RE: [PHP-DEV] Unicode and fetch class

Looks OK to me. Can we call it ZEND_FETCH_CLASS_NO_LOWERCASE instead of ZEND_FETCH_CLASS_NO_NORMALIZE. No biggy but I always think of normalizing as in taking a Unicode string and normalizing it. Or is that what you're actually doing here? Asking because we agreed we won't normalize automagically in PHP. Andi > -----Original Message----- > From: Marcus Boerger [mailto:helly@php.net] > Sent: Monday, July 17, 2006 2:49 PM > To: Andrei Zmievski > Cc: Marcus Boerger; Andrei Zmievski; Zeev Suraski; Andi > Gutmans; internals@lists.php.net > Subject: Re: [PHP-DEV] Unicode and fetch class > > Hello Andrei, > > looks pretty fine to me, after your treatment, go ahead > commit - Andi, Zeev? > > best regards > marcus > > Monday, July 17, 2006, 9:50:44 PM, you wrote: > > > Index: Zend/zend_API.c > > =================================================================== > > RCS file: /repository/ZendEngine2/zend_API.c,v > > retrieving revision 1.378 > > diff -p -u -r1.378 zend_API.c > > --- Zend/zend_API.c 13 Jul 2006 21:27:48 -0000 1.378 > > +++ Zend/zend_API.c 17 Jul 2006 19:47:30 -0000 > > @@ -2578,10 +2578,9 @@ ZEND_API int zend_disable_class(char *cl > > static int zend_is_callable_check_func(int check_flags, zval > > ***zobj_ptr_ptr, zend_class_entry *ce_org, zval *callable, > > zend_class_entry **ce_ptr, zend_function **fptr_ptr TSRMLS_DC) { > > int retval; > > - zstr lcname, lmname, mname, colon; > > + zstr lmname, mname, colon; > > unsigned int clen, mlen; > > zend_function *fptr; > > - zend_class_entry **pce; > > HashTable *ftable; > > > > *ce_ptr = NULL; > > @@ -2601,18 +2600,7 @@ static int zend_is_callable_check_func(i > > } > > } > > if (colon.v != NULL) { > > - lcname = zend_u_str_case_fold(Z_TYPE_P(callable), > > Z_UNIVAL_P(callable), clen, 0, &clen); > > - /* caution: lcname is not '\0' terminated */ > > - if (clen == sizeof("self") - 1 && > > - ZEND_U_EQUAL(Z_TYPE_P(callable), > lcname, clen, "self", sizeof("self")-1)) { > > - *ce_ptr = EG(scope); > > - } else if (clen == sizeof("parent") - 1 && > > - ZEND_U_EQUAL(Z_TYPE_P(callable), > lcname, clen, "parent", sizeof("parent")-1)) { > > - *ce_ptr = EG(scope) ? > EG(scope)->parent : NULL; > > - } else if (zend_u_lookup_class(Z_TYPE_P(callable), > > Z_UNIVAL_P(callable), clen, &pce TSRMLS_CC) == SUCCESS) { > > - *ce_ptr = *pce; > > - } > > - efree(lcname.v); > > + *ce_ptr = zend_u_fetch_class(Z_TYPE_P(callable), > > Z_UNIVAL_P(callable), clen, ZEND_FETCH_CLASS_AUTO TSRMLS_CC); > > if (!*ce_ptr) { > > return 0; > > } > > @@ -2620,9 +2608,9 @@ static int zend_is_callable_check_func(i > > if (ce_org && !instanceof_function(ce_org, > *ce_ptr TSRMLS_CC)) { > > return 0; > > } > > - lmname = > zend_u_str_case_fold(Z_TYPE_P(callable), mname, mlen, 0, &mlen); > > + lmname = zend_u_str_case_fold(Z_TYPE_P(callable), > > + mname, mlen, 1, &mlen); > > } else { > > - lmname = zend_u_str_case_fold(Z_TYPE_P(callable), > > Z_UNIVAL_P(callable), Z_UNILEN_P(callable), 0, &mlen); > > + lmname = zend_u_str_case_fold(Z_TYPE_P(callable), > > Z_UNIVAL_P(callable), Z_UNILEN_P(callable), 1, &mlen); > > if (ce_org) { > > ftable = &ce_org->function_table; > > *ce_ptr = ce_org; > > Index: Zend/zend_builtin_functions.c > > =================================================================== > > RCS file: /repository/ZendEngine2/zend_builtin_functions.c,v > > retrieving revision 1.317 > > diff -p -u -r1.317 zend_builtin_functions.c > > --- Zend/zend_builtin_functions.c 9 Jul 2006 22:40:10 > -0000 1.317 > > +++ Zend/zend_builtin_functions.c 17 Jul 2006 19:47:30 -0000 > > @@ -702,7 +702,7 @@ static void is_a_impl(INTERNAL_FUNCTION_ > > > > convert_to_text_ex(class_name); > > > > - if (zend_u_lookup_class_ex(Z_TYPE_PP(class_name), > > Z_UNIVAL_PP(class_name), Z_UNILEN_PP(class_name), (instance_ce != > > NULL), &ce TSRMLS_CC) == FAILURE) { > > + if (zend_u_lookup_class_ex(Z_TYPE_PP(class_name), > > Z_UNIVAL_PP(class_name), Z_UNILEN_PP(class_name), > (instance_ce != NULL) ? > > 1 : 0, 1, &ce TSRMLS_CC) == FAILURE) { > > retval = 0; > > } else { > > if (only_subclass) { > > Index: Zend/zend_compile.h > > =================================================================== > > RCS file: /repository/ZendEngine2/zend_compile.h,v > > retrieving revision 1.343 > > diff -p -u -r1.343 zend_compile.h > > --- Zend/zend_compile.h 13 Jun 2006 12:56:20 -0000 1.343 > > +++ Zend/zend_compile.h 17 Jul 2006 19:47:30 -0000 > > @@ -615,7 +615,9 @@ int zendlex(znode *zendlval TSRMLS_DC); > > #define ZEND_FETCH_CLASS_GLOBAL 4 > > #define ZEND_FETCH_CLASS_AUTO 5 > > #define ZEND_FETCH_CLASS_INTERFACE 6 > > -#define ZEND_FETCH_CLASS_NO_AUTOLOAD 0x80 > > +#define ZEND_FETCH_CLASS_FLAGS 0xF0 > > +#define ZEND_FETCH_CLASS_NO_NORMALIZE 0x10 #define > > +ZEND_FETCH_CLASS_NO_AUTOLOAD 0x80 > > > > /* variable parsing type (compile-time) */ > > #define ZEND_PARSED_MEMBER (1<<0) > > Index: Zend/zend_execute.c > > =================================================================== > > RCS file: /repository/ZendEngine2/zend_execute.c,v > > retrieving revision 1.748 > > diff -p -u -r1.748 zend_execute.c > > --- Zend/zend_execute.c 10 Jul 2006 00:32:23 -0000 1.748 > > +++ Zend/zend_execute.c 17 Jul 2006 19:47:30 -0000 > > @@ -944,7 +944,7 @@ fetch_string_dim: > > UChar *norm; > > int norm_len; > > > > - if > (!zend_normalize_identifier(&norm, > > &norm_len, offset_key.u, offset_key_length, 0)) { > > + if (zend_normalize_identifier(&norm, > > &norm_len, offset_key.u, offset_key_length, 0) == FAILURE) { > > > zend_error(E_WARNING, "Could > > not normalize identifier: %r", offset_key); > > } else if (norm != offset_key.u) { > > offset_key.u = norm; > > Index: Zend/zend_execute.h > > =================================================================== > > RCS file: /repository/ZendEngine2/zend_execute.h,v > > retrieving revision 1.100 > > diff -p -u -r1.100 zend_execute.h > > --- Zend/zend_execute.h 31 May 2006 13:02:15 -0000 1.100 > > +++ Zend/zend_execute.h 17 Jul 2006 19:47:30 -0000 > > @@ -71,7 +71,7 @@ static inline void safe_free_zval_ptr_re } > > ZEND_API int zend_lookup_class(char *name, int name_length, > > zend_class_entry ***ce TSRMLS_DC); ZEND_API int > > zend_u_lookup_class(zend_uchar type, zstr name, int name_length, > > zend_class_entry ***ce TSRMLS_DC); -ZEND_API int > > zend_u_lookup_class_ex(zend_uchar type, zstr name, int name_length, > > int use_autoload, zend_class_entry ***ce TSRMLS_DC); > > +ZEND_API int zend_u_lookup_class_ex(zend_uchar type, zstr name, int > > name_length, int use_autoload, int do_normalize, zend_class_entry > > ***ce TSRMLS_DC); ZEND_API int zend_eval_string(char *str, zval > > *retval_ptr, char *string_name TSRMLS_DC); ZEND_API int > > zend_eval_string_ex(char *str, zval *retval_ptr, char *string_name, > > int handle_exceptions TSRMLS_DC); ZEND_API int > > zend_u_eval_string(zend_uchar type, zstr str, zval > *retval_ptr, char > > *string_name TSRMLS_DC); > > Index: Zend/zend_execute_API.c > > =================================================================== > > RCS file: /repository/ZendEngine2/zend_execute_API.c,v > > retrieving revision 1.375 > > diff -p -u -r1.375 zend_execute_API.c > > --- Zend/zend_execute_API.c 12 Jul 2006 07:54:18 -0000 > 1.375 > > +++ Zend/zend_execute_API.c 17 Jul 2006 19:47:30 -0000 > > @@ -784,7 +784,7 @@ int zend_call_function(zend_fcall_info * > > } > > > > if (Z_TYPE_P(fci->function_name) == IS_UNICODE) { > > - if ((colon.u = > > u_strstr(Z_USTRVAL_P(fci->function_name), > (UChar*)":\0:\0")) != NULL) > > { > > + if ((colon.u = > > u_strstr(Z_USTRVAL_P(fci->function_name), u_doublecolon)) != NULL) { > > fname_len = u_strlen(colon.u+2); > > clen = > Z_USTRLEN_P(fci->function_name) - fname_len - 2; > > fname.u = colon.u + 2; @@ -799,7 > > +799,7 @@ int zend_call_function(zend_fcall_info * > > if (colon.v != NULL) { > > zend_class_entry **pce, *ce_child = NULL; > > > > - lcname = > > zend_u_str_case_fold(Z_TYPE_P(fci->function_name), > > Z_UNIVAL_P(fci->function_name), clen, 0, &clen); > > + lcname = > > zend_u_str_case_fold(Z_TYPE_P(fci->function_name), > > Z_UNIVAL_P(fci->function_name), clen, 1, &clen); > > /* caution: lcname is not '\0' terminated */ > > if (calling_scope && clen == > sizeof("self") - 1 && > > > ZEND_U_EQUAL(Z_TYPE_P(fci->function_name), > > lcname, clen, "self", sizeof("self")-1)) { @@ -807,7 +807,7 @@ int > > zend_call_function(zend_fcall_info * > > } else if (calling_scope && clen == > sizeof("parent") - 1 && > > > ZEND_U_EQUAL(Z_TYPE_P(fci->function_name), > > lcname, clen, "parent", sizeof("parent")-1)) { > > ce_child = EG(active_op_array) && > > EG(active_op_array)->scope ? EG(scope)->parent : NULL; > > - } else if > > (zend_u_lookup_class(Z_TYPE_P(fci->function_name), > > Z_UNIVAL_P(fci->function_name), clen, &pce TSRMLS_CC) == SUCCESS) { > > + } else if > > (zend_u_lookup_class_ex(Z_TYPE_P(fci->function_name), > lcname, clen, 1, 0, &pce TSRMLS_CC) == SUCCESS) { > > ce_child = *pce; > > } > > efree(lcname.v); @@ -1062,7 +1062,7 @@ int > > zend_call_function(zend_fcall_info * } > > > > > > -ZEND_API int zend_u_lookup_class_ex(zend_uchar type, zstr > name, int > > name_length, int