1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/intl/icu/source/common/propname.cpp Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,327 @@
1.4 +/*
1.5 +**********************************************************************
1.6 +* Copyright (c) 2002-2011, International Business Machines
1.7 +* Corporation and others. All Rights Reserved.
1.8 +**********************************************************************
1.9 +* Author: Alan Liu
1.10 +* Created: October 30 2002
1.11 +* Since: ICU 2.4
1.12 +* 2010nov19 Markus Scherer Rewrite for formatVersion 2.
1.13 +**********************************************************************
1.14 +*/
1.15 +#include "propname.h"
1.16 +#include "unicode/uchar.h"
1.17 +#include "unicode/udata.h"
1.18 +#include "unicode/uscript.h"
1.19 +#include "umutex.h"
1.20 +#include "cmemory.h"
1.21 +#include "cstring.h"
1.22 +#include "ucln_cmn.h"
1.23 +#include "uarrsort.h"
1.24 +#include "uinvchar.h"
1.25 +
1.26 +#define INCLUDED_FROM_PROPNAME_CPP
1.27 +#include "propname_data.h"
1.28 +
1.29 +U_CDECL_BEGIN
1.30 +
1.31 +/**
1.32 + * Get the next non-ignorable ASCII character from a property name
1.33 + * and lowercases it.
1.34 + * @return ((advance count for the name)<<8)|character
1.35 + */
1.36 +static inline int32_t
1.37 +getASCIIPropertyNameChar(const char *name) {
1.38 + int32_t i;
1.39 + char c;
1.40 +
1.41 + /* Ignore delimiters '-', '_', and ASCII White_Space */
1.42 + for(i=0;
1.43 + (c=name[i++])==0x2d || c==0x5f ||
1.44 + c==0x20 || (0x09<=c && c<=0x0d);
1.45 + ) {}
1.46 +
1.47 + if(c!=0) {
1.48 + return (i<<8)|(uint8_t)uprv_asciitolower((char)c);
1.49 + } else {
1.50 + return i<<8;
1.51 + }
1.52 +}
1.53 +
1.54 +/**
1.55 + * Get the next non-ignorable EBCDIC character from a property name
1.56 + * and lowercases it.
1.57 + * @return ((advance count for the name)<<8)|character
1.58 + */
1.59 +static inline int32_t
1.60 +getEBCDICPropertyNameChar(const char *name) {
1.61 + int32_t i;
1.62 + char c;
1.63 +
1.64 + /* Ignore delimiters '-', '_', and EBCDIC White_Space */
1.65 + for(i=0;
1.66 + (c=name[i++])==0x60 || c==0x6d ||
1.67 + c==0x40 || c==0x05 || c==0x15 || c==0x25 || c==0x0b || c==0x0c || c==0x0d;
1.68 + ) {}
1.69 +
1.70 + if(c!=0) {
1.71 + return (i<<8)|(uint8_t)uprv_ebcdictolower((char)c);
1.72 + } else {
1.73 + return i<<8;
1.74 + }
1.75 +}
1.76 +
1.77 +/**
1.78 + * Unicode property names and property value names are compared "loosely".
1.79 + *
1.80 + * UCD.html 4.0.1 says:
1.81 + * For all property names, property value names, and for property values for
1.82 + * Enumerated, Binary, or Catalog properties, use the following
1.83 + * loose matching rule:
1.84 + *
1.85 + * LM3. Ignore case, whitespace, underscore ('_'), and hyphens.
1.86 + *
1.87 + * This function does just that, for (char *) name strings.
1.88 + * It is almost identical to ucnv_compareNames() but also ignores
1.89 + * C0 White_Space characters (U+0009..U+000d, and U+0085 on EBCDIC).
