1/*
2 * The Lean Mean C++ Option Parser
3 *
4 * Copyright (C) 2012 Matthias S. Benkmann
5 *
6 * The "Software" in the following 2 paragraphs refers to this file containing
7 * the code to The Lean Mean C++ Option Parser.
8 * The "Software" does NOT refer to any other files which you
9 * may have received alongside this file (e.g. as part of a larger project that
10 * incorporates The Lean Mean C++ Option Parser).
11 *
12 * Permission is hereby granted, free of charge, to any person obtaining a copy
13 * of this software, to deal in the Software without restriction, including
14 * without limitation the rights to use, copy, modify, merge, publish,
15 * distribute, sublicense, and/or sell copies of the Software, and to permit
16 * persons to whom the Software is furnished to do so, subject to the following
17 * conditions:
18 * The above copyright notice and this permission notice shall be included in
19 * all copies or substantial portions of the Software.
20 *
21 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
22 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
23 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
24 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
25 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
26 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
27 * SOFTWARE.
28 */
29
30/*
31 * NOTE: It is recommended that you read the processed HTML doxygen documentation
32 * rather than this source. If you don't know doxygen, it's like javadoc for C++.
33 * If you don't want to install doxygen you can find a copy of the processed
34 * documentation at
35 *
36 * http://optionparser.sourceforge.net/
37 *
38 */
39
40/**
41 * @file
42 *
43 * @brief This is the only file required to use The Lean Mean C++ Option Parser.
44 * Just \#include it and you're set.
45 *
46 * The Lean Mean C++ Option Parser handles the program's command line arguments
47 * (argc, argv).
48 * It supports the short and long option formats of getopt(), getopt_long()
49 * and getopt_long_only() but has a more convenient interface.
50 * The following features set it apart from other option parsers:
51 *
52 * @par Highlights:
53 * <ul style="padding-left:1em;margin-left:0">
54 * <li> It is a header-only library. Just <code>\#include "optionparser.h"</code> and you're set.
55 * <li> It is freestanding. There are no dependencies whatsoever, not even the
56 * C or C++ standard library.
57 * <li> It has a usage message formatter that supports column alignment and
58 * line wrapping. This aids localization because it adapts to
59 * translated strings that are shorter or longer (even if they contain
60 * Asian wide characters).
61 * <li> Unlike getopt() and derivatives it doesn't force you to loop through
62 * options sequentially. Instead you can access options directly like this:
63 * <ul style="margin-top:.5em">
64 * <li> Test for presence of a switch in the argument vector:
65 * @code if ( options[QUIET] ) ... @endcode
66 * <li> Evaluate --enable-foo/--disable-foo pair where the last one used wins:
67 * @code if ( options[FOO].last()->type() == DISABLE ) ... @endcode
68 * <li> Cumulative option (-v verbose, -vv more verbose, -vvv even more verbose):
69 * @code int verbosity = options[VERBOSE].count(); @endcode
70 * <li> Iterate over all --file=&lt;fname> arguments:
71 * @code for (Option* opt = options[FILE]; opt; opt = opt->next())
72 * fname = opt->arg; ... @endcode
73 * <li> If you really want to, you can still process all arguments in order:
74 * @code
75 * for (int i = 0; i < p.optionsCount(); ++i) {
76 * Option& opt = buffer[i];
77 * switch(opt.index()) {
78 * case HELP: ...
79 * case VERBOSE: ...
80 * case FILE: fname = opt.arg; ...
81 * case UNKNOWN: ...
82 * @endcode
83 * </ul>
84 * </ul> @n
85 * Despite these features the code size remains tiny.
86 * It is smaller than <a href="http://uclibc.org">uClibc</a>'s GNU getopt() and just a
87 * couple 100 bytes larger than uClibc's SUSv3 getopt(). @n
88 * (This does not include the usage formatter, of course. But you don't have to use that.)
89 *
90 * @par Download:
91 * Tarball with examples and test programs:
92 * <a style="font-size:larger;font-weight:bold" href="http://sourceforge.net/projects/optionparser/files/optionparser-1.4.tar.gz/download">optionparser-1.4.tar.gz</a> @n
93 * Just the header (this is all you really need):
94 * <a style="font-size:larger;font-weight:bold" href="http://optionparser.sourceforge.net/optionparser.h">optionparser.h</a>
95 *
96 * @par Changelog:
97 * <b>Version 1.4:</b> Fixed 2 printUsage() bugs that messed up output with small COLUMNS values @n
98 * <b>Version 1.3:</b> Compatible with Microsoft Visual C++. @n
99 * <b>Version 1.2:</b> Added @ref option::Option::namelen "Option::namelen" and removed the extraction
100 * of short option characters into a special buffer. @n
101 * Changed @ref option::Arg::Optional "Arg::Optional" to accept arguments if they are attached
102 * rather than separate. This is what GNU getopt() does and how POSIX recommends
103 * utilities should interpret their arguments.@n
104 * <b>Version 1.1:</b> Optional mode with argument reordering as done by GNU getopt(), so that
105 * options and non-options can be mixed. See
106 * @ref option::Parser::parse() "Parser::parse()".
107 *
108 * @par Feedback:
109 * Send questions, bug reports, feature requests etc. to: <tt><b>optionparser-feedback<span id="antispam">&nbsp;(a)&nbsp;</span>lists.sourceforge.net</b></tt>
110 * @htmlonly <script type="text/javascript">document.getElementById("antispam").innerHTML="@"</script> @endhtmlonly
111 *
112 *
113 * @par Example program:
114 * (Note: @c option::* identifiers are links that take you to their documentation.)
115 * @code
116 * #include <iostream>
117 * #include "optionparser.h"
118 *
119 * enum optionIndex { UNKNOWN, HELP, PLUS };
120 * const option::Descriptor usage[] =
121 * {
122 * {UNKNOWN, 0,"" , "" ,option::Arg::None, "USAGE: example [options]\n\n"
123 * "Options:" },
124 * {HELP, 0,"" , "help",option::Arg::None, " --help \tPrint usage and exit." },
125 * {PLUS, 0,"p", "plus",option::Arg::None, " --plus, -p \tIncrement count." },
126 * {UNKNOWN, 0,"" , "" ,option::Arg::None, "\nExamples:\n"
127 * " example --unknown -- --this_is_no_option\n"
128 * " example -unk --plus -ppp file1 file2\n" },
129 * {0,0,0,0,0,0}
130 * };
131 *
132 * int main(int argc, char* argv[])
133 * {
134 * argc-=(argc>0); argv+=(argc>0); // skip program name argv[0] if present
135 * option::Stats stats(usage, argc, argv);
136 * option::Option options[stats.options_max], buffer[stats.buffer_max];
137 * option::Parser parse(usage, argc, argv, options, buffer);
138 *
139 * if (parse.error())
140 * return 1;
141 *
142 * if (options[HELP] || argc == 0) {
143 * option::printUsage(std::cout, usage);
144 * return 0;
145 * }
146 *
147 * std::cout << "--plus count: " <<
148 * options[PLUS].count() << "\n";
149 *
150 * for (option::Option* opt = options[UNKNOWN]; opt; opt = opt->next())
151 * std::cout << "Unknown option: " << opt->name << "\n";
152 *
153 * for (int i = 0; i < parse.nonOptionsCount(); ++i)
154 * std::cout << "Non-option #" << i << ": " << parse.nonOption(i) << "\n";
155 * }
156 * @endcode
157 *
158 * @par Option syntax:
159 * @li The Lean Mean C++ Option Parser follows POSIX <code>getopt()</code> conventions and supports
160 * GNU-style <code>getopt_long()</code> long options as well as Perl-style single-minus
161 * long options (<code>getopt_long_only()</code>).
162 * @li short options have the format @c -X where @c X is any character that fits in a char.
163 * @li short options can be grouped, i.e. <code>-X -Y</code> is equivalent to @c -XY.
164 * @li a short option may take an argument either separate (<code>-X foo</code>) or
165 * attached (@c -Xfoo). You can make the parser accept the additional format @c -X=foo by
166 * registering @c X as a long option (in addition to being a short option) and
167 * enabling single-minus long options.
168 * @li an argument-taking short option may be grouped if it is the last in the group, e.g.
169 * @c -ABCXfoo or <code> -ABCX foo </code> (@c foo is the argument to the @c -X option).
170 * @li a lone minus character @c '-' is not treated as an option. It is customarily used where
171 * a file name is expected to refer to stdin or stdout.
172 * @li long options have the format @c --option-name.
173 * @li the option-name of a long option can be anything and include any characters.
174 * Even @c = characters will work, but don't do that.
175 * @li [optional] long options may be abbreviated as long as the abbreviation is unambiguous.
176 * You can set a minimum length for abbreviations.
177 * @li [optional] long options may begin with a single minus. The double minus form is always
178 * accepted, too.
179 * @li a long option may take an argument either separate (<code> --option arg </code>) or
180 * attached (<code> --option=arg </code>). In the attached form the equals sign is mandatory.
181 * @li an empty string can be passed as an attached long option argument: <code> --option-name= </code>.
182 * Note the distinction between an empty string as argument and no argument at all.
183 * @li an empty string is permitted as separate argument to both long and short options.
184 * @li Arguments to both short and long options may start with a @c '-' character. E.g.
185 * <code> -X-X </code>, <code>-X -X</code> or <code> --long-X=-X </code>. If @c -X
186 * and @c --long-X take an argument, that argument will be @c "-X" in all 3 cases.
187 * @li If using the built-in @ref option::Arg::Optional "Arg::Optional", optional arguments must
188 * be attached.
189 * @li the special option @c -- (i.e. without a name) terminates the list of
190 * options. Everything that follows is a non-option argument, even if it starts with
191 * a @c '-' character. The @c -- itself will not appear in the parse results.
192 * @li the first argument that doesn't start with @c '-' or @c '--' and does not belong to
193 * a preceding argument-taking option, will terminate the option list and is the
194 * first non-option argument. All following command line arguments are treated as
195 * non-option arguments, even if they start with @c '-' . @n
196 * NOTE: This behaviour is mandated by POSIX, but GNU getopt() only honours this if it is
197 * explicitly requested (e.g. by setting POSIXLY_CORRECT). @n
198 * You can enable the GNU behaviour by passing @c true as first argument to
199 * e.g. @ref option::Parser::parse() "Parser::parse()".
200 * @li Arguments that look like options (i.e. @c '-' followed by at least 1 character) but
201 * aren't, are NOT treated as non-option arguments. They are treated as unknown options and
202 * are collected into a list of unknown options for error reporting. @n
203 * This means that in order to pass a first non-option
204 * argument beginning with the minus character it is required to use the
205 * @c -- special option, e.g.
206 * @code
207 * program -x -- --strange-filename
208 * @endcode
209 * In this example, @c --strange-filename is a non-option argument. If the @c --
210 * were omitted, it would be treated as an unknown option. @n
211 * See @ref option::Descriptor::longopt for information on how to collect unknown options.
212 *
213 */
214
215#ifndef OPTIONPARSER_H_
216#define OPTIONPARSER_H_
217
218/** @brief The namespace of The Lean Mean C++ Option Parser. */
219namespace option
220{
221
222#ifdef _MSC_VER
223#include <intrin.h>
224#pragma intrinsic(_BitScanReverse)
225struct MSC_Builtin_CLZ
226{
227 static int builtin_clz(unsigned x)
228 {
229 unsigned long index;
230 _BitScanReverse(&index, x);
231 return 32-index; // int is always 32bit on Windows, even for target x64
232 }
233};
234#define __builtin_clz(x) MSC_Builtin_CLZ::builtin_clz(x)
235#endif
236
237class Option;
238
239/**
240 * @brief Possible results when checking if an argument is valid for a certain option.
241 *
242 * In the case that no argument is provided for an option that takes an
243 * optional argument, return codes @c ARG_OK and @c ARG_IGNORE are equivalent.
244 */
245enum ArgStatus
246{
247 //! The option does not take an argument.
248 ARG_NONE,
249 //! The argument is acceptable for the option.
250 ARG_OK,
251 //! The argument is not acceptable but that's non-fatal because the option's argument is optional.
252 ARG_IGNORE,
253 //! The argument is not acceptable and that's fatal.
254 ARG_ILLEGAL
255};
256
257/**
258 * @brief Signature of functions that check if an argument is valid for a certain type of option.
259 *
260 * Every Option has such a function assigned in its Descriptor.
261 * @code
262 * Descriptor usage[] = { {UNKNOWN, 0, "", "", Arg::None, ""}, ... };
263 * @endcode
264 *
265 * A CheckArg function has the following signature:
266 * @code ArgStatus CheckArg(const Option& option, bool msg); @endcode
267 *
268 * It is used to check if a potential argument would be acceptable for the option.
269 * It will even be called if there is no argument. In that case @c option.arg will be @c NULL.
270 *
271 * If @c msg is @c true and the function determines that an argument is not acceptable and
272 * that this is a fatal error, it should output a message to the user before
273 * returning @ref ARG_ILLEGAL. If @c msg is @c false the function should remain silent (or you
274 * will get duplicate messages).
275 *
276 * See @ref ArgStatus for the meaning of the return values.
