gecko-dev/intl/lwbrk/jisx4051pairtable.txt
Birunthan Mohanathas 790694c449 Bug 1038537 - Part 5: Flatten intl/lwbrk/{idl,public,src}/ directories. r=smontagu
--HG--
rename : intl/lwbrk/src/Makefile.in => intl/lwbrk/Makefile.in
rename : intl/lwbrk/src/crashtests/416721.html => intl/lwbrk/crashtests/416721.html
rename : intl/lwbrk/src/crashtests/crashtests.list => intl/lwbrk/crashtests/crashtests.list
rename : intl/lwbrk/src/jisx4051class.h => intl/lwbrk/jisx4051class.h
rename : intl/lwbrk/src/jisx4051pairtable.txt => intl/lwbrk/jisx4051pairtable.txt
rename : intl/lwbrk/src/nsCarbonBreaker.cpp => intl/lwbrk/nsCarbonBreaker.cpp
rename : intl/lwbrk/src/nsComplexBreaker.h => intl/lwbrk/nsComplexBreaker.h
rename : intl/lwbrk/public/nsILineBreaker.h => intl/lwbrk/nsILineBreaker.h
rename : intl/lwbrk/idl/nsISemanticUnitScanner.idl => intl/lwbrk/nsISemanticUnitScanner.idl
rename : intl/lwbrk/public/nsIWordBreaker.h => intl/lwbrk/nsIWordBreaker.h
rename : intl/lwbrk/src/nsJISx4051LineBreaker.cpp => intl/lwbrk/nsJISx4051LineBreaker.cpp
rename : intl/lwbrk/src/nsJISx4051LineBreaker.h => intl/lwbrk/nsJISx4051LineBreaker.h
rename : intl/lwbrk/public/nsLWBrkCIID.h => intl/lwbrk/nsLWBrkCIID.h
rename : intl/lwbrk/src/nsPangoBreaker.cpp => intl/lwbrk/nsPangoBreaker.cpp
rename : intl/lwbrk/src/nsRuleBreaker.cpp => intl/lwbrk/nsRuleBreaker.cpp
rename : intl/lwbrk/src/nsSampleWordBreaker.cpp => intl/lwbrk/nsSampleWordBreaker.cpp
rename : intl/lwbrk/src/nsSampleWordBreaker.h => intl/lwbrk/nsSampleWordBreaker.h
rename : intl/lwbrk/src/nsSemanticUnitScanner.cpp => intl/lwbrk/nsSemanticUnitScanner.cpp
rename : intl/lwbrk/src/nsSemanticUnitScanner.h => intl/lwbrk/nsSemanticUnitScanner.h
rename : intl/lwbrk/src/nsUniscribeBreaker.cpp => intl/lwbrk/nsUniscribeBreaker.cpp
rename : intl/lwbrk/src/rulebrk.c => intl/lwbrk/rulebrk.c
rename : intl/lwbrk/src/rulebrk.h => intl/lwbrk/rulebrk.h
rename : intl/lwbrk/src/th_char.h => intl/lwbrk/th_char.h
2014-07-24 10:56:39 -07:00

287 lines
9.0 KiB
Plaintext

/*
Simplification of Pair Table in JIS X 4051
1. The Origion Table - in 4.1.3
In JIS x 4051. The pair table is defined as below
Class of
Leading Class of Trailing Char Class
Char
1 2 3 4 5 6 7 8 9 10 11 12 13 13 14 14 15 16 17 18 19 20
* # * #
1 X X X X X X X X X X X X X X X X X X X X X E
2 X X X X X X
3 X X X X X X
4 X X X X X X
5 X X X X X X
6 X X X X X X
7 X X X X X X X
8 X X X X X X E
9 X X X X X X
10 X X X X X X
11 X X X X X X
12 X X X X X X
13 X X X X X X X
14 X X X X X X X
15 X X X X X X X X X
16 X X X X X X X X
17 X X X X X E
18 X X X X X X X X X
19 X E E E E E X X X X X X X X X X X X E X E E
20 X X X X X E
* Same Char
# Other Char
2. Simplified by remove the class which we do not care
