1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/intl/icu/source/i18n/unicode/rbnf.h Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,992 @@
1.4 +/*
1.5 +*******************************************************************************
1.6 +* Copyright (C) 1997-2013, International Business Machines Corporation and others.
1.7 +* All Rights Reserved.
1.8 +*******************************************************************************
1.9 +*/
1.10 +
1.11 +#ifndef RBNF_H
1.12 +#define RBNF_H
1.13 +
1.14 +#include "unicode/utypes.h"
1.15 +
1.16 +/**
1.17 + * \file
1.18 + * \brief C++ API: Rule Based Number Format
1.19 + */
1.20 +
1.21 +/**
1.22 + * \def U_HAVE_RBNF
1.23 + * This will be 0 if RBNF support is not included in ICU
1.24 + * and 1 if it is.
1.25 + *
1.26 + * @stable ICU 2.4
1.27 + */
1.28 +#if UCONFIG_NO_FORMATTING
1.29 +#define U_HAVE_RBNF 0
1.30 +#else
1.31 +#define U_HAVE_RBNF 1
1.32 +
1.33 +#include "unicode/coll.h"
1.34 +#include "unicode/dcfmtsym.h"
1.35 +#include "unicode/fmtable.h"
1.36 +#include "unicode/locid.h"
1.37 +#include "unicode/numfmt.h"
1.38 +#include "unicode/unistr.h"
1.39 +#include "unicode/strenum.h"
1.40 +
1.41 +U_NAMESPACE_BEGIN
1.42 +
1.43 +class NFRuleSet;
1.44 +class LocalizationInfo;
1.45 +
1.46 +/**
1.47 + * Tags for the predefined rulesets.
1.48 + *
1.49 + * @stable ICU 2.2
1.50 + */
1.51 +enum URBNFRuleSetTag {
1.52 + URBNF_SPELLOUT,
1.53 + URBNF_ORDINAL,
1.54 + URBNF_DURATION,
1.55 + URBNF_NUMBERING_SYSTEM,
1.56 + URBNF_COUNT
1.57 +};
1.58 +
1.59 +#if UCONFIG_NO_COLLATION
1.60 +class Collator;
1.61 +#endif
1.62 +
1.63 +/**
1.64 + * The RuleBasedNumberFormat class formats numbers according to a set of rules. This number formatter is
1.65 + * typically used for spelling out numeric values in words (e.g., 25,3476 as
1.66 + * "twenty-five thousand three hundred seventy-six" or "vingt-cinq mille trois
1.67 + * cents soixante-seize" or
1.68 + * "fünfundzwanzigtausenddreihundertsechsundsiebzig"), but can also be used for
1.69 + * other complicated formatting tasks, such as formatting a number of seconds as hours,
1.70 + * minutes and seconds (e.g., 3,730 as "1:02:10").
1.71 + *
1.72 + * <p>The resources contain three predefined formatters for each locale: spellout, which
1.73 + * spells out a value in words (123 is "one hundred twenty-three"); ordinal, which
1.74 + * appends an ordinal suffix to the end of a numeral (123 is "123rd"); and
1.75 + * duration, which shows a duration in seconds as hours, minutes, and seconds (123 is
1.76 + * "2:03"). The client can also define more specialized <tt>RuleBasedNumberFormat</tt>s
1.77 + * by supplying programmer-defined rule sets.</p>
1.78 + *
1.79 + * <p>The behavior of a <tt>RuleBasedNumberFormat</tt> is specified by a textual description
1.80 + * that is either passed to the constructor as a <tt>String</tt> or loaded from a resource
1.81 + * bundle. In its simplest form, the description consists of a semicolon-delimited list of <em>rules.</em>
1.82 + * Each rule has a string of output text and a value or range of values it is applicable to.
1.83 + * In a typical spellout rule set, the first twenty rules are the words for the numbers from
1.84 + * 0 to 19:</p>
1.85 + *
1.86 + * <pre>zero; one; two; three; four; five; six; seven; eight; nine;
1.87 + * ten; eleven; twelve; thirteen; fourteen; fifteen; sixteen; seventeen; eighteen; nineteen;</pre>
1.88 + *
1.89 + * <p>For larger numbers, we can use the preceding set of rules to format the ones place, and
1.90 + * we only have to supply the words for the multiples of 10:</p>
1.91 + *
1.92 + * <pre> 20: twenty[->>];
1.93 + * 30: thirty[->>];
1.94 + * 40: forty[->>];
1.95 + * 50: fifty[->>];
1.96 + * 60: sixty[->>];
1.97 + * 70: seventy[->>];
1.98 + * 80: eighty[->>];
1.99 + * 90: ninety[->>];</pre>
1.100 + *
1.101 + * <p>In these rules, the <em>base value</em> is spelled out explicitly and set off from the
1.102 + * rule's output text with a colon. The rules are in a sorted list, and a rule is applicable
1.103 + * to all numbers from its own base value to one less than the next rule's base value. The
1.104 + * ">>" token is called a <em>substitution</em> and tells the fomatter to
1.105 + * isolate the number's ones digit, format it using this same set of rules, and place the
1.106 + * result at the position of the ">>" token. Text in brackets is omitted if
1.107 + * the number being formatted is an even multiple of 10 (the hyphen is a literal hyphen; 24
1.108 + * is "twenty-four," not "twenty four").</p>
1.109 + *
1.110 + * <p>For even larger numbers, we can actually look up several parts of the number in the
1.111 + * list:</p>
1.112 + *
1.113 + * <pre>100: << hundred[ >>];</pre>
1.114 + *
1.115 + * <p>The "<<" represents a new kind of substitution. The << isolates
1.116 + * the hundreds digit (and any digits to its left), formats it using this same rule set, and
1.117 + * places the result where the "<<" was. Notice also that the meaning of
1.118 + * >> has changed: it now refers to both the tens and the ones digits. The meaning of
1.119 + * both substitutions depends on the rule's base value. The base value determines the rule's <em>divisor,</em>
1.120 + * which is the highest power of 10 that is less than or equal to the base value (the user
1.121 + * can change this). To fill in the substitutions, the formatter divides the number being
1.122 + * formatted by the divisor. The integral quotient is used to fill in the <<
1.123 + * substitution, and the remainder is used to fill in the >> substitution. The meaning
1.124 + * of the brackets changes similarly: text in brackets is omitted if the value being
1.125 + * formatted is an even multiple of the rule's divisor. The rules are applied recursively, so
1.126 + * if a substitution is filled in with text that includes another substitution, that
1.127 + * substitution is also filled in.</p>
1.128 + *
1.129 + * <p>This rule covers values up to 999, at which point we add another rule:</p>
1.130 + *
1.131 + * <pre>1000: << thousand[ >>];</pre>
1.132 + *
1.133 + * <p>Again, the meanings of the brackets and substitution tokens shift because the rule's
1.134 + * base value is a higher power of 10, changing the rule's divisor. This rule can actually be
1.135 + * used all the way up to 999,999. This allows us to finish out the rules as follows:</p>
1.136 + *
1.137 + * <pre> 1,000,000: << million[ >>];
1.138 + * 1,000,000,000: << billion[ >>];
1.139 + * 1,000,000,000,000: << trillion[ >>];
1.140 + * 1,000,000,000,000,000: OUT OF RANGE!;</pre>
1.141 + *
1.142 + * <p>Commas, periods, and spaces can be used in the base values to improve legibility and
1.143 + * are ignored by the rule parser. The last rule in the list is customarily treated as an
1.144 + * "overflow rule," applying to everything from its base value on up, and often (as
1.145 + * in this example) being used to print out an error message or default representation.
