 * The **cl::desc** attribute specifies a description for the option to be
885 shown in the ``-help`` output for the program.
889 * The **cl::value_desc** attribute specifies a string that can be used to
890 fine tune the ``-help`` output for a command line option. Look `here`_ for an
895 * The **cl::init** attribute specifies an initial value for a `scalar`_
896 option. If this attribute is not specified then the command line option value
897 defaults to the value created by the default constructor for the
902 If you specify both **cl::init** and **cl::location** for an option, you
903 must specify **cl::location** first, so that when the command-line parser
904 sees **cl::init**, it knows where to put the initial value. (You will get an
905 error at runtime if you don't put them in the right order.)
909 * The **cl::location** attribute where to store the value for a parsed command
910 line option if using external storage. See the section on `Internal vs
911 External Storage`_ for more information.
915 * The **cl::aliasopt** attribute specifies which option a `cl::alias`_ option is
920 * The **cl::values** attribute specifies the string-to-value mapping to be used
921 by the generic parser. It takes a **clEnumValEnd terminated** list of
922 (option, value, description) triplets that specify the option name, the value
923 mapped to, and the description shown in the ``-help`` for the tool. Because
924 the generic parser is used most frequently with enum values, two macros are
927 #. The **clEnumVal** macro is used as a nice simple way to specify a triplet
928 for an enum. This macro automatically makes the option name be the same as
929 the enum name. The first option to the macro is the enum, the second is
930 the description for the command line option.
932 #. The **clEnumValN** macro is used to specify macro options where the option
933 name doesn't equal the enum name. For this macro, the first argument is
934 the enum value, the second is the flag name, and the second is the
937 You will get a compile time error if you try to use cl::values with a parser
938 that does not support it.
942 * The **cl::multi_val** attribute specifies that this option takes has multiple
943 values (example: ``-sectalign segname sectname sectvalue``). This attribute
944 takes one unsigned argument - the number of values for the option. This
945 attribute is valid only on ``cl::list`` options (and will fail with compile
946 error if you try to use it with other option types). It is allowed to use all
947 of the usual modifiers on multi-valued options (besides
948 ``cl::ValueDisallowed``, obviously).
953 Option modifiers are the flags and expressions that you pass into the
954 constructors for `cl::opt`_ and `cl::list`_. These modifiers give you the
955 ability to tweak how options are parsed and how ``-help`` output is generated to
956 fit your application well.
958 These options fall into five main categories:
960 #. Hiding an option from ``-help`` output
962 #. Controlling the number of occurrences required and allowed
964 #. Controlling whether or not a value must be specified
966 #. Controlling other formatting options
968 #. Miscellaneous option modifiers
970 It is not possible to specify two options from the same category (you'll get a
971 runtime error) to a single option, except for options in the miscellaneous
972 category. The CommandLine library specifies defaults for all of these settings
973 that are the most useful in practice and the most common, which mean that you
974 usually shouldn't have to worry about these.
976 Hiding an option from ``-help`` output
977 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
979 The ``cl::NotHidden``, ``cl::Hidden``, and ``cl::ReallyHidden`` modifiers are
980 used to control whether or not an option appears in the ``-help`` and
981 ``-help-hidden`` output for the compiled program:
985 * The **cl::NotHidden** modifier (which is the default for `cl::opt`_ and
986 `cl::list`_ options) indicates the option is to appear in both help
991 * The **cl::Hidden** modifier (which is the default for `cl::alias`_ options)
992 indicates that the option should not appear in the ``-help`` output, but
993 should appear in the ``-help-hidden`` output.
995 .. _cl::ReallyHidden:
997 * The **cl::ReallyHidden** modifier indicates that the option should not appear
1000 Controlling the number of occurrences required and allowed
1001 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1003 This group of options is used to control how many time an option is allowed (or
1004 required) to be specified on the command line of your program. Specifying a
1005 value for this setting allows the CommandLine library to do error checking for
1008 The allowed values for this option group are:
1012 * The **cl::Optional** modifier (which is the default for the `cl::opt`_ and
1013 `cl::alias`_ classes) indicates that your program will allow either zero or
1014 one occurrence of the option to be specified.