However, since we do not care about class 13(Subscript), 14(Ruby),
19(split line note begin quote), and 20(split line note end quote)
we can simplify this par table into the following
Class of
Leading Class of Trailing Char Class
Char
1 2 3 4 5 6 7 8 9 10 11 12 15 16 17 18
1 X X X X X X X X X X X X X X X X
2 X X X X X
3 X X X X X
4 X X X X X
5 X X X X X
6 X X X X X
7 X X X X X X
8 X X X X X X
9 X X X X X
10 X X X X X
11 X X X X X
12 X X X X X
15 X X X X X X X X
16 X X X X X X X
17 X X X X X
18 X X X X X X X X
3. Simplified by merged classes
After the 2 simplification, the pair table have some duplication
a. class 2, 3, 4, 5, 6, are the same- we can merged them
b. class 10, 11, 12, 17 are the same- we can merged them
Class of
Leading Class of Trailing Char Class
Char
1 [a] 7 8 9 [b]15 16 18
1 X X X X X X X X X
[a] X
7 X X
8 X X
9 X
[b] X
15 X X X X
16 X X X
18 X X X X
4. Now we use one bit to encode weather it is breakable, and use 2 bytes
for one row, then the bit table will look like:
18 <- 1
1 0000 0001 1111 1111 = 0x01FF
[a] 0000 0000 0000 0010 = 0x0002
7 0000 0000 0000 0110 = 0x0006
8 0000 0000 0100 0010 = 0x0042
9 0000 0000 0000 0010 = 0x0002
[b] 0000 0000 0000 0010 = 0x0042
15 0000 0001 0101 0010 = 0x0152
16 0000 0001 1000 0010 = 0x0182
17 0000 0001 1100 0010 = 0x01C2
*/
static uint16_t gJISx4051SimplifiedPair[9] = {
0x01FF, 0x0002, 0x0006, 0x0042, 0x0002, 0x0042, 0x0152, 0x0182, 0x01C2
};
PRBool XXXX::ClassesToPair(nsJISx4051Cls aCls1, nsJISx4051Cls aCls1)
{
NS_ASSERTION( (aCls1 < 9) "invalid class");
NS_ASSERTION( (aCls2 < 9) "invalid class");
return ( 0 != (gJISx4051SimplifiedPair[aCls1] & (1L << aCls2) ));
}
#define X4051_IS_DIGIT(u) ((0x0030 >= (u)) && ((u) >= 0x0039))
nsJISx4051Cls XXXX::GetClass(
PRUnichar aChar, PRUnichar aBefore = 0, PRUnichar aAfter = 0)
{
// take care the special case in cls 15
if( ((0x2C == aChar) || (0x2E == aChar)) &&
(X4051_IS_DIGIT(aBefore)) && X4051_IS_DIGIT(aAfter)))
{
return kJISx4051Cls_15;
}
nsJISx4051Cls cls;
if(gSingle->Lookup(aChar, &cls))
return cls;
if(gRange->Lookup(aChar, &cls))
return cls;
return kJISx4051Cls_15;
}
typedef enum {
kJISx4051Cls_1 = 0,
kJISx4051Cls_2 = 1,
kJISx4051Cls_3 = 1,
kJISx4051Cls_4 = 1,
kJISx4051Cls_5 = 1,
kJISx4051Cls_6 = 1,
kJISx4051Cls_7 = 2,
kJISx4051Cls_8 = 3,
kJISx4051Cls_9 = 4,
kJISx4051Cls_10 = 5,
kJISx4051Cls_11 = 5,
kJISx4051Cls_12 = 5,
// kJISx4051Cls_13 = 0,
// kJISx4051Cls_14 = 0,
kJISx4051Cls_15 = 6,
kJISx4051Cls_16 = 7,
kJISx4051Cls_17 = 5,
kJISx4051Cls_18 = 8,
// kJISx4051Cls_19 = 0,
// kJISx4051Cls_20 = 0
} nsJISx4051Cls;
// Table 2
YYYY(kJISx4051Cls_1 , 0x0028),
YYYY(kJISx4051Cls_1 , 0x005B),
YYYY(kJISx4051Cls_1 , 0x007B),
YYYY(kJISx4051Cls_1 , 0x2018),
YYYY(kJISx4051Cls_1 , 0x201B),
YYYY(kJISx4051Cls_1 , 0x201C),