277 *
278 * While you can provide your own functions,
279 * often the following pre-defined checks (which never return @ref ARG_ILLEGAL) will suffice:
280 *
281 * @li @c Arg::None @copybrief Arg::None
282 * @li @c Arg::Optional @copybrief Arg::Optional
283 *
284 */
285typedef ArgStatus (*CheckArg)(const Option& option, bool msg);
286
287/**
288 * @brief Describes an option, its help text (usage) and how it should be parsed.
289 *
290 * The main input when constructing an option::Parser is an array of Descriptors.
291
292 * @par Example:
293 * @code
294 * enum OptionIndex {CREATE, ...};
295 * enum OptionType {DISABLE, ENABLE, OTHER};
296 *
297 * const option::Descriptor usage[] = {
298 * { CREATE, // index
299 * OTHER, // type
300 * "c", // shortopt
301 * "create", // longopt
302 * Arg::None, // check_arg
303 * "--create Tells the program to create something." // help
304 * }
305 * , ...
306 * };
307 * @endcode
308 */
309struct Descriptor
310{
311 /**
312 * @brief Index of this option's linked list in the array filled in by the parser.
313 *
314 * Command line options whose Descriptors have the same index will end up in the same
315 * linked list in the order in which they appear on the command line. If you have
316 * multiple long option aliases that refer to the same option, give their descriptors
317 * the same @c index.
318 *
319 * If you have options that mean exactly opposite things
320 * (e.g. @c --enable-foo and @c --disable-foo ), you should also give them the same
321 * @c index, but distinguish them through different values for @ref type.
322 * That way they end up in the same list and you can just take the last element of the
323 * list and use its type. This way you get the usual behaviour where switches later
324 * on the command line override earlier ones without having to code it manually.
325 *
326 * @par Tip:
327 * Use an enum rather than plain ints for better readability, as shown in the example
328 * at Descriptor.
329 */
330 const unsigned index;
331
332 /**
333 * @brief Used to distinguish between options with the same @ref index.
334 * See @ref index for details.
335 *
336 * It is recommended that you use an enum rather than a plain int to make your
337 * code more readable.
338 */
339 const int type;
340
341 /**
342 * @brief Each char in this string will be accepted as a short option character.
343 *
344 * The string must not include the minus character @c '-' or you'll get undefined
345 * behaviour.
346 *
347 * If this Descriptor should not have short option characters, use the empty
348 * string "". NULL is not permitted here!
349 *
350 * See @ref longopt for more information.
351 */
352 const char* const shortopt;
353
354 /**
355 * @brief The long option name (without the leading @c -- ).
356 *
357 * If this Descriptor should not have a long option name, use the empty
358 * string "". NULL is not permitted here!
359 *
360 * While @ref shortopt allows multiple short option characters, each
361 * Descriptor can have only a single long option name. If you have multiple
362 * long option names referring to the same option use separate Descriptors
363 * that have the same @ref index and @ref type. You may repeat
364 * short option characters in such an alias Descriptor but there's no need to.
365 *
366 * @par Dummy Descriptors:
367 * You can use dummy Descriptors with an
368 * empty string for both @ref shortopt and @ref longopt to add text to
369 * the usage that is not related to a specific option. See @ref help.
370 * The first dummy Descriptor will be used for unknown options (see below).
371 *
372 * @par Unknown Option Descriptor:
373 * The first dummy Descriptor in the list of Descriptors,
374 * whose @ref shortopt and @ref longopt are both the empty string, will be used
375 * as the Descriptor for unknown options. An unknown option is a string in
376 * the argument vector that is not a lone minus @c '-' but starts with a minus
377 * character and does not match any Descriptor's @ref shortopt or @ref longopt. @n
378 * Note that the dummy descriptor's @ref check_arg function @e will be called and
379 * its return value will be evaluated as usual. I.e. if it returns @ref ARG_ILLEGAL
380 * the parsing will be aborted with <code>Parser::error()==true</code>. @n
381 * if @c check_arg does not return @ref ARG_ILLEGAL the descriptor's
382 * @ref index @e will be used to pick the linked list into which
383 * to put the unknown option. @n
384 * If there is no dummy descriptor, unknown options will be dropped silently.
385 *
386 */
387 const char* const longopt;
388
389 /**
390 * @brief For each option that matches @ref shortopt or @ref longopt this function
391 * will be called to check a potential argument to the option.
392 *
393 * This function will be called even if there is no potential argument. In that case
394 * it will be passed @c NULL as @c arg parameter. Do not confuse this with the empty
395 * string.
396 *
397 * See @ref CheckArg for more information.
398 */
399 const CheckArg check_arg;
400
401 /**
402 * @brief The usage text associated with the options in this Descriptor.
403 *
404 * You can use option::printUsage() to format your usage message based on
405 * the @c help texts. You can use dummy Descriptors where
406 * @ref shortopt and @ref longopt are both the empty string to add text to
407 * the usage that is not related to a specific option.
408 *
409 * See option::printUsage() for special formatting characters you can use in
410 * @c help to get a column layout.
411 *
412 * @attention
413 * Must be UTF-8-encoded. If your compiler supports C++11 you can use the "u8"
414 * prefix to make sure string literals are properly encoded.
415 */
416 const char* help;
417};
418
419/**
420 * @brief A parsed option from the command line together with its argument if it has one.
421 *
422 * The Parser chains all parsed options with the same Descriptor::index together
423 * to form a linked list. This allows you to easily implement all of the common ways
424 * of handling repeated options and enable/disable pairs.
425 *
426 * @li Test for presence of a switch in the argument vector:
427 * @code if ( options[QUIET] ) ... @endcode
428 * @li Evaluate --enable-foo/--disable-foo pair where the last one used wins:
429 * @code if ( options[FOO].last()->type() == DISABLE ) ... @endcode
430 * @li Cumulative option (-v verbose, -vv more verbose, -vvv even more verbose):
431 * @code int verbosity = options[VERBOSE].count(); @endcode
432 * @li Iterate over all --file=&lt;fname> arguments:
433 * @code for (Option* opt = options[FILE]; opt; opt = opt->next())
434 * fname = opt->arg; ... @endcode
435 */
436class Option
437{
438 Option* next_;
439 Option* prev_;
440public:
441 /**
442 * @brief Pointer to this Option's Descriptor.
443 *
444 * Remember that the first dummy descriptor (see @ref Descriptor::longopt) is used
445 * for unknown options.
446 *
447 * @attention
448 * @c desc==NULL signals that this Option is unused. This is the default state of
449 * elements in the result array. You don't need to test @c desc explicitly. You
450 * can simply write something like this:
451 * @code
452 * if (options[CREATE])
453 * {
454 * ...
455 * }
456 * @endcode
457 * This works because of <code> operator const Option*() </code>.
458 */
459 const Descriptor* desc;
460
461 /**
462 * @brief The name of the option as used on the command line.
463 *
464 * The main purpose of this string is to be presented to the user in messages.
465 *
466 * In the case of a long option, this is the actual @c argv pointer, i.e. the first
467 * character is a '-'. In the case of a short option this points to the option
468 * character within the @c argv string.
469 *
470 * Note that in the case of a short option group or an attached option argument, this
471 * string will contain additional characters following the actual name. Use @ref namelen
472 * to filter out the actual option name only.
473 *
474 */
475 const char* name;
476
477 /**
478 * @brief Pointer to this Option's argument (if any).
479 *
480 * NULL if this option has no argument. Do not confuse this with the empty string which
481 * is a valid argument.
482 */
483 const char* arg;
484
485 /**
486 * @brief The length of the option @ref name.
487 *
488 * Because @ref name points into the actual @c argv string, the option name may be
489 * followed by more characters (e.g. other short options in the same short option group).
490 * This value is the number of bytes (not characters!) that are part of the actual name.
491 *
492 * For a short option, this length is always 1. For a long option this length is always
493 * at least 2 if single minus long options are permitted and at least 3 if they are disabled.
494 *
495 * @note
496 * In the pathological case of a minus within a short option group (e.g. @c -xf-z), this
497 * length is incorrect, because this case will be misinterpreted as a long option and the
498 * name will therefore extend to the string's 0-terminator or a following '=" character
499 * if there is one. This is irrelevant for most uses of @ref name and @c namelen. If you
500 * really need to distinguish the case of a long and a short option, compare @ref name to
501 * the @c argv pointers. A long option's @c name is always identical to one of them,
502 * whereas a short option's is never.
503 */
504 int namelen;
505
506 /**
507 * @brief Returns Descriptor::type of this Option's Descriptor, or 0 if this Option
508 * is invalid (unused).
509 *
510 * Because this method (and last(), too) can be used even on unused Options with desc==0, you can (provided
511 * you arrange your types properly) switch on type() without testing validity first.
512 * @code
513 * enum OptionType { UNUSED=0, DISABLED=0, ENABLED=1 };
514 * enum OptionIndex { FOO };
515 * const Descriptor usage[] = {
516 * { FOO, ENABLED, "", "enable-foo", Arg::None, 0 },
517 * { FOO, DISABLED, "", "disable-foo", Arg::None, 0 },
518 * { 0, 0, 0, 0, 0, 0 } };
519 * ...
520 * switch(options[FOO].last()->type()) // no validity check required!
521 * {
522 * case ENABLED: ...
523 * case DISABLED: ... // UNUSED==DISABLED !
524 * }
525 * @endcode
526 */
527 int type() const
528 {
529 return desc == 0 ? 0 : desc->type;
530 }
531
532 /**
533 * @brief Returns Descriptor::index of this Option's Descriptor, or -1 if this Option
534 * is invalid (unused).
535 */
536 int index() const
537 {
538 return desc == 0 ? -1 : (int)desc->index;
539 }
540
541 /**
542 * @brief Returns the number of times this Option (or others with the same Descriptor::index)
543 * occurs in the argument vector.
544 *
545 * This corresponds to the number of elements in the linked list this Option is part of.
546 * It doesn't matter on which element you call count(). The return value is always the same.
547 *
548 * Use this to implement cumulative options, such as -v, -vv, -vvv for
549 * different verbosity levels.
550 *
551 * Returns 0 when called for an unused/invalid option.
552 */
553 int count()
554 {
555 int c = (desc == 0 ? 0 : 1);
556 Option* p = first();
557 while (!p->isLast())
558 {
559 ++c;
560 p = p->next_;
561 };
562 return c;
563 }
564
565 /**
566 * @brief Returns true iff this is the first element of the linked list.
567 *
568 * The first element in the linked list is the first option on the command line
569 * that has the respective Descriptor::index value.
570 *
571 * Returns true for an unused/invalid option.
572 */
573 bool isFirst() const
574 {
575 return isTagged(prev_);
576 }
577
578 /**
579 * @brief Returns true iff this is the last element of the linked list.
580 *
581 * The last element in the linked list is the last option on the command line
582 * that has the respective Descriptor::index value.
583 *
584 * Returns true for an unused/invalid option.
585 */
586 bool isLast() const
587 {
588 return isTagged(next_);
589 }
590
591 /**
592 * @brief Returns a pointer to the first element of the linked list.
593 *
594 * Use this when you want the first occurrence of an option on the command line to
595 * take precedence. Note that this is not the way most programs handle options.
596 * You should probably be using last() instead.
597 *
598 * @note
599 * This method may be called on an unused/invalid option and will return a pointer to the
600 * option itself.
601 */
602 Option* first()
603 {
604 Option* p = this;
605 while (!p->isFirst())
606 p = p->prev_;
607 return p;
608 }
609
610 /**
611 * @brief Returns a pointer to the last element of the linked list.
612 *
613 * Use this when you want the last occurrence of an option on the command line to
614 * take precedence. This is the most common way of handling conflicting options.
615 *
616 * @note
617 * This method may be called on an unused/invalid option and will return a pointer to the
618 * option itself.
619 *
620 * @par Tip:
621 * If you have options with opposite meanings (e.g. @c --enable-foo and @c --disable-foo), you
622 * can assign them the same Descriptor::index to get them into the same list. Distinguish them by
623 * Descriptor::type and all you have to do is check <code> last()->type() </code> to get
624 * the state listed last on the command line.
625 */
626 Option* last()
627 {
628 return first()->prevwrap();
629 }
630
631 /**
632 * @brief Returns a pointer to the previous element of the linked list or NULL if
633 * called on first().
634 *
635 * If called on first() this method returns NULL. Otherwise it will return the
636 * option with the same Descriptor::index that precedes this option on the command
637 * line.
638 */
639 Option* prev()
640 {
641 return isFirst() ? 0 : prev_;
642 }
643
644 /**
645 * @brief Returns a pointer to the previous element of the linked list with wrap-around from
646 * first() to last().
647 *
648 * If called on first() this method returns last(). Otherwise it will return the
649 * option with the same Descriptor::index that precedes this option on the command
650 * line.
651 */
652 Option* prevwrap()
653 {
654 return untag(prev_);
655 }
656
657 /**
658 * @brief Returns a pointer to the next element of the linked list or NULL if called
659 * on last().
660 *
661 * If called on last() this method returns NULL. Otherwise it will return the
662 * option with the same Descriptor::index that follows this option on the command
663 * line.
