1 /*
2 *******************************************************************************
3 * Copyright (C) 1997-2014, International Business Machines Corporation and others.
4 * All Rights Reserved.
5 *******************************************************************************
6 */
7
8 #ifndef RBNF_H
9 #define RBNF_H
10
11 #include "unicode/utypes.h"
12
13 /**
14 * \file
15 * \brief C++ API: Rule Based Number Format
16 */
17
18 /**
19 * \def U_HAVE_RBNF
20 * This will be 0 if RBNF support is not included in ICU
21 * and 1 if it is.
22 *
23 * @stable ICU 2.4
24 */
25 #if UCONFIG_NO_FORMATTING
26 #define U_HAVE_RBNF 0
27 #else
28 #define U_HAVE_RBNF 1
29
30 #include "unicode/dcfmtsym.h"
31 #include "unicode/fmtable.h"
32 #include "unicode/locid.h"
33 #include "unicode/numfmt.h"
34 #include "unicode/unistr.h"
35 #include "unicode/strenum.h"
36 #include "unicode/brkiter.h"
37
38 U_NAMESPACE_BEGIN
39
40 class NFRuleSet;
41 class LocalizationInfo;
42 class RuleBasedCollator;
43
44 /**
45 * Tags for the predefined rulesets.
46 *
47 * @stable ICU 2.2
48 */
49 enum URBNFRuleSetTag {
50 URBNF_SPELLOUT,
51 URBNF_ORDINAL,
52 URBNF_DURATION,
53 URBNF_NUMBERING_SYSTEM,
54 URBNF_COUNT
55 };
56
57 /**
58 * The RuleBasedNumberFormat class formats numbers according to a set of rules. This number formatter is
59 * typically used for spelling out numeric values in words (e.g., 25,3476 as
60 * "twenty-five thousand three hundred seventy-six" or "vingt-cinq mille trois
61 * cents soixante-seize" or
62 * "fünfundzwanzigtausenddreihundertsechsundsiebzig"), but can also be used for
63 * other complicated formatting tasks, such as formatting a number of seconds as hours,
64 * minutes and seconds (e.g., 3,730 as "1:02:10").
65 *
66 * <p>The resources contain three predefined formatters for each locale: spellout, which
67 * spells out a value in words (123 is "one hundred twenty-three"); ordinal, which
68 * appends an ordinal suffix to the end of a numeral (123 is "123rd"); and
69 * duration, which shows a duration in seconds as hours, minutes, and seconds (123 is
70 * "2:03"). The client can also define more specialized <tt>RuleBasedNumberFormat</tt>s
71 * by supplying programmer-defined rule sets.</p>
72 *
73 * <p>The behavior of a <tt>RuleBasedNumberFormat</tt> is specified by a textual description
74 * that is either passed to the constructor as a <tt>String</tt> or loaded from a resource
75 * bundle. In its simplest form, the description consists of a semicolon-delimited list of <em>rules.</em>
76 * Each rule has a string of output text and a value or range of values it is applicable to.
77 * In a typical spellout rule set, the first twenty rules are the words for the numbers from
78 * 0 to 19:</p>
79 *
80 * <pre>zero; one; two; three; four; five; six; seven; eight; nine;
81 * ten; eleven; twelve; thirteen; fourteen; fifteen; sixteen; seventeen; eighteen; nineteen;</pre>
82 *
83 * <p>For larger numbers, we can use the preceding set of rules to format the ones place, and
84 * we only have to supply the words for the multiples of 10:</p>
85 *
86 * <pre> 20: twenty[->>];
87 * 30: thirty[->>];
88 * 40: forty[->>];
89 * 50: fifty[->>];
90 * 60: sixty[->>];
91 * 70: seventy[->>];
92 * 80: eighty[->>];
93 * 90: ninety[->>];</pre>
94 *
95 * <p>In these rules, the <em>base value</em> is spelled out explicitly and set off from the
96 * rule's output text with a colon. The rules are in a sorted list, and a rule is applicable
97 * to all numbers from its own base value to one less than the next rule's base value. The
98 * ">>" token is called a <em>substitution</em> and tells the fomatter to
99 * isolate the number's ones digit, format it using this same set of rules, and place the
100 * result at the position of the ">>" token. Text in brackets is omitted if
101 * the number being formatted is an even multiple of 10 (the hyphen is a literal hyphen; 24
102 * is "twenty-four," not "twenty four").</p>
103 *
104 * <p>For even larger numbers, we can actually look up several parts of the number in the
105 * list:</p>
106 *
107 * <pre>100: << hundred[ >>];</pre>
108 *
109 * <p>The "<<" represents a new kind of substitution. The << isolates
110 * the hundreds digit (and any digits to its left), formats it using this same rule set, and
111 * places the result where the "<<" was. Notice also that the meaning of
112 * >> has changed: it now refers to both the tens and the ones digits. The meaning of
113 * both substitutions depends on the rule's base value. The base value determines the rule's <em>divisor,</em>
114 * which is the highest power of 10 that is less than or equal to the base value (the user
115 * can change this). To fill in the substitutions, the formatter divides the number being
116 * formatted by the divisor. The integral quotient is used to fill in the <<
117 * substitution, and the remainder is used to fill in the >> substitution. The meaning
118 * of the brackets changes similarly: text in brackets is omitted if the value being
119 * formatted is an even multiple of the rule's divisor. The rules are applied recursively, so
120 * if a substitution is filled in with text that includes another substitution, that
121 * substitution is also filled in.</p>
122 *
123 * <p>This rule covers values up to 999, at which point we add another rule:</p>
124 *
125 * <pre>1000: << thousand[ >>];</pre>
126 *
127 * <p>Again, the meanings of the brackets and substitution tokens shift because the rule's
128 * base value is a higher power of 10, changing the rule's divisor. This rule can actually be
129 * used all the way up to 999,999. This allows us to finish out the rules as follows:</p>
130 *
131 * <pre> 1,000,000: << million[ >>];
132 * 1,000,000,000: << billion[ >>];
133 * 1,000,000,000,000: << trillion[ >>];
134 * 1,000,000,000,000,000: OUT OF RANGE!;</pre>
135 *
136 * <p>Commas, periods, and spaces can be used in the base values to improve legibility and
137 * are ignored by the rule parser. The last rule in the list is customarily treated as an
138 * "overflow rule," applying to everything from its base value on up, and often (as
139 * in this example) being used to print out an error message or default representation.