1018 * The **cl::ZeroOrMore** modifier (which is the default for the `cl::list`_
1019 class) indicates that your program will allow the option to be specified zero
1024 * The **cl::Required** modifier indicates that the specified option must be
1025 specified exactly one time.
1029 * The **cl::OneOrMore** modifier indicates that the option must be specified at
1032 * The **cl::ConsumeAfter** modifier is described in the `Positional arguments
1035 If an option is not specified, then the value of the option is equal to the
1036 value specified by the `cl::init`_ attribute. If the ``cl::init`` attribute is
1037 not specified, the option value is initialized with the default constructor for
1040 If an option is specified multiple times for an option of the `cl::opt`_ class,
1041 only the last value will be retained.
1043 Controlling whether or not a value must be specified
1044 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1046 This group of options is used to control whether or not the option allows a
1047 value to be present. In the case of the CommandLine library, a value is either
1048 specified with an equal sign (e.g. '``-index-depth=17``') or as a trailing
1049 string (e.g. '``-o a.out``').
1051 The allowed values for this option group are:
1053 .. _cl::ValueOptional:
1055 * The **cl::ValueOptional** modifier (which is the default for ``bool`` typed
1056 options) specifies that it is acceptable to have a value, or not. A boolean
1057 argument can be enabled just by appearing on the command line, or it can have
1058 an explicit '``-foo=true``'. If an option is specified with this mode, it is
1059 illegal for the value to be provided without the equal sign. Therefore
1060 '``-foo true``' is illegal. To get this behavior, you must use
1061 the `cl::ValueRequired`_ modifier.
1063 .. _cl::ValueRequired:
1065 * The **cl::ValueRequired** modifier (which is the default for all other types
1066 except for `unnamed alternatives using the generic parser`_) specifies that a
1067 value must be provided. This mode informs the command line library that if an
1068 option is not provides with an equal sign, that the next argument provided
1069 must be the value. This allows things like '``-o a.out``' to work.
1071 .. _cl::ValueDisallowed:
1073 * The **cl::ValueDisallowed** modifier (which is the default for `unnamed
1074 alternatives using the generic parser`_) indicates that it is a runtime error
1075 for the user to specify a value. This can be provided to disallow users from
1076 providing options to boolean options (like '``-foo=true``').
1078 In general, the default values for this option group work just like you would
1079 want them to. As mentioned above, you can specify the `cl::ValueDisallowed`_
1080 modifier to a boolean argument to restrict your command line parser. These
1081 options are mostly useful when `extending the library`_.
1083 .. _formatting option:
1085 Controlling other formatting options
1086 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1088 The formatting option group is used to specify that the command line option has
1089 special abilities and is otherwise different from other command line arguments.
1090 As usual, you can only specify one of these arguments at most.
1092 .. _cl::NormalFormatting:
1094 * The **cl::NormalFormatting** modifier (which is the default all options)
1095 specifies that this option is "normal".
1099 * The **cl::Positional** modifier specifies that this is a positional argument
1100 that does not have a command line option associated with it. See the
1101 `Positional Arguments`_ section for more information.
1103 * The **cl::ConsumeAfter** modifier specifies that this option is used to
1104 capture "interpreter style" arguments. See `this section for more
1110 * The **cl::Prefix** modifier specifies that this option prefixes its value.
1111 With 'Prefix' options, the equal sign does not separate the value from the
1112 option name specified. Instead, the value is everything after the prefix,
1113 including any equal sign if present. This is useful for processing odd
1114 arguments like ``-lmalloc`` and ``-L/usr/lib`` in a linker tool or
1115 ``-DNAME=value`` in a compiler tool. Here, the '``l``', '``D``' and '``L``'
1116 options are normal string (or list) options, that have the **cl::Prefix**
1117 modifier added to allow the CommandLine library to recognize them. Note that
1118 **cl::Prefix** options must not have the **cl::ValueDisallowed** modifier
1124 * The **cl::Grouping** modifier is used to implement Unix-style tools (like
1125 ``ls``) that have lots of single letter arguments, but only require a single
1126 dash. For example, the '``ls -labF``' command actually enables four different
1127 options, all of which are single letters. Note that **cl::Grouping** options
1130 The CommandLine library does not restrict how you use the **cl::Prefix** or
1131 **cl::Grouping** modifiers, but it is possible to specify ambiguous argument
1132 settings. Thus, it is possible to have multiple letter options that are prefix
1133 or grouping options, and they will still work as designed.
