icuSources/i18n/unicode/stsearch.h

   1 /*
   2 **********************************************************************
   3 *   Copyright (C) 2001-2006 IBM and others. All rights reserved.
   4 **********************************************************************
   5 *   Date        Name        Description
   6 *  03/22/2000   helena      Creation.
   7 **********************************************************************
   8 */
   9
  10 #ifndef STSEARCH_H
  11 #define STSEARCH_H
  12
  13 #include "unicode/utypes.h"
  14
  15 /**
  16  * \file
  17  * \brief C++ API: Service for searching text based on RuleBasedCollator.
  18  */
  19
  20 #if !UCONFIG_NO_COLLATION
  21
  22 #include "unicode/tblcoll.h"
  23 #include "unicode/coleitr.h"
  24 #include "unicode/search.h"
  25
  26 U_NAMESPACE_BEGIN
  27
  28 /**
  29  *
  30  * <tt>StringSearch</tt> is a <tt>SearchIterator</tt> that provides
  31  * language-sensitive text searching based on the comparison rules defined
  32  * in a {@link RuleBasedCollator} object.
  33  * StringSearch ensures that language eccentricity can be
  34  * handled, e.g. for the German collator, characters &szlig; and SS will be matched
  35  * if case is chosen to be ignored.
  36  * See the <a href="http://dev.icu-project.org/cgi-bin/viewcvs.cgi/~checkout~/icuhtml/design/collation/ICU_collation_design.htm">
  37  * "ICU Collation Design Document"</a> for more information.
  38  * <p>
  39  * The algorithm implemented is a modified form of the Boyer Moore's search.
  40  * For more information  see
  41  * <a href="http://icu.sourceforge.net/docs/papers/efficient_text_searching_in_java.html">
  42  * "Efficient Text Searching in Java"</a>, published in <i>Java Report</i>
  43  * in February, 1999, for further information on the algorithm.
  44  * <p>
  45  * There are 2 match options for selection:<br>
  46  * Let S' be the sub-string of a text string S between the offsets start and
  47  * end <start, end>.
  48  * <br>
  49  * A pattern string P matches a text string S at the offsets <start, end>
  50  * if
  51  * <pre>
  52  * option 1. Some canonical equivalent of P matches some canonical equivalent
  53  *           of S'
  54  * option 2. P matches S' and if P starts or ends with a combining mark,
  55  *           there exists no non-ignorable combining mark before or after S?
  56  *           in S respectively.
  57  * </pre>
  58  * Option 2. will be the default.
  59  * <p>
  60  * This search has APIs similar to that of other text iteration mechanisms
  61  * such as the break iterators in <tt>BreakIterator</tt>. Using these
  62  * APIs, it is easy to scan through text looking for all occurances of
  63  * a given pattern. This search iterator allows changing of direction by
  64  * calling a <tt>reset</tt> followed by a <tt>next</tt> or <tt>previous</tt>.
  65  * Though a direction change can occur without calling <tt>reset</tt> first,
  66  * this operation comes with some speed penalty.
  67  * Match results in the forward direction will match the result matches in
  68  * the backwards direction in the reverse order
  69  * <p>
  70  * <tt>SearchIterator</tt> provides APIs to specify the starting position
  71  * within the text string to be searched, e.g. <tt>setOffset</tt>,
  72  * <tt>preceding</tt> and <tt>following</tt>. Since the
  73  * starting position will be set as it is specified, please take note that
  74  * there are some danger points which the search may render incorrect
  75  * results:
  76  * <ul>
  77  * <li> The midst of a substring that requires normalization.
  78  * <li> If the following match is to be found, the position should not be the
  79  *      second character which requires to be swapped with the preceding
  80  *      character. Vice versa, if the preceding match is to be found,
  81  *      position to search from should not be the first character which
  82  *      requires to be swapped with the next character. E.g certain Thai and
  83  *      Lao characters require swapping.
  84  * <li> If a following pattern match is to be found, any position within a
  85  *      contracting sequence except the first will fail. Vice versa if a
  86  *      preceding pattern match is to be found, a invalid starting point
  87  *      would be any character within a contracting sequence except the last.
  88  * </ul>
  89  * <p>
  90  * A breakiterator can be used if only matches at logical breaks are desired.
  91  * Using a breakiterator will only give you results that exactly matches the
  92  * boundaries given by the breakiterator. For instance the pattern "e" will
  93  * not be found in the string "\u00e9" if a character break iterator is used.