140 * Notice also that the size of the major groupings in large numbers is controlled by the
141 * spacing of the rules: because in English we group numbers by thousand, the higher rules
142 * are separated from each other by a factor of 1,000.</p>
143 *
144 * <p>To see how these rules actually work in practice, consider the following example:
145 * Formatting 25,430 with this rule set would work like this:</p>
146 *
147 * <table border="0" width="100%">
148 * <tr>
149 * <td><strong><< thousand >></strong></td>
150 * <td>[the rule whose base value is 1,000 is applicable to 25,340]</td>
151 * </tr>
152 * <tr>
153 * <td><strong>twenty->></strong> thousand >></td>
154 * <td>[25,340 over 1,000 is 25. The rule for 20 applies.]</td>
155 * </tr>
156 * <tr>
157 * <td>twenty-<strong>five</strong> thousand >></td>
158 * <td>[25 mod 10 is 5. The rule for 5 is "five."</td>
159 * </tr>
160 * <tr>
161 * <td>twenty-five thousand <strong><< hundred >></strong></td>
162 * <td>[25,340 mod 1,000 is 340. The rule for 100 applies.]</td>
163 * </tr>
164 * <tr>
165 * <td>twenty-five thousand <strong>three</strong> hundred >></td>
166 * <td>[340 over 100 is 3. The rule for 3 is "three."]</td>
167 * </tr>
168 * <tr>
169 * <td>twenty-five thousand three hundred <strong>forty</strong></td>
170 * <td>[340 mod 100 is 40. The rule for 40 applies. Since 40 divides
171 * evenly by 10, the hyphen and substitution in the brackets are omitted.]</td>
172 * </tr>
173 * </table>
174 *
175 * <p>The above syntax suffices only to format positive integers. To format negative numbers,
176 * we add a special rule:</p>
177 *
178 * <pre>-x: minus >>;</pre>
179 *
180 * <p>This is called a <em>negative-number rule,</em> and is identified by "-x"
181 * where the base value would be. This rule is used to format all negative numbers. the
182 * >> token here means "find the number's absolute value, format it with these
183 * rules, and put the result here."</p>
184 *
185 * <p>We also add a special rule called a <em>fraction rule </em>for numbers with fractional
186 * parts:</p>
187 *
188 * <pre>x.x: << point >>;</pre>
189 *
190 * <p>This rule is used for all positive non-integers (negative non-integers pass through the
191 * negative-number rule first and then through this rule). Here, the << token refers to
192 * the number's integral part, and the >> to the number's fractional part. The
193 * fractional part is formatted as a series of single-digit numbers (e.g., 123.456 would be
194 * formatted as "one hundred twenty-three point four five six").</p>
195 *
196 * <p>To see how this rule syntax is applied to various languages, examine the resource data.</p>
197 *
198 * <p>There is actually much more flexibility built into the rule language than the
199 * description above shows. A formatter may own multiple rule sets, which can be selected by
200 * the caller, and which can use each other to fill in their substitutions. Substitutions can
201 * also be filled in with digits, using a DecimalFormat object. There is syntax that can be
202 * used to alter a rule's divisor in various ways. And there is provision for much more
203 * flexible fraction handling. A complete description of the rule syntax follows:</p>
204 *
205 * <hr>
206 *
207 * <p>The description of a <tt>RuleBasedNumberFormat</tt>'s behavior consists of one or more <em>rule
208 * sets.</em> Each rule set consists of a name, a colon, and a list of <em>rules.</em> A rule
209 * set name must begin with a % sign. Rule sets with names that begin with a single % sign
210 * are <em>public:</em> the caller can specify that they be used to format and parse numbers.
211 * Rule sets with names that begin with %% are <em>private:</em> they exist only for the use
212 * of other rule sets. If a formatter only has one rule set, the name may be omitted.</p>
213 *
214 * <p>The user can also specify a special "rule set" named <tt>%%lenient-parse</tt>.
