Skip to content

Implement token_get_all() TOKEN_AS_OBJECT mode #2430

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 1 commit into from
Closed

Implement token_get_all() TOKEN_AS_OBJECT mode #2430

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
155 changes: 155 additions & 0 deletions ext/tokenizer/tests/token_get_all_TOKEN_AS_OBJECT_001.phpt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,155 @@
--TEST--
TOKEN_AS_OBJECT mode
--FILE--
<?php

$code = <<<'PHP'
<?php
function foo() {
echo "bar";
}
PHP;
var_dump(token_get_all($code, TOKEN_AS_OBJECT));

?>
--EXPECT--
array(15) {
[0]=>
object(PhpToken)#1 (3) {
["type"]=>
int(379)
["text"]=>
string(6) "<?php
"
["line"]=>
int(1)
}
[1]=>
object(PhpToken)#2 (3) {
["type"]=>
int(346)
["text"]=>
string(8) "function"
["line"]=>
int(2)
}
[2]=>
object(PhpToken)#3 (3) {
["type"]=>
int(382)
["text"]=>
string(1) " "
["line"]=>
int(2)
}
[3]=>
object(PhpToken)#4 (3) {
["type"]=>
int(319)
["text"]=>
string(3) "foo"
["line"]=>
int(2)
}
[4]=>
object(PhpToken)#5 (3) {
["type"]=>
int(40)
["text"]=>
string(1) "("
["line"]=>
int(2)
}
[5]=>
object(PhpToken)#6 (3) {
["type"]=>
int(41)
["text"]=>
string(1) ")"
["line"]=>
int(2)
}
[6]=>
object(PhpToken)#7 (3) {
["type"]=>
int(382)
["text"]=>
string(1) " "
["line"]=>
int(2)
}
[7]=>
object(PhpToken)#8 (3) {
["type"]=>
int(123)
["text"]=>
string(1) "{"
["line"]=>
int(2)
}
[8]=>
object(PhpToken)#9 (3) {
["type"]=>
int(382)
["text"]=>
string(5) "
"
["line"]=>
int(2)
}
[9]=>
object(PhpToken)#10 (3) {
["type"]=>
int(328)
["text"]=>
string(4) "echo"
["line"]=>
int(3)
}
[10]=>
object(PhpToken)#11 (3) {
["type"]=>
int(382)
["text"]=>
string(1) " "
["line"]=>
int(3)
}
[11]=>
object(PhpToken)#12 (3) {
["type"]=>
int(323)
["text"]=>
string(5) ""bar""
["line"]=>
int(3)
}
[12]=>
object(PhpToken)#13 (3) {
["type"]=>
int(59)
["text"]=>
string(1) ";"
["line"]=>
int(3)
}
[13]=>
object(PhpToken)#14 (3) {
["type"]=>
int(382)
["text"]=>
string(1) "
"
["line"]=>
int(3)
}
[14]=>
object(PhpToken)#15 (3) {
["type"]=>
int(125)
["text"]=>
string(1) "}"
["line"]=>
int(4)
}
}
82 changes: 60 additions & 22 deletions ext/tokenizer/tokenizer.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,10 +38,14 @@
#define zendcursor LANG_SCNG(yy_cursor)
#define zendlimit LANG_SCNG(yy_limit)

#define TOKEN_PARSE 1
#define TOKEN_PARSE (1 << 0)
#define TOKEN_AS_OBJECT (1 << 1)

zend_class_entry *php_token_ce;

void tokenizer_token_get_all_register_constants(INIT_FUNC_ARGS) {
REGISTER_LONG_CONSTANT("TOKEN_PARSE", TOKEN_PARSE, CONST_CS|CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("TOKEN_AS_OBJECT", TOKEN_AS_OBJECT, CONST_CS|CONST_PERSISTENT);
}

/* {{{ arginfo */
Expand Down Expand Up @@ -90,8 +94,17 @@ ZEND_GET_MODULE(tokenizer)
*/
PHP_MINIT_FUNCTION(tokenizer)
{
zend_class_entry ce;

tokenizer_register_constants(INIT_FUNC_ARGS_PASSTHRU);
tokenizer_token_get_all_register_constants(INIT_FUNC_ARGS_PASSTHRU);

INIT_CLASS_ENTRY(ce, "PhpToken", NULL);
php_token_ce = zend_register_internal_class(&ce);
zend_declare_property_null(php_token_ce, "type", sizeof("type")-1, ZEND_ACC_PUBLIC);
zend_declare_property_null(php_token_ce, "text", sizeof("text")-1, ZEND_ACC_PUBLIC);
zend_declare_property_null(php_token_ce, "line", sizeof("line")-1, ZEND_ACC_PUBLIC);

return SUCCESS;
}
/* }}} */
Expand All @@ -106,25 +119,37 @@ PHP_MINFO_FUNCTION(tokenizer)
}
/* }}} */

static inline zend_string *make_str(unsigned char *text, size_t leng) {
if (leng == 1) {
return ZSTR_CHAR(text[0]);
} else {
return zend_string_init((char *) text, leng, 0);
}
}

