diff options
Diffstat (limited to 'ext/intl/breakiterator')
-rw-r--r-- | ext/intl/breakiterator/breakiterator_class.cpp | 397 | ||||
-rw-r--r-- | ext/intl/breakiterator/breakiterator_class.h | 71 | ||||
-rw-r--r-- | ext/intl/breakiterator/breakiterator_iterators.cpp | 346 | ||||
-rw-r--r-- | ext/intl/breakiterator/breakiterator_iterators.h | 42 | ||||
-rw-r--r-- | ext/intl/breakiterator/breakiterator_methods.cpp | 452 | ||||
-rw-r--r-- | ext/intl/breakiterator/breakiterator_methods.h | 64 | ||||
-rw-r--r-- | ext/intl/breakiterator/codepointiterator_internal.cpp | 291 | ||||
-rw-r--r-- | ext/intl/breakiterator/codepointiterator_internal.h | 98 | ||||
-rw-r--r-- | ext/intl/breakiterator/codepointiterator_methods.cpp | 44 | ||||
-rw-r--r-- | ext/intl/breakiterator/codepointiterator_methods.h | 24 | ||||
-rw-r--r-- | ext/intl/breakiterator/rulebasedbreakiterator_methods.cpp | 221 | ||||
-rw-r--r-- | ext/intl/breakiterator/rulebasedbreakiterator_methods.h | 32 |
12 files changed, 2082 insertions, 0 deletions
diff --git a/ext/intl/breakiterator/breakiterator_class.cpp b/ext/intl/breakiterator/breakiterator_class.cpp new file mode 100644 index 0000000000..7bf271a344 --- /dev/null +++ b/ext/intl/breakiterator/breakiterator_class.cpp @@ -0,0 +1,397 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ +*/ + +#ifdef HAVE_CONFIG_H +#include "config.h" +#endif + +#include <unicode/brkiter.h> +#include <unicode/rbbi.h> +#include "codepointiterator_internal.h" + +#include "breakiterator_iterators.h" + +#include <typeinfo> + +extern "C" { +#define USE_BREAKITERATOR_POINTER 1 +#include "breakiterator_class.h" +#include "breakiterator_methods.h" +#include "rulebasedbreakiterator_methods.h" +#include "codepointiterator_methods.h" +#include <zend_exceptions.h> +#include <zend_interfaces.h> +#include <assert.h> +} + +using PHP::CodePointBreakIterator; + +/* {{{ Global variables */ +zend_class_entry *BreakIterator_ce_ptr; +zend_class_entry *RuleBasedBreakIterator_ce_ptr; +zend_class_entry *CodePointBreakIterator_ce_ptr; +zend_object_handlers BreakIterator_handlers; +/* }}} */ + +U_CFUNC void breakiterator_object_create(zval *object, + BreakIterator *biter TSRMLS_DC) +{ + UClassID classId = biter->getDynamicClassID(); + zend_class_entry *ce; + + if (classId == RuleBasedBreakIterator::getStaticClassID()) { + ce = RuleBasedBreakIterator_ce_ptr; + } else if (classId == CodePointBreakIterator::getStaticClassID()) { + ce = CodePointBreakIterator_ce_ptr; + } else { + ce = BreakIterator_ce_ptr; + } + + object_init_ex(object, ce); + breakiterator_object_construct(object, biter TSRMLS_CC); +} + +U_CFUNC void breakiterator_object_construct(zval *object, + BreakIterator *biter TSRMLS_DC) +{ + BreakIterator_object *bio; + + BREAKITER_METHOD_FETCH_OBJECT_NO_CHECK; //populate to from object + assert(bio->biter == NULL); + bio->biter = biter; +} + +/* {{{ compare handler for BreakIterator */ +static int BreakIterator_compare_objects(zval *object1, + zval *object2 TSRMLS_DC) +{ + BreakIterator_object *bio1, + *bio2; + + bio1 = (BreakIterator_object*)zend_object_store_get_object(object1 TSRMLS_CC); + bio2 = (BreakIterator_object*)zend_object_store_get_object(object2 TSRMLS_CC); + + if (bio1->biter == NULL || bio2->biter == NULL) { + return bio1->biter == bio2->biter ? 0 : 1; + } + + return *bio1->biter == *bio2->biter ? 0 : 1; +} +/* }}} */ + +/* {{{ clone handler for BreakIterator */ +static zend_object_value BreakIterator_clone_obj(zval *object TSRMLS_DC) +{ + BreakIterator_object *bio_orig, + *bio_new; + zend_object_value ret_val; + + bio_orig = (BreakIterator_object*)zend_object_store_get_object(object TSRMLS_CC); + intl_errors_reset(INTL_DATA_ERROR_P(bio_orig) TSRMLS_CC); + + ret_val = BreakIterator_ce_ptr->create_object(Z_OBJCE_P(object) TSRMLS_CC); + bio_new = (BreakIterator_object*)zend_object_store_get_object_by_handle( + ret_val.handle TSRMLS_CC); + + zend_objects_clone_members(&bio_new->zo, ret_val, + &bio_orig->zo, Z_OBJ_HANDLE_P(object) TSRMLS_CC); + + if (bio_orig->biter != NULL) { + BreakIterator *new_biter; + + new_biter = bio_orig->biter->clone(); + if (!new_biter) { + char *err_msg; + intl_errors_set_code(BREAKITER_ERROR_P(bio_orig), + U_MEMORY_ALLOCATION_ERROR TSRMLS_CC); + intl_errors_set_custom_msg(BREAKITER_ERROR_P(bio_orig), + "Could not clone BreakIterator", 0 TSRMLS_CC); + err_msg = intl_error_get_message(BREAKITER_ERROR_P(bio_orig) TSRMLS_CC); + zend_throw_exception(NULL, err_msg, 0 TSRMLS_CC); + efree(err_msg); + } else { + bio_new->biter = new_biter; + bio_new->text = bio_orig->text; + if (bio_new->text) { + zval_add_ref(&bio_new->text); + } + } + } else { + zend_throw_exception(NULL, "Cannot clone unconstructed BreakIterator", 0 TSRMLS_CC); + } + + return ret_val; +} +/* }}} */ + +/* {{{ get_debug_info handler for BreakIterator */ +static HashTable *BreakIterator_get_debug_info(zval *object, int *is_temp TSRMLS_DC) +{ + zval zv = zval_used_for_init; + BreakIterator_object *bio; + const BreakIterator *biter; + + *is_temp = 1; + + array_init_size(&zv, 8); + + bio = (BreakIterator_object*)zend_object_store_get_object(object TSRMLS_CC); + biter = bio->biter; + + if (biter == NULL) { + add_assoc_bool_ex(&zv, "valid", sizeof("valid"), 0); + return Z_ARRVAL(zv); + } + add_assoc_bool_ex(&zv, "valid", sizeof("valid"), 1); + + if (bio->text == NULL) { + add_assoc_null_ex(&zv, "text", sizeof("text")); + } else { + zval_add_ref(&bio->text); + add_assoc_zval_ex(&zv, "text", sizeof("text"), bio->text); + } + + add_assoc_string_ex(&zv, "type", sizeof("type"), + const_cast<char*>(typeid(*biter).name()), 1); + + return Z_ARRVAL(zv); +} +/* }}} */ + +/* {{{ void breakiterator_object_init(BreakIterator_object* to) + * Initialize internals of BreakIterator_object not specific to zend standard objects. + */ +static void breakiterator_object_init(BreakIterator_object *bio TSRMLS_DC) +{ + intl_error_init(BREAKITER_ERROR_P(bio) TSRMLS_CC); + bio->biter = NULL; + bio->text = NULL; +} +/* }}} */ + +/* {{{ BreakIterator_objects_dtor */ +static void BreakIterator_objects_dtor(void *object, + zend_object_handle handle TSRMLS_DC) +{ + zend_objects_destroy_object((zend_object*)object, handle TSRMLS_CC); +} +/* }}} */ + +/* {{{ BreakIterator_objects_free */ +static void BreakIterator_objects_free(zend_object *object TSRMLS_DC) +{ + BreakIterator_object* bio = (BreakIterator_object*) object; + + if (bio->text) { + zval_ptr_dtor(&bio->text); + } + if (bio->biter) { + delete bio->biter; + bio->biter = NULL; + } + intl_error_reset(BREAKITER_ERROR_P(bio) TSRMLS_CC); + + zend_object_std_dtor(&bio->zo TSRMLS_CC); + + efree(bio); +} +/* }}} */ + +/* {{{ BreakIterator_object_create */ +static zend_object_value BreakIterator_object_create(zend_class_entry *ce TSRMLS_DC) +{ + zend_object_value retval; + BreakIterator_object* intern; + + intern = (BreakIterator_object*)ecalloc(1, sizeof(BreakIterator_object)); + + zend_object_std_init(&intern->zo, ce TSRMLS_CC); +#if PHP_VERSION_ID < 50399 + zend_hash_copy(intern->zo.properties, &(ce->default_properties), + (copy_ctor_func_t) zval_add_ref, NULL, sizeof(zval*)); +#else + object_properties_init((zend_object*) intern, ce); +#endif + breakiterator_object_init(intern TSRMLS_CC); + + retval.handle = zend_objects_store_put( + intern, + BreakIterator_objects_dtor, + (zend_objects_free_object_storage_t) BreakIterator_objects_free, + NULL TSRMLS_CC); + + retval.handlers = &BreakIterator_handlers; + + return retval; +} +/* }}} */ + +/* {{{ BreakIterator/RuleBasedBreakIterator methods arguments info */ + +ZEND_BEGIN_ARG_INFO_EX(ainfo_biter_void, 0, 0, 0) +ZEND_END_ARG_INFO() + +ZEND_BEGIN_ARG_INFO_EX(ainfo_biter_locale, 0, 0, 0) + ZEND_ARG_INFO(0, "locale") +ZEND_END_ARG_INFO() + +ZEND_BEGIN_ARG_INFO_EX(ainfo_biter_setText, 0, 0, 1) + ZEND_ARG_INFO(0, "text") +ZEND_END_ARG_INFO() + +ZEND_BEGIN_ARG_INFO_EX(ainfo_biter_next, 0, 0, 0) + ZEND_ARG_INFO(0, "offset") +ZEND_END_ARG_INFO() + +ZEND_BEGIN_ARG_INFO_EX(ainfo_biter_offset, 0, 0, 1) + ZEND_ARG_INFO(0, "offset") +ZEND_END_ARG_INFO() + +ZEND_BEGIN_ARG_INFO_EX(ainfo_biter_get_locale, 0, 0, 1) + ZEND_ARG_INFO(0, "locale_type") +ZEND_END_ARG_INFO() + +ZEND_BEGIN_ARG_INFO_EX(ainfo_biter_getPartsIterator, 0, 0, 0) + ZEND_ARG_INFO(0, "key_type") +ZEND_END_ARG_INFO() + +ZEND_BEGIN_ARG_INFO_EX(ainfo_rbbi___construct, 0, 0, 1) + ZEND_ARG_INFO(0, "rules") + ZEND_ARG_INFO(0, "areCompiled") +ZEND_END_ARG_INFO() + +/* }}} */ + +/* {{{ BreakIterator_class_functions + * Every 'BreakIterator' class method has an entry in this table + */ +static const zend_function_entry BreakIterator_class_functions[] = { + PHP_ME(BreakIterator, __construct, ainfo_biter_void, ZEND_ACC_PRIVATE) + PHP_ME_MAPPING(createWordInstance, breakiter_create_word_instance, ainfo_biter_locale, ZEND_ACC_STATIC | ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(createLineInstance, breakiter_create_line_instance, ainfo_biter_locale, ZEND_ACC_STATIC | ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(createCharacterInstance, breakiter_create_character_instance, ainfo_biter_locale, ZEND_ACC_STATIC | ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(createSentenceInstance, breakiter_create_sentence_instance, ainfo_biter_locale, ZEND_ACC_STATIC | ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(createTitleInstance, breakiter_create_title_instance, ainfo_biter_locale, ZEND_ACC_STATIC | ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(createCodePointInstance, breakiter_create_code_point_instance, ainfo_biter_void, ZEND_ACC_STATIC | ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(getText, breakiter_get_text, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(setText, breakiter_set_text, ainfo_biter_setText, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(first, breakiter_first, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(last, breakiter_last, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(previous, breakiter_previous, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(next, breakiter_next, ainfo_biter_next, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(current, breakiter_current, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(following, breakiter_following, ainfo_biter_offset, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(preceding, breakiter_preceding, ainfo_biter_offset, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(isBoundary, breakiter_is_boundary, ainfo_biter_offset, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(getLocale, breakiter_get_locale, ainfo_biter_get_locale, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(getPartsIterator, breakiter_get_parts_iterator, ainfo_biter_getPartsIterator, ZEND_ACC_PUBLIC) + + PHP_ME_MAPPING(getErrorCode, breakiter_get_error_code, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(getErrorMessage, breakiter_get_error_message, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_FE_END +}; +/* }}} */ + +/* {{{ RuleBasedBreakIterator_class_functions + */ +static const zend_function_entry RuleBasedBreakIterator_class_functions[] = { + PHP_ME(IntlRuleBasedBreakIterator, __construct, ainfo_rbbi___construct, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(getRules, rbbi_get_rules, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(getRuleStatus, rbbi_get_rule_status, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_ME_MAPPING(getRuleStatusVec, rbbi_get_rule_status_vec, ainfo_biter_void, ZEND_ACC_PUBLIC) +#if U_ICU_VERSION_MAJOR_NUM * 10 + U_ICU_VERSION_MINOR_NUM >= 48 + PHP_ME_MAPPING(getBinaryRules, rbbi_get_binary_rules, ainfo_biter_void, ZEND_ACC_PUBLIC) +#endif + PHP_FE_END +}; +/* }}} */ + +/* {{{ CodePointBreakIterator_class_functions + */ +static const zend_function_entry CodePointBreakIterator_class_functions[] = { + PHP_ME_MAPPING(getLastCodePoint, cpbi_get_last_code_point, ainfo_biter_void, ZEND_ACC_PUBLIC) + PHP_FE_END +}; +/* }}} */ + + +/* {{{ breakiterator_register_BreakIterator_class + * Initialize 'BreakIterator' class + */ +U_CFUNC void breakiterator_register_BreakIterator_class(TSRMLS_D) +{ + zend_class_entry ce; + + /* Create and register 'BreakIterator' class. */ + INIT_CLASS_ENTRY(ce, "IntlBreakIterator", BreakIterator_class_functions); + ce.create_object = BreakIterator_object_create; + ce.get_iterator = _breakiterator_get_iterator; + BreakIterator_ce_ptr = zend_register_internal_class(&ce TSRMLS_CC); + + memcpy(&BreakIterator_handlers, zend_get_std_object_handlers(), + sizeof BreakIterator_handlers); + BreakIterator_handlers.compare_objects = BreakIterator_compare_objects; + BreakIterator_handlers.clone_obj = BreakIterator_clone_obj; + BreakIterator_handlers.get_debug_info = BreakIterator_get_debug_info; + + zend_class_implements(BreakIterator_ce_ptr TSRMLS_CC, 1, + zend_ce_traversable); + + zend_declare_class_constant_long(BreakIterator_ce_ptr, + "DONE", sizeof("DONE") - 1, BreakIterator::DONE TSRMLS_CC ); + + /* Declare constants that are defined in the C header */ +#define BREAKITER_DECL_LONG_CONST(name) \ + zend_declare_class_constant_long(BreakIterator_ce_ptr, #name, \ + sizeof(#name) - 1, UBRK_ ## name TSRMLS_CC) + + BREAKITER_DECL_LONG_CONST(WORD_NONE); + BREAKITER_DECL_LONG_CONST(WORD_NONE_LIMIT); + BREAKITER_DECL_LONG_CONST(WORD_NUMBER); + BREAKITER_DECL_LONG_CONST(WORD_NUMBER_LIMIT); + BREAKITER_DECL_LONG_CONST(WORD_LETTER); + BREAKITER_DECL_LONG_CONST(WORD_LETTER_LIMIT); + BREAKITER_DECL_LONG_CONST(WORD_KANA); + BREAKITER_DECL_LONG_CONST(WORD_KANA_LIMIT); + BREAKITER_DECL_LONG_CONST(WORD_IDEO); + BREAKITER_DECL_LONG_CONST(WORD_IDEO_LIMIT); + + BREAKITER_DECL_LONG_CONST(LINE_SOFT); + BREAKITER_DECL_LONG_CONST(LINE_SOFT_LIMIT); + BREAKITER_DECL_LONG_CONST(LINE_HARD); + BREAKITER_DECL_LONG_CONST(LINE_HARD_LIMIT); + + BREAKITER_DECL_LONG_CONST(SENTENCE_TERM); + BREAKITER_DECL_LONG_CONST(SENTENCE_TERM_LIMIT); + BREAKITER_DECL_LONG_CONST(SENTENCE_SEP); + BREAKITER_DECL_LONG_CONST(SENTENCE_SEP_LIMIT); + +#undef BREAKITER_DECL_LONG_CONST + + + /* Create and register 'RuleBasedBreakIterator' class. */ + INIT_CLASS_ENTRY(ce, "IntlRuleBasedBreakIterator", + RuleBasedBreakIterator_class_functions); + RuleBasedBreakIterator_ce_ptr = zend_register_internal_class_ex(&ce, + BreakIterator_ce_ptr, NULL TSRMLS_CC); + + /* Create and register 'CodePointBreakIterator' class. */ + INIT_CLASS_ENTRY(ce, "IntlCodePointBreakIterator", + CodePointBreakIterator_class_functions); + CodePointBreakIterator_ce_ptr = zend_register_internal_class_ex(&ce, + BreakIterator_ce_ptr, NULL TSRMLS_CC); +} +/* }}} */ diff --git a/ext/intl/breakiterator/breakiterator_class.h b/ext/intl/breakiterator/breakiterator_class.h new file mode 100644 index 0000000000..cc5d51256f --- /dev/null +++ b/ext/intl/breakiterator/breakiterator_class.h @@ -0,0 +1,71 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ + */ + +#ifndef BREAKITERATOR_CLASS_H +#define BREAKITERATOR_CLASS_H + +//redefinition of inline in PHP headers causes problems, so include this before +#include <math.h> + +#include <php.h> +#include "../intl_error.h" +#include "../intl_data.h" + +#ifndef USE_BREAKITERATOR_POINTER +typedef void BreakIterator; +#endif + +typedef struct { + zend_object zo; + + // error handling + intl_error err; + + // ICU break iterator + BreakIterator* biter; + + // current text + zval *text; +} BreakIterator_object; + +#define BREAKITER_ERROR(bio) (bio)->err +#define BREAKITER_ERROR_P(bio) &(BREAKITER_ERROR(bio)) + +#define BREAKITER_ERROR_CODE(bio) INTL_ERROR_CODE(BREAKITER_ERROR(bio)) +#define BREAKITER_ERROR_CODE_P(bio) &(INTL_ERROR_CODE(BREAKITER_ERROR(bio))) + +#define BREAKITER_METHOD_INIT_VARS INTL_METHOD_INIT_VARS(BreakIterator, bio) +#define BREAKITER_METHOD_FETCH_OBJECT_NO_CHECK INTL_METHOD_FETCH_OBJECT(BreakIterator, bio) +#define BREAKITER_METHOD_FETCH_OBJECT \ + BREAKITER_METHOD_FETCH_OBJECT_NO_CHECK; \ + if (bio->biter == NULL) \ + { \ + intl_errors_set(&bio->err, U_ILLEGAL_ARGUMENT_ERROR, "Found unconstructed BreakIterator", 0 TSRMLS_CC); \ + RETURN_FALSE; \ + } + +void breakiterator_object_create(zval *object, BreakIterator *break_iter TSRMLS_DC); + +void breakiterator_object_construct(zval *object, BreakIterator *break_iter TSRMLS_DC); + +void breakiterator_register_BreakIterator_class(TSRMLS_D); + +extern zend_class_entry *BreakIterator_ce_ptr, + *RuleBasedBreakIterator_ce_ptr; + +extern zend_object_handlers BreakIterator_handlers; + +#endif /* #ifndef BREAKITERATOR_CLASS_H */ diff --git a/ext/intl/breakiterator/breakiterator_iterators.cpp b/ext/intl/breakiterator/breakiterator_iterators.cpp new file mode 100644 index 0000000000..d88ad8a712 --- /dev/null +++ b/ext/intl/breakiterator/breakiterator_iterators.cpp @@ -0,0 +1,346 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ +*/ + +#ifdef HAVE_CONFIG_H +#include "config.h" +#endif + +#include <unicode/brkiter.h> + +#include "breakiterator_iterators.h" +#include "../common/common_enum.h" + +extern "C" { +#define USE_BREAKITERATOR_POINTER +#include "breakiterator_class.h" +#include "../intl_convert.h" +#include "../locale/locale.h" +#include <zend_exceptions.h> +} + +static zend_class_entry *IntlPartsIterator_ce_ptr; +static zend_object_handlers IntlPartsIterator_handlers; + +/* BreakIterator's iterator */ + +inline BreakIterator *_breakiter_prolog(zend_object_iterator *iter TSRMLS_DC) +{ + BreakIterator_object *bio; + bio = (BreakIterator_object*)zend_object_store_get_object( + (const zval*)iter->data TSRMLS_CC); + intl_errors_reset(BREAKITER_ERROR_P(bio) TSRMLS_CC); + if (bio->biter == NULL) { + intl_errors_set(BREAKITER_ERROR_P(bio), U_INVALID_STATE_ERROR, + "The BreakIterator object backing the PHP iterator is not " + "properly constructed", 0 TSRMLS_CC); + } + return bio->biter; +} + +static void _breakiterator_destroy_it(zend_object_iterator *iter TSRMLS_DC) +{ + zval_ptr_dtor((zval**)&iter->data); +} + +static void _breakiterator_move_forward(zend_object_iterator *iter TSRMLS_DC) +{ + BreakIterator *biter = _breakiter_prolog(iter TSRMLS_CC); + zoi_with_current *zoi_iter = (zoi_with_current*)iter; + + iter->funcs->invalidate_current(iter TSRMLS_CC); + + if (biter == NULL) { + return; + } + + int32_t pos = biter->next(); + if (pos != BreakIterator::DONE) { + MAKE_STD_ZVAL(zoi_iter->current); + ZVAL_LONG(zoi_iter->current, (long)pos); + } //else we've reached the end of the enum, nothing more is required +} + +static void _breakiterator_rewind(zend_object_iterator *iter TSRMLS_DC) +{ + BreakIterator *biter = _breakiter_prolog(iter TSRMLS_CC); + zoi_with_current *zoi_iter = (zoi_with_current*)iter; + + int32_t pos = biter->first(); + MAKE_STD_ZVAL(zoi_iter->current); + ZVAL_LONG(zoi_iter->current, (long)pos); +} + +static zend_object_iterator_funcs breakiterator_iterator_funcs = { + zoi_with_current_dtor, + zoi_with_current_valid, + zoi_with_current_get_current_data, + NULL, + _breakiterator_move_forward, + _breakiterator_rewind, + zoi_with_current_invalidate_current +}; + +U_CFUNC zend_object_iterator *_breakiterator_get_iterator( + zend_class_entry *ce, zval *object, int by_ref TSRMLS_DC) +{ + BreakIterator_object *bio; + if (by_ref) { + zend_throw_exception(NULL, + "Iteration by reference is not supported", 0 TSRMLS_CC); + return NULL; + } + + bio = (BreakIterator_object*)zend_object_store_get_object(object TSRMLS_CC); + BreakIterator *biter = bio->biter; + + if (biter == NULL) { + zend_throw_exception(NULL, + "The BreakIterator is not properly constructed", 0 TSRMLS_CC); + return NULL; + } + + zoi_with_current *zoi_iter = + static_cast<zoi_with_current*>(emalloc(sizeof *zoi_iter)); + zoi_iter->zoi.data = static_cast<void*>(object); + zoi_iter->zoi.funcs = &breakiterator_iterator_funcs; + zoi_iter->zoi.index = 0; + zoi_iter->destroy_it = _breakiterator_destroy_it; + zoi_iter->wrapping_obj = NULL; /* not used; object is in zoi.data */ + zoi_iter->current = NULL; + + zval_add_ref(&object); + + return reinterpret_cast<zend_object_iterator *>(zoi_iter); +} + +/* BreakIterator parts iterator */ + +typedef struct zoi_break_iter_parts { + zoi_with_current zoi_cur; + parts_iter_key_type key_type; + BreakIterator_object *bio; /* so we don't have to fetch it all the time */ +} zoi_break_iter_parts; + +static void _breakiterator_parts_destroy_it(zend_object_iterator *iter TSRMLS_DC) +{ + zval_ptr_dtor(reinterpret_cast<zval**>(&iter->data)); +} + +static int _breakiterator_parts_get_current_key(zend_object_iterator *iter, + char **str_key, + uint *str_key_len, + ulong *int_key TSRMLS_DC) +{ + /* the actual work is done in move_forward and rewind */ + *int_key = iter->index; + return HASH_KEY_IS_LONG; +} + +static void _breakiterator_parts_move_forward(zend_object_iterator *iter TSRMLS_DC) +{ + zoi_break_iter_parts *zoi_bit = (zoi_break_iter_parts*)iter; + BreakIterator_object *bio = zoi_bit->bio; + + iter->funcs->invalidate_current(iter TSRMLS_CC); + + int32_t cur, + next; + + cur = bio->biter->current(); + if (cur == BreakIterator::DONE) { + return; + } + next = bio->biter->next(); + if (next == BreakIterator::DONE) { + return; + } + + if (zoi_bit->key_type == PARTS_ITERATOR_KEY_LEFT) { + iter->index = cur; + } else if (zoi_bit->key_type == PARTS_ITERATOR_KEY_RIGHT) { + iter->index = next; + } + /* else zoi_bit->key_type == PARTS_ITERATOR_KEY_SEQUENTIAL + * No need to do anything, the engine increments ->index */ + + const char *s = Z_STRVAL_P(bio->text); + int32_t slen = Z_STRLEN_P(bio->text), + len; + char *res; + + if (next == BreakIterator::DONE) { + next = slen; + } + assert(next <= slen && next >= cur); + len = next - cur; + res = static_cast<char*>(emalloc(len + 1)); + + memcpy(res, &s[cur], len); + res[len] = '\0'; + + MAKE_STD_ZVAL(zoi_bit->zoi_cur.current); + ZVAL_STRINGL(zoi_bit->zoi_cur.current, res, len, 0); +} + +static void _breakiterator_parts_rewind(zend_object_iterator *iter TSRMLS_DC) +{ + zoi_break_iter_parts *zoi_bit = (zoi_break_iter_parts*)iter; + BreakIterator_object *bio = zoi_bit->bio; + + if (zoi_bit->zoi_cur.current) { + iter->funcs->invalidate_current(iter TSRMLS_CC); + } + + bio->biter->first(); + + iter->funcs->move_forward(iter TSRMLS_CC); +} + +static zend_object_iterator_funcs breakiterator_parts_it_funcs = { + zoi_with_current_dtor, + zoi_with_current_valid, + zoi_with_current_get_current_data, + _breakiterator_parts_get_current_key, + _breakiterator_parts_move_forward, + _breakiterator_parts_rewind, + zoi_with_current_invalidate_current +}; + +void IntlIterator_from_BreakIterator_parts(zval *break_iter_zv, + zval *object, + parts_iter_key_type key_type TSRMLS_DC) +{ + IntlIterator_object *ii; + + zval_add_ref(&break_iter_zv); + + object_init_ex(object, IntlPartsIterator_ce_ptr); + ii = (IntlIterator_object*)zend_object_store_get_object(object TSRMLS_CC); + + ii->iterator = (zend_object_iterator*)emalloc(sizeof(zoi_break_iter_parts)); + ii->iterator->data = break_iter_zv; + ii->iterator->funcs = &breakiterator_parts_it_funcs; + ii->iterator->index = 0; + ((zoi_with_current*)ii->iterator)->destroy_it = _breakiterator_parts_destroy_it; + ((zoi_with_current*)ii->iterator)->wrapping_obj = object; + ((zoi_with_current*)ii->iterator)->current = NULL; + + ((zoi_break_iter_parts*)ii->iterator)->bio = (BreakIterator_object*) + zend_object_store_get_object(break_iter_zv TSRMLS_CC); + assert(((zoi_break_iter_parts*)ii->iterator)->bio->biter != NULL); + ((zoi_break_iter_parts*)ii->iterator)->key_type = key_type; +} + +U_CFUNC zend_object_value IntlPartsIterator_object_create(zend_class_entry *ce TSRMLS_DC) +{ + zend_object_value retval; + + retval = IntlIterator_ce_ptr->create_object(ce TSRMLS_CC); + retval.handlers = &IntlPartsIterator_handlers; + + return retval; +} + +U_CFUNC zend_function *IntlPartsIterator_get_method(zval **object_ptr, + char *method, int method_len, const zend_literal *key TSRMLS_DC) +{ + zend_literal local_literal = {0}; + zend_function *ret; + ALLOCA_FLAG(use_heap) + + if (key == NULL) { + Z_STRVAL(local_literal.constant) = static_cast<char*>( + do_alloca(method_len + 1, use_heap)); + zend_str_tolower_copy(Z_STRVAL(local_literal.constant), + method, method_len); + local_literal.hash_value = zend_hash_func( + Z_STRVAL(local_literal.constant), method_len + 1); + key = &local_literal; + } + + if ((key->hash_value & 0xFFFFFFFF) == 0xA2B486A1 /* hash of getrulestatus\0 */ + && method_len == sizeof("getrulestatus") - 1 + && memcmp("getrulestatus", Z_STRVAL(key->constant), method_len) == 0) { + IntlIterator_object *obj = (IntlIterator_object*) + zend_object_store_get_object(*object_ptr TSRMLS_CC); + if (obj->iterator && obj->iterator->data) { + zval *break_iter_zv = static_cast<zval*>(obj->iterator->data); + *object_ptr = break_iter_zv; + ret = Z_OBJ_HANDLER_P(break_iter_zv, get_method)(object_ptr, + method, method_len, key TSRMLS_CC); + goto end; + } + } + + ret = std_object_handlers.get_method(object_ptr, + method, method_len, key TSRMLS_CC); + +end: + if (key == &local_literal) { + free_alloca(Z_STRVAL(local_literal.constant), use_heap); + } + + return ret; +} + +U_CFUNC PHP_METHOD(IntlPartsIterator, getBreakIterator) +{ + INTLITERATOR_METHOD_INIT_VARS; + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "IntlPartsIterator::getBreakIterator: bad arguments", 0 TSRMLS_CC); + return; + } + + INTLITERATOR_METHOD_FETCH_OBJECT; + + zval *biter_zval = static_cast<zval*>(ii->iterator->data); + RETURN_ZVAL(biter_zval, 1, 0); +} + +ZEND_BEGIN_ARG_INFO_EX(ainfo_parts_it_void, 0, 0, 0) +ZEND_END_ARG_INFO() + +static const zend_function_entry IntlPartsIterator_class_functions[] = { + PHP_ME(IntlPartsIterator, getBreakIterator, ainfo_parts_it_void, ZEND_ACC_PUBLIC) + PHP_FE_END +}; + +U_CFUNC void breakiterator_register_IntlPartsIterator_class(TSRMLS_D) +{ + zend_class_entry ce; + + /* Create and register 'BreakIterator' class. */ + INIT_CLASS_ENTRY(ce, "IntlPartsIterator", IntlPartsIterator_class_functions); + IntlPartsIterator_ce_ptr = zend_register_internal_class_ex(&ce, + IntlIterator_ce_ptr, NULL TSRMLS_CC); + IntlPartsIterator_ce_ptr->create_object = IntlPartsIterator_object_create; + + memcpy(&IntlPartsIterator_handlers, &IntlIterator_handlers, + sizeof IntlPartsIterator_handlers); + IntlPartsIterator_handlers.get_method = IntlPartsIterator_get_method; + +#define PARTSITER_DECL_LONG_CONST(name) \ + zend_declare_class_constant_long(IntlPartsIterator_ce_ptr, #name, \ + sizeof(#name) - 1, PARTS_ITERATOR_ ## name TSRMLS_CC) + + PARTSITER_DECL_LONG_CONST(KEY_SEQUENTIAL); + PARTSITER_DECL_LONG_CONST(KEY_LEFT); + PARTSITER_DECL_LONG_CONST(KEY_RIGHT); + +#undef PARTSITER_DECL_LONG_CONST +}
\ No newline at end of file diff --git a/ext/intl/breakiterator/breakiterator_iterators.h b/ext/intl/breakiterator/breakiterator_iterators.h new file mode 100644 index 0000000000..7162072414 --- /dev/null +++ b/ext/intl/breakiterator/breakiterator_iterators.h @@ -0,0 +1,42 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ +*/ +#ifndef INTL_BREAKITERATOR_ITERATORS_H +#define INTL_BREAKITERATOR_ITERATORS_H + +#include <unicode/umachine.h> + +U_CDECL_BEGIN +#include <math.h> +#include <php.h> +U_CDECL_END + +typedef enum { + PARTS_ITERATOR_KEY_SEQUENTIAL, + PARTS_ITERATOR_KEY_LEFT, + PARTS_ITERATOR_KEY_RIGHT, +} parts_iter_key_type; + +#ifdef __cplusplus +void IntlIterator_from_BreakIterator_parts(zval *break_iter_zv, + zval *object, + parts_iter_key_type key_type TSRMLS_DC); +#endif + +U_CFUNC zend_object_iterator *_breakiterator_get_iterator( + zend_class_entry *ce, zval *object, int by_ref TSRMLS_DC); +U_CFUNC void breakiterator_register_IntlPartsIterator_class(TSRMLS_D); + +#endif
\ No newline at end of file diff --git a/ext/intl/breakiterator/breakiterator_methods.cpp b/ext/intl/breakiterator/breakiterator_methods.cpp new file mode 100644 index 0000000000..6a61f8cb93 --- /dev/null +++ b/ext/intl/breakiterator/breakiterator_methods.cpp @@ -0,0 +1,452 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ +*/ + +#ifdef HAVE_CONFIG_H +#include "config.h" +#endif + +#include <unicode/brkiter.h> +#include "codepointiterator_internal.h" + +#include "breakiterator_iterators.h" + +extern "C" { +#include "../php_intl.h" +#define USE_BREAKITERATOR_POINTER 1 +#include "breakiterator_class.h" +#include "../locale/locale.h" +#include <zend_exceptions.h> +} + +using PHP::CodePointBreakIterator; + +U_CFUNC PHP_METHOD(BreakIterator, __construct) +{ + zend_throw_exception( NULL, + "An object of this type cannot be created with the new operator", + 0 TSRMLS_CC ); +} + +static void _breakiter_factory(const char *func_name, + BreakIterator *(*func)(const Locale&, UErrorCode&), + INTERNAL_FUNCTION_PARAMETERS) +{ + BreakIterator *biter; + const char *locale_str = NULL; + int dummy; + char *msg; + UErrorCode status = UErrorCode(); + intl_error_reset(NULL TSRMLS_CC); + + if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "s!", + &locale_str, &dummy) == FAILURE) { + spprintf(&msg, NULL, "%s: bad arguments", func_name); + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, msg, 1 TSRMLS_CC); + efree(msg); + RETURN_NULL(); + } + + if (locale_str == NULL) { + locale_str = intl_locale_get_default(TSRMLS_C); + } + + biter = func(Locale::createFromName(locale_str), status); + intl_error_set_code(NULL, status TSRMLS_CC); + if (U_FAILURE(status)) { + spprintf(&msg, NULL, "%s: error creating BreakIterator", + func_name); + intl_error_set_custom_msg(NULL, msg, 1 TSRMLS_CC); + efree(msg); + RETURN_NULL(); + } + + breakiterator_object_create(return_value, biter TSRMLS_CC); +} + +U_CFUNC PHP_FUNCTION(breakiter_create_word_instance) +{ + _breakiter_factory("breakiter_create_word_instance", + &BreakIterator::createWordInstance, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_create_line_instance) +{ + _breakiter_factory("breakiter_create_line_instance", + &BreakIterator::createLineInstance, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_create_character_instance) +{ + _breakiter_factory("breakiter_create_character_instance", + &BreakIterator::createCharacterInstance, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_create_sentence_instance) +{ + _breakiter_factory("breakiter_create_sentence_instance", + &BreakIterator::createSentenceInstance, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_create_title_instance) +{ + _breakiter_factory("breakiter_create_title_instance", + &BreakIterator::createTitleInstance, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_create_code_point_instance) +{ + UErrorCode status = UErrorCode(); + intl_error_reset(NULL TSRMLS_CC); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_create_code_point_instance: bad arguments", 0 TSRMLS_CC); + RETURN_NULL(); + } + + CodePointBreakIterator *cpbi = new CodePointBreakIterator(); + breakiterator_object_create(return_value, cpbi TSRMLS_CC); +} + +U_CFUNC PHP_FUNCTION(breakiter_get_text) +{ + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_get_text: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + if (bio->text == NULL) { + RETURN_NULL(); + } else { + RETURN_ZVAL(bio->text, 1, 0); + } +} + +U_CFUNC PHP_FUNCTION(breakiter_set_text) +{ + char *text; + int text_len; + UText *ut = NULL; + zval **textzv; + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "s", + &text, &text_len) == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_set_text: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + int res = zend_get_parameters_ex(1, &textzv); + assert(res == SUCCESS); + + BREAKITER_METHOD_FETCH_OBJECT; + + /* assert it's safe to use text and text_len because zpp changes the + * arguments in the stack */ + assert(text == Z_STRVAL_PP(textzv)); + + ut = utext_openUTF8(ut, text, text_len, BREAKITER_ERROR_CODE_P(bio)); + INTL_CTOR_CHECK_STATUS(bio, "breakiter_set_text: error opening UText"); + + bio->biter->setText(ut, BREAKITER_ERROR_CODE(bio)); + utext_close(ut); /* ICU shallow clones the UText */ + INTL_CTOR_CHECK_STATUS(bio, "breakiter_set_text: error calling " + "BreakIterator::setText()"); + + /* When ICU clones the UText, it does not copy the buffer, so we have to + * keep the string buffer around by holding a reference to its zval. This + * also allows a faste implementation of getText() */ + if (bio->text != NULL) { + zval_ptr_dtor(&bio->text); + } + bio->text = *textzv; + zval_add_ref(&bio->text); + + RETURN_TRUE; +} + +static void _breakiter_no_args_ret_int32( + const char *func_name, + int32_t (BreakIterator::*func)(), + INTERNAL_FUNCTION_PARAMETERS) +{ + char *msg; + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + spprintf(&msg, NULL, "%s: bad arguments", func_name); + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, msg, 1 TSRMLS_CC); + efree(msg); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + int32_t res = (bio->biter->*func)(); + + RETURN_LONG((long)res); +} + +static void _breakiter_int32_ret_int32( + const char *func_name, + int32_t (BreakIterator::*func)(int32_t), + INTERNAL_FUNCTION_PARAMETERS) +{ + char *msg; + long arg; + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "l", &arg) == FAILURE) { + spprintf(&msg, NULL, "%s: bad arguments", func_name); + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, msg, 1 TSRMLS_CC); + efree(msg); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + if (arg < INT32_MIN || arg > INT32_MAX) { + spprintf(&msg, NULL, "%s: offset argument is outside bounds of " + "a 32-bit wide integer", func_name); + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, msg, 1 TSRMLS_CC); + efree(msg); + RETURN_FALSE; + } + + int32_t res = (bio->biter->*func)((int32_t)arg); + + RETURN_LONG((long)res); +} + +U_CFUNC PHP_FUNCTION(breakiter_first) +{ + _breakiter_no_args_ret_int32("breakiter_first", + &BreakIterator::first, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_last) +{ + _breakiter_no_args_ret_int32("breakiter_last", + &BreakIterator::last, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_previous) +{ + _breakiter_no_args_ret_int32("breakiter_previous", + &BreakIterator::previous, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_next) +{ + bool no_arg_version = false; + + if (ZEND_NUM_ARGS() == 0) { + no_arg_version = true; + } else if (ZEND_NUM_ARGS() == 1) { + zval **arg; + int res = zend_get_parameters_ex(1, &arg); + assert(res == SUCCESS); + if (Z_TYPE_PP(arg) == IS_NULL) { + no_arg_version = true; + ht = 0; /* pretend we don't have any argument */ + } else { + no_arg_version = false; + } + } + + if (no_arg_version) { + _breakiter_no_args_ret_int32("breakiter_next", + &BreakIterator::next, + INTERNAL_FUNCTION_PARAM_PASSTHRU); + } else { + _breakiter_int32_ret_int32("breakiter_next", + &BreakIterator::next, + INTERNAL_FUNCTION_PARAM_PASSTHRU); + } +} + +U_CFUNC PHP_FUNCTION(breakiter_current) +{ + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_current: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + int32_t res = bio->biter->current(); + + RETURN_LONG((long)res); +} + +U_CFUNC PHP_FUNCTION(breakiter_following) +{ + _breakiter_int32_ret_int32("breakiter_following", + &BreakIterator::following, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_preceding) +{ + _breakiter_int32_ret_int32("breakiter_preceding", + &BreakIterator::preceding, + INTERNAL_FUNCTION_PARAM_PASSTHRU); +} + +U_CFUNC PHP_FUNCTION(breakiter_is_boundary) +{ + long offset; + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "l", + &offset) == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_is_boundary: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + if (offset < INT32_MIN || offset > INT32_MAX) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_is_boundary: offset argument is outside bounds of " + "a 32-bit wide integer", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + UBool res = bio->biter->isBoundary((int32_t)offset); + + RETURN_BOOL((long)res); +} + +U_CFUNC PHP_FUNCTION(breakiter_get_locale) +{ + long locale_type; + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "l", &locale_type) == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_get_locale: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + if (locale_type != ULOC_ACTUAL_LOCALE && locale_type != ULOC_VALID_LOCALE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_get_locale: invalid locale type", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + Locale locale = bio->biter->getLocale((ULocDataLocaleType)locale_type, + BREAKITER_ERROR_CODE(bio)); + INTL_METHOD_CHECK_STATUS(bio, + "breakiter_get_locale: Call to ICU method has failed"); + + RETURN_STRING(locale.getName(), 1); +} + +U_CFUNC PHP_FUNCTION(breakiter_get_parts_iterator) +{ + long key_type = 0; + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "|l", &key_type) == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_get_parts_iterator: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + if (key_type != PARTS_ITERATOR_KEY_SEQUENTIAL + && key_type != PARTS_ITERATOR_KEY_LEFT + && key_type != PARTS_ITERATOR_KEY_RIGHT) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_get_parts_iterator: bad key type", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + IntlIterator_from_BreakIterator_parts( + object, return_value, (parts_iter_key_type)key_type TSRMLS_CC); +} + +U_CFUNC PHP_FUNCTION(breakiter_get_error_code) +{ + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_get_error_code: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + /* Fetch the object (without resetting its last error code ). */ + bio = (BreakIterator_object*)zend_object_store_get_object(object TSRMLS_CC); + if (bio == NULL) + RETURN_FALSE; + + RETURN_LONG((long)BREAKITER_ERROR_CODE(bio)); +} + +U_CFUNC PHP_FUNCTION(breakiter_get_error_message) +{ + const char* message = NULL; + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set( NULL, U_ILLEGAL_ARGUMENT_ERROR, + "breakiter_get_error_message: bad arguments", 0 TSRMLS_CC ); + RETURN_FALSE; + } + + + /* Fetch the object (without resetting its last error code ). */ + bio = (BreakIterator_object*)zend_object_store_get_object(object TSRMLS_CC); + if (bio == NULL) + RETURN_FALSE; + + /* Return last error message. */ + message = intl_error_get_message(BREAKITER_ERROR_P(bio) TSRMLS_CC); + RETURN_STRING(message, 0); +} diff --git a/ext/intl/breakiterator/breakiterator_methods.h b/ext/intl/breakiterator/breakiterator_methods.h new file mode 100644 index 0000000000..a479ac92e8 --- /dev/null +++ b/ext/intl/breakiterator/breakiterator_methods.h @@ -0,0 +1,64 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ + */ + +#ifndef BREAKITERATOR_METHODS_H +#define BREAKITERATOR_METHODS_H + +#include <php.h> + +PHP_METHOD(BreakIterator, __construct); + +PHP_FUNCTION(breakiter_create_word_instance); + +PHP_FUNCTION(breakiter_create_line_instance); + +PHP_FUNCTION(breakiter_create_character_instance); + +PHP_FUNCTION(breakiter_create_sentence_instance); + +PHP_FUNCTION(breakiter_create_title_instance); + +PHP_FUNCTION(breakiter_create_code_point_instance); + +PHP_FUNCTION(breakiter_get_text); + +PHP_FUNCTION(breakiter_set_text); + +PHP_FUNCTION(breakiter_first); + +PHP_FUNCTION(breakiter_last); + +PHP_FUNCTION(breakiter_previous); + +PHP_FUNCTION(breakiter_next); + +PHP_FUNCTION(breakiter_current); + +PHP_FUNCTION(breakiter_following); + +PHP_FUNCTION(breakiter_preceding); + +PHP_FUNCTION(breakiter_is_boundary); + +PHP_FUNCTION(breakiter_get_locale); + +PHP_FUNCTION(breakiter_get_parts_iterator); + +PHP_FUNCTION(breakiter_get_error_code); + +PHP_FUNCTION(breakiter_get_error_message); + +#endif
\ No newline at end of file diff --git a/ext/intl/breakiterator/codepointiterator_internal.cpp b/ext/intl/breakiterator/codepointiterator_internal.cpp new file mode 100644 index 0000000000..bf9239d531 --- /dev/null +++ b/ext/intl/breakiterator/codepointiterator_internal.cpp @@ -0,0 +1,291 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ + */ + +#include "codepointiterator_internal.h" +#include <unicode/uchriter.h> +#include <typeinfo> + +//copied from cmemory.h, which is not public +typedef union { + long t1; + double t2; + void *t3; +} UAlignedMemory; + +#define U_POINTER_MASK_LSB(ptr, mask) (((ptrdiff_t)(char *)(ptr)) & (mask)) +#define U_ALIGNMENT_OFFSET(ptr) U_POINTER_MASK_LSB(ptr, sizeof(UAlignedMemory) - 1) +#define U_ALIGNMENT_OFFSET_UP(ptr) (sizeof(UAlignedMemory) - U_ALIGNMENT_OFFSET(ptr)) + +using namespace PHP; + +UOBJECT_DEFINE_RTTI_IMPLEMENTATION(CodePointBreakIterator); + +CodePointBreakIterator::CodePointBreakIterator() +: BreakIterator(), fCharIter(NULL), lastCodePoint(U_SENTINEL) +{ + UErrorCode uec = UErrorCode(); + this->fText = utext_openUChars(NULL, NULL, 0, &uec); +} + +CodePointBreakIterator::CodePointBreakIterator(const PHP::CodePointBreakIterator &other) +: BreakIterator(other), fText(NULL), fCharIter(NULL), lastCodePoint(U_SENTINEL) +{ + *this = other; +} + +CodePointBreakIterator& CodePointBreakIterator::operator=(const CodePointBreakIterator& that) +{ + UErrorCode uec = UErrorCode(); + UText *ut_clone = NULL; + + if (this == &that) { + return *this; + } + + this->fText = utext_clone(this->fText, that.fText, FALSE, TRUE, &uec); + + //don't bother copying the character iterator, getText() is deprecated + clearCurrentCharIter(); + + this->lastCodePoint = that.lastCodePoint; + return *this; +} + +CodePointBreakIterator::~CodePointBreakIterator() +{ + if (this->fText) { + utext_close(this->fText); + } + clearCurrentCharIter(); +} + +UBool CodePointBreakIterator::operator==(const BreakIterator& that) const +{ + if (typeid(*this) != typeid(that)) { + return FALSE; + } + + const CodePointBreakIterator& that2 = + static_cast<const CodePointBreakIterator&>(that); + + if (!utext_equals(this->fText, that2.fText)) { + return FALSE; + } + + return TRUE; +} + +CodePointBreakIterator* CodePointBreakIterator::clone(void) const +{ + return new CodePointBreakIterator(*this); +} + +CharacterIterator& CodePointBreakIterator::getText(void) const +{ + if (this->fCharIter == NULL) { + //this method is deprecated anyway; setup bogus iterator + static const UChar c = 0; + this->fCharIter = new UCharCharacterIterator(&c, 0); + } + + return *this->fCharIter; +} + +UText *CodePointBreakIterator::getUText(UText *fillIn, UErrorCode &status) const +{ + return utext_clone(fillIn, this->fText, FALSE, TRUE, &status); +} + +void CodePointBreakIterator::setText(const UnicodeString &text) +{ + UErrorCode uec = UErrorCode(); + + //this closes the previous utext, if any + this->fText = utext_openConstUnicodeString(this->fText, &text, &uec); + + clearCurrentCharIter(); +} + +void CodePointBreakIterator::setText(UText *text, UErrorCode &status) +{ + if (U_FAILURE(status)) { + return; + } + + this->fText = utext_clone(this->fText, text, FALSE, TRUE, &status); + + clearCurrentCharIter(); +} + +void CodePointBreakIterator::adoptText(CharacterIterator* it) +{ + UErrorCode uec = UErrorCode(); + clearCurrentCharIter(); + + this->fCharIter = it; + this->fText = utext_openCharacterIterator(this->fText, it, &uec); +} + +int32_t CodePointBreakIterator::first(void) +{ + UTEXT_SETNATIVEINDEX(this->fText, 