664 */
665 Option* next()
666 {
667 return isLast() ? 0 : next_;
668 }
669
670 /**
671 * @brief Returns a pointer to the next element of the linked list with wrap-around from
672 * last() to first().
673 *
674 * If called on last() this method returns first(). Otherwise it will return the
675 * option with the same Descriptor::index that follows this option on the command
676 * line.
677 */
678 Option* nextwrap()
679 {
680 return untag(next_);
681 }
682
683 /**
684 * @brief Makes @c new_last the new last() by chaining it into the list after last().
685 *
686 * It doesn't matter which element you call append() on. The new element will always
687 * be appended to last().
688 *
689 * @attention
690 * @c new_last must not yet be part of a list, or that list will become corrupted, because
691 * this method does not unchain @c new_last from an existing list.
692 */
693 void append(Option* new_last)
694 {
695 Option* p = last();
696 Option* f = first();
697 p->next_ = new_last;
698 new_last->prev_ = p;
699 new_last->next_ = tag(f);
700 f->prev_ = tag(new_last);
701 }
702
703 /**
704 * @brief Casts from Option to const Option* but only if this Option is valid.
705 *
706 * If this Option is valid (i.e. @c desc!=NULL), returns this.
707 * Otherwise returns NULL. This allows testing an Option directly
708 * in an if-clause to see if it is used:
709 * @code
710 * if (options[CREATE])
711 * {
712 * ...
713 * }
714 * @endcode
715 * It also allows you to write loops like this:
716 * @code for (Option* opt = options[FILE]; opt; opt = opt->next())
717 * fname = opt->arg; ... @endcode
718 */
719 operator const Option*() const
720 {
721 return desc ? this : 0;
722 }
723
724 /**
725 * @brief Casts from Option to Option* but only if this Option is valid.
726 *
727 * If this Option is valid (i.e. @c desc!=NULL), returns this.
728 * Otherwise returns NULL. This allows testing an Option directly
729 * in an if-clause to see if it is used:
730 * @code
731 * if (options[CREATE])
732 * {
733 * ...
734 * }
735 * @endcode
736 * It also allows you to write loops like this:
737 * @code for (Option* opt = options[FILE]; opt; opt = opt->next())
738 * fname = opt->arg; ... @endcode
739 */
740 operator Option*()
741 {
742 return desc ? this : 0;
743 }
744
745 /**
746 * @brief Creates a new Option that is a one-element linked list and has NULL
747 * @ref desc, @ref name, @ref arg and @ref namelen.
748 */
749 Option() :
750 desc(0), name(0), arg(0), namelen(0)
751 {
752 prev_ = tag(this);
753 next_ = tag(this);
754 }
755
756 /**
757 * @brief Creates a new Option that is a one-element linked list and has the given
758 * values for @ref desc, @ref name and @ref arg.
759 *
760 * If @c name_ points at a character other than '-' it will be assumed to refer to a
761 * short option and @ref namelen will be set to 1. Otherwise the length will extend to
762 * the first '=' character or the string's 0-terminator.
763 */
764 Option(const Descriptor* desc_, const char* name_, const char* arg_)
765 {
766 init(desc_, name_, arg_);
767 }
768
769 /**
770 * @brief Makes @c *this a copy of @c orig except for the linked list pointers.
771 *
772 * After this operation @c *this will be a one-element linked list.
773 */
774 void operator=(const Option& orig)
775 {
776 init(orig.desc, orig.name, orig.arg);
777 }
778
779 /**
780 * @brief Makes @c *this a copy of @c orig except for the linked list pointers.
781 *
782 * After this operation @c *this will be a one-element linked list.
783 */
784 Option(const Option& orig)
785 {
786 init(orig.desc, orig.name, orig.arg);
787 }
788
789private:
790 /**
791 * @internal
792 * @brief Sets the fields of this Option to the given values (extracting @c name if necessary).
793 *
794 * If @c name_ points at a character other than '-' it will be assumed to refer to a
795 * short option and @ref namelen will be set to 1. Otherwise the length will extend to
796 * the first '=' character or the string's 0-terminator.
797 */
798 void init(const Descriptor* desc_, const char* name_, const char* arg_)
799 {
800 desc = desc_;
801 name = name_;
802 arg = arg_;
803 prev_ = tag(this);
804 next_ = tag(this);
805 namelen = 0;
806 if (name == 0)
807 return;
808 namelen = 1;
809 if (name[0] != '-')
810 return;
811 while (name[namelen] != 0 && name[namelen] != '=')
812 ++namelen;
813 }
814
815 static Option* tag(Option* ptr)
816 {
817 return (Option*) ((unsigned long long) ptr | 1);
818 }
819
820 static Option* untag(Option* ptr)
821 {
822 return (Option*) ((unsigned long long) ptr & ~1ull);
823 }
824
825 static bool isTagged(Option* ptr)
826 {
827 return ((unsigned long long) ptr & 1);
828 }
829};
830
831/**
832 * @brief Functions for checking the validity of option arguments.
833 *
834 * @copydetails CheckArg
835 *
836 * The following example code
837 * can serve as starting place for writing your own more complex CheckArg functions:
838 * @code
839 * struct Arg: public option::Arg
840 * {
841 * static void printError(const char* msg1, const option::Option& opt, const char* msg2)
842 * {
843 * fprintf(stderr, "ERROR: %s", msg1);
844 * fwrite(opt.name, opt.namelen, 1, stderr);
845 * fprintf(stderr, "%s", msg2);
846 * }
847 *
848 * static option::ArgStatus Unknown(const option::Option& option, bool msg)
849 * {
850 * if (msg) printError("Unknown option '", option, "'\n");
851 * return option::ARG_ILLEGAL;
852 * }
853 *
854 * static option::ArgStatus Required(const option::Option& option, bool msg)
855 * {
856 * if (option.arg != 0)
857 * return option::ARG_OK;
858 *
859 * if (msg) printError("Option '", option, "' requires an argument\n");
860 * return option::ARG_ILLEGAL;
861 * }
862 *
863 * static option::ArgStatus NonEmpty(const option::Option& option, bool msg)
864 * {
865 * if (option.arg != 0 && option.arg[0] != 0)
866 * return option::ARG_OK;
867 *
868 * if (msg) printError("Option '", option, "' requires a non-empty argument\n");
869 * return option::ARG_ILLEGAL;
870 * }
871 *
872 * static option::ArgStatus Numeric(const option::Option& option, bool msg)
873 * {
874 * char* endptr = 0;
875 * if (option.arg != 0 && strtol(option.arg, &endptr, 10)){};
876 * if (endptr != option.arg && *endptr == 0)
877 * return option::ARG_OK;
878 *
879 * if (msg) printError("Option '", option, "' requires a numeric argument\n");
880 * return option::ARG_ILLEGAL;
881 * }
882 * };
883 * @endcode
884 */
885struct Arg
886{
887 //! @brief For options that don't take an argument: Returns ARG_NONE.
888 static ArgStatus None(const Option&, bool)
889 {
890 return ARG_NONE;
891 }
892
893 //! @brief Returns ARG_OK if the argument is attached and ARG_IGNORE otherwise.
894 static ArgStatus Optional(const Option& option, bool)
895 {
896 if (option.arg && option.name[option.namelen] != 0)
897 return ARG_OK;
898 else
899 return ARG_IGNORE;
900 }
901};
902
903/**
904 * @brief Determines the minimum lengths of the buffer and options arrays used for Parser.
905 *
906 * Because Parser doesn't use dynamic memory its output arrays have to be pre-allocated.
907 * If you don't want to use fixed size arrays (which may turn out too small, causing
908 * command line arguments to be dropped), you can use Stats to determine the correct sizes.
909 * Stats work cumulative. You can first pass in your default options and then the real
910 * options and afterwards the counts will reflect the union.
911 */
912struct Stats
913{
914 /**
915 * @brief Number of elements needed for a @c buffer[] array to be used for
916 * @ref Parser::parse() "parsing" the same argument vectors that were fed
917 * into this Stats object.
918 *
919 * @note
920 * This number is always 1 greater than the actual number needed, to give
921 * you a sentinel element.
922 */
923 unsigned buffer_max;
924
925 /**
926 * @brief Number of elements needed for an @c options[] array to be used for
927 * @ref Parser::parse() "parsing" the same argument vectors that were fed
928 * into this Stats object.
929 *
930 * @note
931 * @li This number is always 1 greater than the actual number needed, to give
932 * you a sentinel element.
933 * @li This number depends only on the @c usage, not the argument vectors, because
934 * the @c options array needs exactly one slot for each possible Descriptor::index.
935 */
936 unsigned options_max;
937
938 /**
939 * @brief Creates a Stats object with counts set to 1 (for the sentinel element).
940 */
941 Stats() :
942 buffer_max(1), options_max(1) // 1 more than necessary as sentinel
943 {
944 }
945
946 /**
947 * @brief Creates a new Stats object and immediately updates it for the
948 * given @c usage and argument vector. You may pass 0 for @c argc and/or @c argv,
949 * if you just want to update @ref options_max.
950 *
951 * @note
952 * The calls to Stats methods must match the later calls to Parser methods.
953 * See Parser::parse() for the meaning of the arguments.
954 */
955 Stats(bool gnu, const Descriptor usage[], int argc, const char** argv, int min_abbr_len = 0, //
956 bool single_minus_longopt = false) :
957 buffer_max(1), options_max(1) // 1 more than necessary as sentinel
958 {
959 add(gnu, usage, argc, argv, min_abbr_len, single_minus_longopt);
960 }
961
962 //! @brief Stats(...) with non-const argv.
963 Stats(bool gnu, const Descriptor usage[], int argc, char** argv, int min_abbr_len = 0, //
964 bool single_minus_longopt = false) :
965 buffer_max(1), options_max(1) // 1 more than necessary as sentinel
966 {
967 add(gnu, usage, argc, (const char**) argv, min_abbr_len, single_minus_longopt);
968 }
969
970 //! @brief POSIX Stats(...) (gnu==false).
971 Stats(const Descriptor usage[], int argc, const char** argv, int min_abbr_len = 0, //
972 bool single_minus_longopt = false) :
973 buffer_max(1), options_max(1) // 1 more than necessary as sentinel
974 {
975 add(false, usage, argc, argv, min_abbr_len, single_minus_longopt);
976 }
977
978 //! @brief POSIX Stats(...) (gnu==false) with non-const argv.
979 Stats(const Descriptor usage[], int argc, char** argv, int min_abbr_len = 0, //
980 bool single_minus_longopt = false) :
981 buffer_max(1), options_max(1) // 1 more than necessary as sentinel
982 {
983 add(false, usage, argc, (const char**) argv, min_abbr_len, single_minus_longopt);
984 }
985
986 /**
987 * @brief Updates this Stats object for the
988 * given @c usage and argument vector. You may pass 0 for @c argc and/or @c argv,
989 * if you just want to update @ref options_max.
990 *
991 * @note
992 * The calls to Stats methods must match the later calls to Parser methods.
993 * See Parser::parse() for the meaning of the arguments.
994 */
995 void add(bool gnu, const Descriptor usage[], int argc, const char** argv, int min_abbr_len = 0, //
996 bool single_minus_longopt = false);
997
998 //! @brief add() with non-const argv.
999 void add(bool gnu, const Descriptor usage[], int argc, char** argv, int min_abbr_len = 0, //
1000 bool single_minus_longopt = false)
1001 {
1002 add(gnu, usage, argc, (const char**) argv, min_abbr_len, single_minus_longopt);
1003 }
1004
1005 //! @brief POSIX add() (gnu==false).
1006 void add(const Descriptor usage[], int argc, const char** argv, int min_abbr_len = 0, //
1007 bool single_minus_longopt = false)
1008 {
1009 add(false, usage, argc, argv, min_abbr_len, single_minus_longopt);
1010 }
1011
1012 //! @brief POSIX add() (gnu==false) with non-const argv.
1013 void add(const Descriptor usage[], int argc, char** argv, int min_abbr_len = 0, //
1014 bool single_minus_longopt = false)
1015 {
1016 add(false, usage, argc, (const char**) argv, min_abbr_len, single_minus_longopt);
1017 }
1018private:
1019 class CountOptionsAction;
1020};
1021
1022/**
1023 * @brief Checks argument vectors for validity and parses them into data
1024 * structures that are easier to work with.
1025 *
1026 * @par Example:
1027 * @code
1028 * int main(int argc, char* argv[])
1029 * {
1030 * argc-=(argc>0); argv+=(argc>0); // skip program name argv[0] if present
1031 * option::Stats stats(usage, argc, argv);
1032 * option::Option options[stats.options_max], buffer[stats.buffer_max];
1033 * option::Parser parse(usage, argc, argv, options, buffer);
1034 *
1035 * if (parse.error())
1036 * return 1;
1037 *
1038 * if (options[HELP])
1039 * ...
1040 * @endcode
1041 */
1042class Parser
1043{
1044 int op_count; //!< @internal @brief see optionsCount()
1045 int nonop_count; //!< @internal @brief see nonOptionsCount()
1046 const char** nonop_args; //!< @internal @brief see nonOptions()
1047 bool err; //!< @internal @brief see error()
1048public:
1049
1050 /**
1051 * @brief Creates a new Parser.