1.146 + * Notice also that the size of the major groupings in large numbers is controlled by the
1.147 + * spacing of the rules: because in English we group numbers by thousand, the higher rules
1.148 + * are separated from each other by a factor of 1,000.</p>
1.149 + *
1.150 + * <p>To see how these rules actually work in practice, consider the following example:
1.151 + * Formatting 25,430 with this rule set would work like this:</p>
1.152 + *
1.153 + * <table border="0" width="100%">
1.154 + * <tr>
1.155 + * <td><strong><< thousand >></strong></td>
1.156 + * <td>[the rule whose base value is 1,000 is applicable to 25,340]</td>
1.157 + * </tr>
1.158 + * <tr>
1.159 + * <td><strong>twenty->></strong> thousand >></td>
1.160 + * <td>[25,340 over 1,000 is 25. The rule for 20 applies.]</td>
1.161 + * </tr>
1.162 + * <tr>
1.163 + * <td>twenty-<strong>five</strong> thousand >></td>
1.164 + * <td>[25 mod 10 is 5. The rule for 5 is "five."</td>
1.165 + * </tr>
1.166 + * <tr>
1.167 + * <td>twenty-five thousand <strong><< hundred >></strong></td>
1.168 + * <td>[25,340 mod 1,000 is 340. The rule for 100 applies.]</td>
1.169 + * </tr>
1.170 + * <tr>
1.171 + * <td>twenty-five thousand <strong>three</strong> hundred >></td>
1.172 + * <td>[340 over 100 is 3. The rule for 3 is "three."]</td>
1.173 + * </tr>
1.174 + * <tr>
1.175 + * <td>twenty-five thousand three hundred <strong>forty</strong></td>
1.176 + * <td>[340 mod 100 is 40. The rule for 40 applies. Since 40 divides
1.177 + * evenly by 10, the hyphen and substitution in the brackets are omitted.]</td>
1.178 + * </tr>
1.179 + * </table>
1.180 + *
1.181 + * <p>The above syntax suffices only to format positive integers. To format negative numbers,
1.182 + * we add a special rule:</p>
1.183 + *
1.184 + * <pre>-x: minus >>;</pre>
1.185 + *
1.186 + * <p>This is called a <em>negative-number rule,</em> and is identified by "-x"
1.187 + * where the base value would be. This rule is used to format all negative numbers. the
1.188 + * >> token here means "find the number's absolute value, format it with these
1.189 + * rules, and put the result here."</p>
1.190 + *
1.191 + * <p>We also add a special rule called a <em>fraction rule </em>for numbers with fractional
1.192 + * parts:</p>
1.193 + *
1.194 + * <pre>x.x: << point >>;</pre>
1.195 + *
1.196 + * <p>This rule is used for all positive non-integers (negative non-integers pass through the
1.197 + * negative-number rule first and then through this rule). Here, the << token refers to
1.198 + * the number's integral part, and the >> to the number's fractional part. The
1.199 + * fractional part is formatted as a series of single-digit numbers (e.g., 123.456 would be
1.200 + * formatted as "one hundred twenty-three point four five six").</p>
1.201 + *
1.202 + * <p>To see how this rule syntax is applied to various languages, examine the resource data.</p>
1.203 + *
1.204 + * <p>There is actually much more flexibility built into the rule language than the
1.205 + * description above shows. A formatter may own multiple rule sets, which can be selected by
1.206 + * the caller, and which can use each other to fill in their substitutions. Substitutions can
1.207 + * also be filled in with digits, using a DecimalFormat object. There is syntax that can be
1.208 + * used to alter a rule's divisor in various ways. And there is provision for much more
1.209 + * flexible fraction handling. A complete description of the rule syntax follows:</p>
1.210 + *
1.211 + * <hr>
1.212 + *
1.213 + * <p>The description of a <tt>RuleBasedNumberFormat</tt>'s behavior consists of one or more <em>rule
1.214 + * sets.</em> Each rule set consists of a name, a colon, and a list of <em>rules.</em> A rule
1.215 + * set name must begin with a % sign. Rule sets with names that begin with a single % sign
1.216 + * are <em>public:</em> the caller can specify that they be used to format and parse numbers.
1.217 + * Rule sets with names that begin with %% are <em>private:</em> they exist only for the use
1.218 + * of other rule sets. If a formatter only has one rule set, the name may be omitted.</p>
1.219 + *
1.220 + * <p>The user can also specify a special "rule set" named <tt>%%lenient-parse</tt>.
