include/xapian/unicode.h

Go to the documentation of this file.
00001 
00004 /* Copyright (C) 2006,2007 Olly Betts
00005  *
00006  * This program is free software; you can redistribute it and/or modify
00007  * it under the terms of the GNU General Public License as published by
00008  * the Free Software Foundation; either version 2 of the License, or
00009  * (at your option) any later version.
00010  *
00011  * This program is distributed in the hope that it will be useful,
00012  * but WITHOUT ANY WARRANTY; without even the implied warranty of
00013  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
00014  * GNU General Public License for more details.
00015  *
00016  * You should have received a copy of the GNU General Public License
00017  * along with this program; if not, write to the Free Software
00018  * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301 USA
00019  */
00020 
00021 // FIXME: Nail down API details (what to expose; what names to use).
00022 
00023 #ifndef XAPIAN_INCLUDED_UNICODE_H
00024 #define XAPIAN_INCLUDED_UNICODE_H
00025 
00026 #include <xapian/visibility.h>
00027 
00028 #include <string>
00029 
00030 namespace Xapian {
00031 
00035 class XAPIAN_VISIBILITY_DEFAULT Utf8Iterator {
00036     const unsigned char *p;
00037     const unsigned char *end;
00038     mutable unsigned seqlen;
00039 
00040     void calculate_sequence_length() const;
00041 
00042     unsigned get_char() const;
00043 
00044     Utf8Iterator(const unsigned char *p_, const unsigned char *end_, unsigned seqlen_)
00045         : p(p_), end(end_), seqlen(seqlen_) { }
00046 
00047   public:
00049     const char * raw() const {
00050         return reinterpret_cast<const char *>(p ? p : end);
00051     }
00052 
00054     size_t left() const { return p ? end - p : 0; }
00055 
00067     void assign(const char *p_, size_t len) {
00068         if (len) {
00069             p = reinterpret_cast<const unsigned char*>(p_);
00070             end = p + len;
00071             seqlen = 0;
00072         } else {
00073             p = NULL;
00074         }
00075     }
00076 
00087     void assign(const std::string &s) { assign(s.data(), s.size()); }
00088 
00097     explicit Utf8Iterator(const char *p_);
00098 
00109     Utf8Iterator(const char *p_, size_t len) { assign(p_, len); }
00110 
00120     Utf8Iterator(const std::string &s) { assign(s.data(), s.size()); }
00121 
00127     Utf8Iterator() : p(NULL), end(0), seqlen(0) { }
00128 
00133     unsigned operator*() const;
00134 
00139     Utf8Iterator operator++(int) {
00140         // If we've not calculated seqlen yet, do so.
00141         if (seqlen == 0) calculate_sequence_length();
00142         const unsigned char *old_p = p;
00143         unsigned old_seqlen = seqlen;
00144         p += seqlen;
00145         if (p == end) p = NULL;
00146         seqlen = 0;
00147         return Utf8Iterator(old_p, end, old_seqlen);
00148     }
00149 
00154     Utf8Iterator & operator++() {
00155         this->operator++(0);
00156         return *this;
00157     }
00158 
00163     bool operator==(const Utf8Iterator &other) const { return p == other.p; }
00164 
00169     bool operator!=(const Utf8Iterator &other) const { return p != other.p; }
00170 
00172 
00173     typedef std::input_iterator_tag iterator_category;
00174     typedef unsigned value_type;
00175     typedef size_t difference_type;
00176     typedef const unsigned * pointer;
00177     typedef const unsigned & reference;
00179 };
00180 
00181 namespace Unicode {
00182 
00184 typedef enum {
00185     UNASSIGNED,
00186     UPPERCASE_LETTER,
00187     LOWERCASE_LETTER,
00188     TITLECASE_LETTER,
00189     MODIFIER_LETTER,
00190     OTHER_LETTER,
00191     NON_SPACING_MARK,
00192     ENCLOSING_MARK,
00193     COMBINING_SPACING_MARK,
00194     DECIMAL_DIGIT_NUMBER,
00195     LETTER_NUMBER,
00196     OTHER_NUMBER,
00197     SPACE_SEPARATOR,
00198     LINE_SEPARATOR,
00199     PARAGRAPH_SEPARATOR,
00200     CONTROL,
00201     FORMAT,
00202     PRIVATE_USE,
00203     SURROGATE,
00204     CONNECTOR_PUNCTUATION,
00205     DASH_PUNCTUATION,
00206     OPEN_PUNCTUATION,
00207     CLOSE_PUNCTUATION,
00208     INITIAL_QUOTE_PUNCTUATION,
00209     FINAL_QUOTE_PUNCTUATION,
00210     OTHER_PUNCTUATION,
00211     MATH_SYMBOL,
00212     CURRENCY_SYMBOL,
00213     MODIFIER_SYMBOL,
00214     OTHER_SYMBOL
00215 } category;
00216 
00217 namespace Internal {
00218     /* Extract the information about a character from the Unicode character
00219      * tables.
