diff options
Diffstat (limited to 'lib/uniwbrk/wbrktable.c')
| -rw-r--r-- | lib/uniwbrk/wbrktable.c | 54 |
1 files changed, 29 insertions, 25 deletions
diff --git a/lib/uniwbrk/wbrktable.c b/lib/uniwbrk/wbrktable.c index ff25fb3..18c10da 100644 --- a/lib/uniwbrk/wbrktable.c +++ b/lib/uniwbrk/wbrktable.c @@ -1,5 +1,5 @@ /* Word break auxiliary table. - Copyright (C) 2009-2010 Free Software Foundation, Inc. + Copyright (C) 2009-2015 Free Software Foundation, Inc. Written by Bruno Haible <bruno@clisp.org>, 2009. This program is free software: you can redistribute it and/or modify it @@ -22,31 +22,35 @@ /* This table contains the following rules (see UAX #29): - last current - - ALetter × ALetter (WB5) - ALetter × Numeric (WB9) - Numeric × ALetter (WB10) - Numeric × Numeric (WB8) - Katakana × Katakana (WB13) - (ALetter | Numeric | Katakana) × ExtendNumLet (WB13a) - ExtendNumLet × ExtendNumLet (WB13a) - ExtendNumLet × (ALetter | Numeric | Katakana) (WB13b) + last current + + (ALetter | HL) × (ALetter | HL) (WB5) + (ALetter | HL) × Numeric (WB9) + HL × SQ (WB7a) + Numeric × (ALetter | HL) (WB10) + Numeric × Numeric (WB8) + Katakana × Katakana (WB13) +(ALetter | HL | Numeric | Katakana) × ExtendNumLet (WB13a) + ExtendNumLet × ExtendNumLet (WB13a) + ExtendNumLet × (ALetter | HL | Numeric | Katakana) (WB13b) + Regional_Indicator × Regional_Indicator (WB13c) */ -const unsigned char uniwbrk_table[10][8] = -{ /* current: OTHER MIDNUMLET NUMERIC */ - /* KATAKANA MIDLETTER EXTENDNUMLET */ - /* ALETTER MIDNUM */ +const unsigned char uniwbrk_table[12][12] = +{ /* current: OTHER MIDNUMLET NUMERIC DQ */ + /* KATAKANA MIDLETTER EXTENDNUMLET SQ */ + /* ALETTER MIDNUM RI HL */ /* last */ - /* WBP_OTHER */ { 1, 1, 1, 1, 1, 1, 1, 1 }, - /* WBP_KATAKANA */ { 1, 0, 1, 1, 1, 1, 1, 0 }, - /* WBP_ALETTER */ { 1, 1, 0, 1, 1, 1, 0, 0 }, - /* WBP_MIDNUMLET */ { 1, 1, 1, 1, 1, 1, 1, 1 }, - /* WBP_MIDLETTER */ { 1, 1, 1, 1, 1, 1, 1, 1 }, - /* WBP_MIDNUM */ { 1, 1, 1, 1, 1, 1, 1, 1 }, - /* WBP_NUMERIC */ { 1, 1, 0, 1, 1, 1, 0, 0 }, - /* WBP_EXTENDNUMLET */ { 1, 0, 0, 1, 1, 1, 0, 0 }, - /* WBP_EXTEND */ { 1, 1, 1, 1, 1, 1, 1, 1 }, - /* WBP_FORMAT */ { 1, 1, 1, 1, 1, 1, 1, 1 } + /* WBP_OTHER */ { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 }, + /* WBP_KATAKANA */ { 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1 }, + /* WBP_ALETTER */ { 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 1, 0 }, + /* WBP_MIDNUMLET */ { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 }, + /* WBP_MIDLETTER */ { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 }, + /* WBP_MIDNUM */ { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 }, + /* WBP_NUMERIC */ { 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 1, 0 }, + /* WBP_EXTENDNUMLET */ { 1, 0, 0, 1, 1, 1, 0, 0, 1, 1, 1, 0 }, + /* WBP_RI */ { 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1 }, + /* WBP_DQ */ { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 }, + /* WBP_SQ */ { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 }, + /* WBP_HL */ { 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 0 } }; |