1.221 + * The body of <tt>%%lenient-parse</tt> isn't a set of number-formatting rules, but a <tt>RuleBasedCollator</tt>
1.222 + * description which is used to define equivalences for lenient parsing. For more information
1.223 + * on the syntax, see <tt>RuleBasedCollator</tt>. For more information on lenient parsing,
1.224 + * see <tt>setLenientParse()</tt>. <em>Note:</em> symbols that have syntactic meaning
1.225 + * in collation rules, such as '&', have no particular meaning when appearing outside
1.226 + * of the <tt>lenient-parse</tt> rule set.</p>
1.227 + *
1.228 + * <p>The body of a rule set consists of an ordered, semicolon-delimited list of <em>rules.</em>
1.229 + * Internally, every rule has a base value, a divisor, rule text, and zero, one, or two <em>substitutions.</em>
1.230 + * These parameters are controlled by the description syntax, which consists of a <em>rule
1.231 + * descriptor,</em> a colon, and a <em>rule body.</em></p>
1.232 + *
1.233 + * <p>A rule descriptor can take one of the following forms (text in <em>italics</em> is the
1.234 + * name of a token):</p>
1.235 + *
1.236 + * <table border="0" width="100%">
1.237 + * <tr>
1.238 + * <td><em>bv</em>:</td>
1.239 + * <td><em>bv</em> specifies the rule's base value. <em>bv</em> is a decimal
1.240 + * number expressed using ASCII digits. <em>bv</em> may contain spaces, period, and commas,
1.241 + * which are ignored. The rule's divisor is the highest power of 10 less than or equal to
1.242 + * the base value.</td>
1.243 + * </tr>
1.244 + * <tr>
1.245 + * <td><em>bv</em>/<em>rad</em>:</td>
1.246 + * <td><em>bv</em> specifies the rule's base value. The rule's divisor is the
1.247 + * highest power of <em>rad</em> less than or equal to the base value.</td>
1.248 + * </tr>
1.249 + * <tr>
1.250 + * <td><em>bv</em>>:</td>
1.251 + * <td><em>bv</em> specifies the rule's base value. To calculate the divisor,
1.252 + * let the radix be 10, and the exponent be the highest exponent of the radix that yields a
1.253 + * result less than or equal to the base value. Every > character after the base value
1.254 + * decreases the exponent by 1. If the exponent is positive or 0, the divisor is the radix
1.255 + * raised to the power of the exponent; otherwise, the divisor is 1.</td>
1.256 + * </tr>
1.257 + * <tr>
1.258 + * <td><em>bv</em>/<em>rad</em>>:</td>
1.259 + * <td><em>bv</em> specifies the rule's base value. To calculate the divisor,
1.260 + * let the radix be <em>rad</em>, and the exponent be the highest exponent of the radix that
1.261 + * yields a result less than or equal to the base value. Every > character after the radix
1.262 + * decreases the exponent by 1. If the exponent is positive or 0, the divisor is the radix
1.263 + * raised to the power of the exponent; otherwise, the divisor is 1.</td>
1.264 + * </tr>
1.265 + * <tr>
1.266 + * <td>-x:</td>
1.267 + * <td>The rule is a negative-number rule.</td>
1.268 + * </tr>
1.269 + * <tr>
1.270 + * <td>x.x:</td>
1.271 + * <td>The rule is an <em>improper fraction rule.</em></td>
1.272 + * </tr>
1.273 + * <tr>
1.274 + * <td>0.x:</td>
1.275 + * <td>The rule is a <em>proper fraction rule.</em></td>
1.276 + * </tr>
1.277 + * <tr>
1.278 + * <td>x.0:</td>
1.279 + * <td>The rule is a <em>master rule.</em></td>
1.280 + * </tr>
1.281 + * <tr>
1.282 + * <td><em>nothing</em></td>
1.283 + * <td>If the rule's rule descriptor is left out, the base value is one plus the
1.284 + * preceding rule's base value (or zero if this is the first rule in the list) in a normal
1.285 + * rule set. In a fraction rule set, the base value is the same as the preceding rule's
1.286 + * base value.</td>
1.287 + * </tr>
1.288 + * </table>
1.289 + *
1.290 + * <p>A rule set may be either a regular rule set or a <em>fraction rule set,</em> depending
1.291 + * on whether it is used to format a number's integral part (or the whole number) or a
1.292 + * number's fractional part. Using a rule set to format a rule's fractional part makes it a
1.293 + * fraction rule set.</p>
1.294 + *
1.295 + * <p>Which rule is used to format a number is defined according to one of the following
1.296 + * algorithms: If the rule set is a regular rule set, do the following:
1.297 + *
1.298 + * <ul>
1.299 + * <li>If the rule set includes a master rule (and the number was passed in as a <tt>double</tt>),
1.300 + * use the master rule. (If the number being formatted was passed in as a <tt>long</tt>,
1.301 + * the master rule is ignored.)</li>
1.302 + * <li>If the number is negative, use the negative-number rule.</li>
1.303 + * <li>If the number has a fractional part and is greater than 1, use the improper fraction
1.304 + * rule.</li>
1.305 + * <li>If the number has a fractional part and is between 0 and 1, use the proper fraction
1.306 + * rule.</li>
1.307 + * <li>Binary-search the rule list for the rule with the highest base value less than or equal
1.308 + * to the number. If that rule has two substitutions, its base value is not an even multiple
1.309 + * of its divisor, and the number <em>is</em> an even multiple of the rule's divisor, use the
1.310 + * rule that precedes it in the rule list. Otherwise, use the rule itself.</li>
1.311 + * </ul>
1.312 + *
1.313 + * <p>If the rule set is a fraction rule set, do the following:
1.314 + *
1.315 + * <ul>
1.316 + * <li>Ignore negative-number and fraction rules.</li>
1.317 + * <li>For each rule in the list, multiply the number being formatted (which will always be
1.318 + * between 0 and 1) by the rule's base value. Keep track of the distance between the result
1.319 + * the nearest integer.</li>
1.320 + * <li>Use the rule that produced the result closest to zero in the above calculation. In the
1.321 + * event of a tie or a direct hit, use the first matching rule encountered. (The idea here is
1.322 + * to try each rule's base value as a possible denominator of a fraction. Whichever
1.323 + * denominator produces the fraction closest in value to the number being formatted wins.) If
1.324 + * the rule following the matching rule has the same base value, use it if the numerator of
1.325 + * the fraction is anything other than 1; if the numerator is 1, use the original matching
1.326 + * rule. (This is to allow singular and plural forms of the rule text without a lot of extra
1.327 + * hassle.)</li>
1.328 + * </ul>
1.329 + *
1.330 + * <p>A rule's body consists of a string of characters terminated by a semicolon. The rule
1.331 + * may include zero, one, or two <em>substitution tokens,</em> and a range of text in
1.332 + * brackets. The brackets denote optional text (and may also include one or both
1.333 + * substitutions). The exact meanings of the substitution tokens, and under what conditions
1.334 + * optional text is omitted, depend on the syntax of the substitution token and the context.