0); + this->lastCodePoint = U_SENTINEL; + + return 0; +} + +int32_t CodePointBreakIterator::last(void) +{ + int32_t pos = (int32_t)utext_nativeLength(this->fText); + UTEXT_SETNATIVEINDEX(this->fText, pos); + this->lastCodePoint = U_SENTINEL; + + return pos; +} + +int32_t CodePointBreakIterator::previous(void) +{ + this->lastCodePoint = UTEXT_PREVIOUS32(this->fText); + if (this->lastCodePoint == U_SENTINEL) { + return BreakIterator::DONE; + } + + return (int32_t)UTEXT_GETNATIVEINDEX(this->fText); +} + +int32_t CodePointBreakIterator::next(void) +{ + this->lastCodePoint = UTEXT_NEXT32(this->fText); + if (this->lastCodePoint == U_SENTINEL) { + return BreakIterator::DONE; + } + + return (int32_t)UTEXT_GETNATIVEINDEX(this->fText); +} + +int32_t CodePointBreakIterator::current(void) const +{ + return (int32_t)UTEXT_GETNATIVEINDEX(this->fText); +} + +int32_t CodePointBreakIterator::following(int32_t offset) +{ + this->lastCodePoint = utext_next32From(this->fText, offset); + if (this->lastCodePoint == U_SENTINEL) { + return BreakIterator::DONE; + } + + return (int32_t)UTEXT_GETNATIVEINDEX(this->fText); +} + +int32_t CodePointBreakIterator::preceding(int32_t offset) +{ + this->lastCodePoint = utext_previous32From(this->fText, offset); + if (this->lastCodePoint == U_SENTINEL) { + return BreakIterator::DONE; + } + + return (int32_t)UTEXT_GETNATIVEINDEX(this->fText); +} + +UBool CodePointBreakIterator::isBoundary(int32_t offset) +{ + //this function has side effects, and it's supposed to + utext_setNativeIndex(this->fText, offset); + return (offset == utext_getNativeIndex(this->fText)); +} + +int32_t CodePointBreakIterator::next(int32_t n) +{ + UBool res = utext_moveIndex32(this->fText, n); + +#ifndef UTEXT_CURRENT32 +#define UTEXT_CURRENT32 utext_current32 +#endif + + if (res) { + this->lastCodePoint = UTEXT_CURRENT32(this->fText); + return (int32_t)UTEXT_GETNATIVEINDEX(this->fText); + } else { + this->lastCodePoint = U_SENTINEL; + return BreakIterator::DONE; + } +} + +CodePointBreakIterator *CodePointBreakIterator::createBufferClone( + void *stackBuffer, int32_t &bufferSize, UErrorCode &status) +{ + //see implementation of RuleBasedBreakIterator::createBufferClone() + if (U_FAILURE(status)) { + return NULL; + } + + if (bufferSize <= 0) { + bufferSize = sizeof(CodePointBreakIterator) + U_ALIGNMENT_OFFSET_UP(0); + return NULL; + } + + char *buf = (char*)stackBuffer; + uint32_t s = bufferSize; + + if (stackBuffer == NULL) { + s = 0; + } + + if (U_ALIGNMENT_OFFSET(stackBuffer) != 0) { + uint32_t offsetUp = (uint32_t)U_ALIGNMENT_OFFSET_UP(buf); + s -= offsetUp; + buf += offsetUp; + } + + if (s < sizeof(CodePointBreakIterator)) { + CodePointBreakIterator *clonedBI = new CodePointBreakIterator(*this); + if (clonedBI == NULL) { + status = U_MEMORY_ALLOCATION_ERROR; + } else { + status = U_SAFECLONE_ALLOCATED_WARNING; + } + + return clonedBI; + } + + return new(buf) CodePointBreakIterator(*this); +} + +CodePointBreakIterator &CodePointBreakIterator::refreshInputText(UText *input, UErrorCode &status) +{ + //see implementation of RuleBasedBreakIterator::createBufferClone() + if (U_FAILURE(status)) { + return *this; + } + if (input == NULL) { + status = U_ILLEGAL_ARGUMENT_ERROR; + return *this; + } + + int64_t pos = utext_getNativeIndex(this->fText); + this->fText = utext_clone(this->fText, input, FALSE, TRUE, &status); + if (U_FAILURE(status)) { + return *this; + } + + utext_setNativeIndex(this->fText, pos); + if (utext_getNativeIndex(fText) != pos) { + status = U_ILLEGAL_ARGUMENT_ERROR; + } + + return *this; +} diff --git a/ext/intl/breakiterator/codepointiterator_internal.h b/ext/intl/breakiterator/codepointiterator_internal.h new file mode 100644 index 0000000000..988b91c200 --- /dev/null +++ b/ext/intl/breakiterator/codepointiterator_internal.h @@ -0,0 +1,98 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ + */ + +#ifndef CODEPOINTITERATOR_INTERNAL_H +#define CODEPOINTITERATOR_INTERNAL_H + +#include <unicode/brkiter.h> + +using U_ICU_NAMESPACE::BreakIterator; + +namespace PHP { + + class CodePointBreakIterator : public BreakIterator { + + public: + static UClassID getStaticClassID(); + + CodePointBreakIterator(); + + CodePointBreakIterator(const CodePointBreakIterator &other); + + CodePointBreakIterator& operator=(const CodePointBreakIterator& that); + + virtual ~CodePointBreakIterator(); + + virtual UBool operator==(const BreakIterator& that) const; + + virtual CodePointBreakIterator* clone(void) const; + + virtual UClassID getDynamicClassID(void) const; + + virtual CharacterIterator& getText(void) const; + + virtual UText *getUText(UText *fillIn, UErrorCode &status) const; + + virtual void setText(const UnicodeString &text); + + virtual void setText(UText *text, UErrorCode &status); + + virtual void adoptText(CharacterIterator* it); + + virtual int32_t first(void); + + virtual int32_t last(void); + + virtual int32_t previous(void); + + virtual int32_t next(void); + + virtual int32_t current(void) const; + + virtual int32_t following(int32_t offset); + + virtual int32_t preceding(int32_t offset); + + virtual UBool isBoundary(int32_t offset); + + virtual int32_t next(int32_t n); + + virtual CodePointBreakIterator *createBufferClone(void *stackBuffer, + int32_t &BufferSize, + UErrorCode &status); + + virtual CodePointBreakIterator &refreshInputText(UText *input, UErrorCode &status); + + inline UChar32 getLastCodePoint() + { + return this->lastCodePoint; + } + + private: + UText *fText; + UChar32 lastCodePoint; + mutable CharacterIterator *fCharIter; + + inline void clearCurrentCharIter() + { + delete this->fCharIter; + this->fCharIter = NULL; + this->lastCodePoint = U_SENTINEL; + } + }; +} + +#endif
\ No newline at end of file diff --git a/ext/intl/breakiterator/codepointiterator_methods.cpp b/ext/intl/breakiterator/codepointiterator_methods.cpp new file mode 100644 index 0000000000..ae7e526ead --- /dev/null +++ b/ext/intl/breakiterator/codepointiterator_methods.cpp @@ -0,0 +1,44 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ + */ + +#include "codepointiterator_internal.h" + +extern "C" { +#define USE_BREAKITERATOR_POINTER 1 +#include "breakiterator_class.h" +} + +using PHP::CodePointBreakIterator; + +static inline CodePointBreakIterator *fetch_cpbi(BreakIterator_object *bio) { + return (CodePointBreakIterator*)bio->biter; +} + +U_CFUNC PHP_FUNCTION(cpbi_get_last_code_point) +{ + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "cpbi_get_last_code_point: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + RETURN_LONG(fetch_cpbi(bio)->getLastCodePoint()); +}
\ No newline at end of file diff --git a/ext/intl/breakiterator/codepointiterator_methods.h b/ext/intl/breakiterator/codepointiterator_methods.h new file mode 100644 index 0000000000..d34e5b61e2 --- /dev/null +++ b/ext/intl/breakiterator/codepointiterator_methods.h @@ -0,0 +1,24 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ + */ + +#ifndef CODEPOINTITERATOR_METHODS_H +#define CODEPOINTITERATOR_METHODS_H + +#include <php.h> + +PHP_FUNCTION(cpbi_get_last_code_point); + +#endif