YYYY(kJISx4051Cls_1 , 0x201F),
YYYY(kJISx4051Cls_1 , 0x3008),
YYYY(kJISx4051Cls_1 , 0x300A),
YYYY(kJISx4051Cls_1 , 0x300C),
YYYY(kJISx4051Cls_1 , 0x300E),
YYYY(kJISx4051Cls_1 , 0x3010),
YYYY(kJISx4051Cls_1 , 0x3014),
YYYY(kJISx4051Cls_1 , 0x3016),
YYYY(kJISx4051Cls_1 , 0x3018),
YYYY(kJISx4051Cls_1 , 0x301A),
YYYY(kJISx4051Cls_1 , 0x301D),
// Table 3
YYYY(kJISx4051Cls_2 , 0x0029),
YYYY(kJISx4051Cls_2 , 0x002C),
YYYY(kJISx4051Cls_2 , 0x005D),
YYYY(kJISx4051Cls_2 , 0x007D),
YYYY(kJISx4051Cls_2 , 0x2019),
YYYY(kJISx4051Cls_2 , 0x201A),
YYYY(kJISx4051Cls_2 , 0x201D),
YYYY(kJISx4051Cls_2 , 0x201E),
YYYY(kJISx4051Cls_2 , 0x3001),
YYYY(kJISx4051Cls_2 , 0x3009),
YYYY(kJISx4051Cls_2 , 0x300B),
YYYY(kJISx4051Cls_2 , 0x300D),
YYYY(kJISx4051Cls_2 , 0x300F),
YYYY(kJISx4051Cls_2 , 0x3011),
YYYY(kJISx4051Cls_2 , 0x3015),
YYYY(kJISx4051Cls_2 , 0x3017),
YYYY(kJISx4051Cls_2 , 0x3019),
YYYY(kJISx4051Cls_2 , 0x301B),
YYYY(kJISx4051Cls_2 , 0x301E),
YYYY(kJISx4051Cls_2 , 0x301F),
// Table 4
YYYY(kJISx4051Cls_3 , 0x203C),
YYYY(kJISx4051Cls_3 , 0x2044),
YYYY(kJISx4051Cls_3 , 0x301C),
YYYY(kJISx4051Cls_3 , 0x3041),
YYYY(kJISx4051Cls_3 , 0x3043),
YYYY(kJISx4051Cls_3 , 0x3045),
YYYY(kJISx4051Cls_3 , 0x3047),
YYYY(kJISx4051Cls_3 , 0x3049),
YYYY(kJISx4051Cls_3 , 0x3063),
YYYY(kJISx4051Cls_3 , 0x3083),
YYYY(kJISx4051Cls_3 , 0x3085),
YYYY(kJISx4051Cls_3 , 0x3087),
YYYY(kJISx4051Cls_3 , 0x308E),
YYYY(kJISx4051Cls_3 , 0x309D),
YYYY(kJISx4051Cls_3 , 0x309E),
YYYY(kJISx4051Cls_3 , 0x30A1),
YYYY(kJISx4051Cls_3 , 0x30A3),
YYYY(kJISx4051Cls_3 , 0x30A5),
YYYY(kJISx4051Cls_3 , 0x30A7),
YYYY(kJISx4051Cls_3 , 0x30A9),
YYYY(kJISx4051Cls_3 , 0x30C3),
YYYY(kJISx4051Cls_3 , 0x30E3),
YYYY(kJISx4051Cls_3 , 0x30E5),
YYYY(kJISx4051Cls_3 , 0x30E7),
YYYY(kJISx4051Cls_3 , 0x30EE),
YYYY(kJISx4051Cls_3 , 0x30F5),
YYYY(kJISx4051Cls_3 , 0x30F6),
YYYY(kJISx4051Cls_3 , 0x30FC),
YYYY(kJISx4051Cls_3 , 0x30FD),
YYYY(kJISx4051Cls_3 , 0x30FE),
// Table 5
YYYY(kJISx4051Cls_4 , 0x0021),
YYYY(kJISx4051Cls_4 , 0x003F),
// Table 6
YYYY(kJISx4051Cls_5 , 0x003A),
YYYY(kJISx4051Cls_5 , 0x003B),
YYYY(kJISx4051Cls_5 , 0x30FB),
// Table 7
YYYY(kJISx4051Cls_6 , 0x002E),
YYYY(kJISx4051Cls_6 , 0x3002),
// Table 8
YYYY(kJISx4051Cls_7 , 0x2014),
YYYY(kJISx4051Cls_7 , 0x2024),
YYYY(kJISx4051Cls_7 , 0x2025),
YYYY(kJISx4051Cls_7 , 0x2026),
// Table 9
YYYY(kJISx4051Cls_8 , 0x0024),
YYYY(kJISx4051Cls_8 , 0x00A3),
YYYY(kJISx4051Cls_8 , 0x00A5),
YYYY(kJISx4051Cls_8 , 0x2116),
// Table 10
YYYY(kJISx4051Cls_9 , 0x0025),
YYYY(kJISx4051Cls_9 , 0x00A2),
YYYY(kJISx4051Cls_9 , 0x00B0),
YYYY(kJISx4051Cls_9 , 0x2030),
YYYY(kJISx4051Cls_9 , 0x2031),
YYYY(kJISx4051Cls_9 , 0x2032),
YYYY(kJISx4051Cls_9 , 0x2033),
// Table 1
YYYY(kJISx4051Cls_10, 0x3000),
// Table 1
ZZZZ(kJISx4051Cls_11, 0x3000),