1.335 + * The rest of the text in a rule body is literal text that is output when the rule matches
1.336 + * the number being formatted.</p>
1.337 + *
1.338 + * <p>A substitution token begins and ends with a <em>token character.</em> The token
1.339 + * character and the context together specify a mathematical operation to be performed on the
1.340 + * number being formatted. An optional <em>substitution descriptor </em>specifies how the
1.341 + * value resulting from that operation is used to fill in the substitution. The position of
1.342 + * the substitution token in the rule body specifies the location of the resultant text in
1.343 + * the original rule text.</p>
1.344 + *
1.345 + * <p>The meanings of the substitution token characters are as follows:</p>
1.346 + *
1.347 + * <table border="0" width="100%">
1.348 + * <tr>
1.349 + * <td>>></td>
1.350 + * <td>in normal rule</td>
1.351 + * <td>Divide the number by the rule's divisor and format the remainder</td>
1.352 + * </tr>
1.353 + * <tr>
1.354 + * <td></td>
1.355 + * <td>in negative-number rule</td>
1.356 + * <td>Find the absolute value of the number and format the result</td>
1.357 + * </tr>
1.358 + * <tr>
1.359 + * <td></td>
1.360 + * <td>in fraction or master rule</td>
1.361 + * <td>Isolate the number's fractional part and format it.</td>
1.362 + * </tr>
1.363 + * <tr>
1.364 + * <td></td>
1.365 + * <td>in rule in fraction rule set</td>
1.366 + * <td>Not allowed.</td>
1.367 + * </tr>
1.368 + * <tr>
1.369 + * <td>>>></td>
1.370 + * <td>in normal rule</td>
1.371 + * <td>Divide the number by the rule's divisor and format the remainder,
1.372 + * but bypass the normal rule-selection process and just use the
1.373 + * rule that precedes this one in this rule list.</td>
1.374 + * </tr>
1.375 + * <tr>
1.376 + * <td></td>
1.377 + * <td>in all other rules</td>
1.378 + * <td>Not allowed.</td>
1.379 + * </tr>
1.380 + * <tr>
1.381 + * <td><<</td>
1.382 + * <td>in normal rule</td>
1.383 + * <td>Divide the number by the rule's divisor and format the quotient</td>
1.384 + * </tr>
1.385 + * <tr>
1.386 + * <td></td>
1.387 + * <td>in negative-number rule</td>
1.388 + * <td>Not allowed.</td>
1.389 + * </tr>
1.390 + * <tr>
1.391 + * <td></td>
1.392 + * <td>in fraction or master rule</td>
1.393 + * <td>Isolate the number's integral part and format it.</td>
1.394 + * </tr>
1.395 + * <tr>
1.396 + * <td></td>
1.397 + * <td>in rule in fraction rule set</td>
1.398 + * <td>Multiply the number by the rule's base value and format the result.</td>
1.399 + * </tr>
1.400 + * <tr>
1.401 + * <td>==</td>
1.402 + * <td>in all rule sets</td>
1.403 + * <td>Format the number unchanged</td>
1.404 + * </tr>
1.405 + * <tr>
1.406 + * <td>[]</td>
1.407 + * <td>in normal rule</td>
1.408 + * <td>Omit the optional text if the number is an even multiple of the rule's divisor</td>
1.409 + * </tr>
1.410 + * <tr>
1.411 + * <td></td>
1.412 + * <td>in negative-number rule</td>
1.413 + * <td>Not allowed.</td>
1.414 + * </tr>
1.415 + * <tr>
1.416 + * <td></td>
1.417 + * <td>in improper-fraction rule</td>
1.418 + * <td>Omit the optional text if the number is between 0 and 1 (same as specifying both an
1.419 + * x.x rule and a 0.x rule)</td>
1.420 + * </tr>
1.421 + * <tr>
1.422 + * <td></td>
1.423 + * <td>in master rule</td>
1.424 + * <td>Omit the optional text if the number is an integer (same as specifying both an x.x
1.425 + * rule and an x.0 rule)</td>
1.426 + * </tr>
1.427 + * <tr>
1.428 + * <td></td>
1.429 + * <td>in proper-fraction rule</td>
1.430 + * <td>Not allowed.</td>
1.431 + * </tr>
1.432 + * <tr>
1.433 + * <td></td>
1.434 + * <td>in rule in fraction rule set</td>
1.435 + * <td>Omit the optional text if multiplying the number by the rule's base value yields 1.</td>
1.436 + * </tr>
1.437 + * </table>
1.438 + *
1.439 + * <p>The substitution descriptor (i.e., the text between the token characters) may take one
1.440 + * of three forms:</p>
1.441 + *
1.442 + * <table border="0" width="100%">
1.443 + * <tr>
1.444 + * <td>a rule set name</td>
1.445 + * <td>Perform the mathematical operation on the number, and format the result using the
1.446 + * named rule set.</td>
1.447 + * </tr>
1.448 + * <tr>
1.449 + * <td>a DecimalFormat pattern</td>
1.450 + * <td>Perform the mathematical operation on the number, and format the result using a
1.451 + * DecimalFormat with the specified pattern. The pattern must begin with 0 or #.</td>
1.452 + * </tr>
1.453 + * <tr>
1.454 + * <td>nothing</td>
1.455 + * <td>Perform the mathematical operation on the number, and format the result using the rule
1.456 + * set containing the current rule, except:
1.457 + * <ul>
1.458 + * <li>You can't have an empty substitution descriptor with a == substitution.</li>
1.459 + * <li>If you omit the substitution descriptor in a >> substitution in a fraction rule,
1.460 + * format the result one digit at a time using the rule set containing the current rule.</li>
1.461 + * <li>If you omit the substitution descriptor in a << substitution in a rule in a
1.462 + * fraction rule set, format the result using the default rule set for this formatter.</li>
1.463 + * </ul>
1.464 + * </td>
1.465 + * </tr>
1.466 + * </table>
1.467 + *
1.468 + * <p>Whitespace is ignored between a rule set name and a rule set body, between a rule
1.469 + * descriptor and a rule body, or between rules. If a rule body begins with an apostrophe,
1.470 + * the apostrophe is ignored, but all text after it becomes significant (this is how you can
1.471 + * have a rule's rule text begin with whitespace). There is no escape function: the semicolon
1.472 + * is not allowed in rule set names or in rule text, and the colon is not allowed in rule set
1.473 + * names. The characters beginning a substitution token are always treated as the beginning
1.474 + * of a substitution token.</p>
1.475 + *
1.476 + * <p>See the resource data and the demo program for annotated examples of real rule sets
1.477 + * using these features.</p>
1.478 + *
1.479 + * <p><em>User subclasses are not supported.</em> While clients may write
1.480 + * subclasses, such code will not necessarily work and will not be
1.481 + * guaranteed to work stably from release to release.