215 * The body of <tt>%%lenient-parse</tt> isn't a set of number-formatting rules, but a <tt>RuleBasedCollator</tt>
216 * description which is used to define equivalences for lenient parsing. For more information
217 * on the syntax, see <tt>RuleBasedCollator</tt>. For more information on lenient parsing,
218 * see <tt>setLenientParse()</tt>. <em>Note:</em> symbols that have syntactic meaning
219 * in collation rules, such as '&', have no particular meaning when appearing outside
220 * of the <tt>lenient-parse</tt> rule set.</p>
221 *
222 * <p>The body of a rule set consists of an ordered, semicolon-delimited list of <em>rules.</em>
223 * Internally, every rule has a base value, a divisor, rule text, and zero, one, or two <em>substitutions.</em>
224 * These parameters are controlled by the description syntax, which consists of a <em>rule
225 * descriptor,</em> a colon, and a <em>rule body.</em></p>
226 *
227 * <p>A rule descriptor can take one of the following forms (text in <em>italics</em> is the
228 * name of a token):</p>
229 *
230 * <table border="0" width="100%">
231 * <tr>
232 * <td><em>bv</em>:</td>
233 * <td><em>bv</em> specifies the rule's base value. <em>bv</em> is a decimal
234 * number expressed using ASCII digits. <em>bv</em> may contain spaces, period, and commas,
235 * which are ignored. The rule's divisor is the highest power of 10 less than or equal to
236 * the base value.</td>
237 * </tr>
238 * <tr>
239 * <td><em>bv</em>/<em>rad</em>:</td>
240 * <td><em>bv</em> specifies the rule's base value. The rule's divisor is the
241 * highest power of <em>rad</em> less than or equal to the base value.</td>
242 * </tr>
243 * <tr>
244 * <td><em>bv</em>>:</td>
245 * <td><em>bv</em> specifies the rule's base value. To calculate the divisor,
246 * let the radix be 10, and the exponent be the highest exponent of the radix that yields a
247 * result less than or equal to the base value. Every > character after the base value
248 * decreases the exponent by 1. If the exponent is positive or 0, the divisor is the radix
249 * raised to the power of the exponent; otherwise, the divisor is 1.</td>
250 * </tr>
251 * <tr>
252 * <td><em>bv</em>/<em>rad</em>>:</td>
253 * <td><em>bv</em> specifies the rule's base value. To calculate the divisor,
254 * let the radix be <em>rad</em>, and the exponent be the highest exponent of the radix that
255 * yields a result less than or equal to the base value. Every > character after the radix
256 * decreases the exponent by 1. If the exponent is positive or 0, the divisor is the radix
257 * raised to the power of the exponent; otherwise, the divisor is 1.</td>
258 * </tr>
259 * <tr>
260 * <td>-x:</td>
261 * <td>The rule is a negative-number rule.</td>
262 * </tr>
263 * <tr>
264 * <td>x.x:</td>
265 * <td>The rule is an <em>improper fraction rule.</em></td>
266 * </tr>
267 * <tr>
268 * <td>0.x:</td>
269 * <td>The rule is a <em>proper fraction rule.</em></td>
270 * </tr>
271 * <tr>
272 * <td>x.0:</td>
273 * <td>The rule is a <em>master rule.</em></td>
274 * </tr>
275 * <tr>
276 * <td><em>nothing</em></td>
277 * <td>If the rule's rule descriptor is left out, the base value is one plus the
278 * preceding rule's base value (or zero if this is the first rule in the list) in a normal
279 * rule set. In a fraction rule set, the base value is the same as the preceding rule's
280 * base value.</td>
281 * </tr>
282 * </table>
283 *
284 * <p>A rule set may be either a regular rule set or a <em>fraction rule set,</em> depending
285 * on whether it is used to format a number's integral part (or the whole number) or a
286 * number's fractional part. Using a rule set to format a rule's fractional part makes it a
287 * fraction rule set.</p>
288 *
289 * <p>Which rule is used to format a number is defined according to one of the following
290 * algorithms: If the rule set is a regular rule set, do the following:
291 *
292 * <ul>
293 * <li>If the rule set includes a master rule (and the number was passed in as a <tt>double</tt>),
294 * use the master rule. (If the number being formatted was passed in as a <tt>long</tt>,
295 * the master rule is ignored.)</li>
296 * <li>If the number is negative, use the negative-number rule.</li>
297 * <li>If the number has a fractional part and is greater than 1, use the improper fraction
298 * rule.</li>
299 * <li>If the number has a fractional part and is between 0 and 1, use the proper fraction
300 * rule.</li>
301 * <li>Binary-search the rule list for the rule with the highest base value less than or equal
302 * to the number. If that rule has two substitutions, its base value is not an even multiple
303 * of its divisor, and the number <em>is</em> an even multiple of the rule's divisor, use the
304 * rule that precedes it in the rule list. Otherwise, use the rule itself.</li>
305 * </ul>
306 *
307 * <p>If the rule set is a fraction rule set, do the following:
308 *
309 * <ul>
310 * <li>Ignore negative-number and fraction rules.</li>
311 * <li>For each rule in the list, multiply the number being formatted (which will always be
312 * between 0 and 1) by the rule's base value. Keep track of the distance between the result
313 * the nearest integer.</li>
314 * <li>Use the rule that produced the result closest to zero in the above calculation. In the
315 * event of a tie or a direct hit, use the first matching rule encountered. (The idea here is
316 * to try each rule's base value as a possible denominator of a fraction. Whichever
317 * denominator produces the fraction closest in value to the number being formatted wins.) If
318 * the rule following the matching rule has the same base value, use it if the numerator of
319 * the fraction is anything other than 1; if the numerator is 1, use the original matching
320 * rule. (This is to allow singular and plural forms of the rule text without a lot of extra
321 * hassle.)</li>
322 * </ul>
323 *
324 * <p>A rule's body consists of a string of characters terminated by a semicolon. The rule
325 * may include zero, one, or two <em>substitution tokens,</em> and a range of text in
326 * brackets. The brackets denote optional text (and may also include one or both
327 * substitutions). The exact meanings of the substitution tokens, and under what conditions
328 * optional text is omitted, depend on the syntax of the substitution token and the context.