1052 */
1053 Parser() :
1054 op_count(0), nonop_count(0), nonop_args(0), err(false)
1055 {
1056 }
1057
1058 /**
1059 * @brief Creates a new Parser and immediately parses the given argument vector.
1060 * @copydetails parse()
1061 */
1062 Parser(bool gnu, const Descriptor usage[], int argc, const char** argv, Option options[], Option buffer[],
1063 int min_abbr_len = 0, bool single_minus_longopt = false, int bufmax = -1) :
1064 op_count(0), nonop_count(0), nonop_args(0), err(false)
1065 {
1066 parse(gnu, usage, argc, argv, options, buffer, min_abbr_len, single_minus_longopt, bufmax);
1067 }
1068
1069 //! @brief Parser(...) with non-const argv.
1070 Parser(bool gnu, const Descriptor usage[], int argc, char** argv, Option options[], Option buffer[],
1071 int min_abbr_len = 0, bool single_minus_longopt = false, int bufmax = -1) :
1072 op_count(0), nonop_count(0), nonop_args(0), err(false)
1073 {
1074 parse(gnu, usage, argc, (const char**) argv, options, buffer, min_abbr_len, single_minus_longopt, bufmax);
1075 }
1076
1077 //! @brief POSIX Parser(...) (gnu==false).
1078 Parser(const Descriptor usage[], int argc, const char** argv, Option options[], Option buffer[], int min_abbr_len = 0,
1079 bool single_minus_longopt = false, int bufmax = -1) :
1080 op_count(0), nonop_count(0), nonop_args(0), err(false)
1081 {
1082 parse(false, usage, argc, argv, options, buffer, min_abbr_len, single_minus_longopt, bufmax);
1083 }
1084
1085 //! @brief POSIX Parser(...) (gnu==false) with non-const argv.
1086 Parser(const Descriptor usage[], int argc, char** argv, Option options[], Option buffer[], int min_abbr_len = 0,
1087 bool single_minus_longopt = false, int bufmax = -1) :
1088 op_count(0), nonop_count(0), nonop_args(0), err(false)
1089 {
1090 parse(false, usage, argc, (const char**) argv, options, buffer, min_abbr_len, single_minus_longopt, bufmax);
1091 }
1092
1093 /**
1094 * @brief Parses the given argument vector.
1095 *
1096 * @param gnu if true, parse() will not stop at the first non-option argument. Instead it will
1097 * reorder arguments so that all non-options are at the end. This is the default behaviour
1098 * of GNU getopt() but is not conforming to POSIX. @n
1099 * Note, that once the argument vector has been reordered, the @c gnu flag will have
1100 * no further effect on this argument vector. So it is enough to pass @c gnu==true when
1101 * creating Stats.
1102 * @param usage Array of Descriptor objects that describe the options to support. The last entry
1103 * of this array must have 0 in all fields.
1104 * @param argc The number of elements from @c argv that are to be parsed. If you pass -1, the number
1105 * will be determined automatically. In that case the @c argv list must end with a NULL
1106 * pointer.
1107 * @param argv The arguments to be parsed. If you pass -1 as @c argc the last pointer in the @c argv
1108 * list must be NULL to mark the end.
1109 * @param options Each entry is the first element of a linked list of Options. Each new option
1110 * that is parsed will be appended to the list specified by that Option's
1111 * Descriptor::index. If an entry is not yet used (i.e. the Option is invalid),
1112 * it will be replaced rather than appended to. @n
1113 * The minimum length of this array is the greatest Descriptor::index value that
1114 * occurs in @c usage @e PLUS ONE.
1115 * @param buffer Each argument that is successfully parsed (including unknown arguments, if they
1116 * have a Descriptor whose CheckArg does not return @ref ARG_ILLEGAL) will be stored in this
1117 * array. parse() scans the array for the first invalid entry and begins writing at that
1118 * index. You can pass @c bufmax to limit the number of options stored.
1119 * @param min_abbr_len Passing a value <code> min_abbr_len > 0 </code> enables abbreviated long
1120 * options. The parser will match a prefix of a long option as if it was
1121 * the full long option (e.g. @c --foob=10 will be interpreted as if it was
1122 * @c --foobar=10 ), as long as the prefix has at least @c min_abbr_len characters
1123 * (not counting the @c -- ) and is unambiguous.
1124 * @n Be careful if combining @c min_abbr_len=1 with @c single_minus_longopt=true
1125 * because the ambiguity check does not consider short options and abbreviated
1126 * single minus long options will take precedence over short options.
1127 * @param single_minus_longopt Passing @c true for this option allows long options to begin with
1128 * a single minus. The double minus form will still be recognized. Note that
1129 * single minus long options take precedence over short options and short option
1130 * groups. E.g. @c -file would be interpreted as @c --file and not as
1131 * <code> -f -i -l -e </code> (assuming a long option named @c "file" exists).
1132 * @param bufmax The greatest index in the @c buffer[] array that parse() will write to is
1133 * @c bufmax-1. If there are more options, they will be processed (in particular
1134 * their CheckArg will be called) but not stored. @n
1135 * If you used Stats::buffer_max to dimension this array, you can pass
1136 * -1 (or not pass @c bufmax at all) which tells parse() that the buffer is
1137 * "large enough".
1138 * @attention
1139 * Remember that @c options and @c buffer store Option @e objects, not pointers. Therefore it
1140 * is not possible for the same object to be in both arrays. For those options that are found in
1141 * both @c buffer[] and @c options[] the respective objects are independent copies. And only the
1142 * objects in @c options[] are properly linked via Option::next() and Option::prev().
1143 * You can iterate over @c buffer[] to
1144 * process all options in the order they appear in the argument vector, but if you want access to
1145 * the other Options with the same Descriptor::index, then you @e must access the linked list via
1146 * @c options[]. You can get the linked list in options from a buffer object via something like
1147 * @c options[buffer[i].index()].
1148 */
1149 void parse(bool gnu, const Descriptor usage[], int argc, const char** argv, Option options[], Option buffer[],
1150 int min_abbr_len = 0, bool single_minus_longopt = false, int bufmax = -1);
1151
1152 //! @brief parse() with non-const argv.
1153 void parse(bool gnu, const Descriptor usage[], int argc, char** argv, Option options[], Option buffer[],
1154 int min_abbr_len = 0, bool single_minus_longopt = false, int bufmax = -1)
1155 {
1156 parse(gnu, usage, argc, (const char**) argv, options, buffer, min_abbr_len, single_minus_longopt, bufmax);
1157 }
1158
1159 //! @brief POSIX parse() (gnu==false).
1160 void parse(const Descriptor usage[], int argc, const char** argv, Option options[], Option buffer[],
1161 int min_abbr_len = 0, bool single_minus_longopt = false, int bufmax = -1)
1162 {
1163 parse(false, usage, argc, argv, options, buffer, min_abbr_len, single_minus_longopt, bufmax);
1164 }
1165
1166 //! @brief POSIX parse() (gnu==false) with non-const argv.
1167 void parse(const Descriptor usage[], int argc, char** argv, Option options[], Option buffer[], int min_abbr_len = 0,
1168 bool single_minus_longopt = false, int bufmax = -1)
1169 {
1170 parse(false, usage, argc, (const char**) argv, options, buffer, min_abbr_len, single_minus_longopt, bufmax);
1171 }
1172
1173 /**
1174 * @brief Returns the number of valid Option objects in @c buffer[].
1175 *
1176 * @note
1177 * @li The returned value always reflects the number of Options in the buffer[] array used for
1178 * the most recent call to parse().
1179 * @li The count (and the buffer[]) includes unknown options if they are collected
1180 * (see Descriptor::longopt).
1181 */
1182 int optionsCount()
1183 {
1184 return op_count;
1185 }
1186
1187 /**
1188 * @brief Returns the number of non-option arguments that remained at the end of the
1189 * most recent parse() that actually encountered non-option arguments.
1190 *
1191 * @note
1192 * A parse() that does not encounter non-option arguments will leave this value
1193 * as well as nonOptions() undisturbed. This means you can feed the Parser a
1194 * default argument vector that contains non-option arguments (e.g. a default filename).
1195 * Then you feed it the actual arguments from the user. If the user has supplied at
1196 * least one non-option argument, all of the non-option arguments from the default
1197 * disappear and are replaced by the user's non-option arguments. However, if the
1198 * user does not supply any non-option arguments the defaults will still be in
1199 * effect.
1200 */
1201 int nonOptionsCount()
1202 {
1203 return nonop_count;
1204 }
1205
1206 /**
1207 * @brief Returns a pointer to an array of non-option arguments (only valid
1208 * if <code>nonOptionsCount() >0 </code>).
1209 *
1210 * @note
1211 * @li parse() does not copy arguments, so this pointer points into the actual argument
1212 * vector as passed to parse().
1213 * @li As explained at nonOptionsCount() this pointer is only changed by parse() calls
1214 * that actually encounter non-option arguments. A parse() call that encounters only
1215 * options, will not change nonOptions().
1216 */
1217 const char** nonOptions()
1218 {
1219 return nonop_args;
1220 }
1221
1222 /**
1223 * @brief Returns <b><code>nonOptions()[i]</code></b> (@e without checking if i is in range!).
1224 */
1225 const char* nonOption(int i)
1226 {
1227 return nonOptions()[i];
1228 }
1229
1230 /**
1231 * @brief Returns @c true if an unrecoverable error occurred while parsing options.
1232 *
1233 * An illegal argument to an option (i.e. CheckArg returns @ref ARG_ILLEGAL) is an
1234 * unrecoverable error that aborts the parse. Unknown options are only an error if
1235 * their CheckArg function returns @ref ARG_ILLEGAL. Otherwise they are collected.
1236 * In that case if you want to exit the program if either an illegal argument
1237 * or an unknown option has been passed, use code like this
1238 *
1239 * @code
1240 * if (parser.error() || options[UNKNOWN])
1241 * exit(1);
1242 * @endcode
1243 *
1244 */
1245 bool error()
1246 {
1247 return err;
1248 }
1249
1250private:
1251 friend struct Stats;
1252 class StoreOptionAction;
1253 struct Action;
1254
1255 /**
1256 * @internal
1257 * @brief This is the core function that does all the parsing.
1258 * @retval false iff an unrecoverable error occurred.
1259 */
1260 static bool workhorse(bool gnu, const Descriptor usage[], int numargs, const char** args, Action& action,
1261 bool single_minus_longopt, bool print_errors, int min_abbr_len);
1262
1263 /**
1264 * @internal
1265 * @brief Returns true iff @c st1 is a prefix of @c st2 and
1266 * in case @c st2 is longer than @c st1, then
1267 * the first additional character is '='.
1268 *
1269 * @par Examples:
1270 * @code
1271 * streq("foo", "foo=bar") == true
1272 * streq("foo", "foobar") == false
1273 * streq("foo", "foo") == true
1274 * streq("foo=bar", "foo") == false
1275 * @endcode
1276 */
1277 static bool streq(const char* st1, const char* st2)
1278 {
1279 while (*st1 != 0)
1280 if (*st1++ != *st2++)
1281 return false;
1282 return (*st2 == 0 || *st2 == '=');
1283 }
1284
1285 /**
1286 * @internal
1287 * @brief Like streq() but handles abbreviations.
1288 *
1289 * Returns true iff @c st1 and @c st2 have a common
1290 * prefix with the following properties:
1291 * @li (if min > 0) its length is at least @c min characters or the same length as @c st1 (whichever is smaller).
1292 * @li (if min <= 0) its length is the same as that of @c st1
1293 * @li within @c st2 the character following the common prefix is either '=' or end-of-string.
1294 *
1295 * Examples:
1296 * @code
1297 * streqabbr("foo", "foo=bar",<anything>) == true
1298 * streqabbr("foo", "fo=bar" , 2) == true
1299 * streqabbr("foo", "fo" , 2) == true
1300 * streqabbr("foo", "fo" , 0) == false
1301 * streqabbr("foo", "f=bar" , 2) == false
1302 * streqabbr("foo", "f" , 2) == false
1303 * streqabbr("fo" , "foo=bar",<anything>) == false
1304 * streqabbr("foo", "foobar" ,<anything>) == false
1305 * streqabbr("foo", "fobar" ,<anything>) == false
1306 * streqabbr("foo", "foo" ,<anything>) == true
1307 * @endcode
1308 */
1309 static bool streqabbr(const char* st1, const char* st2, long long min)
1310 {
1311 const char* st1start = st1;
1312 while (*st1 != 0 && (*st1 == *st2))
1313 {
1314 ++st1;
1315 ++st2;
1316 }
1317
1318 return (*st1 == 0 || (min > 0 && (st1 - st1start) >= min)) && (*st2 == 0 || *st2 == '=');
1319 }
1320
1321 /**
1322 * @internal
1323 * @brief Returns true iff character @c ch is contained in the string @c st.
1324 *
1325 * Returns @c true for @c ch==0 .
1326 */
1327 static bool instr(char ch, const char* st)
1328 {
1329 while (*st != 0 && *st != ch)
1330 ++st;
1331 return *st == ch;
1332 }
1333
1334 /**
1335 * @internal
1336 * @brief Rotates <code>args[-count],...,args[-1],args[0]</code> to become
1337 * <code>args[0],args[-count],...,args[-1]</code>.