1135 To do this, the CommandLine library uses a greedy algorithm to parse the input
1136 option into (potentially multiple) prefix and grouping options. The strategy
1137 basically looks like this:
1141 parse(string OrigInput) {
1143 1. string input = OrigInput;
1144 2. if (isOption(input)) return getOption(input).parse(); // Normal option
1145 3. while (!isOption(input) && !input.empty()) input.pop_back(); // Remove the last letter
1146 4. if (input.empty()) return error(); // No matching option
1147 5. if (getOption(input).isPrefix())
1148 return getOption(input).parse(input);
1149 6. while (!input.empty()) { // Must be grouping options
1150 getOption(input).parse();
1151 OrigInput.erase(OrigInput.begin(), OrigInput.begin()+input.length());
1153 while (!isOption(input) && !input.empty()) input.pop_back();
1155 7. if (!OrigInput.empty()) error();
1159 Miscellaneous option modifiers
1160 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1162 The miscellaneous option modifiers are the only flags where you can specify more
1163 than one flag from the set: they are not mutually exclusive. These flags
1164 specify boolean properties that modify the option.
1166 .. _cl::CommaSeparated:
1168 * The **cl::CommaSeparated** modifier indicates that any commas specified for an
1169 option's value should be used to split the value up into multiple values for
1170 the option. For example, these two options are equivalent when
1171 ``cl::CommaSeparated`` is specified: "``-foo=a -foo=b -foo=c``" and
1172 "``-foo=a,b,c``". This option only makes sense to be used in a case where the
1173 option is allowed to accept one or more values (i.e. it is a `cl::list`_
1176 .. _cl::PositionalEatsArgs:
1178 * The **cl::PositionalEatsArgs** modifier (which only applies to positional
1179 arguments, and only makes sense for lists) indicates that positional argument
1180 should consume any strings after it (including strings that start with a "-")
1181 up until another recognized positional argument. For example, if you have two
1182 "eating" positional arguments, "``pos1``" and "``pos2``", the string "``-pos1
1183 -foo -bar baz -pos2 -bork``" would cause the "``-foo -bar -baz``" strings to
1184 be applied to the "``-pos1``" option and the "``-bork``" string to be applied
1185 to the "``-pos2``" option.
1189 * The **cl::Sink** modifier is used to handle unknown options. If there is at
1190 least one option with ``cl::Sink`` modifier specified, the parser passes
1191 unrecognized option strings to it as values instead of signaling an error. As
1192 with ``cl::CommaSeparated``, this modifier only makes sense with a `cl::list`_
1195 So far, these are the only three miscellaneous option modifiers.
1202 Some systems, such as certain variants of Microsoft Windows and some older
1203 Unices have a relatively low limit on command-line length. It is therefore
1204 customary to use the so-called 'response files' to circumvent this
1205 restriction. These files are mentioned on the command-line (using the "@file")
1206 syntax. The program reads these files and inserts the contents into argv,
1207 thereby working around the command-line length limits. Response files are
1208 enabled by an optional fourth argument to `cl::ParseEnvironmentOptions`_ and
1209 `cl::ParseCommandLineOptions`_.
1211 Top-Level Classes and Functions
1212 -------------------------------
1214 Despite all of the built-in flexibility, the CommandLine option library really
1215 only consists of one function `cl::ParseCommandLineOptions`_) and three main
1216 classes: `cl::opt`_, `cl::list`_, and `cl::alias`_. This section describes
1217 these three classes in detail.
1219 .. _cl::ParseCommandLineOptions:
1221 The ``cl::ParseCommandLineOptions`` function
1222 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1224 The ``cl::ParseCommandLineOptions`` function is designed to be called directly
1225 from ``main``, and is used to fill in the values of all of the command line
1226 option variables once ``argc`` and ``argv`` are available.
1228 The ``cl::ParseCommandLineOptions`` function requires two parameters (``argc``
1229 and ``argv``), but may also take an optional third parameter which holds
1230 `additional extra text`_ to emit when the ``-help`` option is invoked, and a
1231 fourth boolean parameter that enables `response files`_.
