Bug 393356. Spellchecker should not use private copy(ies) of the Unicode category tables. r=smontagu,a=damon

extensions/spellcheck/src/mozEnglishWordUtils.cpp

@@ -41,7 +41,6 @@
 #include "nsIServiceManager.h"
 #include "nsUnicharUtilCIID.h"
 #include "nsCRT.h"
-#include "cattable.h"
 
 NS_IMPL_ISUPPORTS1(mozEnglishWordUtils, mozISpellI18NUtil)
 
@@ -51,6 +50,8 @@ mozEnglishWordUtils::mozEnglishWordUtils()
 
   nsresult rv;
   mURLDetector = do_CreateInstance(MOZ_TXTTOHTMLCONV_CONTRACTID, &rv);
+  mCaseConv = do_GetService(NS_UNICHARUTIL_CONTRACTID);
+  mCategories = do_GetService(NS_UNICHARCATEGORY_CONTRACTID);
 }
 
 mozEnglishWordUtils::~mozEnglishWordUtils()
@@ -78,12 +79,6 @@ NS_IMETHODIMP mozEnglishWordUtils::GetRootForm(const PRUnichar *aWord, PRUint32
 
   *count = 0;
 
-  if (!mCaseConv) {
-    mCaseConv = do_GetService(NS_UNICHARUTIL_CONTRACTID);
-    if (!mCaseConv)
-      return NS_ERROR_FAILURE;
-  }
-
   mozEnglishWordUtils::myspCapitalization ct = captype(word);
   switch (ct)
     {
@@ -159,10 +154,10 @@ NS_IMETHODIMP mozEnglishWordUtils::GetRootForm(const PRUnichar *aWord, PRUint32
 }
 
 // This needs vast improvement
-static PRBool ucIsAlpha(PRUnichar c)
+PRBool mozEnglishWordUtils::ucIsAlpha(PRUnichar aChar)
 {
   // XXX we have to fix callers to handle the full Unicode range
-  return (5 == GetCat(PRUint32(c)));
+  return nsIUGenCategory::kLetter == mCategories->Get(PRUint32(aChar));
 }
 
 /* void FindNextWord (in wstring word, in PRUint32 length, in PRUint32 offset, out PRUint32 begin, out PRUint32 end); */

extensions/spellcheck/src/mozEnglishWordUtils.h

@@ -44,6 +44,7 @@
 #include "nsIUnicodeDecoder.h"
 #include "nsString.h"
 #include "nsICaseConversion.h"
+#include "nsIUGenCategory.h"
 
 #include "mozITXTToHTMLConv.h" 
 
@@ -62,10 +63,12 @@ public:
 
 protected:
   mozEnglishWordUtils::myspCapitalization captype(const nsString &word);
+  PRBool ucIsAlpha(PRUnichar aChar);
 
   nsString mLanguage;
   nsString mCharset;
   nsCOMPtr<nsICaseConversion> mCaseConv;
+  nsCOMPtr<nsIUGenCategory>   mCategories;
   nsCOMPtr<mozITXTToHTMLConv> mURLDetector; // used to detect urls so the spell checker can skip them.
 };
 

extensions/spellcheck/src/mozInlineSpellWordUtil.cpp

@@ -36,7 +36,6 @@
  *
  * ***** END LICENSE BLOCK ***** */
 
-#include "cattable.h"
 #include "mozInlineSpellWordUtil.h"
 #include "nsDebug.h"
 #include "nsIAtom.h"
@@ -49,14 +48,8 @@
 #include "nsIEditor.h"
 #include "nsIDOMNode.h"
 #include "nsIDOMHTMLBRElement.h"
-
-// some character categories we care about from GetCat()
-#define CHAR_CAT_NUMBER 2
-#define CHAR_CAT_SPACE 3
-#define CHAR_CAT_CONTROL 4
-#define CHAR_CAT_WORD 5
-#define CHAR_CAT_PUNCTUATION1 6
-#define CHAR_CAT_PUNCTUATION2 7
+#include "nsUnicharUtilCIID.h"
+#include "nsServiceManagerUtils.h"
 
 // IsIgnorableCharacter
 //
@@ -87,6 +80,10 @@ mozInlineSpellWordUtil::Init(nsWeakPtr aWeakEditor)
 {
   nsresult rv;
 
