cvs: php-src /ext/standard string.c type.c /ext/standard/tests/strings strtr2.phpt ucfirst.phpt

derick Fri Mar 17 14:29:05 2006 UTC Added files: /php-src/ext/standard/tests/strings strtr2.phpt ucfirst.phpt Modified files: /php-src/ext/standard string.c type.c Log: - Fixed bug in ucfirst() implementation. If the tmp_len = 0 then it will not write the uppercased character to the buffer, but only returns the length of the uppercased letter as per http://icu.sourceforge.net/apiref/icu4c/ustring_8h.html#a50. - Updated is_string(): If Unicode semantics is turned on, return "true" for Unicode strings only. If Unicode semantics is turned off, return "true" for native strings only. - Added is_binary() function that returns "true" for native strings only. - Added first implementation of upgraded strtr function. It works except if combining sequences or surrogates are used in the non-array method of calling this function.

http://cvs.php.net/viewcvs.cgi/php-src/ext/standard/string.c?r1=1.529&r2=1.530&diff_format=u Index: php-src/ext/standard/string.c diff -u php-src/ext/standard/string.c:1.529 php-src/ext/standard/string.c:1.530 --- php-src/ext/standard/string.c:1.529 Wed Mar 15 12:20:49 2006 +++ php-src/ext/standard/string.c Fri Mar 17 14:29:05 2006 @@ -18,7 +18,7 @@ +----------------------------------------------------------------------+ */ -/* $Id: string.c,v 1.529 2006/03/15 12:20:49 derick Exp $ */ +/* $Id: string.c,v 1.530 2006/03/17 14:29:05 derick Exp $ */ /* Synced with php 3.0 revision 1.193 1999-06-16 [ssb] */ @@ -3263,8 +3263,8 @@ Makes an Unicode string's first character uppercase */ static void php_u_ucfirst(zval *ustr, zval *return_value TSRMLS_DC) { - UChar tmp[3] = { 0,}; /* UChar32 will be converted to upto 2 UChar units ? */ - int tmp_len = 0; + UChar tmp[3] = { 0, 0, 0 }; /* UChar32 will be converted to upto 2 UChar units ? */ + int tmp_len = 2; int pos = 0; UErrorCode status = U_ZERO_ERROR; @@ -3427,6 +3427,181 @@ } /* }}} */ +/* {{{ php_u_strtr + */ +PHPAPI UChar *php_u_strtr(UChar *str, int len, UChar *str_from, int str_from_len, UChar *str_to, int str_to_len, int trlen) +{ + int i, j; + int can_optimize = 1; + + if ((trlen < 1) || (len < 1)) { + return str; + } + + /* First loop to see if we can use the optimized version */ + for (i = 0; i < trlen; i++) { + if (str_from[i] > 255 || str_to[i] > 255) { + can_optimize = 0; + break; + } + } + if (can_optimize) { + for (i = trlen; i < str_from_len; i++) { + if (str_from[i] > 255) { + can_optimize = 0; + break; + } + } + } + if (can_optimize) { + for (i = trlen; i < str_to_len; i++) { + if (str_from[i] > 255) { + can_optimize = 0; + break; + } + } + } + + if (can_optimize) { + UChar xlat[256]; + + for (i = 0; i < 256; xlat[i] = i, i++); + + for (i = 0; i < trlen; i++) { + xlat[str_from[i]] = str_to[i]; + } + + for (i = 0; i < len; i++) { + str[i] = xlat[str[i]]; + } + + return str; + } else { + /* UTODO: We're quite fucked... this is *extremely* slow, better + * algorithm wanted here! It also doesn't handle combining sequences, I + * asked the icu-support list for good algorithms. */ + for (i = 0; i < len; i++) { + for (j = 0; j < trlen; j++) { + if (str[i] == str_from[j]) { + str[i] = str_to[j]; + } + } + } + return str; + } +} +/* }}} */ + +/* {{{ php_u_strtr_array + */ +static void php_u_strtr_array(zval *return_value, UChar *str, int slen, HashTable *hash) +{ + zval **entry; + zstr string_key; + uint string_key_len; + zval **trans; + zval ctmp; + ulong num_key; + int minlen = 128*1024; + int maxlen = 0, pos, len, found; + UChar *key; + HashPosition hpos; + smart_str result = {0}; + HashTable tmp_hash; + + zend_hash_init(&tmp_hash, 0, NULL, NULL, 0); + zend_hash_internal_pointer_reset_ex(hash, &hpos); + while (zend_hash_get_current_data_ex(hash, (void **)&entry, &hpos) == SUCCESS) { + switch (zend_hash_get_current_key_ex(hash, &string_key, &string_key_len, &num_key, 0, &hpos)) { + case HASH_KEY_IS_UNICODE: + len = string_key_len-1; + if (len < 1) { + zend_hash_destroy(&tmp_hash); + RETURN_FALSE; + } + zend_u_hash_add(&tmp_hash, IS_UNICODE, string_key, string_key_len, entry, sizeof(zval*), NULL); + if (len > maxlen) { + maxlen = len; + } + if (len < minlen) { + minlen = len; + } + break; + + case HASH_KEY_IS_LONG: + Z_TYPE(ctmp) = IS_LONG; + Z_LVAL(ctmp) = num_key; + + convert_to_unicode(&ctmp); + len = Z_USTRLEN(ctmp); + zend_u_hash_add(&tmp_hash, IS_UNICODE, Z_UNIVAL(ctmp), len+1, entry, sizeof(zval*), NULL); + zval_dtor(&ctmp); + + if (len > maxlen) { + maxlen = len; + } + if (len < minlen) { + minlen = len; + } + break; + } + zend_hash_move_forward_ex(hash, &hpos); + } + + key = eumalloc(maxlen+1); + pos = 0; + + while (pos < slen) { + if ((pos + maxlen) > slen) { + maxlen = slen - pos; + } + + found = 0; + memcpy(key, str+pos, UBYTES(maxlen)); + + for (len = maxlen; len >= minlen; len--) { + key[len] = 0; + + if (zend_u_hash_find(&tmp_hash, IS_UNICODE, ZSTR(key), len+1, (void**)&trans) == SUCCESS) { + UChar *tval; + int tlen; + zval tmp; + + if (Z_TYPE_PP(trans) != IS_UNICODE) { + tmp = **trans; + zval_copy_ctor(&tmp); + convert_to_string(&tmp); + tval = Z_USTRVAL(tmp); + tlen = Z_USTRLEN(tmp); + } else { + tval = Z_USTRVAL_PP(trans); + tlen = Z_USTRLEN_PP(trans); + } + + smart_str_appendl(&result, tval, UBYTES(tlen)); + pos += len; + found = 1; + + if (Z_TYPE_PP(trans) != IS_UNICODE) { + zval_dtor(&tmp); + } + break; + } + } + + if (! found) { + smart_str_append2c(&result, str[pos]); + pos++; + } + } + + efree(key); + zend_hash_destroy(&tmp_hash); + smart_str_0(&result); + RETVAL_UNICODEL((UChar *) result.c, result.len >> 1, 0); +} +/* }}} */ + /* {{{ php_strtr_array */ static void php_strtr_array(zval *return_value, char *str, int slen, HashTable *hash) @@ -3552,27 +3727,52 @@ RETURN_FALSE; } - convert_to_string_ex(str); + if (Z_TYPE_PP(str) != IS_UNICODE && Z_TYPE_PP(str) != IS_STRING) { + convert_to_text_ex(str); + } /* shortcut for empty string */ - if (Z_STRLEN_PP(str) == 0) { + if (Z_TYPE_PP(str) == IS_UNICODE && !Z_USTRLEN_PP(str)) { + RETURN_EMPTY_UNICODE(); + } else if (!Z_STRLEN_PP(str)) { RETURN_EMPTY_STRING(); } - if (ac == 2) { - php_strtr_array(return_value, Z_STRVAL_PP(str), Z_STRLEN_PP(str), HASH_OF(*from)); + if (Z_TYPE_PP(str) == IS_UNICODE) { + if (ac == 2) { + php_u_strtr_array(return_value, Z_USTRVAL_PP(str), Z_USTRLEN_PP(str), HASH_OF(*from)); + Z_TYPE_P(return_value) = IS_UNICODE; + } else { + convert_to_unicode_ex(from); + convert_to_unicode_ex(to); + + ZVAL_UNICODEL(return_value, Z_USTRVAL_PP(str), Z_USTRLEN_PP(str), 