1233 .. _cl::ParseEnvironmentOptions:
1235 The ``cl::ParseEnvironmentOptions`` function
1236 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1238 The ``cl::ParseEnvironmentOptions`` function has mostly the same effects as
1239 `cl::ParseCommandLineOptions`_, except that it is designed to take values for
1240 options from an environment variable, for those cases in which reading the
1241 command line is not convenient or desired. It fills in the values of all the
1242 command line option variables just like `cl::ParseCommandLineOptions`_ does.
1244 It takes four parameters: the name of the program (since ``argv`` may not be
1245 available, it can't just look in ``argv[0]``), the name of the environment
1246 variable to examine, the optional `additional extra text`_ to emit when the
1247 ``-help`` option is invoked, and the boolean switch that controls whether
1248 `response files`_ should be read.
1250 ``cl::ParseEnvironmentOptions`` will break the environment variable's value up
1251 into words and then process them using `cl::ParseCommandLineOptions`_.
1252 **Note:** Currently ``cl::ParseEnvironmentOptions`` does not support quoting, so
1253 an environment variable containing ``-option "foo bar"`` will be parsed as three
1254 words, ``-option``, ``"foo``, and ``bar"``, which is different from what you
1255 would get from the shell with the same input.
1257 The ``cl::SetVersionPrinter`` function
1258 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1260 The ``cl::SetVersionPrinter`` function is designed to be called directly from
1261 ``main`` and *before* ``cl::ParseCommandLineOptions``. Its use is optional. It
1262 simply arranges for a function to be called in response to the ``--version``
1263 option instead of having the ``CommandLine`` library print out the usual version
1264 string for LLVM. This is useful for programs that are not part of LLVM but wish
1265 to use the ``CommandLine`` facilities. Such programs should just define a small
1266 function that takes no arguments and returns ``void`` and that prints out
1267 whatever version information is appropriate for the program. Pass the address of
1268 that function to ``cl::SetVersionPrinter`` to arrange for it to be called when
1269 the ``--version`` option is given by the user.
1274 The ``cl::opt`` class
1275 ^^^^^^^^^^^^^^^^^^^^^
1277 The ``cl::opt`` class is the class used to represent scalar command line
1278 options, and is the one used most of the time. It is a templated class which
1279 can take up to three arguments (all except for the first have default values
1285 template <class DataType, bool ExternalStorage = false,
1286 class ParserClass = parser<DataType> >
1290 The first template argument specifies what underlying data type the command line
1291 argument is, and is used to select a default parser implementation. The second
1292 template argument is used to specify whether the option should contain the
1293 storage for the option (the default) or whether external storage should be used
1294 to contain the value parsed for the option (see `Internal vs External Storage`_
1295 for more information).
1297 The third template argument specifies which parser to use. The default value
1298 selects an instantiation of the ``parser`` class based on the underlying data
1299 type of the option. In general, this default works well for most applications,
1300 so this option is only used when using a `custom parser`_.
1302 .. _lists of arguments:
1305 The ``cl::list`` class
1306 ^^^^^^^^^^^^^^^^^^^^^^
1308 The ``cl::list`` class is the class used to represent a list of command line
1309 options. It too is a templated class which can take up to three arguments:
1314 template <class DataType, class Storage = bool,
1315 class ParserClass = parser<DataType> >
1319 This class works the exact same as the `cl::opt`_ class, except that the second
1320 argument is the **type** of the external storage, not a boolean value. For this
1321 class, the marker type '``bool``' is used to indicate that internal storage
1326 The ``cl::bits`` class
1327 ^^^^^^^^^^^^^^^^^^^^^^
1329 The ``cl::bits`` class is the class used to represent a list of command line
1330 options in the form of a bit vector. It is also a templated class which can
1331 take up to three arguments:
1336 template <class DataType, class Storage = bool,
1337 class ParserClass = parser<DataType> >
1341 This class works the exact same as the `cl::list`_ class, except that the second
1342 argument must be of **type** ``unsigned`` if external storage is used.
1346 The ``cl::alias`` class
1347 ^^^^^^^^^^^^^^^^^^^^^^^
1349 The ``cl::alias`` class is a nontemplated class that is used to form aliases for
1358 The `cl::aliasopt`_ attribute should be used to specify which option this is an
1359 alias for. Alias arguments default to being `cl::Hidden`_, and use the aliased
1360 options parser to do the conversion from string to data.