+  mCategories = do_GetService(NS_UNICHARCATEGORY_CONTRACTID, &rv);
+  if (NS_FAILED(rv))
+    return rv;
+  
   // getting the editor can fail commonly because the editor was detached, so
   // don't assert
   nsCOMPtr<nsIEditor> editor = do_QueryReferent(aWeakEditor, &rv);
@@ -841,8 +838,9 @@ WordSplitState::ClassifyCharacter(PRInt32 aIndex, PRBool aRecurse) const
 
   // this will classify the character, we want to treat "ignorable" characters
   // such as soft hyphens as word characters.
-  PRInt32 charCategory = GetCat(mDOMWordText[aIndex]);
-  if (charCategory == CHAR_CAT_WORD ||
+  nsIUGenCategory::nsUGenCategory
+    charCategory = mWordUtil->GetCategories()->Get(PRUint32(mDOMWordText[aIndex]));
+  if (charCategory == nsIUGenCategory::kLetter ||
       IsIgnorableCharacter(mDOMWordText[aIndex]))
     return CHAR_CLASS_WORD;
 
@@ -871,10 +869,10 @@ WordSplitState::ClassifyCharacter(PRInt32 aIndex, PRBool aRecurse) const
   }
 
   // all other punctuation
-  if (charCategory == CHAR_CAT_SPACE ||
-      charCategory == CHAR_CAT_CONTROL ||
-      charCategory == CHAR_CAT_PUNCTUATION1 ||
-      charCategory == CHAR_CAT_PUNCTUATION2)
+  if (charCategory == nsIUGenCategory::kSeparator ||
+      charCategory == nsIUGenCategory::kOther ||
+      charCategory == nsIUGenCategory::kPunctuation ||
+      charCategory == nsIUGenCategory::kSymbol)
     return CHAR_CLASS_SEPARATOR;
 
   // any other character counts as a word

extensions/spellcheck/src/mozInlineSpellWordUtil.h

@@ -42,6 +42,7 @@
 #include "nsIDocument.h"
 #include "nsString.h"
 #include "nsTArray.h"
+#include "nsIUGenCategory.h"
 
 //#define DEBUG_SPELLCHECK
 
@@ -118,13 +119,15 @@ public:
   nsIDOMDocumentRange* GetDocumentRange() const { return mDOMDocumentRange; }
   nsIDocument* GetDocument() const { return mDocument; }
   nsIDOMNode* GetRootNode() { return mRootNode; }
-
+  nsIUGenCategory* GetCategories() { return mCategories; }
+  
 private:
 
   // cached stuff for the editor, set by Init
   nsCOMPtr<nsIDOMDocumentRange> mDOMDocumentRange;
   nsCOMPtr<nsIDocument>         mDocument;
   nsCOMPtr<nsIDOMViewCSS>       mCSSView;
+  nsCOMPtr<nsIUGenCategory>     mCategories;
 
   // range to check, see SetRange
   nsIDOMNode* mRootNode;

intl/build/nsI18nModule.cpp

@@ -69,6 +69,8 @@ static nsModuleComponentInfo components[] =
  // unicharutil
   { "Unichar Utility", NS_UNICHARUTIL_CID, 
       NS_UNICHARUTIL_CONTRACTID, nsCaseConversionImp2Constructor},
+  { "Unichar Category Table", NS_UNICHARCATEGORY_CID, 
+      NS_UNICHARCATEGORY_CONTRACTID, nsCategoryImpConstructor},
   { "Unicode To Entity Converter", NS_ENTITYCONVERTER_CID, 
       NS_ENTITYCONVERTER_CONTRACTID, nsEntityConverterConstructor },
   { "Unicode To Charset Converter", NS_SAVEASCHARSET_CID, 

intl/unicharutil/public/Makefile.in

@@ -48,7 +48,7 @@ EXPORTS		= \
 		nsICaseConversion.h \
 		nsIOrderIdFormater.h \
 		nsITextTransform.h \
-		nsIUGenDetailCategory.h \
+		nsIUGenCategory.h \
 		nsUnicharUtilCIID.h \
 		nsHankakuToZenkakuCID.h \
 		$(NULL)

intl/unicharutil/public/nsIUGenCategory.h

@@ -41,10 +41,10 @@
 #include "nsISupports.h"
 #include "nscore.h"
 
