1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/extensions/spellcheck/src/mozEnglishWordUtils.cpp Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,286 @@
1.4 +/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
1.5 +/* This Source Code Form is subject to the terms of the Mozilla Public
1.6 + * License, v. 2.0. If a copy of the MPL was not distributed with this
1.7 + * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
1.8 +
1.9 +#include "mozEnglishWordUtils.h"
1.10 +#include "nsReadableUtils.h"
1.11 +#include "nsIServiceManager.h"
1.12 +#include "nsUnicharUtils.h"
1.13 +#include "nsUnicharUtilCIID.h"
1.14 +#include "nsUnicodeProperties.h"
1.15 +#include "nsCRT.h"
1.16 +#include "mozilla/Likely.h"
1.17 +
1.18 +NS_IMPL_CYCLE_COLLECTING_ADDREF(mozEnglishWordUtils)
1.19 +NS_IMPL_CYCLE_COLLECTING_RELEASE(mozEnglishWordUtils)
1.20 +
1.21 +NS_INTERFACE_MAP_BEGIN(mozEnglishWordUtils)
1.22 + NS_INTERFACE_MAP_ENTRY(mozISpellI18NUtil)
1.23 + NS_INTERFACE_MAP_ENTRY_AMBIGUOUS(nsISupports, mozISpellI18NUtil)
1.24 + NS_INTERFACE_MAP_ENTRIES_CYCLE_COLLECTION(mozEnglishWordUtils)
1.25 +NS_INTERFACE_MAP_END
1.26 +
1.27 +NS_IMPL_CYCLE_COLLECTION(mozEnglishWordUtils,
1.28 + mURLDetector)
1.29 +
1.30 +mozEnglishWordUtils::mozEnglishWordUtils()
1.31 +{
1.32 + mLanguage.AssignLiteral("en");
1.33 +
1.34 + nsresult rv;
1.35 + mURLDetector = do_CreateInstance(MOZ_TXTTOHTMLCONV_CONTRACTID, &rv);
1.36 +}
1.37 +
1.38 +mozEnglishWordUtils::~mozEnglishWordUtils()
1.39 +{
1.40 +}
1.41 +
1.42 +/* attribute wstring language; */
1.43 +NS_IMETHODIMP mozEnglishWordUtils::GetLanguage(char16_t * *aLanguage)
1.44 +{
1.45 + nsresult rv = NS_OK;
1.46 + NS_ENSURE_ARG_POINTER(aLanguage);
1.47 +
1.48 + *aLanguage = ToNewUnicode(mLanguage);
1.49 + if(!aLanguage) rv = NS_ERROR_OUT_OF_MEMORY;
1.50 + return rv;
1.51 + }
1.52 +
1.53 +/* void GetRootForm (in wstring aWord, in uint32_t type, [array, size_is (count)] out wstring words, out uint32_t count); */
1.54 +// return the possible root forms of aWord.
1.55 +NS_IMETHODIMP mozEnglishWordUtils::GetRootForm(const char16_t *aWord, uint32_t type, char16_t ***words, uint32_t *count)
1.56 +{
1.57 + nsAutoString word(aWord);
1.58 + char16_t **tmpPtr;
1.59 + int32_t length = word.Length();
1.60 +
1.61 + *count = 0;
1.62 +
1.63 + mozEnglishWordUtils::myspCapitalization ct = captype(word);
1.64 + switch (ct)
1.65 + {
1.66 + case HuhCap:
1.67 + case NoCap:
1.68 + tmpPtr = (char16_t **)nsMemory::Alloc(sizeof(char16_t *));
1.69 + if (!tmpPtr)
1.70 + return NS_ERROR_OUT_OF_MEMORY;
1.71 + tmpPtr[0] = ToNewUnicode(word);
1.72 + if (!tmpPtr[0]) {
1.73 + NS_FREE_XPCOM_ALLOCATED_POINTER_ARRAY(0, tmpPtr);
1.74 + return NS_ERROR_OUT_OF_MEMORY;
1.75 + }
1.76 + *words = tmpPtr;
1.77 + *count = 1;
1.78 + break;
1.79 +
1.80 +
1.81 + case AllCap:
1.82 + tmpPtr = (char16_t **)nsMemory::Alloc(sizeof(char16_t *) * 3);
1.83 + if (!tmpPtr)
1.84 + return NS_ERROR_OUT_OF_MEMORY;
1.85 + tmpPtr[0] = ToNewUnicode(word);
1.86 + if (!tmpPtr[0]) {
1.87 + NS_FREE_XPCOM_ALLOCATED_POINTER_ARRAY(0, tmpPtr);
1.88 + return NS_ERROR_OUT_OF_MEMORY;
1.89 + }
1.90 + ToLowerCase(tmpPtr[0], tmpPtr[0], length);