1.482 + *
1.483 + * <p><b>Localizations</b></p>
1.484 + * <p>Constructors are available that allow the specification of localizations for the
1.485 + * public rule sets (and also allow more control over what public rule sets are available).
1.486 + * Localization data is represented as a textual description. The description represents
1.487 + * an array of arrays of string. The first element is an array of the public rule set names,
1.488 + * each of these must be one of the public rule set names that appear in the rules. Only
1.489 + * names in this array will be treated as public rule set names by the API. Each subsequent
1.490 + * element is an array of localizations of these names. The first element of one of these
1.491 + * subarrays is the locale name, and the remaining elements are localizations of the
1.492 + * public rule set names, in the same order as they were listed in the first arrray.</p>
1.493 + * <p>In the syntax, angle brackets '<', '>' are used to delimit the arrays, and comma ',' is used
1.494 + * to separate elements of an array. Whitespace is ignored, unless quoted.</p>
1.495 + * <p>For example:<pre>
1.496 + * < < %foo, %bar, %baz >,
1.497 + * < en, Foo, Bar, Baz >,
1.498 + * < fr, 'le Foo', 'le Bar', 'le Baz' >
1.499 + * < zh, \\u7532, \\u4e59, \\u4e19 > >
1.500 + * </pre></p>
1.501 + * @author Richard Gillam
1.502 + * @see NumberFormat
1.503 + * @see DecimalFormat
1.504 + * @stable ICU 2.0
1.505 + */
1.506 +class U_I18N_API RuleBasedNumberFormat : public NumberFormat {
1.507 +public:
1.508 +
1.509 + //-----------------------------------------------------------------------
1.510 + // constructors
1.511 + //-----------------------------------------------------------------------
1.512 +
1.513 + /**
1.514 + * Creates a RuleBasedNumberFormat that behaves according to the description
1.515 + * passed in. The formatter uses the default locale.
1.516 + * @param rules A description of the formatter's desired behavior.
1.517 + * See the class documentation for a complete explanation of the description
1.518 + * syntax.
1.519 + * @param perror The parse error if an error was encountered.
1.520 + * @param status The status indicating whether the constructor succeeded.
1.521 + * @stable ICU 3.2
1.522 + */
1.523 + RuleBasedNumberFormat(const UnicodeString& rules, UParseError& perror, UErrorCode& status);
1.524 +
1.525 + /**
1.526 + * Creates a RuleBasedNumberFormat that behaves according to the description
1.527 + * passed in. The formatter uses the default locale.
1.528 + * <p>
1.529 + * The localizations data provides information about the public
1.530 + * rule sets and their localized display names for different
1.531 + * locales. The first element in the list is an array of the names
1.532 + * of the public rule sets. The first element in this array is
1.533 + * the initial default ruleset. The remaining elements in the
1.534 + * list are arrays of localizations of the names of the public
1.535 + * rule sets. Each of these is one longer than the initial array,
1.536 + * with the first String being the ULocale ID, and the remaining
1.537 + * Strings being the localizations of the rule set names, in the
1.538 + * same order as the initial array. Arrays are NULL-terminated.
1.539 + * @param rules A description of the formatter's desired behavior.
1.540 + * See the class documentation for a complete explanation of the description
1.541 + * syntax.
1.542 + * @param localizations the localization information.
1.543 + * names in the description. These will be copied by the constructor.
1.544 + * @param perror The parse error if an error was encountered.
1.545 + * @param status The status indicating whether the constructor succeeded.
1.546 + * @stable ICU 3.2
1.547 + */
1.548 + RuleBasedNumberFormat(const UnicodeString& rules, const UnicodeString& localizations,
1.549 + UParseError& perror, UErrorCode& status);
1.550 +
1.551 + /**
1.552 + * Creates a RuleBasedNumberFormat that behaves according to the rules
1.553 + * passed in. The formatter uses the specified locale to determine the
1.554 + * characters to use when formatting numerals, and to define equivalences
1.555 + * for lenient parsing.
1.556 + * @param rules The formatter rules.
1.557 + * See the class documentation for a complete explanation of the rule
1.558 + * syntax.
1.559 + * @param locale A locale that governs which characters are used for
1.560 + * formatting values in numerals and which characters are equivalent in
1.561 + * lenient parsing.
1.562 + * @param perror The parse error if an error was encountered.
1.563 + * @param status The status indicating whether the constructor succeeded.
1.564 + * @stable ICU 2.0
1.565 + */
1.566 + RuleBasedNumberFormat(const UnicodeString& rules, const Locale& locale,
1.567 + UParseError& perror, UErrorCode& status);
1.568 +
1.569 + /**
1.570 + * Creates a RuleBasedNumberFormat that behaves according to the description
1.571 + * passed in. The formatter uses the default locale.
1.572 + * <p>
1.573 + * The localizations data provides information about the public
1.574 + * rule sets and their localized display names for different
1.575 + * locales. The first element in the list is an array of the names
1.576 + * of the public rule sets. The first element in this array is
1.577 + * the initial default ruleset. The remaining elements in the
1.578 + * list are arrays of localizations of the names of the public
1.579 + * rule sets. Each of these is one longer than the initial array,
1.580 + * with the first String being the ULocale ID, and the remaining
1.581 + * Strings being the localizations of the rule set names, in the
1.582 + * same order as the initial array. Arrays are NULL-terminated.