  94  * <p>
  95  * Options are provided to handle overlapping matches.
  96  * E.g. In English, overlapping matches produces the result 0 and 2
  97  * for the pattern "abab" in the text "ababab", where else mutually
  98  * exclusive matches only produce the result of 0.
  99  * <p>
 100  * Though collator attributes will be taken into consideration while
 101  * performing matches, there are no APIs here for setting and getting the
 102  * attributes. These attributes can be set by getting the collator
 103  * from <tt>getCollator</tt> and using the APIs in <tt>coll.h</tt>.
 104  * Lastly to update StringSearch to the new collator attributes,
 105  * reset() has to be called.
 106  * <p>
 107  * Restriction: <br>
 108  * Currently there are no composite characters that consists of a
 109  * character with combining class > 0 before a character with combining
 110  * class == 0. However, if such a character exists in the future,
 111  * StringSearch does not guarantee the results for option 1.
 112  * <p>
 113  * Consult the <tt>SearchIterator</tt> documentation for information on
 114  * and examples of how to use instances of this class to implement text
 115  * searching.
 116  * <pre><code>
 117  * UnicodeString target("The quick brown fox jumped over the lazy fox");
 118  * UnicodeString pattern("fox");
 119  *
 120  * SearchIterator *iter  = new StringSearch(pattern, target);
 121  * UErrorCode      error = U_ZERO_ERROR;
 122  * for (int pos = iter->first(error); pos != USEARCH_DONE;
 123  *                               pos = iter->next(error)) {
 124  *     printf("Found match at %d pos, length is %d\n", pos,
 125  *                                             iter.getMatchLength());
 126  * }
 127  * </code></pre>
 128  * <p>
 129  * Note, StringSearch is not to be subclassed.
 130  * </p>
 131  * @see SearchIterator
 132  * @see RuleBasedCollator
 133  * @since ICU 2.0
 134  */
 135
 136 class U_I18N_API StringSearch : public SearchIterator
 137 {
 138 public:
 139
 140     // public constructors and destructors --------------------------------
 141
 142     /**
 143      * Creating a <tt>StringSearch</tt> instance using the argument locale
 144      * language rule set. A collator will be created in the process, which
 145      * will be owned by this instance and will be deleted during
 146      * destruction
 147      * @param pattern The text for which this object will search.
 148      * @param text    The text in which to search for the pattern.
 149      * @param locale  A locale which defines the language-sensitive
 150      *                comparison rules used to determine whether text in the
 151      *                pattern and target matches.
 152      * @param breakiter A <tt>BreakIterator</tt> object used to constrain
 153      *                the matches that are found. Matches whose start and end
 154      *                indices in the target text are not boundaries as
 155      *                determined by the <tt>BreakIterator</tt> are
 156      *                ignored. If this behavior is not desired,
 157      *                <tt>NULL</tt> can be passed in instead.
 158      * @param status  for errors if any. If pattern or text is NULL, or if
 159      *               either the length of pattern or text is 0 then an
 160      *               U_ILLEGAL_ARGUMENT_ERROR is returned.
 161      * @stable ICU 2.0
 162      */
 163     StringSearch(const UnicodeString &pattern, const UnicodeString &text,
 164                  const Locale        &locale,
 165                        BreakIterator *breakiter,
 166                        UErrorCode    &status);
 167
 168     /**
 169      * Creating a <tt>StringSearch</tt> instance using the argument collator
 170      * language rule set. Note, user retains the ownership of this collator,
 171      * it does not get destroyed during this instance's destruction.
 172      * @param pattern The text for which this object will search.
 173      * @param text    The text in which to search for the pattern.
 174      * @param coll    A <tt>RuleBasedCollator</tt> object which defines
 175      *                the language-sensitive comparison rules used to
 176      *                determine whether text in the pattern and target
 177      *                matches. User is responsible for the clearing of this
 178      *                object.
 179      * @param breakiter A <tt>BreakIterator</tt> object used to constrain
 180      *                the matches that are found. Matches whose start and end
 181      *                indices in the target text are not boundaries as
 182      *                determined by the <tt>BreakIterator</tt> are
 183      *                ignored. If this behavior is not desired,
 184      *                <tt>NULL</tt> can be passed in instead.
 185      * @param status for errors if any. If either the length of pattern or
 186      *               text is 0 then an U_ILLEGAL_ARGUMENT_ERROR is returned.