-// {E86B3371-BF89-11d2-B3AF-00805F8A6670}
+// {671fea05-fcee-4b1c-82a3-6eb03eda8ddc}
 #define NS_IUGENCATEGORY_IID \
-{ 0xe86b3371, 0xbf89, 0x11d2, \
-    { 0xb3, 0xaf, 0x0, 0x80, 0x5f, 0x8a, 0x66, 0x70 } }
+{ 0x671fea05, 0xfcee, 0x4b1c, \
+    { 0x82, 0xa3, 0x6e, 0xb0, 0x3e, 0xda, 0x8d, 0xdc } }
 
 
 class nsIUGenCategory : public nsISupports {
@@ -54,31 +54,24 @@ public:
   NS_DECLARE_STATIC_IID_ACCESSOR(NS_IUGENCATEGORY_IID)
 
    /**
-    *  Read ftp://ftp.unicode.org/Public/UNIDATA/ReadMe-Latest.txt
-    *  section GENERAL CATEGORY
-    *  for the detail defintation of the following categories
+    *  Read http://www.unicode.org/Public/UNIDATA/UCD.html#General_Category_Values
+    *  for the detailed definition of the following categories
     */
    typedef enum {
-     kUGenCategory_Mark         = 1, // Mn, Mc, and Me
-     kUGenCategory_Number       = 2, // Nd, Nl, and No 
-     kUGenCategory_Separator    = 3, // Zs, Zl, and Zp
-     kUGenCategory_Other        = 4, // Cc, Cf, Cs, Co, and Cn
-     kUGenCategory_Letter       = 5, // Lu, Ll, Lt, Lm, and Lo
-     kUGenCategory_Punctuation  = 6, // Pc, Pd, Ps, Pe, Pi, Pf, and Po
-     kUGenCategory_Symbol       = 7  // Sm, Sc, Sk, and So
+     kUndefined    = 0,
+     kMark         = 1, // Mn, Mc, and Me
+     kNumber       = 2, // Nd, Nl, and No 
+     kSeparator    = 3, // Zs, Zl, and Zp
+     kOther        = 4, // Cc, Cf, Cs, Co, and Cn
+     kLetter       = 5, // Lu, Ll, Lt, Lm, and Lo
+     kPunctuation  = 6, // Pc, Pd, Ps, Pe, Pi, Pf, and Po
+     kSymbol       = 7  // Sm, Sc, Sk, and So
    } nsUGenCategory;
 
    /**
     * Give a Unichar, return a nsUGenCategory
     */
-   NS_IMETHOD Get( PRUnichar aChar, nsUGenCategory* oResult) = 0 ;
-    
-   /**
-    * Give a Unichar, and a nsUGenCategory, 
-    * return PR_TRUE if the Unichar is in that category, 
-    * return PR_FALSE, otherwise
-    */
-   NS_IMETHOD Is( PRUnichar aChar, nsUGenCategory aCategory, PRBool* oResult) = 0;
+   virtual nsUGenCategory Get(PRUint32 aChar) = 0;
 };
 
 NS_DEFINE_STATIC_IID_ACCESSOR(nsIUGenCategory, NS_IUGENCATEGORY_IID)

intl/unicharutil/public/nsUnicharUtilCIID.h

@@ -46,4 +46,11 @@
   { 0xb3, 0xae, 0x0, 0x80, 0x5f, 0x8a, 0x66, 0x70 } }
 
 #define NS_UNICHARUTIL_CONTRACTID "@mozilla.org/intl/unicharutil;1"
+
+#define NS_UNICHARCATEGORY_CID \
+{ 0x748a1132, 0x671a, 0x409a, \
+  { 0x8d, 0x1d, 0xf1, 0xcd, 0xf6, 0xb3, 0xa6, 0xb4 } }
+
+#define NS_UNICHARCATEGORY_CONTRACTID "@mozilla.org/intl/unicharcategory;1"
+
 #endif

intl/unicharutil/src/Makefile.in

@@ -57,6 +57,7 @@ EXPORTS = cattable.h
 
 CPPSRCS		= \
 		nsCaseConversionImp2.cpp \
+		nsCategoryImp.cpp \
 		nsHankakuToZenkaku.cpp \
 		nsEntityConverter.cpp \
 		nsSaveAsCharset.cpp \

intl/unicharutil/src/nsCategoryImp.cpp

@@ -42,36 +42,26 @@
 #include "nsCategoryImp.h"
 #include "cattable.h"
 