1.91 +
1.92 + tmpPtr[1] = ToNewUnicode(word);
1.93 + if (!tmpPtr[1]) {
1.94 + NS_FREE_XPCOM_ALLOCATED_POINTER_ARRAY(1, tmpPtr);
1.95 + return NS_ERROR_OUT_OF_MEMORY;
1.96 + }
1.97 + ToLowerCase(tmpPtr[1], tmpPtr[1], length);
1.98 + ToUpperCase(tmpPtr[1], tmpPtr[1], 1);
1.99 +
1.100 + tmpPtr[2] = ToNewUnicode(word);
1.101 + if (!tmpPtr[2]) {
1.102 + NS_FREE_XPCOM_ALLOCATED_POINTER_ARRAY(2, tmpPtr);
1.103 + return NS_ERROR_OUT_OF_MEMORY;
1.104 + }
1.105 +
1.106 + *words = tmpPtr;
1.107 + *count = 3;
1.108 + break;
1.109 +
1.110 + case InitCap:
1.111 + tmpPtr = (char16_t **)nsMemory::Alloc(sizeof(char16_t *) * 2);
1.112 + if (!tmpPtr)
1.113 + return NS_ERROR_OUT_OF_MEMORY;
1.114 +
1.115 + tmpPtr[0] = ToNewUnicode(word);
1.116 + if (!tmpPtr[0]) {
1.117 + NS_FREE_XPCOM_ALLOCATED_POINTER_ARRAY(0, tmpPtr);
1.118 + return NS_ERROR_OUT_OF_MEMORY;
1.119 + }
1.120 + ToLowerCase(tmpPtr[0], tmpPtr[0], length);
1.121 +
1.122 + tmpPtr[1] = ToNewUnicode(word);
1.123 + if (!tmpPtr[1]) {
1.124 + NS_FREE_XPCOM_ALLOCATED_POINTER_ARRAY(1, tmpPtr);
1.125 + return NS_ERROR_OUT_OF_MEMORY;
1.126 + }
1.127 +
1.128 + *words = tmpPtr;
1.129 + *count = 2;
1.130 + break;
1.131 + default:
1.132 + return NS_ERROR_FAILURE; // should never get here;
1.133 + }
1.134 + return NS_OK;
1.135 +}
1.136 +
1.137 +// This needs vast improvement
1.138 +bool mozEnglishWordUtils::ucIsAlpha(char16_t aChar)
1.139 +{
1.140 + // XXX we have to fix callers to handle the full Unicode range
1.141 + return nsIUGenCategory::kLetter == mozilla::unicode::GetGenCategory(aChar);
1.142 +}
1.143 +
1.144 +/* void FindNextWord (in wstring word, in uint32_t length, in uint32_t offset, out uint32_t begin, out uint32_t end); */
1.145 +NS_IMETHODIMP mozEnglishWordUtils::FindNextWord(const char16_t *word, uint32_t length, uint32_t offset, int32_t *begin, int32_t *end)
1.146 +{
1.147 + const char16_t *p = word + offset;
1.148 + const char16_t *endbuf = word + length;
1.149 + const char16_t *startWord=p;
1.150 + if(p<endbuf){
1.151 + // XXX These loops should be modified to handle non-BMP characters.
1.152 + // if previous character is a word character, need to advance out of the word
1.153 + if (offset > 0 && ucIsAlpha(*(p-1))) {
1.154 + while (p < endbuf && ucIsAlpha(*p))
1.155 + p++;
1.156 + }
1.157 + while((p < endbuf) && (!ucIsAlpha(*p)))
1.158 + {
1.159 + p++;
1.160 + }
1.161 + startWord=p;
1.162 + while((p < endbuf) && ((ucIsAlpha(*p))||(*p=='\'')))
1.163 + {
1.164 + p++;
1.165 + }
1.166 +
1.167 + // we could be trying to break down a url, we don't want to break a url into parts,
1.168 + // instead we want to find out if it really is a url and if so, skip it, advancing startWord
1.169 + // to a point after the url.
1.170 +
1.171 + // before we spend more time looking to see if the word is a url, look for a url identifer
1.172 + // and make sure that identifer isn't the last character in the word fragment.
1.173 + if ( (*p == ':' || *p == '@' || *p == '.') && p < endbuf - 1) {
1.174 +
1.175 + // ok, we have a possible url...do more research to find out if we really have one
1.176 + // and determine the length of the url so we can skip over it.