 187      * @stable ICU 2.0
 188      */
 189     StringSearch(const UnicodeString     &pattern,
 190                  const UnicodeString     &text,
 191                        RuleBasedCollator *coll,
 192                        BreakIterator     *breakiter,
 193                        UErrorCode        &status);
 194
 195     /**
 196      * Creating a <tt>StringSearch</tt> instance using the argument locale
 197      * language rule set. A collator will be created in the process, which
 198      * will be owned by this instance and will be deleted during
 199      * destruction
 200      * <p>
 201      * Note: No parsing of the text within the <tt>CharacterIterator</tt>
 202      * will be done during searching for this version. The block of text
 203      * in <tt>CharacterIterator</tt> will be used as it is.
 204      * @param pattern The text for which this object will search.
 205      * @param text    The text iterator in which to search for the pattern.
 206      * @param locale  A locale which defines the language-sensitive
 207      *                comparison rules used to determine whether text in the
 208      *                pattern and target matches. User is responsible for
 209      *                the clearing of this object.
 210      * @param breakiter A <tt>BreakIterator</tt> object used to constrain
 211      *                the matches that are found. Matches whose start and end
 212      *                indices in the target text are not boundaries as
 213      *                determined by the <tt>BreakIterator</tt> are
 214      *                ignored. If this behavior is not desired,
 215      *                <tt>NULL</tt> can be passed in instead.
 216      * @param status for errors if any. If either the length of pattern or
 217      *               text is 0 then an U_ILLEGAL_ARGUMENT_ERROR is returned.
 218      * @stable ICU 2.0
 219      */
 220     StringSearch(const UnicodeString &pattern, CharacterIterator &text,
 221                  const Locale        &locale,
 222                        BreakIterator *breakiter,
 223                        UErrorCode    &status);
 224
 225     /**
 226      * Creating a <tt>StringSearch</tt> instance using the argument collator
 227      * language rule set. Note, user retains the ownership of this collator,
 228      * it does not get destroyed during this instance's destruction.
 229      * <p>
 230      * Note: No parsing of the text within the <tt>CharacterIterator</tt>
 231      * will be done during searching for this version. The block of text
 232      * in <tt>CharacterIterator</tt> will be used as it is.
 233      * @param pattern The text for which this object will search.
 234      * @param text    The text in which to search for the pattern.
 235      * @param coll    A <tt>RuleBasedCollator</tt> object which defines
 236      *                the language-sensitive comparison rules used to
 237      *                determine whether text in the pattern and target
 238      *                matches. User is responsible for the clearing of this
 239      *                object.
 240      * @param breakiter A <tt>BreakIterator</tt> object used to constrain
 241      *                the matches that are found. Matches whose start and end
 242      *                indices in the target text are not boundaries as
 243      *                determined by the <tt>BreakIterator</tt> are
 244      *                ignored. If this behavior is not desired,
 245      *                <tt>NULL</tt> can be passed in instead.
 246      * @param status for errors if any. If either the length of pattern or
 247      *               text is 0 then an U_ILLEGAL_ARGUMENT_ERROR is returned.
 248      * @stable ICU 2.0
 249      */
 250     StringSearch(const UnicodeString     &pattern, CharacterIterator &text,
 251                        RuleBasedCollator *coll,
 252                        BreakIterator     *breakiter,
 253                        UErrorCode        &status);
 254
 255     /**
 256      * Copy constructor that creates a StringSearch instance with the same
 257      * behavior, and iterating over the same text.
 258      * @param that StringSearch instance to be copied.
 259      * @stable ICU 2.0
 260      */
 261     StringSearch(const StringSearch &that);
 262
 263     /**
 264     * Destructor. Cleans up the search iterator data struct.
 265     * If a collator is created in the constructor, it will be destroyed here.
 266     * @stable ICU 2.0
 267     */
 268     virtual ~StringSearch(void);
 269
 270     /**
 271      * Clone this object.
 272      * Clones can be used concurrently in multiple threads.
 273      * If an error occurs, then NULL is returned.
 274      * The caller must delete the clone.
 275      *
 276      * @return a clone of this object
 277      *
 278      * @see getDynamicClassID
 279      * @stable ICU 2.8
 280      */
 281     StringSearch *clone() const;
 282
 283     // operator overloading ---------------------------------------------
 284
 285     /**
 286      * Assignment operator. Sets this iterator to have the same behavior,
 287      * and iterate over the same text, as the one passed in.