329 * The rest of the text in a rule body is literal text that is output when the rule matches
330 * the number being formatted.</p>
331 *
332 * <p>A substitution token begins and ends with a <em>token character.</em> The token
333 * character and the context together specify a mathematical operation to be performed on the
334 * number being formatted. An optional <em>substitution descriptor </em>specifies how the
335 * value resulting from that operation is used to fill in the substitution. The position of
336 * the substitution token in the rule body specifies the location of the resultant text in
337 * the original rule text.</p>
338 *
339 * <p>The meanings of the substitution token characters are as follows:</p>
340 *
341 * <table border="0" width="100%">
342 * <tr>
343 * <td>>></td>
344 * <td>in normal rule</td>
345 * <td>Divide the number by the rule's divisor and format the remainder</td>
346 * </tr>
347 * <tr>
348 * <td></td>
349 * <td>in negative-number rule</td>
350 * <td>Find the absolute value of the number and format the result</td>
351 * </tr>
352 * <tr>
353 * <td></td>
354 * <td>in fraction or master rule</td>
355 * <td>Isolate the number's fractional part and format it.</td>
356 * </tr>
357 * <tr>
358 * <td></td>
359 * <td>in rule in fraction rule set</td>
360 * <td>Not allowed.</td>
361 * </tr>
362 * <tr>
363 * <td>>>></td>
364 * <td>in normal rule</td>
365 * <td>Divide the number by the rule's divisor and format the remainder,
366 * but bypass the normal rule-selection process and just use the
367 * rule that precedes this one in this rule list.</td>
368 * </tr>
369 * <tr>
370 * <td></td>
371 * <td>in all other rules</td>
372 * <td>Not allowed.</td>
373 * </tr>
374 * <tr>
375 * <td><<</td>
376 * <td>in normal rule</td>
377 * <td>Divide the number by the rule's divisor and format the quotient</td>
378 * </tr>
379 * <tr>
380 * <td></td>
381 * <td>in negative-number rule</td>
382 * <td>Not allowed.</td>
383 * </tr>
384 * <tr>
385 * <td></td>
386 * <td>in fraction or master rule</td>
387 * <td>Isolate the number's integral part and format it.</td>
388 * </tr>
389 * <tr>
390 * <td></td>
391 * <td>in rule in fraction rule set</td>
392 * <td>Multiply the number by the rule's base value and format the result.</td>
393 * </tr>
394 * <tr>
395 * <td>==</td>
396 * <td>in all rule sets</td>
397 * <td>Format the number unchanged</td>
398 * </tr>
399 * <tr>
400 * <td>[]</td>
401 * <td>in normal rule</td>
402 * <td>Omit the optional text if the number is an even multiple of the rule's divisor</td>
403 * </tr>
404 * <tr>
405 * <td></td>
406 * <td>in negative-number rule</td>
407 * <td>Not allowed.</td>
408 * </tr>
409 * <tr>
410 * <td></td>
411 * <td>in improper-fraction rule</td>
412 * <td>Omit the optional text if the number is between 0 and 1 (same as specifying both an
413 * x.x rule and a 0.x rule)</td>
414 * </tr>
415 * <tr>
416 * <td></td>
417 * <td>in master rule</td>
418 * <td>Omit the optional text if the number is an integer (same as specifying both an x.x
419 * rule and an x.0 rule)</td>
420 * </tr>
421 * <tr>
422 * <td></td>
423 * <td>in proper-fraction rule</td>
424 * <td>Not allowed.</td>
425 * </tr>
426 * <tr>
427 * <td></td>
428 * <td>in rule in fraction rule set</td>
429 * <td>Omit the optional text if multiplying the number by the rule's base value yields 1.</td>
430 * </tr>
431 * </table>
432 *
433 * <p>The substitution descriptor (i.e., the text between the token characters) may take one
434 * of three forms:</p>
435 *
436 * <table border="0" width="100%">
437 * <tr>
438 * <td>a rule set name</td>
439 * <td>Perform the mathematical operation on the number, and format the result using the
440 * named rule set.</td>
441 * </tr>
442 * <tr>
443 * <td>a DecimalFormat pattern</td>
444 * <td>Perform the mathematical operation on the number, and format the result using a
445 * DecimalFormat with the specified pattern. The pattern must begin with 0 or #.</td>
446 * </tr>
447 * <tr>
448 * <td>nothing</td>
449 * <td>Perform the mathematical operation on the number, and format the result using the rule
450 * set containing the current rule, except:
451 * <ul>
452 * <li>You can't have an empty substitution descriptor with a == substitution.</li>
453 * <li>If you omit the substitution descriptor in a >> substitution in a fraction rule,
454 * format the result one digit at a time using the rule set containing the current rule.</li>
455 * <li>If you omit the substitution descriptor in a << substitution in a rule in a
456 * fraction rule set, format the result using the default rule set for this formatter.</li>
457 * </ul>
458 * </td>
459 * </tr>
460 * </table>
461 *
462 * <p>Whitespace is ignored between a rule set name and a rule set body, between a rule
463 * descriptor and a rule body, or between rules. If a rule body begins with an apostrophe,
464 * the apostrophe is ignored, but all text after it becomes significant (this is how you can
465 * have a rule's rule text begin with whitespace). There is no escape function: the semicolon
466 * is not allowed in rule set names or in rule text, and the colon is not allowed in rule set
467 * names. The characters beginning a substitution token are always treated as the beginning
468 * of a substitution token.</p>
469 *
470 * <p>See the resource data and the demo program for annotated examples of real rule sets
471 * using these features.</p>
472 *
473 * <p><em>User subclasses are not supported.</em> While clients may write
474 * subclasses, such code will not necessarily work and will not be
475 * guaranteed to work stably from release to release.