use_autoload, zend_class_entry ***ce TSRMLS_DC) > > +ZEND_API int zend_u_lookup_class_ex(zend_uchar type, zstr name, int > > name_length, int use_autoload, int do_normalize, zend_class_entry > > ***ce TSRMLS_DC) { > > zval **args[1]; > > zval autoload_function; > > @@ -1080,10 +1080,17 @@ ZEND_API int zend_u_lookup_class_ex(zend > > return FAILURE; > > } > > > > - lc_name = zend_u_str_case_fold(type, name, > name_length, 1, &lc_name_len); > > + if (do_normalize) { > > + lc_name = zend_u_str_case_fold(type, name, > name_length, 1, &lc_name_len); > > + } else { > > + lc_name = name; > > + lc_name_len = name_length; > > + } > > > > if (zend_u_hash_find(EG(class_table), type, lc_name, > > lc_name_len+1, (void **) ce) == SUCCESS) { > > - efree(lc_name.v); > > + if (do_normalize) { > > + efree(lc_name.v); > > + } > > return SUCCESS; > > } > > > > @@ -1091,7 +1098,9 @@ ZEND_API int zend_u_lookup_class_ex(zend > > * (doesn't impact fuctionality of __autoload() > > */ > > if (!use_autoload || zend_is_compiling(TSRMLS_C)) { > > - efree(lc_name.v); > > + if (do_normalize) { > > + efree(lc_name.v); > > + } > > return FAILURE; > > } > > > > @@ -1101,7 +1110,9 @@ ZEND_API int zend_u_lookup_class_ex(zend > > } > > > > if (zend_u_hash_add(EG(in_autoload), type, lc_name, > > lc_name_len+1, (void**)&dummy, sizeof(char), NULL) == FAILURE) { > > - efree(lc_name.v); > > + if (do_normalize) { > > + efree(lc_name.v); > > + } > > return FAILURE; > > } > > > > @@ -1143,12 +1154,16 @@ ZEND_API int zend_u_lookup_class_ex(zend > > > > if (retval == FAILURE) { > > EG(exception) = exception; > > - efree(lc_name.v); > > + if (do_normalize) { > > + efree(lc_name.v); > > + } > > return FAILURE; > > } > > > > if (EG(exception) && exception) { > > - efree(lc_name.v); > > + if (do_normalize) { > > + efree(lc_name.v); > > + } > > zend_error(E_ERROR, "Function %s(%R) threw an > > exception of type '%v'", ZEND_AUTOLOAD_FUNC_NAME, type, > name, Z_OBJCE_P(EG(exception))->name); > > return FAILURE; > > } > > @@ -1160,13 +1175,15 @@ ZEND_API int zend_u_lookup_class_ex(zend > > } > > > > retval = zend_u_hash_find(EG(class_table), type, > lc_name, lc_name_len + 1, (void **) ce); > > - efree(lc_name.v); > > + if (do_normalize) { > > + efree(lc_name.v); > > + } > > return retval; > > } > > > > ZEND_API int zend_u_lookup_class(zend_uchar type, zstr name, int > > name_length, zend_class_entry ***ce TSRMLS_DC) { > > - return zend_u_lookup_class_ex(type, name, > name_length, 1, ce TSRMLS_CC); > > + return zend_u_lookup_class_ex(type, name, > name_length, 1, 1, > > + ce TSRMLS_CC); > > } > > > > ZEND_API int zend_lookup_class(char *name, int name_length, > > zend_class_entry ***ce TSRMLS_DC) @@ -1545,9 +1562,12 @@ void > > zend_unset_timeout(TSRMLS_D) ZEND_API zend_class_entry > > *zend_u_fetch_class(zend_uchar type, zstr class_name, uint > > class_name_len, int fetch_type TSRMLS_DC) { > > zend_class_entry **pce; > > - int use_autoload = (fetch_type & > ZEND_FETCH_CLASS_NO_AUTOLOAD) == 0; > > + int use_autoload = (fetch_type & > ZEND_FETCH_CLASS_NO_AUTOLOAD) ? 