 288      * @param that instance to be copied.
 289      * @stable ICU 2.0
 290      */
 291     StringSearch & operator=(const StringSearch &that);
 292
 293     /**
 294      * Equality operator.
 295      * @param that instance to be compared.
 296      * @return TRUE if both instances have the same attributes,
 297      *         breakiterators, collators and iterate over the same text
 298      *         while looking for the same pattern.
 299      * @stable ICU 2.0
 300      */
 301     virtual UBool operator==(const SearchIterator &that) const;
 302
 303     // public get and set methods ----------------------------------------
 304
 305     /**
 306      * Sets the index to point to the given position, and clears any state
 307      * that's affected.
 308      * <p>
 309      * This method takes the argument index and sets the position in the text
 310      * string accordingly without checking if the index is pointing to a
 311      * valid starting point to begin searching.
 312      * @param position within the text to be set. If position is less
 313      *          than or greater than the text range for searching,
 314      *          an U_INDEX_OUTOFBOUNDS_ERROR will be returned
 315      * @param status for errors if it occurs
 316      * @stable ICU 2.0
 317      */
 318     virtual void setOffset(int32_t position, UErrorCode &status);
 319
 320     /**
 321      * Return the current index in the text being searched.
 322      * If the iteration has gone past the end of the text
 323      * (or past the beginning for a backwards search), USEARCH_DONE
 324      * is returned.
 325      * @return current index in the text being searched.
 326      * @stable ICU 2.0
 327      */
 328     virtual int32_t getOffset(void) const;
 329
 330     /**
 331      * Set the target text to be searched.
 332      * Text iteration will hence begin at the start of the text string.
 333      * This method is
 334      * useful if you want to re-use an iterator to search for the same
 335      * pattern within a different body of text.
 336      * @param text text string to be searched
 337      * @param status for errors if any. If the text length is 0 then an
 338      *        U_ILLEGAL_ARGUMENT_ERROR is returned.
 339      * @stable ICU 2.0
 340      */
 341     virtual void setText(const UnicodeString &text, UErrorCode &status);
 342
 343     /**
 344      * Set the target text to be searched.
 345      * Text iteration will hence begin at the start of the text string.
 346      * This method is
 347      * useful if you want to re-use an iterator to search for the same
 348      * pattern within a different body of text.
 349      * Note: No parsing of the text within the <tt>CharacterIterator</tt>
 350      * will be done during searching for this version. The block of text
 351      * in <tt>CharacterIterator</tt> will be used as it is.
 352      * @param text text string to be searched
 353      * @param status for errors if any. If the text length is 0 then an
 354      *        U_ILLEGAL_ARGUMENT_ERROR is returned.
 355      * @stable ICU 2.0
 356      */
 357     virtual void setText(CharacterIterator &text, UErrorCode &status);
 358
 359     /**
 360      * Gets the collator used for the language rules.
 361      * <p>
 362      * Caller may modify but <b>must not</b> delete the <tt>RuleBasedCollator</tt>!
 363      * Modifications to this collator will affect the original collator passed in to
 364      * the <tt>StringSearch></tt> constructor or to setCollator, if any.
 365      * @return collator used for string search
 366      * @stable ICU 2.0
 367      */
 368     RuleBasedCollator * getCollator() const;
 369
 370     /**
 371      * Sets the collator used for the language rules. User retains the
 372      * ownership of this collator, thus the responsibility of deletion lies
 373      * with the user. This method causes internal data such as Boyer-Moore
 374      * shift tables to be recalculated, but the iterator's position is
 375      * unchanged.
 376      * @param coll    collator
 377      * @param status  for errors if any
 378      * @stable ICU 2.0
 379      */
 380     void setCollator(RuleBasedCollator *coll, UErrorCode &status);
 381
 382     /**
 383      * Sets the pattern used for matching.
 384      * Internal data like the Boyer Moore table will be recalculated, but
 385      * the iterator's position is unchanged.
 386      * @param pattern search pattern to be found
 387      * @param status for errors if any. If the pattern length is 0 then an
 388      *               U_ILLEGAL_ARGUMENT_ERROR is returned.
 389      * @stable ICU 2.0
 390      */
 391     void setPattern(const UnicodeString &pattern, UErrorCode &status);
 392
 393     /**
 394      * Gets the search pattern.
 395      * @return pattern used for matching
 396      * @stable ICU 2.0
 397      */
 398     const UnicodeString & getPattern() const;
 399
 400     // public methods ----------------------------------------------------
 401
 402     /**
 403      * Reset the iteration.
