Timerix/utf8proc
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Reduce the size of the binary.

Browse Source 
Use integers instead of pointers in Unicode tables. Saves 226 kb / 716 kb in the
compiled library.
This commit is contained in:
Michaël Meyer 2015-12-09 19:55:48 +01:00
parent 6b510deff2
commit 26436c9775
4 changed files with 6682 additions and 6677 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
data/data_generator.rb
View File

@ -116,13 +116,14 @@ def str2c(string, prefix)
return "UTF8PROC_#{prefix}_#{string.upcase}" return "UTF8PROC_#{prefix}_#{string.upcase}"
end end
def ary2c(array) def ary2c(array)
return "NULL" if array.nil? return "UINT16_MAX" if array.nil?
unless $int_array_indicies[array] unless $int_array_indicies[array]
$int_array_indicies[array] = $int_array.length $int_array_indicies[array] = $int_array.length
array.each { |entry| $int_array << entry } array.each { |entry| $int_array << entry }
$int_array << -1 $int_array << -1
end end
return "utf8proc_sequences + #{$int_array_indicies[array]}" raise "Array index out of bound" if $int_array_indicies[array] >= 65535
return "#{$int_array_indicies[array]}"
end end
class UnicodeChar class UnicodeChar
@ -305,7 +306,7 @@ end
$stdout << "};\n\n" $stdout << "};\n\n"
$stdout << "const utf8proc_property_t utf8proc_properties[] = {\n" $stdout << "const utf8proc_property_t utf8proc_properties[] = {\n"
$stdout << " {0, 0, 0, 0, NULL, NULL, -1, -1, -1, -1, -1, false,false,false,false, UTF8PROC_BOUNDCLASS_OTHER, 0},\n" $stdout << " {0, 0, 0, 0, UINT16_MAX, UINT16_MAX, -1, -1, -1, -1, -1, false,false,false,false, UTF8PROC_BOUNDCLASS_OTHER, 0},\n"
properties.each { |line| properties.each { |line|
$stdout << line $stdout << line
} }

8
utf8proc.c
View File

@ -357,10 +357,10 @@ UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_decompose_char(utf8proc_int32_t uc,
category == UTF8PROC_CATEGORY_ME) return 0; category == UTF8PROC_CATEGORY_ME) return 0;
} }
if (options & UTF8PROC_CASEFOLD) { if (options & UTF8PROC_CASEFOLD) {
if (property->casefold_mapping) { if (property->casefold_mapping != UINT16_MAX) {
const utf8proc_int32_t *casefold_entry; const utf8proc_int32_t *casefold_entry;
utf8proc_ssize_t written = 0; utf8proc_ssize_t written = 0;
for (casefold_entry = property->casefold_mapping; for (casefold_entry = &utf8proc_sequences[property->casefold_mapping];
*casefold_entry >= 0; casefold_entry++) { *casefold_entry >= 0; casefold_entry++) {
written += utf8proc_decompose_char(*casefold_entry, dst+written, written += utf8proc_decompose_char(*casefold_entry, dst+written,
(bufsize > written) ? (bufsize - written) : 0, options, (bufsize > written) ? (bufsize - written) : 0, options,
@ -371,11 +371,11 @@ UTF8PROC_DLLEXPORT utf8proc_ssize_t utf8proc_decompose_char(utf8proc_int32_t uc,
} }
} }
if (options & (UTF8PROC_COMPOSE|UTF8PROC_DECOMPOSE)) { if (options & (UTF8PROC_COMPOSE|UTF8PROC_DECOMPOSE)) {
if (property->decomp_mapping && if (property->decomp_mapping != UINT16_MAX &&
(!property->decomp_type || (options & UTF8PROC_COMPAT))) { (!property->decomp_type || (options & UTF8PROC_COMPAT))) {
const utf8proc_int32_t *decomp_entry; const utf8proc_int32_t *decomp_entry;
utf8proc_ssize_t written = 0; utf8proc_ssize_t written = 0;
for (decomp_entry = property->decomp_mapping; for (decomp_entry = &utf8proc_sequences[property->decomp_mapping];
*decomp_entry >= 0; decomp_entry++) { *decomp_entry >= 0; decomp_entry++) {
written += utf8proc_decompose_char(*decomp_entry, dst+written, written += utf8proc_decompose_char(*decomp_entry, dst+written,
(bufsize > written) ? (bufsize - written) : 0, options, (bufsize > written) ? (bufsize - written) : 0, options,

8
utf8proc.h
View File

@ -132,6 +132,10 @@ extern "C" {
#define SSIZE_MAX ((size_t)SIZE_MAX/2) #define SSIZE_MAX ((size_t)SIZE_MAX/2)
#endif #endif
#ifndef UINT16_MAX
# define UINT16_MAX ~(utf8proc_uint16_t)0
#endif
/** /**
* Option flags used by several functions in the library. * Option flags used by several functions in the library.
*/ */
@ -238,8 +242,8 @@ typedef struct utf8proc_property_struct {
* @see utf8proc_decomp_type_t. * @see utf8proc_decomp_type_t.
*/ */
utf8proc_propval_t decomp_type; utf8proc_propval_t decomp_type;
const utf8proc_int32_t *decomp_mapping; utf8proc_uint16_t decomp_mapping;
const utf8proc_int32_t *casefold_mapping; utf8proc_uint16_t casefold_mapping;
utf8proc_int32_t uppercase_mapping; utf8proc_int32_t uppercase_mapping;
utf8proc_int32_t lowercase_mapping; utf8proc_int32_t lowercase_mapping;
utf8proc_int32_t titlecase_mapping; utf8proc_int32_t titlecase_mapping;

13336
utf8proc_data.c
View File

File diff suppressed because it is too large Load Diff