1338 */
1339 static void shift(const char** args, int count)
1340 {
1341 for (int i = 0; i > -count; --i)
1342 {
1343 const char* temp = args[i];
1344 args[i] = args[i - 1];
1345 args[i - 1] = temp;
1346 }
1347 }
1348};
1349
1350/**
1351 * @internal
1352 * @brief Interface for actions Parser::workhorse() should perform for each Option it
1353 * parses.
1354 */
1355struct Parser::Action
1356{
1357 /**
1358 * @brief Called by Parser::workhorse() for each Option that has been successfully
1359 * parsed (including unknown
1360 * options if they have a Descriptor whose Descriptor::check_arg does not return
1361 * @ref ARG_ILLEGAL.
1362 *
1363 * Returns @c false iff a fatal error has occured and the parse should be aborted.
1364 */
1365 virtual bool perform(Option&)
1366 {
1367 return true;
1368 }
1369
1370 /**
1371 * @brief Called by Parser::workhorse() after finishing the parse.
1372 * @param numargs the number of non-option arguments remaining
1373 * @param args pointer to the first remaining non-option argument (if numargs > 0).
1374 *
1375 * @return
1376 * @c false iff a fatal error has occurred.
1377 */
1378 virtual bool finished(int numargs, const char** args)
1379 {
1380 (void) numargs;
1381 (void) args;
1382 return true;
1383 }
1384};
1385
1386/**
1387 * @internal
1388 * @brief An Action to pass to Parser::workhorse() that will increment a counter for
1389 * each parsed Option.
1390 */
1391class Stats::CountOptionsAction: public Parser::Action
1392{
1393 unsigned* buffer_max;
1394public:
1395 /**
1396 * Creates a new CountOptionsAction that will increase @c *buffer_max_ for each
1397 * parsed Option.
1398 */
1399 CountOptionsAction(unsigned* buffer_max_) :
1400 buffer_max(buffer_max_)
1401 {
1402 }
1403
1404 bool perform(Option&)
1405 {
1406 if (*buffer_max == 0x7fffffff)
1407 return false; // overflow protection: don't accept number of options that doesn't fit signed int
1408 ++*buffer_max;
1409 return true;
1410 }
1411};
1412
1413/**
1414 * @internal
1415 * @brief An Action to pass to Parser::workhorse() that will store each parsed Option in
1416 * appropriate arrays (see Parser::parse()).
1417 */
1418class Parser::StoreOptionAction: public Parser::Action
1419{
1420 Parser& parser;
1421 Option* options;
1422 Option* buffer;
1423 int bufmax; //! Number of slots in @c buffer. @c -1 means "large enough".
1424public:
1425 /**
1426 * @brief Creates a new StoreOption action.
1427 * @param parser_ the parser whose op_count should be updated.
1428 * @param options_ each Option @c o is chained into the linked list @c options_[o.desc->index]
1429 * @param buffer_ each Option is appended to this array as long as there's a free slot.
1430 * @param bufmax_ number of slots in @c buffer_. @c -1 means "large enough".
1431 */
1432 StoreOptionAction(Parser& parser_, Option options_[], Option buffer_[], int bufmax_) :
1433 parser(parser_), options(options_), buffer(buffer_), bufmax(bufmax_)
1434 {
1435 // find first empty slot in buffer (if any)
1436 int bufidx = 0;
1437 while ((bufmax < 0 || bufidx < bufmax) && buffer[bufidx])
1438 ++bufidx;
1439
1440 // set parser's optionCount
1441 parser.op_count = bufidx;
1442 }
1443
1444 bool perform(Option& option)
1445 {
1446 if (bufmax < 0 || parser.op_count < bufmax)
1447 {
1448 if (parser.op_count == 0x7fffffff)
1449 return false; // overflow protection: don't accept number of options that doesn't fit signed int
1450
1451 buffer[parser.op_count] = option;
1452 int idx = buffer[parser.op_count].desc->index;
1453 if (options[idx])
1454 options[idx].append(buffer[parser.op_count]);
1455 else
1456 options[idx] = buffer[parser.op_count];
1457 ++parser.op_count;
1458 }
1459 return true; // NOTE: an option that is discarded because of a full buffer is not fatal
1460 }
1461
1462 bool finished(int numargs, const char** args)
1463 {
1464 // only overwrite non-option argument list if there's at least 1
1465 // new non-option argument. Otherwise we keep the old list. This
1466 // makes it easy to use default non-option arguments.
1467 if (numargs > 0)
1468 {
1469 parser.nonop_count = numargs;
1470 parser.nonop_args = args;
1471 }
1472
1473 return true;
1474 }
1475};
1476
1477inline void Parser::parse(bool gnu, const Descriptor usage[], int argc, const char** argv, Option options[],
1478 Option buffer[], int min_abbr_len, bool single_minus_longopt, int bufmax)
1479{
1480 StoreOptionAction action(*this, options, buffer, bufmax);
1481 err = !workhorse(gnu, usage, argc, argv, action, single_minus_longopt, true, min_abbr_len);
1482}
1483
1484inline void Stats::add(bool gnu, const Descriptor usage[], int argc, const char** argv, int min_abbr_len,
1485 bool single_minus_longopt)
1486{
1487 // determine size of options array. This is the greatest index used in the usage + 1
1488 int i = 0;
1489 while (usage[i].shortopt != 0)
1490 {
1491 if (usage[i].index + 1 >= options_max)
1492 options_max = (usage[i].index + 1) + 1; // 1 more than necessary as sentinel
1493
1494 ++i;
1495 }
1496
1497 CountOptionsAction action(&buffer_max);
1498 Parser::workhorse(gnu, usage, argc, argv, action, single_minus_longopt, false, min_abbr_len);
1499}
1500
1501inline bool Parser::workhorse(bool gnu, const Descriptor usage[], int numargs, const char** args, Action& action,
1502 bool single_minus_longopt, bool print_errors, int min_abbr_len)
1503{
1504 // protect against NULL pointer
1505 if (args == 0)
1506 numargs = 0;
1507
1508 int nonops = 0;
1509
1510 while (numargs != 0 && *args != 0)
1511 {
1512 const char* param = *args; // param can be --long-option, -srto or non-option argument
1513
1514 // in POSIX mode the first non-option argument terminates the option list
1515 // a lone minus character is a non-option argument
1516 if (param[0] != '-' || param[1] == 0)
1517 {
1518 if (gnu)
1519 {
1520 ++nonops;
1521 ++args;
1522 if (numargs > 0)
1523 --numargs;
1524 continue;
1525 }
1526 else
1527 break;
1528 }
1529
1530 // -- terminates the option list. The -- itself is skipped.
1531 if (param[1] == '-' && param[2] == 0)
1532 {
1533 shift(args, nonops);
1534 ++args;
1535 if (numargs > 0)
1536 --numargs;
1537 break;
1538 }
1539
1540 bool handle_short_options;
1541 const char* longopt_name;
1542 if (param[1] == '-') // if --long-option
1543 {
1544 handle_short_options = false;
1545 longopt_name = param + 2;
1546 }
1547 else
1548 {
1549 handle_short_options = true;
1550 longopt_name = param + 1; //for testing a potential -long-option
1551 }
1552
1553 bool try_single_minus_longopt = single_minus_longopt;
1554 bool have_more_args = (numargs > 1 || numargs < 0); // is referencing argv[1] valid?
1555
1556 do // loop over short options in group, for long options the body is executed only once
1557 {
1558 int idx;
1559
1560 const char* optarg;
1561
1562 /******************** long option **********************/
1563 if (handle_short_options == false || try_single_minus_longopt)
1564 {
1565 idx = 0;
1566 while (usage[idx].longopt != 0 && !streq(usage[idx].longopt, longopt_name))
1567 ++idx;
1568
1569 if (usage[idx].longopt == 0 && min_abbr_len > 0) // if we should try to match abbreviated long options
1570 {
1571 int i1 = 0;
1572 while (usage[i1].longopt != 0 && !streqabbr(usage[i1].longopt, longopt_name, min_abbr_len))
1573 ++i1;
1574 if (usage[i1].longopt != 0)
1575 { // now test if the match is unambiguous by checking for another match
1576 int i2 = i1 + 1;
1577 while (usage[i2].longopt != 0 && !streqabbr(usage[i2].longopt, longopt_name, min_abbr_len))
1578 ++i2;
1579
1580 if (usage[i2].longopt == 0) // if there was no second match it's unambiguous, so accept i1 as idx
1581 idx = i1;
1582 }
1583 }
1584
1585 // if we found something, disable handle_short_options (only relevant if single_minus_longopt)
1586 if (usage[idx].longopt != 0)
1587 handle_short_options = false;
1588
1589 try_single_minus_longopt = false; // prevent looking for longopt in the middle of shortopt group
1590
1591 optarg = longopt_name;
1592 while (*optarg != 0 && *optarg != '=')
1593 ++optarg;
1594 if (*optarg == '=') // attached argument
1595 ++optarg;
1596 else
1597 // possibly detached argument
1598 optarg = (have_more_args ? args[1] : 0);
1599 }
1600
1601 /************************ short option ***********************************/
1602 if (handle_short_options)
1603 {
1604 if (*++param == 0) // point at the 1st/next option character
1605 break; // end of short option group
1606
1607 idx = 0;
1608 while (usage[idx].shortopt != 0 && !instr(*param, usage[idx].shortopt))
1609 ++idx;
1610
1611 if (param[1] == 0) // if the potential argument is separate
1612 optarg = (have_more_args ? args[1] : 0);
1613 else
1614 // if the potential argument is attached
1615 optarg = param + 1;
1616 }
1617
1618 const Descriptor* descriptor = &usage[idx];
1619
1620 if (descriptor->shortopt == 0) /************** unknown option ********************/
1621 {
1622 // look for dummy entry (shortopt == "" and longopt == "") to use as Descriptor for unknown options
1623 idx = 0;
1624 while (usage[idx].shortopt != 0 && (usage[idx].shortopt[0] != 0 || usage[idx].longopt[0] != 0))
1625 ++idx;
1626 descriptor = (usage[idx].shortopt == 0 ? 0 : &usage[idx]);
1627 }
1628
1629 if (descriptor != 0)
1630 {
1631 Option option(descriptor, param, optarg);
1632 switch (descriptor->check_arg(option, print_errors))
1633 {
1634 case ARG_ILLEGAL:
1635 return false; // fatal
1636 case ARG_OK:
1637 // skip one element of the argument vector, if it's a separated argument
1638 if (optarg != 0 && have_more_args && optarg == args[1])
1639 {
1640 shift(args, nonops);
1641 if (numargs > 0)
1642 --numargs;
1643 ++args;
1644 }
1645
1646 // No further short options are possible after an argument
1647 handle_short_options = false;
1648
1649 break;
1650 case ARG_IGNORE:
1651 case ARG_NONE:
1652 option.arg = 0;
1653 break;
1654 }
1655
1656 if (!action.perform(option))
1657 return false;
1658 }
1659
1660 } while (handle_short_options);
1661
1662 shift(args, nonops);
1663 ++args;
1664 if (numargs > 0)
1665 --numargs;
1666
1667 } // while
1668
1669 if (numargs > 0 && *args == 0) // It's a bug in the caller if numargs is greater than the actual number
1670 numargs = 0; // of arguments, but as a service to the user we fix this if we spot it.
1671
1672 if (numargs < 0) // if we don't know the number of remaining non-option arguments
1673 { // we need to count them
1674 numargs = 0;
1675 while (args[numargs] != 0)
1676 ++numargs;
1677 }
1678
1679 return action.finished(numargs + nonops, args - nonops);
1680}
1681
1682/**
1683 * @internal
1684 * @brief The implementation of option::printUsage().
1685 */
1686struct PrintUsageImplementation
1687{
1688 /**
1689 * @internal
1690 * @brief Interface for Functors that write (part of) a string somewhere.
1691 */
1692 struct IStringWriter
1693 {
1694 /**
1695 * @brief Writes the given number of chars beginning at the given pointer somewhere.
1696 */
1697 virtual void operator()(const char*, int)
1698 {
1699 }
1700 };
1701
1702 /**
1703 * @internal
1704 * @brief Encapsulates a function with signature <code>func(string, size)</code> where
1705 * string can be initialized with a const char* and size with an int.
1706 */
1707 template<typename Function>
1708 struct FunctionWriter: public IStringWriter
1709 {
1710 Function* write;
1711
1712 virtual void operator()(const char* str, int size)
1713 {
1714 (*write)(str, size);
1715 }
1716
1717 FunctionWriter(Function* w) :
1718 write(w)
1719 {
1720 }
1721 };
1722
1723 /**
1724 * @internal
1725 * @brief Encapsulates a reference to an object with a <code>write(string, size)</code>
1726 * method like that of @c std::ostream.