1); + + php_u_strtr(Z_USTRVAL_P(return_value), + Z_USTRLEN_P(return_value), + Z_USTRVAL_PP(from), + Z_USTRLEN_PP(from), + Z_USTRVAL_PP(to), + Z_USTRLEN_PP(to), + MIN(Z_USTRLEN_PP(from), Z_USTRLEN_PP(to))); + Z_TYPE_P(return_value) = IS_UNICODE; + } } else { - convert_to_string_ex(from); - convert_to_string_ex(to); + if (ac == 2) { + php_strtr_array(return_value, Z_STRVAL_PP(str), Z_STRLEN_PP(str), HASH_OF(*from)); + } else { + convert_to_string_ex(from); + convert_to_string_ex(to); - ZVAL_STRINGL(return_value, Z_STRVAL_PP(str), Z_STRLEN_PP(str), 1); - - php_strtr(Z_STRVAL_P(return_value), - Z_STRLEN_P(return_value), - Z_STRVAL_PP(from), - Z_STRVAL_PP(to), - MIN(Z_STRLEN_PP(from), - Z_STRLEN_PP(to))); + ZVAL_STRINGL(return_value, Z_STRVAL_PP(str), Z_STRLEN_PP(str), 1); + + php_strtr(Z_STRVAL_P(return_value), + Z_STRLEN_P(return_value), + Z_STRVAL_PP(from), + Z_STRVAL_PP(to), + MIN(Z_STRLEN_PP(from), + Z_STRLEN_PP(to))); + } } } /* }}} */ http://cvs.php.net/viewcvs.cgi/php-src/ext/standard/type.c?r1=1.40&r2=1.41&diff_format=u Index: php-src/ext/standard/type.c diff -u php-src/ext/standard/type.c:1.40 php-src/ext/standard/type.c:1.41 --- php-src/ext/standard/type.c:1.40 Tue Feb 21 20:12:42 2006 +++ php-src/ext/standard/type.c Fri Mar 17 14:29:05 2006 @@ -16,7 +16,7 @@ +----------------------------------------------------------------------+ */ -/* $Id: type.c,v 1.40 2006/02/21 20:12:42 dmitry Exp $ */ +/* $Id: type.c,v 1.41 2006/03/17 14:29:05 derick Exp $ */ #include "php.h" #include "php_incomplete_class.h" @@ -285,11 +285,19 @@ } /* }}} */ +/* {{{ proto bool is_binary(mixed var) + Returns true if variable is a native (binary) string */ +PHP_FUNCTION(is_binary) +{ + php_is_type(INTERNAL_FUNCTION_PARAM_PASSTHRU, IS_STRING); +} +/* }}} */ + /* {{{ proto bool is_string(mixed var) - Returns true if variable is a native string */ + Returns true if variable is a string */ PHP_FUNCTION(is_string) { - php_is_type(INTERNAL_FUNCTION_PARAM_PASSTHRU, IS_STRING); + php_is_type(INTERNAL_FUNCTION_PARAM_PASSTHRU, UG(unicode) ? IS_UNICODE : IS_STRING); } /* }}} */ http://cvs.php.net/viewcvs.cgi/php-src/ext/standard/tests/strings/strtr2.phpt?view=markup&rev=1.1 Index: php-src/ext/standard/tests/strings/strtr2.phpt +++ php-src/ext/standard/tests/strings/strtr2.phpt --TEST-- strtr() function (with unicode chars and combining sequences) --FILE-- <?php declare(encoding="utf8"); $from = "aåаиу"; $to = "12356"; $string = "Dе åkаt krаpt de krуllen van de trap af"; var_dump( strtr( $string, $from, $to ) ); $from = "aeoui"; $to = "12356"; $string = "De akat krapt de krullen van de trap af"; var_dump( strtr( $string, $from, $to ) ); $ar = array( "a" => 1, "e" => "2", "o" => 3, "u" => 5, "i" => 6 ); $string = "De akat krapt de krullen van de trap af"; var_dump( strtr( $string, $ar ) ); // Test with combining chars $from = "åb"; $to = "1"; $string = "xyzabc"; var_dump( strtr( $string, $from, $to ) ); ?> --EXPECT-- unicode(39) "Dе 2k3t kr3pt de kr6llen v1n de tr1p 1f" unicode(39) "D2 1k1t kr1pt d2 kr5ll2n v1n d2 tr1p 1f" unicode(39) "D2 1k1t kr1pt d2 kr5ll2n v1n d2 tr1p 1f" http://cvs.php.net/viewcvs.cgi/php-src/ext/standard/tests/strings/ucfirst.phpt?view=markup&rev=1.1 Index: php-src/ext/standard/tests/strings/ucfirst.phpt +++ php-src/ext/standard/tests/strings/ucfirst.phpt --TEST-- ucfirst() --FILE-- <?php echo ucfirst("peren"), "\n"; echo ucfirst("appelen"), "\n"; echo ucfirst("ßen"), "\n"; ?> --EXPECT-- Peren Appelen SSen