icuSources/common/unicode/ubiditransform.h

   1 /*
   2 ******************************************************************************
   3 *
   4 * © 2016 and later: Unicode, Inc. and others.
   5 * License & terms of use: http://www.unicode.org/copyright.html
   6 *
   7 ******************************************************************************
   8 *   file name:  ubiditransform.h
   9 *   encoding:   UTF-8
  10 *   tab size:   8 (not used)
  11 *   indentation:4
  12 *
  13 *   created on: 2016jul24
  14 *   created by: Lina Kemmel
  15 *
  16 */
  17
  18 #ifndef UBIDITRANSFORM_H
  19 #define UBIDITRANSFORM_H
  20
  21 #include "unicode/utypes.h"
  22 #include "unicode/ubidi.h"
  23 #include "unicode/uchar.h"
  24 #include "unicode/localpointer.h"
  25
  26 /**
  27  * \file
  28  * \brief Bidi Transformations
  29  */
  30
  31 /**
  32  * `UBiDiOrder` indicates the order of text.
  33  *
  34  * This bidi transformation engine supports all possible combinations (4 in
  35  * total) of input and output text order:
  36  *
  37  *   - <logical input, visual output>: unless the output direction is RTL, this
  38  *     corresponds to a normal operation of the Bidi algorithm as described in the
  39  *     Unicode Technical Report and implemented by `UBiDi` when the
  40  *     reordering mode is set to `UBIDI_REORDER_DEFAULT`. Visual RTL
  41  *     mode is not supported by `UBiDi` and is accomplished through
  42  *     reversing a visual LTR string,
  43  *
  44  *   - <visual input, logical output>: unless the input direction is RTL, this
  45  *     corresponds to an "inverse bidi algorithm" in `UBiDi` with the
  46  *     reordering mode set to `UBIDI_REORDER_INVERSE_LIKE_DIRECT`.
  47  *     Visual RTL mode is not not supported by `UBiDi` and is
  48  *     accomplished through reversing a visual LTR string,
  49  *
  50  *   - <logical input, logical output>: if the input and output base directions
  51  *     mismatch, this corresponds to the `UBiDi` implementation with the
  52  *     reordering mode set to `UBIDI_REORDER_RUNS_ONLY`; and if the
  53  *     input and output base directions are identical, the transformation engine
  54  *     will only handle character mirroring and Arabic shaping operations without
  55  *     reordering,
  56  *
  57  *   - <visual input, visual output>: this reordering mode is not supported by
  58  *     the `UBiDi` engine; it implies character mirroring, Arabic
  59  *     shaping, and - if the input/output base directions mismatch -  string
  60  *     reverse operations.
  61  * @see ubidi_setInverse
  62  * @see ubidi_setReorderingMode
  63  * @see UBIDI_REORDER_DEFAULT
  64  * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT
  65  * @see UBIDI_REORDER_RUNS_ONLY
  66  * @stable ICU 58
  67  */
  68 typedef enum {
  69     /** 0: Constant indicating a logical order.
  70       * This is the default for input text.
  71       * @stable ICU 58
  72       */
  73     UBIDI_LOGICAL = 0,
  74     /** 1: Constant indicating a visual order.
  75       * This is a default for output text.
  76       * @stable ICU 58
  77       */
  78     UBIDI_VISUAL
  79 } UBiDiOrder;
  80
  81 /**
  82  * <code>UBiDiMirroring</code> indicates whether or not characters with the
  83  * "mirrored" property in RTL runs should be replaced with their mirror-image
  84  * counterparts.
  85  * @see UBIDI_DO_MIRRORING
  86  * @see ubidi_setReorderingOptions
  87  * @see ubidi_writeReordered
  88  * @see ubidi_writeReverse
  89  * @stable ICU 58
  90  */
  91 typedef enum {
  92     /** 0: Constant indicating that character mirroring should not be
  93       * performed.
  94       * This is the default.
  95       * @stable ICU 58
  96       */
  97     UBIDI_MIRRORING_OFF = 0,
  98     /** 1: Constant indicating that character mirroring should be performed.
