utf8proc
annotate ruby/utf8proc_native.c @ 7:fcfd8c836c64
Version 1.1.1
- Added a new PostgreSQL function 'unistrip', which behaves like 'unifold', but also removes all character marks (e.g. accents).
- Changed license from BSD to MIT style.
- Added a new function 'utf8proc_codepoint_valid' to the C library.
- Changed compiler flags in Makefile from -g -O0 to -O2
- The ruby script, which was used to build the utf8proc_data.c file, is now included in the distribution.
- Added a new PostgreSQL function 'unistrip', which behaves like 'unifold', but also removes all character marks (e.g. accents).
- Changed license from BSD to MIT style.
- Added a new function 'utf8proc_codepoint_valid' to the C library.
- Changed compiler flags in Makefile from -g -O0 to -O2
- The ruby script, which was used to build the utf8proc_data.c file, is now included in the distribution.
author | jbe |
---|---|
date | Sun Jul 22 12:00:00 2007 +0200 (2007-07-22) |
parents | c18366878af9 |
children | 00d2bcbdc945 |
rev | line source |
---|---|
jbe@0 | 1 /* |
jbe@7 | 2 * Copyright (c) 2006-2007 Jan Behrens, FlexiGuided GmbH, Berlin |
jbe@0 | 3 * |
jbe@7 | 4 * Permission is hereby granted, free of charge, to any person obtaining a |
jbe@7 | 5 * copy of this software and associated documentation files (the "Software"), |
jbe@7 | 6 * to deal in the Software without restriction, including without limitation |
jbe@7 | 7 * the rights to use, copy, modify, merge, publish, distribute, sublicense, |
jbe@7 | 8 * and/or sell copies of the Software, and to permit persons to whom the |
jbe@7 | 9 * Software is furnished to do so, subject to the following conditions: |
jbe@0 | 10 * |
jbe@7 | 11 * The above copyright notice and this permission notice shall be included in |
jbe@7 | 12 * all copies or substantial portions of the Software. |
jbe@0 | 13 * |
jbe@7 | 14 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
jbe@7 | 15 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
jbe@7 | 16 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
jbe@7 | 17 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
jbe@7 | 18 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING |
jbe@7 | 19 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER |
jbe@7 | 20 * DEALINGS IN THE SOFTWARE. |
jbe@0 | 21 */ |
jbe@7 | 22 |
jbe@0 | 23 |
jbe@0 | 24 /* |
jbe@0 | 25 * File name: ruby/utf8proc_native.c |
jbe@7 | 26 * Version: 1.1.1 |
jbe@7 | 27 * Last changed: 2007-07-22 |
jbe@0 | 28 * |
jbe@0 | 29 * Description: |
jbe@0 | 30 * Native part of the ruby wrapper for libutf8proc. |
jbe@0 | 31 */ |
jbe@0 | 32 |
jbe@0 | 33 |
jbe@0 | 34 #include "../utf8proc.c" |
jbe@0 | 35 #include "ruby.h" |
jbe@0 | 36 |
jbe@0 | 37 typedef struct utf8proc_ruby_mapenv_struct { |
jbe@0 | 38 int32_t *buffer; |