0 : 1; > > + int do_normalize = (fetch_type & > ZEND_FETCH_CLASS_NO_NORMALIZE) ? 0 : 1; > > + zstr lcname = class_name; > > + > > + fetch_type = fetch_type & ~ZEND_FETCH_CLASS_FLAGS; > > > > - fetch_type = fetch_type & ~ZEND_FETCH_CLASS_NO_AUTOLOAD; > > check_fetch_type: > > switch (fetch_type) { > > case ZEND_FETCH_CLASS_SELF: > > @@ -1564,15 +1584,22 @@ check_fetch_type: > > } > > return EG(scope)->parent; > > case ZEND_FETCH_CLASS_AUTO: { > > - fetch_type = > > zend_get_class_fetch_type(type, class_name, class_name_len); > > + if (do_normalize) { > > + lcname = > > zend_u_str_case_fold(type, class_name, class_name_len, 1, > > &class_name_len); > > + } > > + fetch_type = > > zend_get_class_fetch_type(type, lcname, class_name_len); > > if > > (fetch_type!=ZEND_FETCH_CLASS_DEFAULT) { > > + if (do_normalize) { > > + efree(lcname.v); > > + do_normalize = 0; /* > > we've normalized it already, don't do it twice */ > > + } > > goto check_fetch_type; > > } > > } > > break; > > } > > > > - if (zend_u_lookup_class_ex(type, class_name, class_name_len, > > use_autoload, &pce TSRMLS_CC)==FAILURE) { > > + if (zend_u_lookup_class_ex(type, lcname, class_name_len, > > use_autoload, do_normalize, &pce TSRMLS_CC)==FAILURE) { > > if (use_autoload) { > > if (fetch_type == > ZEND_FETCH_CLASS_INTERFACE) { > > zend_error(E_ERROR, "Interface '%R' > > not found", type, class_name); @@ -1580,8 +1607,14 @@ > check_fetch_type: > > zend_error(E_ERROR, "Class > '%R' not found", type, class_name); > > } > > } > > + if (lcname.v != class_name.v) { > > + efree(lcname.v); > > + } > > return NULL; > > } else { > > + if (lcname.v != class_name.v) { > > + efree(lcname.v); > > + } > > return *pce; > > } > > } > > Index: Zend/zend_language_scanner.l > > =================================================================== > > RCS file: /repository/ZendEngine2/zend_language_scanner.l,v > > retrieving revision 1.155 > > diff -p -u -r1.155 zend_language_scanner.l > > --- Zend/zend_language_scanner.l 12 Jun 2006 > 17:06:39 -0000 1.155 > > +++ Zend/zend_language_scanner.l 17 Jul 2006 19:47:30 -0000 > > @@ -444,7 +444,7 @@ static inline int zend_check_and_normali > > efree(Z_USTRVAL_P(zendlval)); > > return 0; > > } > > - if (!zend_normalize_identifier(&norm, &norm_len, > > Z_USTRVAL_P(zendlval), Z_USTRLEN_P(zendlval), 0)) { > > + if (zend_normalize_identifier(&norm, &norm_len, > > Z_USTRVAL_P(zendlval), Z_USTRLEN_P(zendlval), 0) == FAILURE) { > > zend_error(E_COMPILE_WARNING, "Could not normalize > > identifier: %r", Z_USTRVAL_P(zendlval)); > > efree(Z_USTRVAL_P(zendlval)); > > return 0; > > Index: Zend/zend_operators.c > > =================================================================== > > RCS file: /repository/ZendEngine2/zend_operators.c,v > > retrieving revision 1.248 > > diff -p -u -r1.248 zend_operators.c > > --- Zend/zend_operators.c 12 Jul 2006 19:34:55 -0000 > 1.248 > > +++ Zend/zend_operators.c 17 Jul 2006 19:47:30 -0000 > > @@ -2109,14 +2109,21 @@ ZEND_API zstr zend_u_str_case_fold(zend_ > > zstr ret; > > > > if (type == IS_UNICODE) { > > - int