  99       * This corresponds to calling <code>ubidi_writeReordered</code> or
 100       * <code>ubidi_writeReverse</code> with the
 101       * <code>UBIDI_DO_MIRRORING</code> option bit set.
 102       * @stable ICU 58
 103       */
 104     UBIDI_MIRRORING_ON
 105 } UBiDiMirroring;
 106
 107 /**
 108  * Forward declaration of the <code>UBiDiTransform</code> structure that stores
 109  * information used by the layout transformation engine.
 110  * @stable ICU 58
 111  */
 112 typedef struct UBiDiTransform UBiDiTransform;
 113
 114 /**
 115  * Performs transformation of text from the bidi layout defined by the input
 116  * ordering scheme to the bidi layout defined by the output ordering scheme,
 117  * and applies character mirroring and Arabic shaping operations.<p>
 118  * In terms of <code>UBiDi</code>, such a transformation implies:
 119  * <ul>
 120  * <li>calling <code>ubidi_setReorderingMode</code> as needed (when the
 121  * reordering mode is other than normal),</li>
 122  * <li>calling <code>ubidi_setInverse</code> as needed (when text should be
 123  * transformed from a visual to a logical form),</li>
 124  * <li>resolving embedding levels of each character in the input text by
 125  * calling <code>ubidi_setPara</code>,</li>
 126  * <li>reordering the characters based on the computed embedding levels, also
 127  * performing character mirroring as needed, and streaming the result to the
 128  * output, by calling <code>ubidi_writeReordered</code>,</li>
 129  * <li>performing Arabic digit and letter shaping on the output text by calling
 130  * <code>u_shapeArabic</code>.</li>
 131  * </ul>
 132  * An "ordering scheme" encompasses the base direction and the order of text,
 133  * and these characteristics must be defined by the caller for both input and
 134  * output explicitly .<p>
 135  * There are 36 possible combinations of <input, output> ordering schemes,
 136  * which are partially supported by <code>UBiDi</code> already. Examples of the
 137  * currently supported combinations:
 138  * <ul>
 139  * <li><Logical LTR, Visual LTR>: this is equivalent to calling
 140  * <code>ubidi_setPara</code> with <code>paraLevel == UBIDI_LTR</code>,</li>
 141  * <li><Logical RTL, Visual LTR>: this is equivalent to calling
 142  * <code>ubidi_setPara</code> with <code>paraLevel == UBIDI_RTL</code>,</li>
 143  * <li><Logical Default ("Auto") LTR, Visual LTR>: this is equivalent to
 144  * calling <code>ubidi_setPara</code> with
 145  * <code>paraLevel == UBIDI_DEFAULT_LTR</code>,</li>
 146  * <li><Logical Default ("Auto") RTL, Visual LTR>: this is equivalent to
 147  * calling <code>ubidi_setPara</code> with
 148  * <code>paraLevel == UBIDI_DEFAULT_RTL</code>,</li>
 149  * <li><Visual LTR, Logical LTR>: this is equivalent to
 150  * calling <code>ubidi_setInverse(UBiDi*, TRUE)</code> and then
 151  * <code>ubidi_setPara</code> with <code>paraLevel == UBIDI_LTR</code>,</li>
 152  * <li><Visual LTR, Logical RTL>: this is equivalent to
 153  * calling <code>ubidi_setInverse(UBiDi*, TRUE)</code> and then
 154  * <code>ubidi_setPara</code> with <code>paraLevel == UBIDI_RTL</code>.</li>
 155  * </ul>
 156  * All combinations that involve the Visual RTL scheme are unsupported by
 157  * <code>UBiDi</code>, for instance:
 158  * <ul>
 159  * <li><Logical LTR, Visual RTL>,</li>
 160  * <li><Visual RTL, Logical RTL>.</li>
 161  * </ul>
 162  * <p>Example of usage of the transformation engine:<br>
 163  * <pre>
 164  * \code
 165  * UChar text1[] = {'a', 'b', 'c', 0x0625, '1', 0};
 166  * UChar text2[] = {'a', 'b', 'c', 0x0625, '1', 0};
 167  * UErrorCode errorCode = U_ZERO_ERROR;
 168  * // Run a transformation.