476 *
477 * <p><b>Localizations</b></p>
478 * <p>Constructors are available that allow the specification of localizations for the
479 * public rule sets (and also allow more control over what public rule sets are available).
480 * Localization data is represented as a textual description. The description represents
481 * an array of arrays of string. The first element is an array of the public rule set names,
482 * each of these must be one of the public rule set names that appear in the rules. Only
483 * names in this array will be treated as public rule set names by the API. Each subsequent
484 * element is an array of localizations of these names. The first element of one of these
485 * subarrays is the locale name, and the remaining elements are localizations of the
486 * public rule set names, in the same order as they were listed in the first arrray.</p>
487 * <p>In the syntax, angle brackets '<', '>' are used to delimit the arrays, and comma ',' is used
488 * to separate elements of an array. Whitespace is ignored, unless quoted.</p>
489 * <p>For example:<pre>
490 * < < %foo, %bar, %baz >,
491 * < en, Foo, Bar, Baz >,
492 * < fr, 'le Foo', 'le Bar', 'le Baz' >
493 * < zh, \\u7532, \\u4e59, \\u4e19 > >
494 * </pre></p>
495 * @author Richard Gillam
496 * @see NumberFormat
497 * @see DecimalFormat
498 * @stable ICU 2.0
499 */
500 class U_I18N_API RuleBasedNumberFormat : public NumberFormat {
501 public:
502
503 //-----------------------------------------------------------------------
504 // constructors
505 //-----------------------------------------------------------------------
506
507 /**
508 * Creates a RuleBasedNumberFormat that behaves according to the description
509 * passed in. The formatter uses the default locale.
510 * @param rules A description of the formatter's desired behavior.
511 * See the class documentation for a complete explanation of the description
512 * syntax.
513 * @param perror The parse error if an error was encountered.
514 * @param status The status indicating whether the constructor succeeded.
515 * @stable ICU 3.2
516 */
517 RuleBasedNumberFormat(const UnicodeString& rules, UParseError& perror, UErrorCode& status);
518
519 /**
520 * Creates a RuleBasedNumberFormat that behaves according to the description
521 * passed in. The formatter uses the default locale.
522 * <p>
523 * The localizations data provides information about the public
524 * rule sets and their localized display names for different
525 * locales. The first element in the list is an array of the names
526 * of the public rule sets. The first element in this array is
527 * the initial default ruleset. The remaining elements in the
528 * list are arrays of localizations of the names of the public
529 * rule sets. Each of these is one longer than the initial array,
530 * with the first String being the ULocale ID, and the remaining
531 * Strings being the localizations of the rule set names, in the
532 * same order as the initial array. Arrays are NULL-terminated.
533 * @param rules A description of the formatter's desired behavior.
534 * See the class documentation for a complete explanation of the description
535 * syntax.
536 * @param localizations the localization information.
537 * names in the description. These will be copied by the constructor.
538 * @param perror The parse error if an error was encountered.
539 * @param status The status indicating whether the constructor succeeded.
540 * @stable ICU 3.2
541 */
542 RuleBasedNumberFormat(const UnicodeString& rules, const UnicodeString& localizations,
543 UParseError& perror, UErrorCode& status);
544
545 /**
546 * Creates a RuleBasedNumberFormat that behaves according to the rules
547 * passed in. The formatter uses the specified locale to determine the
548 * characters to use when formatting numerals, and to define equivalences
549 * for lenient parsing.
550 * @param rules The formatter rules.
551 * See the class documentation for a complete explanation of the rule
552 * syntax.
553 * @param locale A locale that governs which characters are used for
554 * formatting values in numerals and which characters are equivalent in
555 * lenient parsing.
556 * @param perror The parse error if an error was encountered.
557 * @param status The status indicating whether the constructor succeeded.
558 * @stable ICU 2.0
559 */
560 RuleBasedNumberFormat(const UnicodeString& rules, const Locale& locale,
561 UParseError& perror, UErrorCode& status);
562
563 /**
564 * Creates a RuleBasedNumberFormat that behaves according to the description
565 * passed in. The formatter uses the default locale.
566 * <p>
567 * The localizations data provides information about the public
568 * rule sets and their localized display names for different
569 * locales. The first element in the list is an array of the names
570 * of the public rule sets. The first element in this array is
571 * the initial default ruleset. The remaining elements in the
572 * list are arrays of localizations of the names of the public
573 * rule sets. Each of these is one longer than the initial array,
574 * with the first String being the ULocale ID, and the remaining
575 * Strings being the localizations of the rule set names, in the
576 * same order as the initial array. Arrays are NULL-terminated.
577 * @param rules A description of the formatter's desired behavior.
578 * See the class documentation for a complete explanation of the description
579 * syntax.
580 * @param localizations a list of localizations for the rule set
581 * names in the description. These will be copied by the constructor.
582 * @param locale A locale that governs which characters are used for
583 * formatting values in numerals and which characters are equivalent in
584 * lenient parsing.