1364 The ``cl::extrahelp`` class
1365 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
1367 The ``cl::extrahelp`` class is a nontemplated class that allows extra help text
1368 to be printed out for the ``-help`` option.
1376 To use the extrahelp, simply construct one with a ``const char*`` parameter to
1377 the constructor. The text passed to the constructor will be printed at the
1378 bottom of the help message, verbatim. Note that multiple ``cl::extrahelp``
1379 **can** be used, but this practice is discouraged. If your tool needs to print
1380 additional help information, put all that help into a single ``cl::extrahelp``
1387 cl::extrahelp("\nADDITIONAL HELP:\n\n This is the extra help\n");
1389 .. _different parser:
1390 .. _discussed previously:
1395 Parsers control how the string value taken from the command line is translated
1396 into a typed value, suitable for use in a C++ program. By default, the
1397 CommandLine library uses an instance of ``parser<type>`` if the command line
1398 option specifies that it uses values of type '``type``'. Because of this,
1399 custom option processing is specified with specializations of the '``parser``'
1402 The CommandLine library provides the following builtin parser specializations,
1403 which are sufficient for most applications. It can, however, also be extended to
1404 work with new data types and new ways of interpreting the same data. See the
1405 `Writing a Custom Parser`_ for more details on this type of library extension.
1410 * The generic ``parser<t>`` parser can be used to map strings values to any data
1411 type, through the use of the `cl::values`_ property, which specifies the
1412 mapping information. The most common use of this parser is for parsing enum
1413 values, which allows you to use the CommandLine library for all of the error
1414 checking to make sure that only valid enum values are specified (as opposed to
1415 accepting arbitrary strings). Despite this, however, the generic parser class
1416 can be used for any data type.
1421 * The **parser<bool> specialization** is used to convert boolean strings to a
1422 boolean value. Currently accepted strings are "``true``", "``TRUE``",
1423 "``True``", "``1``", "``false``", "``FALSE``", "``False``", and "``0``".
1425 * The **parser<boolOrDefault> specialization** is used for cases where the value
1426 is boolean, but we also need to know whether the option was specified at all.
1427 boolOrDefault is an enum with 3 values, BOU_UNSET, BOU_TRUE and BOU_FALSE.
1428 This parser accepts the same strings as **``parser<bool>``**.
1432 * The **parser<string> specialization** simply stores the parsed string into the
1433 string value specified. No conversion or modification of the data is
1439 * The **parser<int> specialization** uses the C ``strtol`` function to parse the
1440 string input. As such, it will accept a decimal number (with an optional '+'
1441 or '-' prefix) which must start with a non-zero digit. It accepts octal
1442 numbers, which are identified with a '``0``' prefix digit, and hexadecimal
1443 numbers with a prefix of '``0x``' or '``0X``'.
1449 * The **parser<double>** and **parser<float> specializations** use the standard
1450 C ``strtod`` function to convert floating point strings into floating point
1451 values. As such, a broad range of string formats is supported, including
1452 exponential notation (ex: ``1.7e15``) and properly supports locales.
1454 .. _Extension Guide:
1455 .. _extending the library:
1460 Although the CommandLine library has a lot of functionality built into it
1461 already (as discussed previously), one of its true strengths lie in its
1462 extensibility. This section discusses how the CommandLine library works under
1463 the covers and illustrates how to do some simple, common, extensions.
1467 .. _Writing a Custom Parser:
1469 Writing a custom parser
1470 -----------------------
1472 One of the simplest and most common extensions is the use of a custom parser.
1473 As `discussed previously`_, parsers are the portion of the CommandLine library
1474 that turns string input from the user into a particular parsed data type,
1475 validating the input in the process.
1477 There are two ways to use a new parser:
1479 #. Specialize the `cl::parser`_ template for your custom data type.
1481 This approach has the advantage that users of your custom data type will
1482 automatically use your custom parser whenever they define an option with a
1483 value type of your data type. The disadvantage of this approach is that it
1484 doesn't work if your fundamental data type is something that is already
1487 #. Write an independent class, using it explicitly from options that need it.