 169  * ubiditransform_transform(pBidiTransform,
 170  *          text1, -1, text2, -1,
 171  *          UBIDI_LTR, UBIDI_VISUAL,
 172  *          UBIDI_RTL, UBIDI_LOGICAL,
 173  *          UBIDI_MIRRORING_OFF,
 174  *          U_SHAPE_DIGITS_AN2EN | U_SHAPE_DIGIT_TYPE_AN_EXTENDED,
 175  *          &errorCode);
 176  * // Do something with text2.
 177  *  text2[4] = '2';
 178  * // Run a reverse transformation.
 179  * ubiditransform_transform(pBidiTransform,
 180  *          text2, -1, text1, -1,
 181  *          UBIDI_RTL, UBIDI_LOGICAL,
 182  *          UBIDI_LTR, UBIDI_VISUAL,
 183  *          UBIDI_MIRRORING_OFF,
 184  *          U_SHAPE_DIGITS_EN2AN | U_SHAPE_DIGIT_TYPE_AN_EXTENDED,
 185  *          &errorCode);
 186  *\endcode
 187  * </pre>
 188  * </p>
 189  *
 190  * @param pBiDiTransform A pointer to a <code>UBiDiTransform</code> object
 191  *        allocated with <code>ubiditransform_open()</code> or
 192  *        <code>NULL</code>.<p>
 193  *        This object serves for one-time setup to amortize initialization
 194  *        overheads. Use of this object is not thread-safe. All other threads
 195  *        should allocate a new <code>UBiDiTransform</code> object by calling
 196  *        <code>ubiditransform_open()</code> before using it. Alternatively,
 197  *        a caller can set this parameter to <code>NULL</code>, in which case
 198  *        the object will be allocated by the engine on the fly.</p>
 199  * @param src A pointer to the text that the Bidi layout transformations will
 200  *        be performed on.
 201  *        <p><strong>Note:</strong> the text must be (at least)
 202  *        <code>srcLength</code> long.</p>
 203  * @param srcLength The length of the text, in number of UChars. If
 204  *        <code>length == -1</code> then the text must be zero-terminated.
 205  * @param dest A pointer to where the processed text is to be copied.
 206  * @param destSize The size of the <code>dest</code> buffer, in number of
 207  *        UChars. If the <code>U_SHAPE_LETTERS_UNSHAPE</code> option is set,
 208  *        then the destination length could be as large as
 209  *        <code>srcLength * 2</code>. Otherwise, the destination length will
 210  *        not exceed <code>srcLength</code>. If the caller reserves the last
 211  *        position for zero-termination, it should be excluded from
 212  *        <code>destSize</code>.
 213  *        <p><code>destSize == -1</code> is allowed and makes sense when
 214  *        <code>dest</code> was holds some meaningful value, e.g. that of
 215  *        <code>src</code>. In this case <code>dest</code> must be
 216  *        zero-terminated.</p>
 217  * @param inParaLevel A base embedding level of the input as defined in
 218  *        <code>ubidi_setPara</code> documentation for the
 219  *        <code>paraLevel</code> parameter.
 220  * @param inOrder An order of the input, which can be one of the
 221  *        <code>UBiDiOrder</code> values.
 222  * @param outParaLevel A base embedding level of the output as defined in
 223  *        <code>ubidi_setPara</code> documentation for the
 224  *        <code>paraLevel</code> parameter.
 225  * @param outOrder An order of the output, which can be one of the
 226  *        <code>UBiDiOrder</code> values.
 227  * @param doMirroring Indicates whether or not to perform character mirroring,
 228  *        and can accept one of the <code>UBiDiMirroring</code> values.
 229  * @param shapingOptions Arabic digit and letter shaping options defined in the
 230  *        ushape.h documentation.
 231  *        <p><strong>Note:</strong> Direction indicator options are computed by
 232  *        the transformation engine based on the effective ordering schemes, so
 233  *        user-defined direction indicators will be ignored.</p>
 234  * @param pErrorCode A pointer to an error code value.