585 * @param perror The parse error if an error was encountered.
586 * @param status The status indicating whether the constructor succeeded.
587 * @stable ICU 3.2
588 */
589 RuleBasedNumberFormat(const UnicodeString& rules, const UnicodeString& localizations,
590 const Locale& locale, UParseError& perror, UErrorCode& status);
591
592 /**
593 * Creates a RuleBasedNumberFormat from a predefined ruleset. The selector
594 * code choosed among three possible predefined formats: spellout, ordinal,
595 * and duration.
596 * @param tag A selector code specifying which kind of formatter to create for that
597 * locale. There are four legal values: URBNF_SPELLOUT, which creates a formatter that
598 * spells out a value in words in the desired language, URBNF_ORDINAL, which attaches
599 * an ordinal suffix from the desired language to the end of a number (e.g. "123rd"),
600 * URBNF_DURATION, which formats a duration in seconds as hours, minutes, and seconds,
601 * and URBNF_NUMBERING_SYSTEM, which is used to invoke rules for alternate numbering
602 * systems such as the Hebrew numbering system, or for Roman Numerals, etc.
603 * @param locale The locale for the formatter.
604 * @param status The status indicating whether the constructor succeeded.
605 * @stable ICU 2.0
606 */
607 RuleBasedNumberFormat(URBNFRuleSetTag tag, const Locale& locale, UErrorCode& status);
608
609 //-----------------------------------------------------------------------
610 // boilerplate
611 //-----------------------------------------------------------------------
612
613 /**
614 * Copy constructor
615 * @param rhs the object to be copied from.
616 * @stable ICU 2.6
617 */
618 RuleBasedNumberFormat(const RuleBasedNumberFormat& rhs);
619
620 /**
621 * Assignment operator
622 * @param rhs the object to be copied from.
623 * @stable ICU 2.6
624 */
625 RuleBasedNumberFormat& operator=(const RuleBasedNumberFormat& rhs);
626
627 /**
628 * Release memory allocated for a RuleBasedNumberFormat when you are finished with it.
629 * @stable ICU 2.6
630 */
631 virtual ~RuleBasedNumberFormat();
632
633 /**
634 * Clone this object polymorphically. The caller is responsible
635 * for deleting the result when done.
636 * @return A copy of the object.
637 * @stable ICU 2.6
638 */
639 virtual Format* clone(void) const;
640
641 /**
642 * Return true if the given Format objects are semantically equal.
643 * Objects of different subclasses are considered unequal.
644 * @param other the object to be compared with.
645 * @return true if the given Format objects are semantically equal.
646 * @stable ICU 2.6
647 */
648 virtual UBool operator==(const Format& other) const;
649
650 //-----------------------------------------------------------------------
651 // public API functions
652 //-----------------------------------------------------------------------
653
654 /**
655 * return the rules that were provided to the RuleBasedNumberFormat.
656 * @return the result String that was passed in
657 * @stable ICU 2.0
658 */
659 virtual UnicodeString getRules() const;
660
661 /**
662 * Return the number of public rule set names.
663 * @return the number of public rule set names.
664 * @stable ICU 2.0
665 */
666 virtual int32_t getNumberOfRuleSetNames() const;
667
668 /**
669 * Return the name of the index'th public ruleSet. If index is not valid,
670 * the function returns null.
671 * @param index the index of the ruleset
672 * @return the name of the index'th public ruleSet.
673 * @stable ICU 2.0
674 */
675 virtual UnicodeString getRuleSetName(int32_t index) const;
676
677 /**
678 * Return the number of locales for which we have localized rule set display names.
679 * @return the number of locales for which we have localized rule set display names.
680 * @stable ICU 3.2
681 */
682 virtual int32_t getNumberOfRuleSetDisplayNameLocales(void) const;
683
684 /**
685 * Return the index'th display name locale.
686 * @param index the index of the locale
687 * @param status set to a failure code when this function fails
688 * @return the locale
689 * @see #getNumberOfRuleSetDisplayNameLocales
690 * @stable ICU 3.2
691 */
692 virtual Locale getRuleSetDisplayNameLocale(int32_t index, UErrorCode& status) const;
693
694 /**
695 * Return the rule set display names for the provided locale. These are in the same order
696 * as those returned by getRuleSetName. The locale is matched against the locales for
697 * which there is display name data, using normal fallback rules. If no locale matches,
698 * the default display names are returned. (These are the internal rule set names minus
699 * the leading '%'.)
700 * @param index the index of the rule set
701 * @param locale the locale (returned by getRuleSetDisplayNameLocales) for which the localized
702 * display name is desired
703 * @return the display name for the given index, which might be bogus if there is an error
704 * @see #getRuleSetName
705 * @stable ICU 3.2
706 */
707 virtual UnicodeString getRuleSetDisplayName(int32_t index,
708 const Locale& locale = Locale::getDefault());
709
710 /**
711 * Return the rule set display name for the provided rule set and locale.
712 * The locale is matched against the locales for which there is display name data, using
713 * normal fallback rules. If no locale matches, the default display name is returned.
714 * @return the display name for the rule set
715 * @stable ICU 3.2
716 * @see #getRuleSetDisplayName
717 */
718 virtual UnicodeString getRuleSetDisplayName(const UnicodeString& ruleSetName,
719 const Locale& locale = Locale::getDefault());
720
721
722 using NumberFormat::format;
723
724 /**
725 * Formats the specified 32-bit number using the default ruleset.