1.583 + * @param rules A description of the formatter's desired behavior.
1.584 + * See the class documentation for a complete explanation of the description
1.585 + * syntax.
1.586 + * @param localizations a list of localizations for the rule set
1.587 + * names in the description. These will be copied by the constructor.
1.588 + * @param locale A locale that governs which characters are used for
1.589 + * formatting values in numerals and which characters are equivalent in
1.590 + * lenient parsing.
1.591 + * @param perror The parse error if an error was encountered.
1.592 + * @param status The status indicating whether the constructor succeeded.
1.593 + * @stable ICU 3.2
1.594 + */
1.595 + RuleBasedNumberFormat(const UnicodeString& rules, const UnicodeString& localizations,
1.596 + const Locale& locale, UParseError& perror, UErrorCode& status);
1.597 +
1.598 + /**
1.599 + * Creates a RuleBasedNumberFormat from a predefined ruleset. The selector
1.600 + * code choosed among three possible predefined formats: spellout, ordinal,
1.601 + * and duration.
1.602 + * @param tag A selector code specifying which kind of formatter to create for that
1.603 + * locale. There are four legal values: URBNF_SPELLOUT, which creates a formatter that
1.604 + * spells out a value in words in the desired language, URBNF_ORDINAL, which attaches
1.605 + * an ordinal suffix from the desired language to the end of a number (e.g. "123rd"),
1.606 + * URBNF_DURATION, which formats a duration in seconds as hours, minutes, and seconds,
1.607 + * and URBNF_NUMBERING_SYSTEM, which is used to invoke rules for alternate numbering
1.608 + * systems such as the Hebrew numbering system, or for Roman Numerals, etc.
1.609 + * @param locale The locale for the formatter.
1.610 + * @param status The status indicating whether the constructor succeeded.
1.611 + * @stable ICU 2.0
1.612 + */
1.613 + RuleBasedNumberFormat(URBNFRuleSetTag tag, const Locale& locale, UErrorCode& status);
1.614 +
1.615 + //-----------------------------------------------------------------------
1.616 + // boilerplate
1.617 + //-----------------------------------------------------------------------
1.618 +
1.619 + /**
1.620 + * Copy constructor
1.621 + * @param rhs the object to be copied from.
1.622 + * @stable ICU 2.6
1.623 + */
1.624 + RuleBasedNumberFormat(const RuleBasedNumberFormat& rhs);
1.625 +
1.626 + /**
1.627 + * Assignment operator
1.628 + * @param rhs the object to be copied from.
1.629 + * @stable ICU 2.6
1.630 + */
1.631 + RuleBasedNumberFormat& operator=(const RuleBasedNumberFormat& rhs);
1.632 +
1.633 + /**
1.634 + * Release memory allocated for a RuleBasedNumberFormat when you are finished with it.
1.635 + * @stable ICU 2.6
1.636 + */
1.637 + virtual ~RuleBasedNumberFormat();
1.638 +
1.639 + /**
1.640 + * Clone this object polymorphically. The caller is responsible
1.641 + * for deleting the result when done.
1.642 + * @return A copy of the object.
1.643 + * @stable ICU 2.6
1.644 + */
1.645 + virtual Format* clone(void) const;
1.646 +
1.647 + /**
1.648 + * Return true if the given Format objects are semantically equal.
1.649 + * Objects of different subclasses are considered unequal.
1.650 + * @param other the object to be compared with.
1.651 + * @return true if the given Format objects are semantically equal.
1.652 + * @stable ICU 2.6
1.653 + */
1.654 + virtual UBool operator==(const Format& other) const;
1.655 +
1.656 +//-----------------------------------------------------------------------
1.657 +// public API functions
1.658 +//-----------------------------------------------------------------------
1.659 +
1.660 + /**
1.661 + * return the rules that were provided to the RuleBasedNumberFormat.
1.662 + * @return the result String that was passed in
1.663 + * @stable ICU 2.0
1.664 + */
1.665 + virtual UnicodeString getRules() const;
1.666 +
1.667 + /**
1.668 + * Return the number of public rule set names.
1.669 + * @return the number of public rule set names.
1.670 + * @stable ICU 2.0
1.671 + */
1.672 + virtual int32_t getNumberOfRuleSetNames() const;
1.673 +
1.674 + /**
1.675 + * Return the name of the index'th public ruleSet. If index is not valid,
1.676 + * the function returns null.
1.677 + * @param index the index of the ruleset
1.678 + * @return the name of the index'th public ruleSet.
1.679 + * @stable ICU 2.0
1.680 + */
1.681 + virtual UnicodeString getRuleSetName(int32_t index) const;
1.682 +
1.683 + /**
1.684 + * Return the number of locales for which we have localized rule set display names.
1.685 + * @return the number of locales for which we have localized rule set display names.
1.686 + * @stable ICU 3.2
1.687 + */
1.688 + virtual int32_t getNumberOfRuleSetDisplayNameLocales(void) const;
1.689 +
1.690 + /**
1.691 + * Return the index'th display name locale.
1.692 + * @param index the index of the locale
1.693 + * @param status set to a failure code when this function fails
1.694 + * @return the locale
1.695 + * @see #getNumberOfRuleSetDisplayNameLocales
1.696 + * @stable ICU 3.2
1.697 + */
1.698 + virtual Locale getRuleSetDisplayNameLocale(int32_t index, UErrorCode& status) const;
1.699 +
1.700 + /**
1.701 + * Return the rule set display names for the provided locale. These are in the same order
1.702 + * as those returned by getRuleSetName. The locale is matched against the locales for
1.703 + * which there is display name data, using normal fallback rules. If no locale matches,
1.704 + * the default display names are returned. (These are the internal rule set names minus
1.705 + * the leading '%'.)
1.706 + * @param index the index of the rule set
1.707 + * @param locale the locale (returned by getRuleSetDisplayNameLocales) for which the localized
1.708 + * display name is desired
1.709 + * @return the display name for the given index, which might be bogus if there is an error
1.710 + * @see #getRuleSetName
1.711 + * @stable ICU 3.2
1.712 + */
1.713 + virtual UnicodeString getRuleSetDisplayName(int32_t index,
1.714 + const Locale& locale = Locale::getDefault());
1.715 +
1.716 + /**
1.717 + * Return the rule set display name for the provided rule set and locale.