1.90 + *
1.91 + * @internal
1.92 + */
1.93 +
1.94 +U_CAPI int32_t U_EXPORT2
1.95 +uprv_compareASCIIPropertyNames(const char *name1, const char *name2) {
1.96 + int32_t rc, r1, r2;
1.97 +
1.98 + for(;;) {
1.99 + r1=getASCIIPropertyNameChar(name1);
1.100 + r2=getASCIIPropertyNameChar(name2);
1.101 +
1.102 + /* If we reach the ends of both strings then they match */
1.103 + if(((r1|r2)&0xff)==0) {
1.104 + return 0;
1.105 + }
1.106 +
1.107 + /* Compare the lowercased characters */
1.108 + if(r1!=r2) {
1.109 + rc=(r1&0xff)-(r2&0xff);
1.110 + if(rc!=0) {
1.111 + return rc;
1.112 + }
1.113 + }
1.114 +
1.115 + name1+=r1>>8;
1.116 + name2+=r2>>8;
1.117 + }
1.118 +}
1.119 +
1.120 +U_CAPI int32_t U_EXPORT2
1.121 +uprv_compareEBCDICPropertyNames(const char *name1, const char *name2) {
1.122 + int32_t rc, r1, r2;
1.123 +
1.124 + for(;;) {
1.125 + r1=getEBCDICPropertyNameChar(name1);
1.126 + r2=getEBCDICPropertyNameChar(name2);
1.127 +
1.128 + /* If we reach the ends of both strings then they match */
1.129 + if(((r1|r2)&0xff)==0) {
1.130 + return 0;
1.131 + }
1.132 +
1.133 + /* Compare the lowercased characters */
1.134 + if(r1!=r2) {
1.135 + rc=(r1&0xff)-(r2&0xff);
1.136 + if(rc!=0) {
1.137 + return rc;
1.138 + }
1.139 + }
1.140 +
1.141 + name1+=r1>>8;
1.142 + name2+=r2>>8;
1.143 + }
1.144 +}
1.145 +
1.146 +U_CDECL_END
1.147 +
1.148 +U_NAMESPACE_BEGIN
1.149 +
1.150 +int32_t PropNameData::findProperty(int32_t property) {
1.151 + int32_t i=1; // valueMaps index, initially after numRanges
1.152 + for(int32_t numRanges=valueMaps[0]; numRanges>0; --numRanges) {
1.153 + // Read and skip the start and limit of this range.
1.154 + int32_t start=valueMaps[i];
1.155 + int32_t limit=valueMaps[i+1];
1.156 + i+=2;
1.157 + if(property<start) {
1.158 + break;
1.159 + }
1.160 + if(property<limit) {
1.161 + return i+(property-start)*2;
1.162 + }
1.163 + i+=(limit-start)*2; // Skip all entries for this range.
1.164 + }
1.165 + return 0;
1.166 +}
1.167 +
1.168 +int32_t PropNameData::findPropertyValueNameGroup(int32_t valueMapIndex, int32_t value) {
1.169 + if(valueMapIndex==0) {
1.170 + return 0; // The property does not have named values.
1.171 + }
1.172 + ++valueMapIndex; // Skip the BytesTrie offset.
1.173 + int32_t numRanges=valueMaps[valueMapIndex++];
1.174 + if(numRanges<0x10) {
1.175 + // Ranges of values.
1.176 + for(; numRanges>0; --numRanges) {
1.177 + // Read and skip the start and limit of this range.
1.178 + int32_t start=valueMaps[valueMapIndex];
1.179 + int32_t limit=valueMaps[valueMapIndex+1];
1.180 + valueMapIndex+=2;
1.181 + if(value<start) {
1.182 + break;
1.183 + }
1.184 + if(value<limit) {
1.185 + return valueMaps[valueMapIndex+value-start];
1.186 + }
1.187 + valueMapIndex+=limit-start; // Skip all entries for this range.
1.188 + }
1.189 + } else {
1.190 + // List of values.
1.191 + int32_t valuesStart=valueMapIndex;
1.192 + int32_t nameGroupOffsetsStart=valueMapIndex+numRanges-0x10;
1.193 + do {
1.194 + int32_t v=valueMaps[valueMapIndex];
1.195 + if(value<v) {
1.196 + break;
1.197 + }
1.198 + if(value==v) {
1.199 + return valueMaps[nameGroupOffsetsStart+valueMapIndex-valuesStart];
1.200 + }
1.201 + } while(++valueMapIndex<nameGroupOffsetsStart);
1.202 + }
1.203 + return 0;
1.204 +}
1.205 +
1.206 +const char *PropNameData::getName(const char *nameGroup, int32_t nameIndex) {
1.207 + int32_t numNames=*nameGroup++;
1.208 + if(nameIndex<0 || numNames<=nameIndex) {
1.209 + return NULL;
1.210 + }
1.211 + // Skip nameIndex names.
1.212 + for(; nameIndex>0; --nameIndex) {
1.213 + nameGroup=uprv_strchr(nameGroup, 0)+1;
1.214 + }
1.215 + if(*nameGroup==0) {
1.216 + return NULL; // no name (Property[Value]Aliases.txt has "n/a")
1.217 + }
1.218 + return nameGroup;
1.219 +}
1.220 +
1.221 +UBool PropNameData::containsName(BytesTrie &trie, const char *name) {
1.222 + if(name==NULL) {
1.223 + return FALSE;
1.224 + }
1.225 + UStringTrieResult result=USTRINGTRIE_NO_VALUE;
1.226 + char c;
1.227 + while((c=*name++)!=0) {
1.228 + c=uprv_invCharToLowercaseAscii(c);
1.229 + // Ignore delimiters '-', '_', and ASCII White_Space.