726 * @param number The number to format.
727 * @param toAppendTo the string that will hold the (appended) result
728 * @param pos the fieldposition
729 * @return A textual representation of the number.
730 * @stable ICU 2.0
731 */
732 virtual UnicodeString& format(int32_t number,
733 UnicodeString& toAppendTo,
734 FieldPosition& pos) const;
735
736 /**
737 * Formats the specified 64-bit number using the default ruleset.
738 * @param number The number to format.
739 * @param toAppendTo the string that will hold the (appended) result
740 * @param pos the fieldposition
741 * @return A textual representation of the number.
742 * @stable ICU 2.1
743 */
744 virtual UnicodeString& format(int64_t number,
745 UnicodeString& toAppendTo,
746 FieldPosition& pos) const;
747 /**
748 * Formats the specified number using the default ruleset.
749 * @param number The number to format.
750 * @param toAppendTo the string that will hold the (appended) result
751 * @param pos the fieldposition
752 * @return A textual representation of the number.
753 * @stable ICU 2.0
754 */
755 virtual UnicodeString& format(double number,
756 UnicodeString& toAppendTo,
757 FieldPosition& pos) const;
758
759 /**
760 * Formats the specified number using the named ruleset.
761 * @param number The number to format.
762 * @param ruleSetName The name of the rule set to format the number with.
763 * This must be the name of a valid public rule set for this formatter.
764 * @param toAppendTo the string that will hold the (appended) result
765 * @param pos the fieldposition
766 * @param status the status
767 * @return A textual representation of the number.
768 * @stable ICU 2.0
769 */
770 virtual UnicodeString& format(int32_t number,
771 const UnicodeString& ruleSetName,
772 UnicodeString& toAppendTo,
773 FieldPosition& pos,
774 UErrorCode& status) const;
775 /**
776 * Formats the specified 64-bit number using the named ruleset.
777 * @param number The number to format.
778 * @param ruleSetName The name of the rule set to format the number with.
779 * This must be the name of a valid public rule set for this formatter.
780 * @param toAppendTo the string that will hold the (appended) result
781 * @param pos the fieldposition
782 * @param status the status
783 * @return A textual representation of the number.
784 * @stable ICU 2.1
785 */
786 virtual UnicodeString& format(int64_t number,
787 const UnicodeString& ruleSetName,
788 UnicodeString& toAppendTo,
789 FieldPosition& pos,
790 UErrorCode& status) const;
791 /**
792 * Formats the specified number using the named ruleset.
793 * @param number The number to format.
794 * @param ruleSetName The name of the rule set to format the number with.
795 * This must be the name of a valid public rule set for this formatter.
796 * @param toAppendTo the string that will hold the (appended) result
797 * @param pos the fieldposition
798 * @param status the status
799 * @return A textual representation of the number.
800 * @stable ICU 2.0
801 */
802 virtual UnicodeString& format(double number,
803 const UnicodeString& ruleSetName,
804 UnicodeString& toAppendTo,
805 FieldPosition& pos,
806 UErrorCode& status) const;
807
808 using NumberFormat::parse;
809
810 /**
811 * Parses the specfied string, beginning at the specified position, according
812 * to this formatter's rules. This will match the string against all of the
813 * formatter's public rule sets and return the value corresponding to the longest
814 * parseable substring. This function's behavior is affected by the lenient
815 * parse mode.
816 * @param text The string to parse
817 * @param result the result of the parse, either a double or a long.
818 * @param parsePosition On entry, contains the position of the first character
819 * in "text" to examine. On exit, has been updated to contain the position
820 * of the first character in "text" that wasn't consumed by the parse.
821 * @see #setLenient
822 * @stable ICU 2.0
823 */
824 virtual void parse(const UnicodeString& text,
825 Formattable& result,
826 ParsePosition& parsePosition) const;
827
828 #if !UCONFIG_NO_COLLATION
829
830 /**
831 * Turns lenient parse mode on and off.
832 *
833 * When in lenient parse mode, the formatter uses a Collator for parsing the text.
834 * Only primary differences are treated as significant. This means that case
835 * differences, accent differences, alternate spellings of the same letter
836 * (e.g., ae and a-umlaut in German), ignorable characters, etc. are ignored in
837 * matching the text. In many cases, numerals will be accepted in place of words
838 * or phrases as well.
839 *
840 * For example, all of the following will correctly parse as 255 in English in
841 * lenient-parse mode:
842 * <br>"two hundred fifty-five"
843 * <br>"two hundred fifty five"
844 * <br>"TWO HUNDRED FIFTY-FIVE"
845 * <br>"twohundredfiftyfive"
846 * <br>"2 hundred fifty-5"
847 *
848 * The Collator used is determined by the locale that was
849 * passed to this object on construction. The description passed to this object
850 * on construction may supply additional collation rules that are appended to the
851 * end of the default collator for the locale, enabling additional equivalences
852 * (such as adding more ignorable characters or permitting spelled-out version of
853 * symbols; see the demo program for examples).
854 *
855 * It's important to emphasize that even strict parsing is relatively lenient: it
856 * will accept some text that it won't produce as output. In English, for example,
857 * it will correctly parse "two hundred zero" and "fifteen hundred".
858 *
859 * @param enabled If true, turns lenient-parse mode on; if false, turns it off.
860 * @see RuleBasedCollator
861 * @stable ICU 2.0
862 */
863 virtual void setLenient(UBool enabled);
864
865 /**
866 * Returns true if lenient-parse mode is turned on. Lenient parsing is off
867 * by default.