 235  *
 236  * @return The destination length, i.e. the number of UChars written to
 237  *         <code>dest</code>. If the transformation fails, the return value
 238  *         will be 0 (and the error code will be written to
 239  *         <code>pErrorCode</code>).
 240  *
 241  * @see UBiDiLevel
 242  * @see UBiDiOrder
 243  * @see UBiDiMirroring
 244  * @see ubidi_setPara
 245  * @see u_shapeArabic
 246  * @stable ICU 58
 247  */
 248 U_STABLE uint32_t U_EXPORT2
 249 ubiditransform_transform(UBiDiTransform *pBiDiTransform,
 250             const UChar *src, int32_t srcLength,
 251             UChar *dest, int32_t destSize,
 252             UBiDiLevel inParaLevel, UBiDiOrder inOrder,
 253             UBiDiLevel outParaLevel, UBiDiOrder outOrder,
 254             UBiDiMirroring doMirroring, uint32_t shapingOptions,
 255             UErrorCode *pErrorCode);
 256
 257 /**
 258  * Allocates a <code>UBiDiTransform</code> object. This object can be reused,
 259  * e.g. with different ordering schemes, mirroring or shaping options.<p>
 260  * <strong>Note:</strong>The object can only be reused in the same thread.
 261  * All other threads should allocate a new <code>UBiDiTransform</code> object
 262  * before using it.<p>
 263  * Example of usage:<p>
 264  * <pre>
 265  * \code
 266  * UErrorCode errorCode = U_ZERO_ERROR;
 267  * // Open a new UBiDiTransform.
 268  * UBiDiTransform* transform = ubiditransform_open(&errorCode);
 269  * // Run a transformation.
 270  * ubiditransform_transform(transform,
 271  *          text1, -1, text2, -1,
 272  *          UBIDI_RTL, UBIDI_LOGICAL,
 273  *          UBIDI_LTR, UBIDI_VISUAL,
 274  *          UBIDI_MIRRORING_ON,
 275  *          U_SHAPE_DIGITS_EN2AN,
 276  *          &errorCode);
 277  * // Do something with the output text and invoke another transformation using
 278  * //   that text as input.
 279  * ubiditransform_transform(transform,
 280  *          text2, -1, text3, -1,
 281  *          UBIDI_LTR, UBIDI_VISUAL,
 282  *          UBIDI_RTL, UBIDI_VISUAL,
 283  *          UBIDI_MIRRORING_ON,
 284  *          0, &errorCode);
 285  *\endcode
 286  * </pre>
 287  * <p>
 288  * The <code>UBiDiTransform</code> object must be deallocated by calling
 289  * <code>ubiditransform_close()</code>.
 290  *
 291  * @return An empty <code>UBiDiTransform</code> object.
 292  * @stable ICU 58
 293  */
 294 U_STABLE UBiDiTransform* U_EXPORT2
 295 ubiditransform_open(UErrorCode *pErrorCode);
 296
 297 /**
 298  * Deallocates the given <code>UBiDiTransform</code> object.
 299  * @stable ICU 58
 300  */
 301 U_STABLE void U_EXPORT2
 302 ubiditransform_close(UBiDiTransform *pBidiTransform);
 303
 304 #if U_SHOW_CPLUSPLUS_API
 305
 306 U_NAMESPACE_BEGIN
 307
 308 /**
 309  * \class LocalUBiDiTransformPointer
 310  * "Smart pointer" class, closes a UBiDiTransform via ubiditransform_close().
 311  * For most methods see the LocalPointerBase base class.
 312  *
 313  * @see LocalPointerBase
 314  * @see LocalPointer
 315  * @stable ICU 58
 316  */
 317 U_DEFINE_LOCAL_OPEN_POINTER(LocalUBiDiTransformPointer, UBiDiTransform, ubiditransform_close);
 318
 319 U_NAMESPACE_END
 320
 321 #endif // U_SHOW_CPLUSPLUS_API
 322
 323 #endif