1.230 + if(c==0x2d || c==0x5f || c==0x20 || (0x09<=c && c<=0x0d)) {
1.231 + continue;
1.232 + }
1.233 + if(!USTRINGTRIE_HAS_NEXT(result)) {
1.234 + return FALSE;
1.235 + }
1.236 + result=trie.next((uint8_t)c);
1.237 + }
1.238 + return USTRINGTRIE_HAS_VALUE(result);
1.239 +}
1.240 +
1.241 +const char *PropNameData::getPropertyName(int32_t property, int32_t nameChoice) {
1.242 + int32_t valueMapIndex=findProperty(property);
1.243 + if(valueMapIndex==0) {
1.244 + return NULL; // Not a known property.
1.245 + }
1.246 + return getName(nameGroups+valueMaps[valueMapIndex], nameChoice);
1.247 +}
1.248 +
1.249 +const char *PropNameData::getPropertyValueName(int32_t property, int32_t value, int32_t nameChoice) {
1.250 + int32_t valueMapIndex=findProperty(property);
1.251 + if(valueMapIndex==0) {
1.252 + return NULL; // Not a known property.
1.253 + }
1.254 + int32_t nameGroupOffset=findPropertyValueNameGroup(valueMaps[valueMapIndex+1], value);
1.255 + if(nameGroupOffset==0) {
1.256 + return NULL;
1.257 + }
1.258 + return getName(nameGroups+nameGroupOffset, nameChoice);
1.259 +}
1.260 +
1.261 +int32_t PropNameData::getPropertyOrValueEnum(int32_t bytesTrieOffset, const char *alias) {
1.262 + BytesTrie trie(bytesTries+bytesTrieOffset);
1.263 + if(containsName(trie, alias)) {
1.264 + return trie.getValue();
1.265 + } else {
1.266 + return UCHAR_INVALID_CODE;
1.267 + }
1.268 +}
1.269 +
1.270 +int32_t PropNameData::getPropertyEnum(const char *alias) {
1.271 + return getPropertyOrValueEnum(0, alias);
1.272 +}
1.273 +
1.274 +int32_t PropNameData::getPropertyValueEnum(int32_t property, const char *alias) {
1.275 + int32_t valueMapIndex=findProperty(property);
1.276 + if(valueMapIndex==0) {
1.277 + return UCHAR_INVALID_CODE; // Not a known property.
1.278 + }
1.279 + valueMapIndex=valueMaps[valueMapIndex+1];
1.280 + if(valueMapIndex==0) {
1.281 + return UCHAR_INVALID_CODE; // The property does not have named values.
1.282 + }
1.283 + // valueMapIndex is the start of the property's valueMap,
1.284 + // where the first word is the BytesTrie offset.
1.285 + return getPropertyOrValueEnum(valueMaps[valueMapIndex], alias);
1.286 +}
1.287 +U_NAMESPACE_END
1.288 +
1.289 +//----------------------------------------------------------------------
1.290 +// Public API implementation
1.291 +
1.292 +U_CAPI const char* U_EXPORT2
1.293 +u_getPropertyName(UProperty property,
1.294 + UPropertyNameChoice nameChoice) {
1.295 + U_NAMESPACE_USE
1.296 + return PropNameData::getPropertyName(property, nameChoice);
1.297 +}
1.298 +
1.299 +U_CAPI UProperty U_EXPORT2
1.300 +u_getPropertyEnum(const char* alias) {
1.301 + U_NAMESPACE_USE
1.302 + return (UProperty)PropNameData::getPropertyEnum(alias);
1.303 +}
1.304 +
1.305 +U_CAPI const char* U_EXPORT2
1.306 +u_getPropertyValueName(UProperty property,
1.307 + int32_t value,
1.308 + UPropertyNameChoice nameChoice) {
1.309 + U_NAMESPACE_USE
1.310 + return PropNameData::getPropertyValueName(property, value, nameChoice);
1.311 +}
1.312 +
1.313 +U_CAPI int32_t U_EXPORT2
1.314 +u_getPropertyValueEnum(UProperty property,
1.315 + const char* alias) {
1.316 + U_NAMESPACE_USE
1.317 + return PropNameData::getPropertyValueEnum(property, alias);
1.318 +}
1.319 +
1.320 +U_CAPI const char* U_EXPORT2
1.321 +uscript_getName(UScriptCode scriptCode){
1.322 + return u_getPropertyValueName(UCHAR_SCRIPT, scriptCode,
1.323 + U_LONG_PROPERTY_NAME);
1.324 +}
1.325 +
1.326 +U_CAPI const char* U_EXPORT2
1.327 +uscript_getShortName(UScriptCode scriptCode){
1.328 + return u_getPropertyValueName(UCHAR_SCRIPT, scriptCode,
1.329 + U_SHORT_PROPERTY_NAME);
1.330 +}