\ No newline at end of file diff --git a/ext/intl/breakiterator/rulebasedbreakiterator_methods.cpp b/ext/intl/breakiterator/rulebasedbreakiterator_methods.cpp new file mode 100644 index 0000000000..454e5249fd --- /dev/null +++ b/ext/intl/breakiterator/rulebasedbreakiterator_methods.cpp @@ -0,0 +1,221 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ + */ + +#include <unicode/rbbi.h> + +extern "C" { +#define USE_BREAKITERATOR_POINTER 1 +#include "breakiterator_class.h" +#include <zend_exceptions.h> +#include <limits.h> +} + +#include "../intl_convertcpp.h" + +static inline RuleBasedBreakIterator *fetch_rbbi(BreakIterator_object *bio) { + return (RuleBasedBreakIterator*)bio->biter; +} + +static void _php_intlrbbi_constructor_body(INTERNAL_FUNCTION_PARAMETERS) +{ + zval *object = getThis(); + char *rules; + int rules_len; + zend_bool compiled = 0; + UErrorCode status = U_ZERO_ERROR; + intl_error_reset(NULL TSRMLS_CC); + + if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "s|b", + &rules, &rules_len, &compiled) == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "rbbi_create_instance: bad arguments", 0 TSRMLS_CC); + RETURN_NULL(); + } + + // instantiation of ICU object + RuleBasedBreakIterator *rbbi; + + if (!compiled) { + UnicodeString rulesStr; + UParseError parseError = UParseError(); + if (intl_stringFromChar(rulesStr, rules, rules_len, &status) + == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "rbbi_create_instance: rules were not a valid UTF-8 string", + 0 TSRMLS_CC); + RETURN_NULL(); + } + + rbbi = new RuleBasedBreakIterator(rulesStr, parseError, status); + intl_error_set_code(NULL, status TSRMLS_CC); + if (U_FAILURE(status)) { + char *msg; + smart_str parse_error_str; + parse_error_str = intl_parse_error_to_string(&parseError); + spprintf(&msg, 0, "rbbi_create_instance: unable to create " + "RuleBasedBreakIterator from rules (%s)", parse_error_str.c); + smart_str_free(&parse_error_str); + intl_error_set_custom_msg(NULL, msg, 1 TSRMLS_CC); + efree(msg); + delete rbbi; + RETURN_NULL(); + } + } else { // compiled +#if U_ICU_VERSION_MAJOR_NUM * 10 + U_ICU_VERSION_MINOR_NUM >= 48 + rbbi = new RuleBasedBreakIterator((uint8_t*)rules, rules_len, status); + if (U_FAILURE(status)) { + intl_error_set(NULL, status, "rbbi_create_instance: unable to " + "create instance from compiled rules", 0 TSRMLS_CC); + delete rbbi; + RETURN_NULL(); + } +#else + intl_error_set(NULL, U_UNSUPPORTED_ERROR, "rbbi_create_instance: " + "compiled rules require ICU >= 4.8", 0 TSRMLS_CC); + RETURN_NULL(); +#endif + } + + breakiterator_object_create(return_value, rbbi TSRMLS_CC); +} + +U_CFUNC PHP_METHOD(IntlRuleBasedBreakIterator, __construct) +{ + zval orig_this = *getThis(); + + return_value = getThis(); + //changes this to IS_NULL (without first destroying) if there's an error + _php_intlrbbi_constructor_body(INTERNAL_FUNCTION_PARAM_PASSTHRU); + + if (Z_TYPE_P(return_value) == IS_NULL) { + zend_object_store_ctor_failed(&orig_this TSRMLS_CC); + zval_dtor(&orig_this); + } +} + +U_CFUNC PHP_FUNCTION(rbbi_get_rules) +{ + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "rbbi_get_rules: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + const UnicodeString rules = fetch_rbbi(bio)->getRules(); + + Z_TYPE_P(return_value) = IS_STRING; + if (intl_charFromString(rules, &Z_STRVAL_P(return_value), + &Z_STRLEN_P(return_value), BREAKITER_ERROR_CODE_P(bio)) == FAILURE) + { + intl_errors_set(BREAKITER_ERROR_P(bio), BREAKITER_ERROR_CODE(bio), + "rbbi_hash_code: Error converting result to UTF-8 string", + 0 TSRMLS_CC); + RETURN_FALSE; + } +} + +U_CFUNC PHP_FUNCTION(rbbi_get_rule_status) +{ + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "rbbi_get_rule_status: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + RETURN_LONG(fetch_rbbi(bio)->getRuleStatus()); +} + +U_CFUNC PHP_FUNCTION(rbbi_get_rule_status_vec) +{ + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "rbbi_get_rule_status_vec: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + int32_t num_rules = fetch_rbbi(bio)->getRuleStatusVec(NULL, 0, + BREAKITER_ERROR_CODE(bio)); + if (BREAKITER_ERROR_CODE(bio) == U_BUFFER_OVERFLOW_ERROR) { + BREAKITER_ERROR_CODE(bio) = U_ZERO_ERROR; + } else { + // should not happen + INTL_METHOD_CHECK_STATUS(bio, "rbbi_get_rule_status_vec: failed " + " determining the number of status values"); + } + int32_t *rules = new int32_t[num_rules]; + num_rules = fetch_rbbi(bio)->getRuleStatusVec(rules, num_rules, + BREAKITER_ERROR_CODE(bio)); + if (U_FAILURE(BREAKITER_ERROR_CODE(bio))) { + delete[] rules; + intl_errors_set(BREAKITER_ERROR_P(bio), BREAKITER_ERROR_CODE(bio), + "rbbi_get_rule_status_vec: failed obtaining the status values", + 0 TSRMLS_CC); + RETURN_FALSE; + } + + array_init_size(return_value, num_rules); + for (int32_t i = 0; i < num_rules; i++) { + add_next_index_long(return_value, rules[i]); + } + delete[] rules; +} + +#if U_ICU_VERSION_MAJOR_NUM * 10 + U_ICU_VERSION_MINOR_NUM >= 48 +U_CFUNC PHP_FUNCTION(rbbi_get_binary_rules) +{ + BREAKITER_METHOD_INIT_VARS; + object = getThis(); + + if (zend_parse_parameters_none() == FAILURE) { + intl_error_set(NULL, U_ILLEGAL_ARGUMENT_ERROR, + "rbbi_get_binary_rules: bad arguments", 0 TSRMLS_CC); + RETURN_FALSE; + } + + BREAKITER_METHOD_FETCH_OBJECT; + + uint32_t rules_len; + const uint8_t *rules = fetch_rbbi(bio)->getBinaryRules(rules_len); + + if (rules_len > INT_MAX - 1) { + intl_errors_set(BREAKITER_ERROR_P(bio), BREAKITER_ERROR_CODE(bio), + "rbbi_get_binary_rules: the rules are too large", + 0 TSRMLS_CC); + RETURN_FALSE; + } + + char *ret_rules = static_cast<char*>(emalloc(rules_len + 1)); + memcpy(ret_rules, rules, rules_len); + ret_rules[rules_len] = '\0'; + + RETURN_STRINGL(ret_rules, rules_len, 0); +} +#endif diff --git a/ext/intl/breakiterator/rulebasedbreakiterator_methods.h b/ext/intl/breakiterator/rulebasedbreakiterator_methods.h new file mode 100644 index 0000000000..edea4ea2a6 --- /dev/null +++ b/ext/intl/breakiterator/rulebasedbreakiterator_methods.h @@ -0,0 +1,32 @@ +/* + +----------------------------------------------------------------------+ + | PHP Version 5 | + +----------------------------------------------------------------------+ + | This source file is subject to version 3.01 of the PHP license, | + | that is bundled with this package in the file LICENSE, and is | + | available through the world-wide-web at the following url: | + | http://www.php.net/license/3_01.txt | + | If you did not receive a copy of the PHP license and are unable to | + | obtain it through the world-wide-web, please send a note to | + | license@php.net so we can mail you a copy immediately. | + +----------------------------------------------------------------------+ + | Authors: Gustavo Lopes <cataphract@php.net> | + +----------------------------------------------------------------------+ + */ + +#ifndef RULEBASEDBREAKITERATOR_METHODS_H +#define RULEBASEDBREAKITERATOR_METHODS_H + +#include <php.h> + +PHP_METHOD(IntlRuleBasedBreakIterator, __construct); + +PHP_FUNCTION(rbbi_get_rules); + +PHP_FUNCTION(rbbi_get_rule_status); + +PHP_FUNCTION(rbbi_get_rule_status_vec); + +PHP_FUNCTION(rbbi_get_binary_rules); + +#endif
\ No newline at end of file |