YourWishes/jerryscript
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Reduce memory consumption of property names.

Browse Source 
Property names were always required a string reference which consumed
a large amount of memory for arrays. This patch reduces this consumption
by directly storing the value part of certain strings.

JerryScript-DCO-1.0-Signed-off-by: Zoltan Herczeg zherczeg.u-szeged@partner.samsung.com
This commit is contained in:
Zoltan Herczeg
2016-11-25 05:44:57 -08:00
parent abaf4c8af8
commit 6d67b8dff9
18 changed files with 595 additions and 393 deletions
Download Patch File Download Diff File Expand all files Collapse all files
jerry-core/lit/lit-magic-strings.c
+105 -44
View File
@@ -23,7 +23,7 @@
* @return number of the strings, if there were registered,
* zero - otherwise.
*/
uint32_t
inline uint32_t __attr_always_inline___
lit_get_magic_string_ex_count (void)
{
return JERRY_CONTEXT (lit_magic_string_ex_count);
@@ -161,28 +161,24 @@ lit_magic_strings_ex_set (const lit_utf8_byte_t **ex_str_items, /**< character a
#endif /* !JERRY_NDEBUG */
} /* lit_magic_strings_ex_set */
/**
* Check if passed cesu-8 string equals to one of magic strings
* and if equal magic string was found, return it's id in 'out_id_p' argument.
* Returns the magic string id of the argument string if it is available.
*
* @return true - if magic string equal to passed string was found,
* false - otherwise.
* @return id - if magic string id is found,
* LIT_MAGIC_STRING__COUNT - otherwise.
*/
bool
lit_magic_string_id_t
lit_is_utf8_string_magic (const lit_utf8_byte_t *string_p, /**< utf-8 string */
lit_utf8_size_t string_size, /**< string size in bytes */
lit_magic_string_id_t *out_id_p) /**< [out] magic string's id */
lit_utf8_size_t string_size) /**< string size in bytes */
{
*out_id_p = LIT_MAGIC_STRING__COUNT;
if (string_size > lit_get_magic_string_size (LIT_MAGIC_STRING__COUNT - 1))
{
return false;
return LIT_MAGIC_STRING__COUNT;
}
int first = (int) lit_get_magic_string_size_block_start (string_size); /**< First magic string element */
int last = (int) (lit_get_magic_string_size_block_start (string_size + 1) - 1); /**< Last magic string element */
/**< The string must be in this id range. */
int first = (int) lit_get_magic_string_size_block_start (string_size);
int last = (int) (lit_get_magic_string_size_block_start (string_size + 1) - 1);
while (first <= last)
{
@@ -191,8 +187,7 @@ lit_is_utf8_string_magic (const lit_utf8_byte_t *string_p, /**< utf-8 string */
if (compare == 0)
{
*out_id_p = (lit_magic_string_id_t) middle;
return true;
return (lit_magic_string_id_t) middle;
}
else if (compare > 0)
{
@@ -204,19 +199,70 @@ lit_is_utf8_string_magic (const lit_utf8_byte_t *string_p, /**< utf-8 string */
}
}
return false;
return LIT_MAGIC_STRING__COUNT;
} /* lit_is_utf8_string_magic */
/**
* Check if passed utf-8 string equals to one of external magic strings
* and if equal magic string was found, return it's id in 'out_id_p' argument.
* Returns the magic string id of the argument string pair if it is available.
*
* @return true - if external magic string equal to passed string was found,
* false - otherwise.
* @return id - if magic string id is found,
* LIT_MAGIC_STRING__COUNT - otherwise.