 404      * Search will begin at the start of the text string if a forward
 405      * iteration is initiated before a backwards iteration. Otherwise if
 406      * a backwards iteration is initiated before a forwards iteration, the
 407      * search will begin at the end of the text string.
 408      * @stable ICU 2.0
 409      */
 410     virtual void reset();
 411
 412     /**
 413      * Returns a copy of StringSearch with the same behavior, and
 414      * iterating over the same text, as this one. Note that all data will be
 415      * replicated, except for the user-specified collator and the
 416      * breakiterator.
 417      * @return cloned object
 418      * @stable ICU 2.0
 419      */
 420     virtual SearchIterator * safeClone(void) const;
 421
 422     /**
 423      * ICU "poor man's RTTI", returns a UClassID for the actual class.
 424      *
 425      * @stable ICU 2.2
 426      */
 427     virtual UClassID getDynamicClassID() const;
 428
 429     /**
 430      * ICU "poor man's RTTI", returns a UClassID for this class.
 431      *
 432      * @stable ICU 2.2
 433      */
 434     static UClassID U_EXPORT2 getStaticClassID();
 435
 436 protected:
 437
 438     // protected method -------------------------------------------------
 439
 440     /**
 441      * Search forward for matching text, starting at a given location.
 442      * Clients should not call this method directly; instead they should
 443      * call {@link SearchIterator#next }.
 444      * <p>
 445      * If a match is found, this method returns the index at which the match
 446      * starts and calls {@link SearchIterator#setMatchLength } with the number
 447      * of characters in the target text that make up the match. If no match
 448      * is found, the method returns <tt>USEARCH_DONE</tt>.
 449      * <p>
 450      * The <tt>StringSearch</tt> is adjusted so that its current index
 451      * (as returned by {@link #getOffset }) is the match position if one was
 452      * found.
 453      * If a match is not found, <tt>USEARCH_DONE</tt> will be returned and
 454      * the <tt>StringSearch</tt> will be adjusted to the index USEARCH_DONE.
 455      * @param position The index in the target text at which the search
 456      *                 starts
 457      * @param status for errors if any occurs
 458      * @return The index at which the matched text in the target starts, or
 459      *         USEARCH_DONE if no match was found.
 460      * @stable ICU 2.0
 461      */
 462     virtual int32_t handleNext(int32_t position, UErrorCode &status);
 463
 464     /**
 465      * Search backward for matching text, starting at a given location.
 466      * Clients should not call this method directly; instead they should call
 467      * <tt>SearchIterator.previous()</tt>, which this method overrides.
 468      * <p>
 469      * If a match is found, this method returns the index at which the match
 470      * starts and calls {@link SearchIterator#setMatchLength } with the number
 471      * of characters in the target text that make up the match. If no match
 472      * is found, the method returns <tt>USEARCH_DONE</tt>.
 473      * <p>
 474      * The <tt>StringSearch</tt> is adjusted so that its current index
 475      * (as returned by {@link #getOffset }) is the match position if one was
 476      * found.
 477      * If a match is not found, <tt>USEARCH_DONE</tt> will be returned and
 478      * the <tt>StringSearch</tt> will be adjusted to the index USEARCH_DONE.
 479      * @param position The index in the target text at which the search
 480      *                 starts.
 481      * @param status for errors if any occurs
 482      * @return The index at which the matched text in the target starts, or
 483      *         USEARCH_DONE if no match was found.
 484      * @stable ICU 2.0
 485      */
 486     virtual int32_t handlePrev(int32_t position, UErrorCode &status);
 487
 488 private :
 489     StringSearch(); // default constructor not implemented
 490
 491     // private data members ----------------------------------------------
 492
 493     /**
 494     * RuleBasedCollator, contains exactly the same UCollator * in m_strsrch_
 495     * @stable ICU 2.0
 496     */
 497     RuleBasedCollator  m_collator_;
 498     /**
 499     * Pattern text
 500     * @stable ICU 2.0
 501     */
 502     UnicodeString      m_pattern_;
 503     /**
 504     * String search struct data
 505     * @stable ICU 2.0
 506     */
 507     UStringSearch     *m_strsrch_;
 508
 509 };
 510
 511 U_NAMESPACE_END
 512
 513 #endif /* #if !UCONFIG_NO_COLLATION */
 514
 515 #endif
 516