1.718 + * The locale is matched against the locales for which there is display name data, using
1.719 + * normal fallback rules. If no locale matches, the default display name is returned.
1.720 + * @return the display name for the rule set
1.721 + * @stable ICU 3.2
1.722 + * @see #getRuleSetDisplayName
1.723 + */
1.724 + virtual UnicodeString getRuleSetDisplayName(const UnicodeString& ruleSetName,
1.725 + const Locale& locale = Locale::getDefault());
1.726 +
1.727 +
1.728 + using NumberFormat::format;
1.729 +
1.730 + /**
1.731 + * Formats the specified 32-bit number using the default ruleset.
1.732 + * @param number The number to format.
1.733 + * @param toAppendTo the string that will hold the (appended) result
1.734 + * @param pos the fieldposition
1.735 + * @return A textual representation of the number.
1.736 + * @stable ICU 2.0
1.737 + */
1.738 + virtual UnicodeString& format(int32_t number,
1.739 + UnicodeString& toAppendTo,
1.740 + FieldPosition& pos) const;
1.741 +
1.742 + /**
1.743 + * Formats the specified 64-bit number using the default ruleset.
1.744 + * @param number The number to format.
1.745 + * @param toAppendTo the string that will hold the (appended) result
1.746 + * @param pos the fieldposition
1.747 + * @return A textual representation of the number.
1.748 + * @stable ICU 2.1
1.749 + */
1.750 + virtual UnicodeString& format(int64_t number,
1.751 + UnicodeString& toAppendTo,
1.752 + FieldPosition& pos) const;
1.753 + /**
1.754 + * Formats the specified number using the default ruleset.
1.755 + * @param number The number to format.
1.756 + * @param toAppendTo the string that will hold the (appended) result
1.757 + * @param pos the fieldposition
1.758 + * @return A textual representation of the number.
1.759 + * @stable ICU 2.0
1.760 + */
1.761 + virtual UnicodeString& format(double number,
1.762 + UnicodeString& toAppendTo,
1.763 + FieldPosition& pos) const;
1.764 +
1.765 + /**
1.766 + * Formats the specified number using the named ruleset.
1.767 + * @param number The number to format.
1.768 + * @param ruleSetName The name of the rule set to format the number with.
1.769 + * This must be the name of a valid public rule set for this formatter.
1.770 + * @param toAppendTo the string that will hold the (appended) result
1.771 + * @param pos the fieldposition
1.772 + * @param status the status
1.773 + * @return A textual representation of the number.
1.774 + * @stable ICU 2.0
1.775 + */
1.776 + virtual UnicodeString& format(int32_t number,
1.777 + const UnicodeString& ruleSetName,
1.778 + UnicodeString& toAppendTo,
1.779 + FieldPosition& pos,
1.780 + UErrorCode& status) const;
1.781 + /**
1.782 + * Formats the specified 64-bit number using the named ruleset.
1.783 + * @param number The number to format.
1.784 + * @param ruleSetName The name of the rule set to format the number with.
1.785 + * This must be the name of a valid public rule set for this formatter.
1.786 + * @param toAppendTo the string that will hold the (appended) result
1.787 + * @param pos the fieldposition
1.788 + * @param status the status
1.789 + * @return A textual representation of the number.
1.790 + * @stable ICU 2.1
1.791 + */
1.792 + virtual UnicodeString& format(int64_t number,
1.793 + const UnicodeString& ruleSetName,
1.794 + UnicodeString& toAppendTo,
1.795 + FieldPosition& pos,
1.796 + UErrorCode& status) const;
1.797 + /**
1.798 + * Formats the specified number using the named ruleset.
1.799 + * @param number The number to format.
1.800 + * @param ruleSetName The name of the rule set to format the number with.
1.801 + * This must be the name of a valid public rule set for this formatter.
1.802 + * @param toAppendTo the string that will hold the (appended) result
1.803 + * @param pos the fieldposition
1.804 + * @param status the status
1.805 + * @return A textual representation of the number.
1.806 + * @stable ICU 2.0
1.807 + */
1.808 + virtual UnicodeString& format(double number,
1.809 + const UnicodeString& ruleSetName,
1.810 + UnicodeString& toAppendTo,
1.811 + FieldPosition& pos,
1.812 + UErrorCode& status) const;
1.813 +
1.814 + using NumberFormat::parse;
1.815 +
1.816 + /**
1.817 + * Parses the specfied string, beginning at the specified position, according
1.818 + * to this formatter's rules. This will match the string against all of the
1.819 + * formatter's public rule sets and return the value corresponding to the longest
1.820 + * parseable substring. This function's behavior is affected by the lenient
1.821 + * parse mode.
1.822 + * @param text The string to parse
1.823 + * @param result the result of the parse, either a double or a long.
1.824 + * @param parsePosition On entry, contains the position of the first character
1.825 + * in "text" to examine. On exit, has been updated to contain the position
1.826 + * of the first character in "text" that wasn't consumed by the parse.
1.827 + * @see #setLenient
1.828 + * @stable ICU 2.0
1.829 + */
1.830 + virtual void parse(const UnicodeString& text,
1.831 + Formattable& result,
1.832 + ParsePosition& parsePosition) const;
1.833 +
1.834 +#if !UCONFIG_NO_COLLATION
1.835 +
1.836 + /**
1.837 + * Turns lenient parse mode on and off.
1.838 + *
1.839 + * When in lenient parse mode, the formatter uses a Collator for parsing the text.
1.840 + * Only primary differences are treated as significant. This means that case
1.841 + * differences, accent differences, alternate spellings of the same letter
1.842 + * (e.g., ae and a-umlaut in German), ignorable characters, etc. are ignored in
1.843 + * matching the text. In many cases, numerals will be accepted in place of words
1.844 + * or phrases as well.