868 * @return true if lenient-parse mode is turned on.
869 * @see #setLenient
870 * @stable ICU 2.0
871 */
872 virtual inline UBool isLenient(void) const;
873
874 #endif
875
876 /**
877 * Override the default rule set to use. If ruleSetName is null, reset
878 * to the initial default rule set. If the rule set is not a public rule set name,
879 * U_ILLEGAL_ARGUMENT_ERROR is returned in status.
880 * @param ruleSetName the name of the rule set, or null to reset the initial default.
881 * @param status set to failure code when a problem occurs.
882 * @stable ICU 2.6
883 */
884 virtual void setDefaultRuleSet(const UnicodeString& ruleSetName, UErrorCode& status);
885
886 /**
887 * Return the name of the current default rule set. If the current rule set is
888 * not public, returns a bogus (and empty) UnicodeString.
889 * @return the name of the current default rule set
890 * @stable ICU 3.0
891 */
892 virtual UnicodeString getDefaultRuleSetName() const;
893
894 /* Cannot use #ifndef U_HIDE_DRAFT_API for the following draft method since it is virtual */
895 /**
896 * Set a particular UDisplayContext value in the formatter, such as
897 * UDISPCTX_CAPITALIZATION_FOR_STANDALONE. Note: For getContext, see
898 * NumberFormat.
899 * @param value The UDisplayContext value to set.
900 * @param status Input/output status. If at entry this indicates a failure
901 * status, the function will do nothing; otherwise this will be
902 * updated with any new status from the function.
903 * @draft ICU 53
904 */
905 virtual void setContext(UDisplayContext value, UErrorCode& status);
906
907 public:
908 /**
909 * ICU "poor man's RTTI", returns a UClassID for this class.
910 *
911 * @stable ICU 2.8
912 */
913 static UClassID U_EXPORT2 getStaticClassID(void);
914
915 /**
916 * ICU "poor man's RTTI", returns a UClassID for the actual class.
917 *
918 * @stable ICU 2.8
919 */
920 virtual UClassID getDynamicClassID(void) const;
921
922 /**
923 * Sets the decimal format symbols, which is generally not changed
924 * by the programmer or user. The formatter takes ownership of
925 * symbolsToAdopt; the client must not delete it.
926 *
927 * @param symbolsToAdopt DecimalFormatSymbols to be adopted.
928 * @stable ICU 49
929 */
930 virtual void adoptDecimalFormatSymbols(DecimalFormatSymbols* symbolsToAdopt);
931
932 /**
933 * Sets the decimal format symbols, which is generally not changed
934 * by the programmer or user. A clone of the symbols is created and
935 * the symbols is _not_ adopted; the client is still responsible for
936 * deleting it.
937 *
938 * @param symbols DecimalFormatSymbols.
939 * @stable ICU 49
940 */
941 virtual void setDecimalFormatSymbols(const DecimalFormatSymbols& symbols);
942
943 private:
944 RuleBasedNumberFormat(); // default constructor not implemented
945
946 // this will ref the localizations if they are not NULL
947 // caller must deref to get adoption
948 RuleBasedNumberFormat(const UnicodeString& description, LocalizationInfo* localizations,
949 const Locale& locale, UParseError& perror, UErrorCode& status);
950
951 void init(const UnicodeString& rules, LocalizationInfo* localizations, UParseError& perror, UErrorCode& status);
952 void initCapitalizationContextInfo(const Locale& thelocale);
953 void dispose();
954 void stripWhitespace(UnicodeString& src);
955 void initDefaultRuleSet();
956 void format(double number, NFRuleSet& ruleSet);
957 NFRuleSet* findRuleSet(const UnicodeString& name, UErrorCode& status) const;
958
959 /* friend access */
960 friend class NFSubstitution;
961 friend class NFRule;
962 friend class FractionalPartSubstitution;
963
964 inline NFRuleSet * getDefaultRuleSet() const;
965 const RuleBasedCollator * getCollator() const;
966 DecimalFormatSymbols * getDecimalFormatSymbols() const;
967 UnicodeString& adjustForCapitalizationContext(int32_t startPos, UnicodeString& currentResult) const;
968
969 private:
970 NFRuleSet **ruleSets;
971 UnicodeString* ruleSetDescriptions;
972 int32_t numRuleSets;
973 NFRuleSet *defaultRuleSet;
974 Locale locale;
975 RuleBasedCollator* collator;
976 DecimalFormatSymbols* decimalFormatSymbols;
977 UBool lenient;
978 UnicodeString* lenientParseRules;
979 LocalizationInfo* localizations;
980 UnicodeString originalDescription;
981 UBool capitalizationInfoSet;
982 UBool capitalizationForUIListMenu;
983 UBool capitalizationForStandAlone;
984 BreakIterator* capitalizationBrkIter;
985 };
986
987 // ---------------
988
989 #if !UCONFIG_NO_COLLATION
990
991 inline UBool
isLenient(void)992 RuleBasedNumberFormat::isLenient(void) const {
993 return lenient;
994 }
995
996 #endif
997
998 inline NFRuleSet*
getDefaultRuleSet()999 RuleBasedNumberFormat::getDefaultRuleSet() const {
1000 return defaultRuleSet;
1001 }
1002
1003 U_NAMESPACE_END
1004
1005 /* U_HAVE_RBNF */
1006 #endif
1007
1008 /* RBNF_H */
1009 #endif
1010