*/
bool lit_is_ex_utf8_string_magic (const lit_utf8_byte_t *string_p, /**< utf-8 string */
lit_utf8_size_t string_size, /**< string size in bytes */
lit_magic_string_ex_id_t *out_id_p) /**< [out] magic string's id */
lit_magic_string_id_t
lit_is_utf8_string_pair_magic (const lit_utf8_byte_t *string1_p, /**< first utf-8 string */
lit_utf8_size_t string1_size, /**< first string size in bytes */
const lit_utf8_byte_t *string2_p, /**< second utf-8 string */
lit_utf8_size_t string2_size) /**< second string size in bytes */
{
lit_utf8_size_t total_string_size = string1_size + string2_size;
if (total_string_size > lit_get_magic_string_size (LIT_MAGIC_STRING__COUNT - 1))
{
return LIT_MAGIC_STRING__COUNT;
}
/**< The string must be in this id range. */
int first = (int) lit_get_magic_string_size_block_start (total_string_size);
int last = (int) (lit_get_magic_string_size_block_start (total_string_size + 1) - 1);
while (first <= last)
{
int middle = ((first + last) / 2); /**< mid point of search */
const lit_utf8_byte_t *middle_string_p = lit_get_magic_string_utf8 (middle);
int compare = memcmp (middle_string_p, string1_p, string1_size);
if (compare == 0)
{
compare = memcmp (middle_string_p + string1_size, string2_p, string2_size);
}
if (compare == 0)
{
return (lit_magic_string_id_t) middle;
}
else if (compare > 0)
{
last = middle - 1;
}
else
{
first = middle + 1;
}
}
return LIT_MAGIC_STRING__COUNT;
} /* lit_is_utf8_string_pair_magic */
/**
* Returns the ex magic string id of the argument string if it is available.
*
* @return id - if magic string id is found,
* lit_get_magic_string_ex_count () - otherwise.
*/
lit_magic_string_ex_id_t
lit_is_ex_utf8_string_magic (const lit_utf8_byte_t *string_p, /**< utf-8 string */
lit_utf8_size_t string_size) /**< string size in bytes */
{
/* TODO: Improve performance of search */
@@ -224,36 +270,51 @@ bool lit_is_ex_utf8_string_magic (const lit_utf8_byte_t *string_p, /**< utf-8 st
id < JERRY_CONTEXT (lit_magic_string_ex_count);
id = (lit_magic_string_ex_id_t) (id + 1))
{
if (lit_compare_utf8_string_and_magic_string_ex (string_p, string_size, id))
if (string_size == lit_get_magic_string_ex_size (id))
{
*out_id_p = id;
return true;
if (memcmp (string_p, lit_get_magic_string_ex_utf8 (id), string_size) == 0)
{
return id;
}
}
}
*out_id_p = JERRY_CONTEXT (lit_magic_string_ex_count);
return false;
return JERRY_CONTEXT (lit_magic_string_ex_count);
} /* lit_is_ex_utf8_string_magic */
/**
* Compare utf-8 string and external magic string for equality
* Returns the ex magic string id of the argument string pair if it is available.
*
* @return true if strings are equal
* false otherwise
* @return id - if magic string id is found,
* lit_get_magic_string_ex_count () - otherwise.
*/
bool
lit_compare_utf8_string_and_magic_string_ex (const lit_utf8_byte_t *string_p, /**< utf-8 string */
lit_utf8_size_t string_size, /**< string size in bytes */
lit_magic_string_ex_id_t magic_string_ex_id) /**< external magic string's
* id */
lit_magic_string_ex_id_t
lit_is_ex_utf8_string_pair_magic (const lit_utf8_byte_t *string1_p, /**< first utf-8 string */
lit_utf8_size_t string1_size, /**< first string size in bytes */
const lit_utf8_byte_t *string2_p, /**< second utf-8 string */
lit_utf8_size_t string2_size) /**< second string size in bytes */
{
return lit_compare_utf8_strings (string_p,
string_size,
lit_get_magic_string_ex_utf8 (magic_string_ex_id),
lit_get_magic_string_ex_size (magic_string_ex_id));
} /* lit_compare_utf8_string_and_magic_string_ex */
/* TODO: Improve performance of search */
lit_utf8_size_t total_string_size = string1_size + string2_size;
for (lit_magic_string_ex_id_t id = (lit_magic_string_ex_id_t) 0;
id < JERRY_CONTEXT (lit_magic_string_ex_count);
id = (lit_magic_string_ex_id_t) (id + 1))
{
if (total_string_size == lit_get_magic_string_ex_size (id))
{
const lit_utf8_byte_t *ex_magic_string_p = lit_get_magic_string_ex_utf8 (id);
if (memcmp (string1_p, ex_magic_string_p, string1_size) == 0
&& memcmp (string2_p, ex_magic_string_p + string1_size, string2_size) == 0)
{
return id;
}
}
}
return JERRY_CONTEXT (lit_magic_string_ex_count);
} /* lit_is_ex_utf8_string_pair_magic */
/**
* Copy magic string to buffer