ret_len; > > + int ret_len = length; > > > > if (normalize) { > > - zend_normalize_identifier(&ret.u, > &ret_len, source.u, length, 1); > > + if > (zend_normalize_identifier(&ret.u, &ret_len, source.u, > length, 1) == FAILURE) { > > + zend_error(E_NOTICE, "Could > not normalize identifier"); > > + ret.u = eustrndup(source.u, length); > > + } > > } else { > > UErrorCode status = U_ZERO_ERROR; > > > > zend_case_fold_string(&ret.u, &ret_len, > > source.u, length, U_FOLD_CASE_DEFAULT, &status); > > + if (U_FAILURE(status)) { > > + zend_error(E_NOTICE, "Could > not case-fold string"); > > + ret.u = eustrndup(source.u, length); > > + } > > } > > > > *new_len = ret_len; > > Index: Zend/zend_unicode.c > > =================================================================== > > RCS file: /repository/ZendEngine2/zend_unicode.c,v > > retrieving revision 1.25 > > diff -p -u -r1.25 zend_unicode.c > > --- Zend/zend_unicode.c 12 Jul 2006 17:35:06 -0000 1.25 > > +++ Zend/zend_unicode.c 17 Jul 2006 19:47:30 -0000 > > @@ -736,7 +736,7 @@ ZEND_API int zend_normalize_identifier(U > > if (unorm_quickCheck(ident, ident_len, UNORM_NFKC, > &status) != UNORM_YES) { > > zend_normalize_string(&buffer, &buffer_len, > ident, ident_len, &status); > > if (U_FAILURE(status)) { > > - return 0; > > + return FAILURE; > > } > > ident = buffer; > > ident_len = buffer_len; @@ -748,7 +748,7 @@ > ZEND_API > > int zend_normalize_identifier(U > > efree(ident); > > } > > if (U_FAILURE(status)) { > > - return 0; > > + return FAILURE; > > } > > ident = buffer; > > ident_len = buffer_len; @@ -759,7 +759,7 @@ > ZEND_API > > int zend_normalize_identifier(U > > efree(ident); > > } > > if (U_FAILURE(status)) { > > - return 0; > > + return FAILURE; > > } > > ident = buffer; > > ident_len = buffer_len; @@ -768,7 +768,7 @@ > > ZEND_API int zend_normalize_identifier(U > > > > *dest = ident; > > *dest_len = ident_len; > > - return 1; > > + return SUCCESS; > > } > > /* }}} */ > > > > Index: ext/reflection/php_reflection.c > > =================================================================== > > RCS file: /repository/php-src/ext/reflection/php_reflection.c,v > > retrieving revision 1.243 > > diff -p -u -r1.243 php_reflection.c > > --- ext/reflection/php_reflection.c 13 Jul 2006 > 12:34:30 -0000 1.243 > > +++ ext/reflection/php_reflection.c 17 Jul 2006 19:47:31 -0000 > > @@ -1972,7 +1972,7 @@ ZEND_METHOD(reflection_parameter, getCla > > GET_REFLECTION_OBJECT_PTR(param); > > > > if (param->arg_info->class_name.v) { > > - if > > (zend_u_lookup_class_ex(UG(unicode)?IS_UNICODE:IS_STRING, > > param->arg_info->class_name, > param->arg_info->class_name_len, 1, &pce > > param->TSRMLS_CC) == FAILURE) { > > + if > > (zend_u_lookup_class_ex(UG(unicode)?IS_UNICODE:IS_STRING, > > param->arg_info->class_name, param->arg_info->class_name_len, 1, 1, > > param->&pce TSRMLS_CC) == FAILURE) { > > > zend_throw_exception_ex(reflection_exception_ptr, 0 TSRMLS_CC, > > "Class %v does not exist", > param->arg_info->class_name); > > return; > > > > > Best regards, > Marcus >