1.845 + *
1.846 + * For example, all of the following will correctly parse as 255 in English in
1.847 + * lenient-parse mode:
1.848 + * <br>"two hundred fifty-five"
1.849 + * <br>"two hundred fifty five"
1.850 + * <br>"TWO HUNDRED FIFTY-FIVE"
1.851 + * <br>"twohundredfiftyfive"
1.852 + * <br>"2 hundred fifty-5"
1.853 + *
1.854 + * The Collator used is determined by the locale that was
1.855 + * passed to this object on construction. The description passed to this object
1.856 + * on construction may supply additional collation rules that are appended to the
1.857 + * end of the default collator for the locale, enabling additional equivalences
1.858 + * (such as adding more ignorable characters or permitting spelled-out version of
1.859 + * symbols; see the demo program for examples).
1.860 + *
1.861 + * It's important to emphasize that even strict parsing is relatively lenient: it
1.862 + * will accept some text that it won't produce as output. In English, for example,
1.863 + * it will correctly parse "two hundred zero" and "fifteen hundred".
1.864 + *
1.865 + * @param enabled If true, turns lenient-parse mode on; if false, turns it off.
1.866 + * @see RuleBasedCollator
1.867 + * @stable ICU 2.0
1.868 + */
1.869 + virtual void setLenient(UBool enabled);
1.870 +
1.871 + /**
1.872 + * Returns true if lenient-parse mode is turned on. Lenient parsing is off
1.873 + * by default.
1.874 + * @return true if lenient-parse mode is turned on.
1.875 + * @see #setLenient
1.876 + * @stable ICU 2.0
1.877 + */
1.878 + virtual inline UBool isLenient(void) const;
1.879 +
1.880 +#endif
1.881 +
1.882 + /**
1.883 + * Override the default rule set to use. If ruleSetName is null, reset
1.884 + * to the initial default rule set. If the rule set is not a public rule set name,
1.885 + * U_ILLEGAL_ARGUMENT_ERROR is returned in status.
1.886 + * @param ruleSetName the name of the rule set, or null to reset the initial default.
1.887 + * @param status set to failure code when a problem occurs.
1.888 + * @stable ICU 2.6
1.889 + */
1.890 + virtual void setDefaultRuleSet(const UnicodeString& ruleSetName, UErrorCode& status);
1.891 +
1.892 + /**
1.893 + * Return the name of the current default rule set. If the current rule set is
1.894 + * not public, returns a bogus (and empty) UnicodeString.
1.895 + * @return the name of the current default rule set
1.896 + * @stable ICU 3.0
1.897 + */
1.898 + virtual UnicodeString getDefaultRuleSetName() const;
1.899 +
1.900 +public:
1.901 + /**
1.902 + * ICU "poor man's RTTI", returns a UClassID for this class.
1.903 + *
1.904 + * @stable ICU 2.8
1.905 + */
1.906 + static UClassID U_EXPORT2 getStaticClassID(void);
1.907 +
1.908 + /**
1.909 + * ICU "poor man's RTTI", returns a UClassID for the actual class.
1.910 + *
1.911 + * @stable ICU 2.8
1.912 + */
1.913 + virtual UClassID getDynamicClassID(void) const;
1.914 +
1.915 + /**
1.916 + * Sets the decimal format symbols, which is generally not changed
1.917 + * by the programmer or user. The formatter takes ownership of
1.918 + * symbolsToAdopt; the client must not delete it.
1.919 + *
1.920 + * @param symbolsToAdopt DecimalFormatSymbols to be adopted.
1.921 + * @stable ICU 49
1.922 + */
1.923 + virtual void adoptDecimalFormatSymbols(DecimalFormatSymbols* symbolsToAdopt);
1.924 +
1.925 + /**
1.926 + * Sets the decimal format symbols, which is generally not changed
1.927 + * by the programmer or user. A clone of the symbols is created and
1.928 + * the symbols is _not_ adopted; the client is still responsible for
1.929 + * deleting it.
1.930 + *
1.931 + * @param symbols DecimalFormatSymbols.
1.932 + * @stable ICU 49
1.933 + */
1.934 + virtual void setDecimalFormatSymbols(const DecimalFormatSymbols& symbols);
1.935 +
1.936 +private:
1.937 + RuleBasedNumberFormat(); // default constructor not implemented
1.938 +
1.939 + // this will ref the localizations if they are not NULL
1.940 + // caller must deref to get adoption
1.941 + RuleBasedNumberFormat(const UnicodeString& description, LocalizationInfo* localizations,
1.942 + const Locale& locale, UParseError& perror, UErrorCode& status);
1.943 +
1.944 + void init(const UnicodeString& rules, LocalizationInfo* localizations, UParseError& perror, UErrorCode& status);
1.945 + void dispose();
1.946 + void stripWhitespace(UnicodeString& src);
1.947 + void initDefaultRuleSet();
1.948 + void format(double number, NFRuleSet& ruleSet);
1.949 + NFRuleSet* findRuleSet(const UnicodeString& name, UErrorCode& status) const;
1.950 +
1.951 + /* friend access */
1.952 + friend class NFSubstitution;
1.953 + friend class NFRule;
1.954 + friend class FractionalPartSubstitution;
1.955 +
1.956 + inline NFRuleSet * getDefaultRuleSet() const;
1.957 + Collator * getCollator() const;
1.958 + DecimalFormatSymbols * getDecimalFormatSymbols() const;
1.959 +
1.960 +private:
1.961 + NFRuleSet **ruleSets;
1.962 + UnicodeString* ruleSetDescriptions;
1.963 + int32_t numRuleSets;
1.964 + NFRuleSet *defaultRuleSet;
1.965 + Locale locale;
1.966 + Collator* collator;
1.967 + DecimalFormatSymbols* decimalFormatSymbols;
1.968 + UBool lenient;
1.969 + UnicodeString* lenientParseRules;
1.970 + LocalizationInfo* localizations;
1.971 +};
1.972 +
1.973 +// ---------------
1.974 +
1.975 +#if !UCONFIG_NO_COLLATION
1.976 +
1.977 +inline UBool
1.978 +RuleBasedNumberFormat::isLenient(void) const {
1.979 + return lenient;
1.980 +}
1.981 +
1.982 +#endif
1.983 +
1.984 +inline NFRuleSet*
1.985 +RuleBasedNumberFormat::getDefaultRuleSet() const {
1.986 + return defaultRuleSet;
1.987 +}
1.988 +
1.989 +U_NAMESPACE_END
1.990 +
1.991 +/* U_HAVE_RBNF */
1.992 +#endif
1.993 +
1.994 +/* RBNF_H */
1.995 +#endif