1489 This approach works well in situations where you would line to parse an
1490 option using special syntax for a not-very-special data-type. The drawback
1491 of this approach is that users of your parser have to be aware that they are
1492 using your parser instead of the builtin ones.
1494 To guide the discussion, we will discuss a custom parser that accepts file
1495 sizes, specified with an optional unit after the numeric size. For example, we
1496 would like to parse "102kb", "41M", "1G" into the appropriate integer value. In
1497 this case, the underlying data type we want to parse into is '``unsigned``'. We
1498 choose approach #2 above because we don't want to make this the default for all
1499 ``unsigned`` options.
1501 To start out, we declare our new ``FileSizeParser`` class:
1505 struct FileSizeParser : public cl::basic_parser<unsigned> {
1506 // parse - Return true on error.
1507 bool parse(cl::Option &O, const char *ArgName, const std::string &ArgValue,
1511 Our new class inherits from the ``cl::basic_parser`` template class to fill in
1512 the default, boiler plate code for us. We give it the data type that we parse
1513 into, the last argument to the ``parse`` method, so that clients of our custom
1514 parser know what object type to pass in to the parse method. (Here we declare
1515 that we parse into '``unsigned``' variables.)
1517 For most purposes, the only method that must be implemented in a custom parser
1518 is the ``parse`` method. The ``parse`` method is called whenever the option is
1519 invoked, passing in the option itself, the option name, the string to parse, and
1520 a reference to a return value. If the string to parse is not well-formed, the
1521 parser should output an error message and return true. Otherwise it should
1522 return false and set '``Val``' to the parsed value. In our example, we
1523 implement ``parse`` as:
1527 bool FileSizeParser::parse(cl::Option &O, const char *ArgName,
1528 const std::string &Arg, unsigned &Val) {
1529 const char *ArgStart = Arg.c_str();
1532 // Parse integer part, leaving 'End' pointing to the first non-integer char
1533 Val = (unsigned)strtol(ArgStart, &End, 0);
1537 case 0: return false; // No error
1538 case 'i': // Ignore the 'i' in KiB if people use that
1539 case 'b': case 'B': // Ignore B suffix
1542 case 'g': case 'G': Val *= 1024*1024*1024; break;
1543 case 'm': case 'M': Val *= 1024*1024; break;
1544 case 'k': case 'K': Val *= 1024; break;
1547 // Print an error message if unrecognized character!
1548 return O.error("'" + Arg + "' value invalid for file size argument!");
1553 This function implements a very simple parser for the kinds of strings we are
1554 interested in. Although it has some holes (it allows "``123KKK``" for example),
1555 it is good enough for this example. Note that we use the option itself to print
1556 out the error message (the ``error`` method always returns true) in order to get
1557 a nice error message (shown below). Now that we have our parser class, we can
1562 static cl::opt<unsigned, false, FileSizeParser>
1563 MFS("max-file-size", cl::desc("Maximum file size to accept"),
1564 cl::value_desc("size"));
1566 Which adds this to the output of our program:
1571 -help - display available options (-help-hidden for more)
1573 -max-file-size=<size> - Maximum file size to accept
1575 And we can test that our parse works correctly now (the test program just prints
1576 out the max-file-size argument value):
1582 $ ./test -max-file-size=123MB
1584 $ ./test -max-file-size=3G
1586 $ ./test -max-file-size=dog
1587 -max-file-size option: 'dog' value invalid for file size argument!
1589 It looks like it works. The error message that we get is nice and helpful, and
1590 we seem to accept reasonable file sizes. This wraps up the "custom parser"
1593 Exploiting external storage
1594 ---------------------------
1596 Several of the LLVM libraries define static ``cl::opt`` instances that will
1597 automatically be included in any program that links with that library. This is
1598 a feature. However, sometimes it is necessary to know the value of the command
1599 line option outside of the library. In these cases the library does or should
1600 provide an external storage location that is accessible to users of the
1601 library. Examples of this include the ``llvm::DebugFlag`` exported by the
1602 ``lib/Support/Debug.cpp`` file and the ``llvm::TimePassesIsEnabled`` flag
1603 exported by the ``lib/VMCore/PassManager.cpp`` file.
1607 TODO: complete this section
1609 .. _dynamically loaded options:
1611 Dynamically adding command line options
1615 TODO: fill in this section