jbe@0 | 39 } utf8proc_ruby_mapenv_t; |
jbe@0 | 40 |
jbe@0 | 41 void utf8proc_ruby_mapenv_free(utf8proc_ruby_mapenv_t *env) { |
jbe@0 | 42 free(env->buffer); |
jbe@0 | 43 free(env); |
jbe@0 | 44 } |
jbe@0 | 45 |
jbe@0 | 46 VALUE utf8proc_ruby_module; |
jbe@0 | 47 VALUE utf8proc_ruby_options; |
jbe@0 | 48 VALUE utf8proc_ruby_eUnicodeError; |
jbe@0 | 49 VALUE utf8proc_ruby_eInvalidUtf8Error; |
jbe@0 | 50 VALUE utf8proc_ruby_eCodeNotAssignedError; |
jbe@0 | 51 |
jbe@0 | 52 VALUE utf8proc_ruby_map_error(ssize_t result) { |
jbe@0 | 53 VALUE excpt_class; |
jbe@0 | 54 switch (result) { |
jbe@0 | 55 case UTF8PROC_ERROR_NOMEM: |
jbe@0 | 56 excpt_class = rb_eNoMemError; break; |
jbe@0 | 57 case UTF8PROC_ERROR_OVERFLOW: |
jbe@3 | 58 case UTF8PROC_ERROR_INVALIDOPTS: |
jbe@0 | 59 excpt_class = rb_eArgError; break; |
jbe@0 | 60 case UTF8PROC_ERROR_INVALIDUTF8: |
jbe@0 | 61 excpt_class = utf8proc_ruby_eInvalidUtf8Error; break; |
jbe@0 | 62 case UTF8PROC_ERROR_NOTASSIGNED: |
jbe@0 | 63 excpt_class = utf8proc_ruby_eCodeNotAssignedError; break; |
jbe@0 | 64 default: |
jbe@0 | 65 excpt_class = rb_eRuntimeError; |
jbe@0 | 66 } |
jbe@0 | 67 rb_raise(excpt_class, "%s", utf8proc_errmsg(result)); |
jbe@0 | 68 return Qnil; |
jbe@0 | 69 } |
jbe@0 | 70 |
jbe@0 | 71 VALUE utf8proc_ruby_map(VALUE self, VALUE str_param, VALUE options_param) { |
jbe@0 | 72 VALUE str; |
jbe@0 | 73 int options; |
jbe@0 | 74 VALUE env_obj; |
jbe@0 | 75 utf8proc_ruby_mapenv_t *env; |
jbe@0 | 76 ssize_t result; |
jbe@0 | 77 VALUE retval; |
jbe@0 | 78 str = StringValue(str_param); |
jbe@0 | 79 options = NUM2INT(options_param) & ~UTF8PROC_NULLTERM; |
jbe@0 | 80 env_obj = Data_Make_Struct(rb_cObject, utf8proc_ruby_mapenv_t, NULL, |
jbe@0 | 81 utf8proc_ruby_mapenv_free, env); |
jbe@0 | 82 result = utf8proc_decompose(RSTRING(str)->ptr, RSTRING(str)->len, |
jbe@0 | 83 NULL, 0, options); |
jbe@2 | 84 if (result < 0) { |
jbe@2 | 85 utf8proc_ruby_map_error(result); |
jbe@2 | 86 return Qnil; // needed to prevent problems with optimization |
jbe@2 | 87 } |
jbe@0 | 88 env->buffer = ALLOC_N(int32_t, result+1); |
jbe@0 | 89 result = utf8proc_decompose(RSTRING(str)->ptr, RSTRING(str)->len, |
jbe@0 | 90 env->buffer, result, options); |
jbe@0 | 91 if (result < 0) { |
jbe@0 | 92 free(env->buffer); |
jbe@0 | 93 env->buffer = 0; |
jbe@1 | 94 utf8proc_ruby_map_error(result); |
jbe@2 | 95 return Qnil; // needed to prevent problems with optimization |
jbe@0 | 96 } |
jbe@0 | 97 result = utf8proc_reencode(env->buffer, result, options); |
jbe@0 | 98 if (result < 0) { |
jbe@0 | 99 free(env->buffer); |
jbe@0 | 100 env->buffer = 0; |
jbe@1 | 101 utf8proc_ruby_map_error(result); |
jbe@2 | 102 return Qnil; // needed to prevent problems with optimization |
jbe@0 | 103 } |