-NS_IMPL_ISUPPORTS1(nsCategoryImp, nsIUGenCategory)
+static nsCategoryImp gCategoryImp;
 
+NS_IMPL_THREADSAFE_QUERY_INTERFACE1(nsCategoryImp, nsIUGenCategory)
 
-nsCategoryImp::nsCategoryImp()
+NS_IMETHODIMP_(nsrefcnt) nsCategoryImp::AddRef(void)
 {
+  return nsrefcnt(1);
 }
 
-nsCategoryImp::~nsCategoryImp()
+NS_IMETHODIMP_(nsrefcnt) nsCategoryImp::Release(void)
 {
+  return nsrefcnt(1);
 }
 
-nsresult nsCategoryImp::Get( PRUnichar aChar, nsUGenCategory* oResult)
+nsCategoryImp* nsCategoryImp::GetInstance()
 {
-   PRUint8 ret = GetCat(aChar);
-   if( 0 == ret)
-      *oResult = kUGenCategory_Other; // treat it as Cn - Other, Not Assigned
-   else 
-      *oResult = (nsUGenCategory)ret;
-   return NS_OK;
+  return &gCategoryImp;
 }
-    
-nsresult nsCategoryImp::Is( PRUnichar aChar, nsUGenCategory aCategory, PRBool* oResult)
 
+nsIUGenCategory::nsUGenCategory nsCategoryImp::Get(PRUint32 aChar)
 {
-   nsUGenCategory cat ;
-   PRUint8 ret = GetCat(aChar);
-   if( 0 == ret)
-      cat = kUGenCategory_Other; // treat it as Cn - Other, Not Assigned
-   else 
-      cat = (nsUGenCategory)ret;
-   *oResult = (aCategory == cat );
-   return NS_OK;
+  return nsUGenCategory(GetCat(aChar));
 }

intl/unicharutil/src/nsCategoryImp.h

@@ -37,29 +37,18 @@
 #ifndef nsCategoryImp_h__
 #define nsCategoryImp_h__
 
-#include "nscore.h"
-#include "nsISupports.h"
 #include "nsIUGenCategory.h"
 
 class nsCategoryImp : public nsIUGenCategory {
    NS_DECL_ISUPPORTS
    
-public: 
-   nsCategoryImp();
-   virtual ~nsCategoryImp();
-
-
+public:
+   static nsCategoryImp* GetInstance();
+    
    /**
     * Give a Unichar, return a nsUGenCategory
     */
-   NS_IMETHOD Get( PRUnichar aChar, nsUGenCategory* oResult);
-    
-   /**
-    * Give a Unichar, and a nsUGenCategory, 
-    * return PR_TRUE if the Unichar is in that category, 
-    * return PR_FALSE, otherwise
-    */
-   NS_IMETHOD Is( PRUnichar aChar, nsUGenCategory aCategory, PRBool* oResult);
+   virtual nsUGenCategory Get(PRUint32 aChar);
 };
 
 #endif  /* nsCategoryImp_h__ */

intl/unicharutil/src/nsUcharUtilConstructors.h

@@ -40,6 +40,7 @@
 
 #include "nsUnicharUtilCIID.h"
 #include "nsCaseConversionImp2.h"
+#include "nsCategoryImp.h"
 #include "nsHankakuToZenkakuCID.h"
 #include "nsTextTransformFactory.h"
 #include "nsICaseConversion.h"
@@ -82,6 +83,8 @@ UNICHARUTIL_MAKE_CTOR(HankakuToZenkaku)
 
 NS_GENERIC_FACTORY_SINGLETON_CONSTRUCTOR(nsCaseConversionImp2,
                                          nsCaseConversionImp2::GetInstance)
+NS_GENERIC_FACTORY_SINGLETON_CONSTRUCTOR(nsCategoryImp,
+                                         nsCategoryImp::GetInstance)
 NS_GENERIC_FACTORY_CONSTRUCTOR(nsEntityConverter)
 NS_GENERIC_FACTORY_CONSTRUCTOR(nsSaveAsCharset)
 NS_GENERIC_FACTORY_CONSTRUCTOR(nsUnicodeNormalizer)