1727 */
1728 template<typename OStream>
1729 struct OStreamWriter: public IStringWriter
1730 {
1731 OStream& ostream;
1732
1733 virtual void operator()(const char* str, int size)
1734 {
1735 ostream.write(str, size);
1736 }
1737
1738 OStreamWriter(OStream& o) :
1739 ostream(o)
1740 {
1741 }
1742 };
1743
1744 /**
1745 * @internal
1746 * @brief Like OStreamWriter but encapsulates a @c const reference, which is
1747 * typically a temporary object of a user class.
1748 */
1749 template<typename Temporary>
1750 struct TemporaryWriter: public IStringWriter
1751 {
1752 const Temporary& userstream;
1753
1754 virtual void operator()(const char* str, int size)
1755 {
1756 userstream.write(str, size);
1757 }
1758
1759 TemporaryWriter(const Temporary& u) :
1760 userstream(u)
1761 {
1762 }
1763 };
1764
1765 /**
1766 * @internal
1767 * @brief Encapsulates a function with the signature <code>func(fd, string, size)</code> (the
1768 * signature of the @c write() system call)
1769 * where fd can be initialized from an int, string from a const char* and size from an int.
1770 */
1771 template<typename Syscall>
1772 struct SyscallWriter: public IStringWriter
1773 {
1774 Syscall* write;
1775 int fd;
1776
1777 virtual void operator()(const char* str, int size)
1778 {
1779 (*write)(fd, str, size);
1780 }
1781
1782 SyscallWriter(Syscall* w, int f) :
1783 write(w), fd(f)
1784 {
1785 }
1786 };
1787
1788 /**
1789 * @internal
1790 * @brief Encapsulates a function with the same signature as @c std::fwrite().
1791 */
1792 template<typename Function, typename Stream>
1793 struct StreamWriter: public IStringWriter
1794 {
1795 Function* fwrite;
1796 Stream* stream;
1797
1798 virtual void operator()(const char* str, int size)
1799 {
1800 (*fwrite)(str, size, 1, stream);
1801 }
1802
1803 StreamWriter(Function* w, Stream* s) :
1804 fwrite(w), stream(s)
1805 {
1806 }
1807 };
1808
1809 /**
1810 * @internal
1811 * @brief Sets <code> i1 = max(i1, i2) </code>
1812 */
1813 static void upmax(int& i1, int i2)
1814 {
1815 i1 = (i1 >= i2 ? i1 : i2);
1816 }
1817
1818 /**
1819 * @internal
1820 * @brief Moves the "cursor" to column @c want_x assuming it is currently at column @c x
1821 * and sets @c x=want_x .
1822 * If <code> x > want_x </code>, a line break is output before indenting.
1823 *
1824 * @param write Spaces and possibly a line break are written via this functor to get
1825 * the desired indentation @c want_x .
1826 * @param[in,out] x the current indentation. Set to @c want_x by this method.
1827 * @param want_x the desired indentation.
1828 */
1829 static void indent(IStringWriter& write, int& x, int want_x)
1830 {
1831 int indent = want_x - x;
1832 if (indent < 0)
1833 {
1834 write("\n", 1);
1835 indent = want_x;
1836 }
1837
1838 if (indent > 0)
1839 {
1840 char space = ' ';
1841 for (int i = 0; i < indent; ++i)
1842 write(&space, 1);
1843 x = want_x;
1844 }
1845 }
1846
1847 /**
1848 * @brief Returns true if ch is the unicode code point of a wide character.
1849 *
1850 * @note
1851 * The following character ranges are treated as wide
1852 * @code
1853 * 1100..115F
1854 * 2329..232A (just 2 characters!)
1855 * 2E80..A4C6 except for 303F
1856 * A960..A97C
1857 * AC00..D7FB
1858 * F900..FAFF
1859 * FE10..FE6B
1860 * FF01..FF60
1861 * FFE0..FFE6
1862 * 1B000......
1863 * @endcode
1864 */
1865 static bool isWideChar(unsigned ch)
1866 {
1867 if (ch == 0x303F)
1868 return false;
1869
1870 return ((0x1100 <= ch && ch <= 0x115F) || (0x2329 <= ch && ch <= 0x232A) || (0x2E80 <= ch && ch <= 0xA4C6)
1871 || (0xA960 <= ch && ch <= 0xA97C) || (0xAC00 <= ch && ch <= 0xD7FB) || (0xF900 <= ch && ch <= 0xFAFF)
1872 || (0xFE10 <= ch && ch <= 0xFE6B) || (0xFF01 <= ch && ch <= 0xFF60) || (0xFFE0 <= ch && ch <= 0xFFE6)
1873 || (0x1B000 <= ch));
1874 }
1875
1876 /**
1877 * @internal
1878 * @brief Splits a @c Descriptor[] array into tables, rows, lines and columns and
1879 * iterates over these components.
1880 *
1881 * The top-level organizational unit is the @e table.
1882 * A table begins at a Descriptor with @c help!=NULL and extends up to
1883 * a Descriptor with @c help==NULL.
1884 *
1885 * A table consists of @e rows. Due to line-wrapping and explicit breaks
1886 * a row may take multiple lines on screen. Rows within the table are separated
1887 * by \\n. They never cross Descriptor boundaries. This means a row ends either
1888 * at \\n or the 0 at the end of the help string.
1889 *
1890 * A row consists of columns/cells. Columns/cells within a row are separated by \\t.
1891 * Line breaks within a cell are marked by \\v.
1892 *
1893 * Rows in the same table need not have the same number of columns/cells. The
1894 * extreme case are interjections, which are rows that contain neither \\t nor \\v.
1895 * These are NOT treated specially by LinePartIterator, but they are treated
1896 * specially by printUsage().
1897 *
1898 * LinePartIterator iterates through the usage at 3 levels: table, row and part.
1899 * Tables and rows are as described above. A @e part is a line within a cell.
1900 * LinePartIterator iterates through 1st parts of all cells, then through the 2nd
1901 * parts of all cells (if any),... @n
1902 * Example: The row <code> "1 \v 3 \t 2 \v 4" </code> has 2 cells/columns and 4 parts.
1903 * The parts will be returned in the order 1, 2, 3, 4.
1904 *
1905 * It is possible that some cells have fewer parts than others. In this case
1906 * LinePartIterator will "fill up" these cells with 0-length parts. IOW, LinePartIterator
1907 * always returns the same number of parts for each column. Note that this is different
1908 * from the way rows and columns are handled. LinePartIterator does @e not guarantee that
1909 * the same number of columns will be returned for each row.
1910 *
1911 */
1912 class LinePartIterator
1913 {
1914 const Descriptor* tablestart; //!< The 1st descriptor of the current table.
1915 const Descriptor* rowdesc; //!< The Descriptor that contains the current row.
1916 const char* rowstart; //!< Ptr to 1st character of current row within rowdesc->help.
1917 const char* ptr; //!< Ptr to current part within the current row.
1918 int col; //!< Index of current column.
1919 int len; //!< Length of the current part (that ptr points at) in BYTES
1920 int screenlen; //!< Length of the current part in screen columns (taking narrow/wide chars into account).
1921 int max_line_in_block; //!< Greatest index of a line within the block. This is the number of \\v within the cell with the most \\vs.
1922 int line_in_block; //!< Line index within the current cell of the current part.
1923 int target_line_in_block; //!< Line index of the parts we should return to the user on this iteration.
1924 bool hit_target_line; //!< Flag whether we encountered a part with line index target_line_in_block in the current cell.
1925
1926 /**
1927 * @brief Determines the byte and character lengths of the part at @ref ptr and
1928 * stores them in @ref len and @ref screenlen respectively.
1929 */
1930 void update_length()
1931 {
1932 screenlen = 0;
1933 for (len = 0; ptr[len] != 0 && ptr[len] != '\v' && ptr[len] != '\t' && ptr[len] != '\n'; ++len)
1934 {
1935 ++screenlen;
1936 unsigned ch = (unsigned char) ptr[len];
1937 if (ch > 0xC1) // everything <= 0xC1 (yes, even 0xC1 itself) is not a valid UTF-8 start byte
1938 {
1939 // int __builtin_clz (unsigned int x)
1940 // Returns the number of leading 0-bits in x, starting at the most significant bit
1941 unsigned mask = (unsigned) -1 >> __builtin_clz(ch ^ 0xff);
1942 ch = ch & mask; // mask out length bits, we don't verify their correctness
1943 while (((unsigned char) ptr[len + 1] ^ 0x80) <= 0x3F) // while next byte is continuation byte
1944 {
1945 ch = (ch << 6) ^ (unsigned char) ptr[len + 1] ^ 0x80; // add continuation to char code
1946 ++len;
1947 }
1948 // ch is the decoded unicode code point
1949 if (ch >= 0x1100 && isWideChar(ch)) // the test for 0x1100 is here to avoid the function call in the Latin case
1950 ++screenlen;
1951 }
1952 }
1953 }
1954
1955 public:
1956 //! @brief Creates an iterator for @c usage.
1957 LinePartIterator(const Descriptor usage[]) :
1958 tablestart(usage), rowdesc(0), rowstart(0), ptr(0), col(-1), len(0), max_line_in_block(0), line_in_block(0),
1959 target_line_in_block(0), hit_target_line(true)
1960 {
1961 }
1962
1963 /**
1964 * @brief Moves iteration to the next table (if any). Has to be called once on a new
1965 * LinePartIterator to move to the 1st table.
1966 * @retval false if moving to next table failed because no further table exists.
1967 */
1968 bool nextTable()
1969 {
1970 // If this is NOT the first time nextTable() is called after the constructor,
1971 // then skip to the next table break (i.e. a Descriptor with help == 0)
1972 if (rowdesc != 0)
1973 {
1974 while (tablestart->help != 0 && tablestart->shortopt != 0)
1975 ++tablestart;
1976 }
1977
1978 // Find the next table after the break (if any)
1979 while (tablestart->help == 0 && tablestart->shortopt != 0)
1980 ++tablestart;
1981
1982 restartTable();
1983 return rowstart != 0;
1984 }
1985
1986 /**
1987 * @brief Reset iteration to the beginning of the current table.
1988 */
1989 void restartTable()
1990 {
1991 rowdesc = tablestart;
1992 rowstart = tablestart->help;
1993 ptr = 0;
1994 }
1995
1996 /**
1997 * @brief Moves iteration to the next row (if any). Has to be called once after each call to
1998 * @ref nextTable() to move to the 1st row of the table.
1999 * @retval false if moving to next row failed because no further row exists.
2000 */
2001 bool nextRow()
2002 {
2003 if (ptr == 0)
2004 {
2005 restartRow();
2006 return rowstart != 0;
2007 }
2008
2009 while (*ptr != 0 && *ptr != '\n')
2010 ++ptr;
2011
2012 if (*ptr == 0)
2013 {
2014 if ((rowdesc + 1)->help == 0) // table break
2015 return false;
2016
2017 ++rowdesc;
2018 rowstart = rowdesc->help;
2019 }
2020 else // if (*ptr == '\n')
2021 {
2022 rowstart = ptr + 1;
2023 }
2024
2025 restartRow();
2026 return true;
2027 }
2028
2029 /**
2030 * @brief Reset iteration to the beginning of the current row.
2031 */
2032 void restartRow()
2033 {
2034 ptr = rowstart;
2035 col = -1;
2036 len = 0;
2037 screenlen = 0;
2038 max_line_in_block = 0;
2039 line_in_block = 0;
2040 target_line_in_block = 0;
2041 hit_target_line = true;
2042 }
2043
2044 /**
2045 * @brief Moves iteration to the next part (if any). Has to be called once after each call to
2046 * @ref nextRow() to move to the 1st part of the row.
2047 * @retval false if moving to next part failed because no further part exists.
2048 *
2049 * See @ref LinePartIterator for details about the iteration.
2050 */
2051 bool next()
2052 {
2053 if (ptr == 0)
2054 return false;
2055
2056 if (col == -1)
2057 {
2058 col = 0;
2059 update_length();
2060 return true;
2061 }
2062
2063 ptr += len;
2064 while (true)
2065 {
2066 switch (*ptr)
2067 {
2068 case '\v':
2069 upmax(max_line_in_block, ++line_in_block);
2070 ++ptr;
2071 break;
2072 case '\t':
2073 if (!hit_target_line) // if previous column did not have the targetline
2074 { // then "insert" a 0-length part
2075 update_length();
2076 hit_target_line = true;
2077 return true;
2078 }
2079
2080 hit_target_line = false;
2081 line_in_block = 0;
2082 ++col;
2083 ++ptr;
2084 break;
2085 case 0:
2086 case '\n':
2087 if (!hit_target_line) // if previous column did not have the targetline
2088 { // then "insert" a 0-length part
2089 update_length();
2090 hit_target_line = true;
2091 return true;
2092 }
2093
2094 if (++target_line_in_block > max_line_in_block)
2095 {
2096 update_length();
2097 return false;
2098 }
2099
2100 hit_target_line = false;
2101 line_in_block = 0;
2102 col = 0;
2103 ptr = rowstart;
2104 continue;
2105 default:
2106 ++ptr;
2107 continue;
2108 } // switch
2109
2110 if (line_in_block == target_line_in_block)
2111 {
2112 update_length();
2113 hit_target_line = true;
2114 return true;
2115 }
2116 } // while
2117 }
2118
2119 /**
2120 * @brief Returns the index (counting from 0) of the column in which
2121 * the part pointed to by @ref data() is located.