00220      *
00221      * ch must be a valid Unicode character value (i.e. < 0x110000)
00222      */
00223     XAPIAN_VISIBILITY_DEFAULT
00224     int get_character_info(unsigned ch);
00225 
00227     inline int get_case_type(int info) { return ((info & 0xe0) >> 5); }
00228 
00230     inline category get_category(int info) { return static_cast<category>(info & 0x1f); }
00231 
00233     inline int get_delta(int info) {
00234         /* It's implementation defined if sign extension happens on right shift
00235          * of a signed int, hence the conditional (hopefully the compiler will
00236          * spot this and optimise it to a sign-extending shift on architectures
00237          * with a suitable instruction).
00238          */
00239         return (info >= 0) ? (info >> 15) : (~(~info >> 15));
00240     }
00241 }
00242 
00252 XAPIAN_VISIBILITY_DEFAULT
00253 unsigned nonascii_to_utf8(unsigned ch, char * buf);
00254 
00262 inline unsigned to_utf8(unsigned ch, char *buf) {
00263     if (ch < 128) {
00264         *buf = static_cast<unsigned char>(ch);
00265         return 1;
00266     }
00267     return Xapian::Unicode::nonascii_to_utf8(ch, buf);
00268 }
00269 
00273 inline void append_utf8(std::string &s, unsigned ch) {
00274     char buf[4];
00275     s.append(buf, to_utf8(ch, buf));
00276 }
00277 
00279 inline category get_category(unsigned ch) {
00280     // Categorise non-Unicode values as UNASSIGNED.
00281     if (ch >= 0x110000) return Xapian::Unicode::UNASSIGNED;
00282     return Internal::get_category(Internal::get_character_info(ch));
00283 }
00284 
00286 inline bool is_wordchar(unsigned ch) {
00287     const unsigned int WORDCHAR_MASK =
00288             (1 << Xapian::Unicode::UPPERCASE_LETTER) |
00289             (1 << Xapian::Unicode::LOWERCASE_LETTER) |
00290             (1 << Xapian::Unicode::TITLECASE_LETTER) |
00291             (1 << Xapian::Unicode::MODIFIER_LETTER) |
00292             (1 << Xapian::Unicode::OTHER_LETTER) |
00293             (1 << Xapian::Unicode::DECIMAL_DIGIT_NUMBER) |
00294             (1 << Xapian::Unicode::LETTER_NUMBER) |
00295             (1 << Xapian::Unicode::OTHER_NUMBER) |
00296             (1 << Xapian::Unicode::CONNECTOR_PUNCTUATION);
00297     return ((WORDCHAR_MASK >> get_category(ch)) & 1);
00298 }
00299 
00301 inline bool is_whitespace(unsigned ch) {
00302     const unsigned int WHITESPACE_MASK =
00303             (1 << Xapian::Unicode::CONTROL) | // For TAB, CR, LF, FF.
00304             (1 << Xapian::Unicode::SPACE_SEPARATOR) |
00305             (1 << Xapian::Unicode::LINE_SEPARATOR) |
00306             (1 << Xapian::Unicode::PARAGRAPH_SEPARATOR);
00307     return ((WHITESPACE_MASK >> get_category(ch)) & 1);
00308 }
00309 
00311 inline bool is_currency(unsigned ch) {
00312     return (get_category(ch) == Xapian::Unicode::CURRENCY_SYMBOL);
00313 }
00314 
00316 inline unsigned tolower(unsigned ch) {
00317     int info;
00318     // Leave non-Unicode values unchanged.
00319     if (ch >= 0x110000 || !(Internal::get_case_type((info = Xapian::Unicode::Internal::get_character_info(ch))) & 2))
00320         return ch;
00321     return ch + Internal::get_delta(info);
00322 }
00323 
00325 inline std::string
00326 tolower(const std::string &term)
00327 {
00328     std::string result;
00329     result.reserve(term.size());
00330     for (Utf8Iterator i(term); i != Utf8Iterator(); ++i) {
00331         append_utf8(result, tolower(*i));
00332     }
00333     return result;
00334 }
00335 
00336 }
00337 
00338 }
00339 
00340 #endif // XAPIAN_INCLUDED_UNICODE_H

Documentation for Xapian (version 1.0.0).
Generated on 18 May 2007 by Doxygen 1.4.6.