static void add_token(zval *return_value, int token_type,
unsigned char *text, size_t leng, int lineno) {
if (token_type >= 256) {
unsigned char *text, size_t leng, int lineno, zend_bool as_object) {
if (as_object) {
zend_object *obj = zend_objects_new(php_token_ce);
zval obj_zv;
ZVAL_OBJ(&obj_zv, obj);
ZVAL_LONG(OBJ_PROP_NUM(obj, 0), token_type);
ZVAL_STR(OBJ_PROP_NUM(obj, 1), make_str(text, leng));
ZVAL_LONG(OBJ_PROP_NUM(obj, 2), lineno);
add_next_index_zval(return_value, &obj_zv);
} else if (token_type >= 256) {
zval keyword;
array_init(&keyword);
add_next_index_long(&keyword, token_type);
add_next_index_stringl(&keyword, (char *) text, leng);
add_next_index_str(&keyword, make_str(text, leng));
add_next_index_long(&keyword, lineno);
add_next_index_zval(return_value, &keyword);
} else {
if (leng == 1) {
add_next_index_str(return_value, ZSTR_CHAR(text[0]));
} else {
add_next_index_stringl(return_value, (char *) text, leng);
}
add_next_index_str(return_value, make_str(text, leng));
}
}

static zend_bool tokenize(zval *return_value, zend_string *source)
static zend_bool tokenize(zval *return_value, zend_string *source, zend_bool as_object)
{
zval source_zval;
zend_lex_state original_lex_state;
Expand All @@ -146,7 +171,7 @@ static zend_bool tokenize(zval *return_value, zend_string *source)

ZVAL_UNDEF(&token);
while ((token_type = lex_scan(&token))) {
add_token(return_value, token_type, zendtext, zendleng, token_line);
add_token(return_value, token_type, zendtext, zendleng, token_line, as_object);

if (Z_TYPE(token) != IS_UNDEF) {
zval_dtor(&token);
Expand All @@ -162,7 +187,7 @@ static zend_bool tokenize(zval *return_value, zend_string *source)
/* fetch the rest into a T_INLINE_HTML */
if (zendcursor != zendlimit) {
add_token(return_value, T_INLINE_HTML,
zendcursor, zendlimit - zendcursor, token_line);
zendcursor, zendlimit - zendcursor, token_line, as_object);
}
break;
}
Expand All @@ -184,34 +209,43 @@ static zend_bool tokenize(zval *return_value, zend_string *source)
return 1;
}

struct event_context {
zval *tokens;
zend_bool as_object;
};

void on_event(zend_php_scanner_event event, int token, int line, void *context)
{
zval *token_stream = (zval *) context;
struct event_context *ctx = context;
HashTable *tokens_ht;
zval *token_zv;

switch (event) {
case ON_TOKEN:
if (token == END) break;
add_token(token_stream, token, LANG_SCNG(yy_text), LANG_SCNG(yy_leng), line);
add_token(ctx->tokens, token,
LANG_SCNG(yy_text), LANG_SCNG(yy_leng), line, ctx->as_object);
break;
case ON_FEEDBACK:
tokens_ht = Z_ARRVAL_P(token_stream);
tokens_ht = Z_ARRVAL_P(ctx->tokens);
token_zv = zend_hash_index_find(tokens_ht, zend_hash_num_elements(tokens_ht) - 1);
if (token_zv && Z_TYPE_P(token_zv) == IS_ARRAY) {
ZEND_ASSERT(token_zv);
if (Z_TYPE_P(token_zv) == IS_ARRAY) {
ZVAL_LONG(zend_hash_index_find(Z_ARRVAL_P(token_zv), 0), token);
} else {
zend_update_property_long(php_token_ce, token_zv, "type", sizeof("type")-1, token);
}
break;
case ON_STOP:
if (LANG_SCNG(yy_cursor) != LANG_SCNG(yy_limit)) {
add_token(token_stream, T_INLINE_HTML, LANG_SCNG(yy_cursor),
LANG_SCNG(yy_limit) - LANG_SCNG(yy_cursor), CG(zend_lineno));
add_token(ctx->tokens, T_INLINE_HTML, LANG_SCNG(yy_cursor),
LANG_SCNG(yy_limit) - LANG_SCNG(yy_cursor), CG(zend_lineno), ctx->as_object);
}
break;
}
}

static zend_bool tokenize_parse(zval *return_value, zend_string *source)
static zend_bool tokenize_parse(zval *return_value, zend_string *source, zend_bool as_object)
{
zval source_zval;
zend_lex_state original_lex_state;
Expand All @@ -225,14 +259,18 @@ static zend_bool tokenize_parse(zval *return_value, zend_string *source)
zend_save_lexical_state(&original_lex_state);

if ((success = (zend_prepare_string_for_scanning(&source_zval, "") == SUCCESS))) {
struct event_context ctx;
zval token_stream;
array_init(&token_stream);

ctx.tokens = &token_stream;
ctx.as_object = as_object;

CG(ast) = NULL;
CG(ast_arena) = zend_arena_create(1024 * 32);
LANG_SCNG(yy_state) = yycINITIAL;
LANG_SCNG(on_event) = on_event;
LANG_SCNG(on_event_context) = &token_stream;
LANG_SCNG(on_event_context) = &ctx;

if((success = (zendparse() == SUCCESS))) {
ZVAL_COPY_VALUE(return_value, &token_stream);
Expand Down Expand Up @@ -270,9 +308,9 @@ PHP_FUNCTION(token_get_all)
ZEND_PARSE_PARAMETERS_END();

if (flags & TOKEN_PARSE) {
success = tokenize_parse(return_value, source);
success = tokenize_parse(return_value, source, (flags & TOKEN_AS_OBJECT) != 0);
} else {
success = tokenize(return_value, source);
success = tokenize(return_value, source, (flags & TOKEN_AS_OBJECT) != 0);
/* Normal token_get_all() should not throw. */
zend_clear_exception();
}
Expand Down