2122 */
2123 int column()
2124 {
2125 return col;
2126 }
2127
2128 /**
2129 * @brief Returns the index (counting from 0) of the line within the current column
2130 * this part belongs to.
2131 */
2132 int line()
2133 {
2134 return target_line_in_block; // NOT line_in_block !!! It would be wrong if !hit_target_line
2135 }
2136
2137 /**
2138 * @brief Returns the length of the part pointed to by @ref data() in raw chars (not UTF-8 characters).
2139 */
2140 int length()
2141 {
2142 return len;
2143 }
2144
2145 /**
2146 * @brief Returns the width in screen columns of the part pointed to by @ref data().
2147 * Takes multi-byte UTF-8 sequences and wide characters into account.
2148 */
2149 int screenLength()
2150 {
2151 return screenlen;
2152 }
2153
2154 /**
2155 * @brief Returns the current part of the iteration.
2156 */
2157 const char* data()
2158 {
2159 return ptr;
2160 }
2161 };
2162
2163 /**
2164 * @internal
2165 * @brief Takes input and line wraps it, writing out one line at a time so that
2166 * it can be interleaved with output from other columns.
2167 *
2168 * The LineWrapper is used to handle the last column of each table as well as interjections.
2169 * The LineWrapper is called once for each line of output. If the data given to it fits
2170 * into the designated width of the last column it is simply written out. If there
2171 * is too much data, an appropriate split point is located and only the data up to this
2172 * split point is written out. The rest of the data is queued for the next line.
2173 * That way the last column can be line wrapped and interleaved with data from
2174 * other columns. The following example makes this clearer:
2175 * @code
2176 * Column 1,1 Column 2,1 This is a long text
2177 * Column 1,2 Column 2,2 that does not fit into
2178 * a single line.
2179 * @endcode
2180 *
2181 * The difficulty in producing this output is that the whole string
2182 * "This is a long text that does not fit into a single line" is the
2183 * 1st and only part of column 3. In order to produce the above
2184 * output the string must be output piecemeal, interleaved with
2185 * the data from the other columns.
2186 */
2187 class LineWrapper
2188 {
2189 static const int bufmask = 15; //!< Must be a power of 2 minus 1.
2190 /**
2191 * @brief Ring buffer for length component of pair (data, length).
2192 */
2193 int lenbuf[bufmask + 1];
2194 /**
2195 * @brief Ring buffer for data component of pair (data, length).
2196 */
2197 const char* datbuf[bufmask + 1];
2198 /**
2199 * @brief The indentation of the column to which the LineBuffer outputs. LineBuffer
2200 * assumes that the indentation has already been written when @ref process()
2201 * is called, so this value is only used when a buffer flush requires writing
2202 * additional lines of output.
2203 */
2204 int x;
2205 /**
2206 * @brief The width of the column to line wrap.
2207 */
2208 int width;
2209 int head; //!< @brief index for next write
2210 int tail; //!< @brief index for next read - 1 (i.e. increment tail BEFORE read)
2211
2212 /**
2213 * @brief Multiple methods of LineWrapper may decide to flush part of the buffer to
2214 * free up space. The contract of process() says that only 1 line is output. So
2215 * this variable is used to track whether something has output a line. It is
2216 * reset at the beginning of process() and checked at the end to decide if
2217 * output has already occurred or is still needed.
2218 */
2219 bool wrote_something;
2220
2221 bool buf_empty()
2222 {
2223 return ((tail + 1) & bufmask) == head;
2224 }
2225
2226 bool buf_full()
2227 {
2228 return tail == head;
2229 }
2230
2231 void buf_store(const char* data, int len)
2232 {
2233 lenbuf[head] = len;
2234 datbuf[head] = data;
2235 head = (head + 1) & bufmask;
2236 }
2237
2238 //! @brief Call BEFORE reading ...buf[tail].
2239 void buf_next()
2240 {
2241 tail = (tail + 1) & bufmask;
2242 }
2243
2244 /**
2245 * @brief Writes (data,len) into the ring buffer. If the buffer is full, a single line
2246 * is flushed out of the buffer into @c write.
2247 */
2248 void output(IStringWriter& write, const char* data, int len)
2249 {
2250 if (buf_full())
2251 write_one_line(write);
2252
2253 buf_store(data, len);
2254 }
2255
2256 /**
2257 * @brief Writes a single line of output from the buffer to @c write.
2258 */
2259 void write_one_line(IStringWriter& write)
2260 {
2261 if (wrote_something) // if we already wrote something, we need to start a new line
2262 {
2263 write("\n", 1);
2264 int _ = 0;
2265 indent(write, _, x);
2266 }
2267
2268 if (!buf_empty())
2269 {
2270 buf_next();
2271 write(datbuf[tail], lenbuf[tail]);
2272 }
2273
2274 wrote_something = true;
2275 }
2276 public:
2277
2278 /**
2279 * @brief Writes out all remaining data from the LineWrapper using @c write.
2280 * Unlike @ref process() this method indents all lines including the first and
2281 * will output a \\n at the end (but only if something has been written).
2282 */
2283 void flush(IStringWriter& write)
2284 {
2285 if (buf_empty())
2286 return;
2287 int _ = 0;
2288 indent(write, _, x);
2289 wrote_something = false;
2290 while (!buf_empty())
2291 write_one_line(write);
2292 write("\n", 1);
2293 }
2294
2295 /**
2296 * @brief Process, wrap and output the next piece of data.
2297 *
2298 * process() will output at least one line of output. This is not necessarily
2299 * the @c data passed in. It may be data queued from a prior call to process().
2300 * If the internal buffer is full, more than 1 line will be output.
2301 *
2302 * process() assumes that the a proper amount of indentation has already been
2303 * output. It won't write any further indentation before the 1st line. If
2304 * more than 1 line is written due to buffer constraints, the lines following
2305 * the first will be indented by this method, though.
2306 *
2307 * No \\n is written by this method after the last line that is written.
2308 *
2309 * @param write where to write the data.
2310 * @param data the new chunk of data to write.
2311 * @param len the length of the chunk of data to write.
2312 */
2313 void process(IStringWriter& write, const char* data, int len)
2314 {
2315 wrote_something = false;
2316
2317 while (len > 0)
2318 {
2319 if (len <= width) // quick test that works because utf8width <= len (all wide chars have at least 2 bytes)
2320 {
2321 output(write, data, len);
2322 len = 0;
2323 }
2324 else // if (len > width) it's possible (but not guaranteed) that utf8len > width
2325 {
2326 int utf8width = 0;
2327 int maxi = 0;
2328 while (maxi < len && utf8width < width)
2329 {
2330 int charbytes = 1;
2331 unsigned ch = (unsigned char) data[maxi];
2332 if (ch > 0xC1) // everything <= 0xC1 (yes, even 0xC1 itself) is not a valid UTF-8 start byte
2333 {
2334 // int __builtin_clz (unsigned int x)
2335 // Returns the number of leading 0-bits in x, starting at the most significant bit
2336 unsigned mask = (unsigned) -1 >> __builtin_clz(ch ^ 0xff);
2337 ch = ch & mask; // mask out length bits, we don't verify their correctness
2338 while ((maxi + charbytes < len) && //
2339 (((unsigned char) data[maxi + charbytes] ^ 0x80) <= 0x3F)) // while next byte is continuation byte
2340 {
2341 ch = (ch << 6) ^ (unsigned char) data[maxi + charbytes] ^ 0x80; // add continuation to char code
2342 ++charbytes;
2343 }
2344 // ch is the decoded unicode code point
2345 if (ch >= 0x1100 && isWideChar(ch)) // the test for 0x1100 is here to avoid the function call in the Latin case
2346 {
2347 if (utf8width + 2 > width)
2348 break;
2349 ++utf8width;
2350 }
2351 }
2352 ++utf8width;
2353 maxi += charbytes;
2354 }
2355
2356 // data[maxi-1] is the last byte of the UTF-8 sequence of the last character that fits
2357 // onto the 1st line. If maxi == len, all characters fit on the line.
2358
2359 if (maxi == len)
2360 {
2361 output(write, data, len);
2362 len = 0;
2363 }
2364 else // if (maxi < len) at least 1 character (data[maxi] that is) doesn't fit on the line
2365 {
2366 int i;
2367 for (i = maxi; i >= 0; --i)
2368 if (data[i] == ' ')
2369 break;
2370
2371 if (i >= 0)
2372 {
2373 output(write, data, i);
2374 data += i + 1;
2375 len -= i + 1;
2376 }
2377 else // did not find a space to split at => split before data[maxi]
2378 { // data[maxi] is always the beginning of a character, never a continuation byte
2379 output(write, data, maxi);
2380 data += maxi;
2381 len -= maxi;
2382 }
2383 }
2384 }
2385 }
2386 if (!wrote_something) // if we didn't already write something to make space in the buffer
2387 write_one_line(write); // write at most one line of actual output
2388 }
2389
2390 /**
2391 * @brief Constructs a LineWrapper that wraps its output to fit into
2392 * screen columns @c x1 (incl.) to @c x2 (excl.).
2393 *
2394 * @c x1 gives the indentation LineWrapper uses if it needs to indent.
2395 */
2396 LineWrapper(int x1, int x2) :
2397 x(x1), width(x2 - x1), head(0), tail(bufmask)
2398 {
2399 if (width < 2) // because of wide characters we need at least width 2 or the code breaks
2400 width = 2;
2401 }
2402 };
2403
2404 /**
2405 * @internal
2406 * @brief This is the implementation that is shared between all printUsage() templates.
2407 * Because all printUsage() templates share this implementation, there is no template bloat.
2408 */
2409 static void printUsage(IStringWriter& write, const Descriptor usage[], int width = 80, //
2410 int last_column_min_percent = 50, int last_column_own_line_max_percent = 75)
2411 {
2412 if (width < 1) // protect against nonsense values
2413 width = 80;
2414
2415 if (width > 10000) // protect against overflow in the following computation
2416 width = 10000;
2417
2418 int last_column_min_width = ((width * last_column_min_percent) + 50) / 100;
2419 int last_column_own_line_max_width = ((width * last_column_own_line_max_percent) + 50) / 100;
2420 if (last_column_own_line_max_width == 0)
2421 last_column_own_line_max_width = 1;
2422
2423 LinePartIterator part(usage);
2424 while (part.nextTable())
2425 {
2426
2427 /***************** Determine column widths *******************************/
2428
2429 const int maxcolumns = 8; // 8 columns are enough for everyone
2430 int col_width[maxcolumns];
2431 int lastcolumn;
2432 int leftwidth;
2433 int overlong_column_threshold = 10000;
2434 do
2435 {
2436 lastcolumn = 0;
2437 for (int i = 0; i < maxcolumns; ++i)
2438 col_width[i] = 0;
2439
2440 part.restartTable();
2441 while (part.nextRow())
2442 {
2443 while (part.next())
2444 {
2445 if (part.column() < maxcolumns)
2446 {
2447 upmax(lastcolumn, part.column());
2448 if (part.screenLength() < overlong_column_threshold)
2449 // We don't let rows that don't use table separators (\t or \v) influence
2450 // the width of column 0. This allows the user to interject section headers
2451 // or explanatory paragraphs that do not participate in the table layout.
2452 if (part.column() > 0 || part.line() > 0 || part.data()[part.length()] == '\t'
2453 || part.data()[part.length()] == '\v')
2454 upmax(col_width[part.column()], part.screenLength());
2455 }
2456 }
2457 }
2458
2459 /*
2460 * If the last column doesn't fit on the same
2461 * line as the other columns, we can fix that by starting it on its own line.
2462 * However we can't do this for any of the columns 0..lastcolumn-1.
2463 * If their sum exceeds the maximum width we try to fix this by iteratively
2464 * ignoring the widest line parts in the width determination until
2465 * we arrive at a series of column widths that fit into one line.
2466 * The result is a layout where everything is nicely formatted
2467 * except for a few overlong fragments.
2468 * */
2469
2470 leftwidth = 0;
2471 overlong_column_threshold = 0;
2472 for (int i = 0; i < lastcolumn; ++i)
2473 {
2474 leftwidth += col_width[i];
2475 upmax(overlong_column_threshold, col_width[i]);
2476 }
2477
2478 } while (leftwidth > width);
2479
2480 /**************** Determine tab stops and last column handling **********************/
2481
2482 int tabstop[maxcolumns];
2483 tabstop[0] = 0;
2484 for (int i = 1; i < maxcolumns; ++i)
2485 tabstop[i] = tabstop[i - 1] + col_width[i - 1];
2486
2487 int rightwidth = width - tabstop[lastcolumn];
2488 bool print_last_column_on_own_line = false;
2489 if (rightwidth < last_column_min_width && // if we don't have the minimum requested width for the last column
2490 ( col_width[lastcolumn] == 0 || // and all last columns are > overlong_column_threshold
2491 rightwidth < col_width[lastcolumn] // or there is at least one last column that requires more than the space available
2492 )
2493 )
2494 {
2495 print_last_column_on_own_line = true;
2496 rightwidth = last_column_own_line_max_width;
2497 }
2498
2499 // If lastcolumn == 0 we must disable print_last_column_on_own_line because
2500 // otherwise 2 copies of the last (and only) column would be output.