jbe@0 | 104 retval = rb_str_new((char *)env->buffer, result); |
jbe@0 | 105 free(env->buffer); |
jbe@0 | 106 env->buffer = 0; |
jbe@0 | 107 return retval; |
jbe@0 | 108 } |
jbe@0 | 109 |
jbe@0 | 110 static VALUE utf8proc_ruby_char(VALUE self, VALUE code_param) { |
jbe@0 | 111 char buffer[4]; |
jbe@0 | 112 ssize_t result; |
jbe@0 | 113 int uc; |
jbe@0 | 114 uc = NUM2INT(code_param); |
jbe@7 | 115 if (!utf8proc_codepoint_valid(uc)) |
jbe@0 | 116 rb_raise(rb_eArgError, "Invalid Unicode code point"); |
jbe@0 | 117 result = utf8proc_encode_char(uc, buffer); |
jbe@0 | 118 return rb_str_new(buffer, result); |
jbe@0 | 119 } |
jbe@0 | 120 |
jbe@0 | 121 #define register_utf8proc_option(sym, field) \ |
jbe@0 | 122 rb_hash_aset(utf8proc_ruby_options, ID2SYM(rb_intern(sym)), INT2FIX(field)) |
jbe@0 | 123 |
jbe@0 | 124 void Init_utf8proc_native() { |
jbe@0 | 125 utf8proc_ruby_module = rb_define_module("Utf8Proc"); |
jbe@0 | 126 rb_define_module_function(utf8proc_ruby_module, "utf8map", |
jbe@0 | 127 utf8proc_ruby_map, 2); |
jbe@0 | 128 rb_define_module_function(utf8proc_ruby_module, "utf8char", |
jbe@0 | 129 utf8proc_ruby_char, 1); |
jbe@0 | 130 utf8proc_ruby_eUnicodeError = rb_define_class_under(utf8proc_ruby_module, |
jbe@0 | 131 "UnicodeError", rb_eStandardError); |
jbe@0 | 132 utf8proc_ruby_eInvalidUtf8Error = rb_define_class_under( |
jbe@0 | 133 utf8proc_ruby_module, "InvalidUtf8Error", utf8proc_ruby_eUnicodeError); |
jbe@0 | 134 utf8proc_ruby_eCodeNotAssignedError = rb_define_class_under( |
jbe@0 | 135 utf8proc_ruby_module, "CodeNotAssignedError", |
jbe@0 | 136 utf8proc_ruby_eUnicodeError); |
jbe@0 | 137 utf8proc_ruby_options = rb_hash_new(); |
jbe@2 | 138 register_utf8proc_option("stable", UTF8PROC_STABLE); |
jbe@2 | 139 register_utf8proc_option("compat", UTF8PROC_COMPAT); |
jbe@2 | 140 register_utf8proc_option("compose", UTF8PROC_COMPOSE); |
jbe@2 | 141 register_utf8proc_option("decompose", UTF8PROC_DECOMPOSE); |
jbe@2 | 142 register_utf8proc_option("ignore", UTF8PROC_IGNORE); |
jbe@2 | 143 register_utf8proc_option("rejectna", UTF8PROC_REJECTNA); |
jbe@2 | 144 register_utf8proc_option("nlf2ls", UTF8PROC_NLF2LS); |
jbe@2 | 145 register_utf8proc_option("nlf2ps", UTF8PROC_NLF2PS); |
jbe@2 | 146 register_utf8proc_option("nlf2lf", UTF8PROC_NLF2LF); |
jbe@2 | 147 register_utf8proc_option("stripcc", UTF8PROC_STRIPCC); |
jbe@2 | 148 register_utf8proc_option("casefold", UTF8PROC_CASEFOLD); |
jbe@2 | 149 register_utf8proc_option("charbound", UTF8PROC_CHARBOUND); |
jbe@3 | 150 register_utf8proc_option("lump", UTF8PROC_LUMP); |
jbe@3 | 151 register_utf8proc_option("stripmark", UTF8PROC_STRIPMARK); |
jbe@0 | 152 OBJ_FREEZE(utf8proc_ruby_options); |
jbe@0 | 153 rb_define_const(utf8proc_ruby_module, "Options", utf8proc_ruby_options); |
jbe@0 | 154 } |
jbe@0 | 155 |