1.177 +
1.178 + if (mURLDetector)
1.179 + {
1.180 + int32_t startPos = -1;
1.181 + int32_t endPos = -1;
1.182 +
1.183 + mURLDetector->FindURLInPlaintext(startWord, endbuf - startWord, p - startWord, &startPos, &endPos);
1.184 +
1.185 + // ok, if we got a url, adjust the array bounds, skip the current url text and find the next word again
1.186 + if (startPos != -1 && endPos != -1) {
1.187 + startWord = p + endPos + 1; // skip over the url
1.188 + p = startWord; // reset p
1.189 +
1.190 + // now recursively call FindNextWord to search for the next word now that we have skipped the url
1.191 + return FindNextWord(word, length, startWord - word, begin, end);
1.192 + }
1.193 + }
1.194 + }
1.195 +
1.196 + while((p > startWord)&&(*(p-1) == '\'')){ // trim trailing apostrophes
1.197 + p--;
1.198 + }
1.199 + }
1.200 + else{
1.201 + startWord = endbuf;
1.202 + }
1.203 + if(startWord == endbuf){
1.204 + *begin = -1;
1.205 + *end = -1;
1.206 + }
1.207 + else{
1.208 + *begin = startWord-word;
1.209 + *end = p-word;
1.210 + }
1.211 + return NS_OK;
1.212 +}
1.213 +
1.214 +mozEnglishWordUtils::myspCapitalization
1.215 +mozEnglishWordUtils::captype(const nsString &word)
1.216 +{
1.217 + char16_t* lword=ToNewUnicode(word);
1.218 + ToUpperCase(lword,lword,word.Length());
1.219 + if(word.Equals(lword)){
1.220 + nsMemory::Free(lword);
1.221 + return AllCap;
1.222 + }
1.223 +
1.224 + ToLowerCase(lword,lword,word.Length());
1.225 + if(word.Equals(lword)){
1.226 + nsMemory::Free(lword);
1.227 + return NoCap;
1.228 + }
1.229 + int32_t length=word.Length();
1.230 + if(Substring(word,1,length-1).Equals(lword+1)){
1.231 + nsMemory::Free(lword);
1.232 + return InitCap;
1.233 + }
1.234 + nsMemory::Free(lword);
1.235 + return HuhCap;
1.236 +}
1.237 +
1.238 +// Convert the list of words in iwords to the same capitalization aWord and
1.239 +// return them in owords.
1.240 +NS_IMETHODIMP mozEnglishWordUtils::FromRootForm(const char16_t *aWord, const char16_t **iwords, uint32_t icount, char16_t ***owords, uint32_t *ocount)
1.241 +{
1.242 + nsAutoString word(aWord);
1.243 + nsresult rv = NS_OK;
1.244 +
1.245 + int32_t length;
1.246 + char16_t **tmpPtr = (char16_t **)nsMemory::Alloc(sizeof(char16_t *)*icount);
1.247 + if (!tmpPtr)
1.248 + return NS_ERROR_OUT_OF_MEMORY;
1.249 +
1.250 + mozEnglishWordUtils::myspCapitalization ct = captype(word);
1.251 + for(uint32_t i = 0; i < icount; ++i) {
1.252 + length = NS_strlen(iwords[i]);
1.253 + tmpPtr[i] = (char16_t *) nsMemory::Alloc(sizeof(char16_t) * (length + 1));
1.254 + if (MOZ_UNLIKELY(!tmpPtr[i])) {
1.255 + NS_FREE_XPCOM_ALLOCATED_POINTER_ARRAY(i, tmpPtr);
1.256 + return NS_ERROR_OUT_OF_MEMORY;
1.257 + }
1.258 + memcpy(tmpPtr[i], iwords[i], (length + 1) * sizeof(char16_t));
1.259 +
1.260 + nsAutoString capTest(tmpPtr[i]);
1.261 + mozEnglishWordUtils::myspCapitalization newCt=captype(capTest);
1.262 + if(newCt == NoCap){
1.263 + switch(ct)
1.264 + {
1.265 + case HuhCap:
1.266 + case NoCap:
1.267 + break;
1.268 + case AllCap:
1.269 + ToUpperCase(tmpPtr[i],tmpPtr[i],length);
1.270 + rv = NS_OK;
1.271 + break;
1.272 + case InitCap:
1.273 + ToUpperCase(tmpPtr[i],tmpPtr[i],1);
1.274 + rv = NS_OK;
1.275 + break;
1.276 + default:
1.277 + rv = NS_ERROR_FAILURE; // should never get here;
1.278 + break;
1.279 +
1.280 + }
1.281 + }
1.282 + }
1.283 + if (NS_SUCCEEDED(rv)){
1.284 + *owords = tmpPtr;
1.285 + *ocount = icount;
1.286 + }
1.287 + return rv;
1.288 +}
1.289 +