2501 // Actually this is just defensive programming. It is currently not
2502 // possible that lastcolumn==0 and print_last_column_on_own_line==true
2503 // at the same time, because lastcolumn==0 => tabstop[lastcolumn] == 0 =>
2504 // rightwidth==width => rightwidth>=last_column_min_width (unless someone passes
2505 // a bullshit value >100 for last_column_min_percent) => the above if condition
2506 // is false => print_last_column_on_own_line==false
2507 if (lastcolumn == 0)
2508 print_last_column_on_own_line = false;
2509
2510 LineWrapper lastColumnLineWrapper(width - rightwidth, width);
2511 LineWrapper interjectionLineWrapper(0, width);
2512
2513 part.restartTable();
2514
2515 /***************** Print out all rows of the table *************************************/
2516
2517 while (part.nextRow())
2518 {
2519 int x = -1;
2520 while (part.next())
2521 {
2522 if (part.column() > lastcolumn)
2523 continue; // drop excess columns (can happen if lastcolumn == maxcolumns-1)
2524
2525 if (part.column() == 0)
2526 {
2527 if (x >= 0)
2528 write("\n", 1);
2529 x = 0;
2530 }
2531
2532 indent(write, x, tabstop[part.column()]);
2533
2534 if ((part.column() < lastcolumn)
2535 && (part.column() > 0 || part.line() > 0 || part.data()[part.length()] == '\t'
2536 || part.data()[part.length()] == '\v'))
2537 {
2538 write(part.data(), part.length());
2539 x += part.screenLength();
2540 }
2541 else // either part.column() == lastcolumn or we are in the special case of
2542 // an interjection that doesn't contain \v or \t
2543 {
2544 // NOTE: This code block is not necessarily executed for
2545 // each line, because some rows may have fewer columns.
2546
2547 LineWrapper& lineWrapper = (part.column() == 0) ? interjectionLineWrapper : lastColumnLineWrapper;
2548
2549 if (!print_last_column_on_own_line || part.column() != lastcolumn)
2550 lineWrapper.process(write, part.data(), part.length());
2551 }
2552 } // while
2553
2554 if (print_last_column_on_own_line)
2555 {
2556 part.restartRow();
2557 while (part.next())
2558 {
2559 if (part.column() == lastcolumn)
2560 {
2561 write("\n", 1);
2562 int _ = 0;
2563 indent(write, _, width - rightwidth);
2564 lastColumnLineWrapper.process(write, part.data(), part.length());
2565 }
2566 }
2567 }
2568
2569 write("\n", 1);
2570 lastColumnLineWrapper.flush(write);
2571 interjectionLineWrapper.flush(write);
2572 }
2573 }
2574 }
2575
2576}
2577;
2578
2579/**
2580 * @brief Outputs a nicely formatted usage string with support for multi-column formatting
2581 * and line-wrapping.
2582 *
2583 * printUsage() takes the @c help texts of a Descriptor[] array and formats them into
2584 * a usage message, wrapping lines to achieve the desired output width.
2585 *
2586 * <b>Table formatting:</b>
2587 *
2588 * Aside from plain strings which are simply line-wrapped, the usage may contain tables. Tables
2589 * are used to align elements in the output.
2590 *
2591 * @code
2592 * // Without a table. The explanatory texts are not aligned.
2593 * -c, --create |Creates something.
2594 * -k, --kill |Destroys something.
2595 *
2596 * // With table formatting. The explanatory texts are aligned.
2597 * -c, --create |Creates something.
2598 * -k, --kill |Destroys something.
2599 * @endcode
2600 *
2601 * Table formatting removes the need to pad help texts manually with spaces to achieve
2602 * alignment. To create a table, simply insert \\t (tab) characters to separate the cells
2603 * within a row.
2604 *
2605 * @code
2606 * const option::Descriptor usage[] = {
2607 * {..., "-c, --create \tCreates something." },
2608 * {..., "-k, --kill \tDestroys something." }, ...
2609 * @endcode
2610 *
2611 * Note that you must include the minimum amount of space desired between cells yourself.
2612 * Table formatting will insert further spaces as needed to achieve alignment.
2613 *
2614 * You can insert line breaks within cells by using \\v (vertical tab).
2615 *
2616 * @code
2617 * const option::Descriptor usage[] = {
2618 * {..., "-c,\v--create \tCreates\vsomething." },
2619 * {..., "-k,\v--kill \tDestroys\vsomething." }, ...
2620 *
2621 * // results in
2622 *
2623 * -c, Creates
2624 * --create something.
2625 * -k, Destroys
2626 * --kill something.
2627 * @endcode
2628 *
2629 * You can mix lines that do not use \\t or \\v with those that do. The plain
2630 * lines will not mess up the table layout. Alignment of the table columns will
2631 * be maintained even across these interjections.
2632 *
2633 * @code
2634 * const option::Descriptor usage[] = {
2635 * {..., "-c, --create \tCreates something." },
2636 * {..., "----------------------------------" },
2637 * {..., "-k, --kill \tDestroys something." }, ...
2638 *
2639 * // results in
2640 *
2641 * -c, --create Creates something.
2642 * ----------------------------------
2643 * -k, --kill Destroys something.
2644 * @endcode
2645 *
2646 * You can have multiple tables within the same usage whose columns are
2647 * aligned independently. Simply insert a dummy Descriptor with @c help==0.
2648 *
2649 * @code
2650 * const option::Descriptor usage[] = {
2651 * {..., "Long options:" },
2652 * {..., "--very-long-option \tDoes something long." },
2653 * {..., "--ultra-super-mega-long-option \tTakes forever to complete." },
2654 * {..., 0 }, // ---------- table break -----------
2655 * {..., "Short options:" },
2656 * {..., "-s \tShort." },
2657 * {..., "-q \tQuick." }, ...
2658 *
2659 * // results in
2660 *
2661 * Long options:
2662 * --very-long-option Does something long.
2663 * --ultra-super-mega-long-option Takes forever to complete.
2664 * Short options:
2665 * -s Short.
2666 * -q Quick.
2667 *
2668 * // Without the table break it would be
2669 *
2670 * Long options:
2671 * --very-long-option Does something long.
2672 * --ultra-super-mega-long-option Takes forever to complete.
2673 * Short options:
2674 * -s Short.
2675 * -q Quick.
2676 * @endcode
2677 *
2678 * <b>Output methods:</b>
2679 *
2680 * Because TheLeanMeanC++Option parser is freestanding, you have to provide the means for
2681 * output in the first argument(s) to printUsage(). Because printUsage() is implemented as
2682 * a set of template functions, you have great flexibility in your choice of output
2683 * method. The following example demonstrates typical uses. Anything that's similar enough
2684 * will work.
2685 *
2686 * @code
2687 * #include <unistd.h> // write()
2688 * #include <iostream> // cout
2689 * #include <sstream> // ostringstream
2690 * #include <cstdio> // fwrite()
2691 * using namespace std;
2692 *
2693 * void my_write(const char* str, int size) {
2694 * fwrite(str, size, 1, stdout);
2695 * }
2696 *
2697 * struct MyWriter {
2698 * void write(const char* buf, size_t size) const {
2699 * fwrite(str, size, 1, stdout);
2700 * }
2701 * };
2702 *
2703 * struct MyWriteFunctor {
2704 * void operator()(const char* buf, size_t size) {
2705 * fwrite(str, size, 1, stdout);
2706 * }
2707 * };
2708 * ...
2709 * printUsage(my_write, usage); // custom write function
2710 * printUsage(MyWriter(), usage); // temporary of a custom class
2711 * MyWriter writer;
2712 * printUsage(writer, usage); // custom class object
2713 * MyWriteFunctor wfunctor;
2714 * printUsage(&wfunctor, usage); // custom functor
2715 * printUsage(write, 1, usage); // write() to file descriptor 1
2716 * printUsage(cout, usage); // an ostream&
2717 * printUsage(fwrite, stdout, usage); // fwrite() to stdout
2718 * ostringstream sstr;
2719 * printUsage(sstr, usage); // an ostringstream&
2720 *
2721 * @endcode
2722 *
2723 * @par Notes:
2724 * @li the @c write() method of a class that is to be passed as a temporary
2725 * as @c MyWriter() is in the example, must be a @c const method, because
2726 * temporary objects are passed as const reference. This only applies to
2727 * temporary objects that are created and destroyed in the same statement.
2728 * If you create an object like @c writer in the example, this restriction
2729 * does not apply.
2730 * @li a functor like @c MyWriteFunctor in the example must be passed as a pointer.
2731 * This differs from the way functors are passed to e.g. the STL algorithms.
2732 * @li All printUsage() templates are tiny wrappers around a shared non-template implementation.
2733 * So there's no penalty for using different versions in the same program.
2734 * @li printUsage() always interprets Descriptor::help as UTF-8 and always produces UTF-8-encoded
2735 * output. If your system uses a different charset, you must do your own conversion. You
2736 * may also need to change the font of the console to see non-ASCII characters properly.
2737 * This is particularly true for Windows.
2738 * @li @b Security @b warning: Do not insert untrusted strings (such as user-supplied arguments)
2739 * into the usage. printUsage() has no protection against malicious UTF-8 sequences.
2740 *
2741 * @param prn The output method to use. See the examples above.
2742 * @param usage the Descriptor[] array whose @c help texts will be formatted.
2743 * @param width the maximum number of characters per output line. Note that this number is
2744 * in actual characters, not bytes. printUsage() supports UTF-8 in @c help and will
2745 * count multi-byte UTF-8 sequences properly. Asian wide characters are counted
2746 * as 2 characters.
2747 * @param last_column_min_percent (0-100) The minimum percentage of @c width that should be available
2748 * for the last column (which typically contains the textual explanation of an option).
2749 * If less space is available, the last column will be printed on its own line, indented
2750 * according to @c last_column_own_line_max_percent.
2751 * @param last_column_own_line_max_percent (0-100) If the last column is printed on its own line due to
2752 * less than @c last_column_min_percent of the width being available, then only
2753 * @c last_column_own_line_max_percent of the extra line(s) will be used for the
2754 * last column's text. This ensures an indentation. See example below.
2755 *
2756 * @code
2757 * // width=20, last_column_min_percent=50 (i.e. last col. min. width=10)
2758 * --3456789 1234567890
2759 * 1234567890
2760 *
2761 * // width=20, last_column_min_percent=75 (i.e. last col. min. width=15)
2762 * // last_column_own_line_max_percent=75
2763 * --3456789
2764 * 123456789012345
2765 * 67890
2766 *
2767 * // width=20, last_column_min_percent=75 (i.e. last col. min. width=15)
2768 * // last_column_own_line_max_percent=33 (i.e. max. 5)
2769 * --3456789
2770 * 12345
2771 * 67890
2772 * 12345
2773 * 67890
2774 * @endcode
2775 */
2776template<typename OStream>
2777void printUsage(OStream& prn, const Descriptor usage[], int width = 80, int last_column_min_percent = 50,
2778 int last_column_own_line_max_percent = 75)
2779{
2780 PrintUsageImplementation::OStreamWriter<OStream> write(prn);
2781 PrintUsageImplementation::printUsage(write, usage, width, last_column_min_percent, last_column_own_line_max_percent);
2782}
2783
2784template<typename Function>
2785void printUsage(Function* prn, const Descriptor usage[], int width = 80, int last_column_min_percent = 50,
2786 int last_column_own_line_max_percent = 75)
2787{
2788 PrintUsageImplementation::FunctionWriter<Function> write(prn);
2789 PrintUsageImplementation::printUsage(write, usage, width, last_column_min_percent, last_column_own_line_max_percent);
2790}
2791
2792template<typename Temporary>
2793void printUsage(const Temporary& prn, const Descriptor usage[], int width = 80, int last_column_min_percent = 50,
2794 int last_column_own_line_max_percent = 75)
2795{
2796 PrintUsageImplementation::TemporaryWriter<Temporary> write(prn);
2797 PrintUsageImplementation::printUsage(write, usage, width, last_column_min_percent, last_column_own_line_max_percent);
2798}
2799
2800template<typename Syscall>
2801void printUsage(Syscall* prn, int fd, const Descriptor usage[], int width = 80, int last_column_min_percent = 50,
2802 int last_column_own_line_max_percent = 75)
2803{
2804 PrintUsageImplementation::SyscallWriter<Syscall> write(prn, fd);
2805 PrintUsageImplementation::printUsage(write, usage, width, last_column_min_percent, last_column_own_line_max_percent);
2806}
2807
2808template<typename Function, typename Stream>
2809void printUsage(Function* prn, Stream* stream, const Descriptor usage[], int width = 80, int last_column_min_percent =
2810 50,
2811 int last_column_own_line_max_percent = 75)
2812{
2813 PrintUsageImplementation::StreamWriter<Function, Stream> write(prn, stream);
2814 PrintUsageImplementation::printUsage(write, usage, width, last_column_min_percent, last_column_own_line_max_percent);
2815}
2816
2817}
2818// namespace option
2819
2820#endif /* OPTIONPARSER_H_ */
2821