ICU-1779 set data object for entire tree of functors under a rule

X-SVN-Rev: 8131

icu4c/source/i18n/funcrepl.cpp

@@ -91,6 +91,13 @@ UnicodeString& FunctionReplacer::toReplacerPattern(UnicodeString& rule,
     return rule;
 }
 
+/**
+ * UnicodeFunctor API
+ */
+void FunctionReplacer::setData(const TransliterationRuleData* d) {
+    replacer->setData(d);
+}
+
 U_NAMESPACE_END
 
 //eof

icu4c/source/i18n/funcrepl.h

@@ -80,6 +80,11 @@ class FunctionReplacer : public UnicodeFunctor, public UnicodeReplacer {
      */
     virtual UnicodeString& toReplacerPattern(UnicodeString& rule,
                                              UBool escapeUnprintable) const;
+
+    /**
+     * UnicodeFunctor API
+     */
+    virtual void setData(const TransliterationRuleData*);
 };
 
 U_NAMESPACE_END

icu4c/source/i18n/quant.cpp

@@ -114,6 +114,13 @@ UBool Quantifier::matchesIndexValue(uint8_t v) const {
     return (minCount == 0) || matcher->toMatcher()->matchesIndexValue(v);
 }
 
+/**
+ * Implement UnicodeFunctor
+ */
+void Quantifier::setData(const TransliterationRuleData* d) {
+    matcher->setData(d);
+}
+
 U_NAMESPACE_END
 
 //eof

icu4c/source/i18n/quant.h

@@ -56,6 +56,11 @@ class Quantifier : public UnicodeFunctor, public UnicodeMatcher {
      */
     virtual UBool matchesIndexValue(uint8_t v) const;
 
+    /**
+     * UnicodeFunctor API
+     */
+    virtual void setData(const TransliterationRuleData*);
+
  private:
 
     static void appendNumber(UnicodeString& result, int32_t n);

icu4c/source/i18n/rbt_data.cpp

@@ -68,16 +68,22 @@ TransliterationRuleData::~TransliterationRuleData() {
     }
 }
 
+UnicodeFunctor*
+TransliterationRuleData::lookup(UChar32 standIn) const {
+    int32_t i = standIn - variablesBase;
+    return (i >= 0 && i < variablesLength) ? variables[i] : 0;
+}
+
 UnicodeMatcher*
 TransliterationRuleData::lookupMatcher(UChar32 standIn) const {
-    int32_t i = standIn - variablesBase;
-    return (i >= 0 && i < variablesLength) ? variables[i]->toMatcher() : 0;
+    UnicodeFunctor *f = lookup(standIn);
+    return (f != 0) ? f->toMatcher() : 0;
 }
 
 UnicodeReplacer*
 TransliterationRuleData::lookupReplacer(UChar32 standIn) const {
-    int32_t i = standIn - variablesBase;
-    return (i >= 0 && i < variablesLength) ? variables[i]->toReplacer() : 0;
+    UnicodeFunctor *f = lookup(standIn);
+    return (f != 0) ? f->toReplacer() : 0;
 }
 
 U_NAMESPACE_END

icu4c/source/i18n/rbt_data.h

@@ -35,7 +35,7 @@ class Hashtable;
  * data structure handles this.  See the parsing code for more
  * details.
  */
-class TransliterationRuleData {
+class U_I18N_API TransliterationRuleData {
 
 public:
 
@@ -88,15 +88,23 @@ public:
 
     ~TransliterationRuleData();
 
+    /**
+     * Given a stand-in character, return the UnicodeFunctor that it
+     * represents, or NULL if it doesn't represent anything.
+     */
+    UnicodeFunctor* lookup(UChar32 standIn) const;
+
     /**
      * Given a stand-in character, return the UnicodeMatcher that it
-     * represents, or NULL.
+     * represents, or NULL if it doesn't represent anything or if it
+     * represents something that is not a matcher.
      */
     UnicodeMatcher* lookupMatcher(UChar32 standIn) const;
 
     /**
      * Given a stand-in character, return the UnicodeReplacer that it
-     * represents, or NULL.
+     * represents, or NULL if it doesn't represent anything or if it
+     * represents something that is not a replacer.
      */
     UnicodeReplacer* lookupReplacer(UChar32 standIn) const;
 };

icu4c/source/i18n/rbt_rule.cpp

@@ -486,6 +486,15 @@ UnicodeString& TransliterationRule::toRule(UnicodeString& rule,
     return rule;
 }
 
+void TransliterationRule::setData(const TransliterationRuleData* d) {
+    data = d;
+    anteContext->setData(d);
+    postContext->setData(d);
+    key->setData(d);
+    output->setData(d);
+    // Don't have to do segments since they are in the context or key
+}
+
 U_NAMESPACE_END
 
 //eof

icu4c/source/i18n/rbt_rule.h

@@ -204,7 +204,7 @@ public:
      * Change the data object that this rule belongs to.  Used
      * internally by the TransliterationRuleData copy constructor.
      */
-    inline void setData(const TransliterationRuleData* data);
+    void setData(const TransliterationRuleData* data);
 
     /**
      * Return the preceding context length.  This method is needed to
@@ -281,10 +281,6 @@ public:
     friend class StringMatcher;
 };
 
-inline void TransliterationRule::setData(const TransliterationRuleData* d) {
-    data = d;
-}
-
 U_NAMESPACE_END
 
 #endif

icu4c/source/i18n/rbt_set.h

@@ -23,7 +23,7 @@ class UnicodeString;
  * A set of rules for a <code>RuleBasedTransliterator</code>.
  * @author Alan Liu
  */
-class TransliterationRuleSet {
+class U_I18N_API TransliterationRuleSet {
     /**
      * Vector of rules, in the order added.  This is used while the
      * rule set is getting built.  After that, freeze() reorders and

icu4c/source/i18n/strmatch.cpp

@@ -19,7 +19,7 @@ StringMatcher::StringMatcher(const UnicodeString& theString,
                              int32_t limit,
                              int32_t segmentNum,
                              const TransliterationRuleData& theData) :
-    data(theData),
+    data(&theData),
     segmentNumber(segmentNum),
     matchStart(-1),
     matchLimit(-1)
@@ -79,7 +79,7 @@ UMatchDegree StringMatcher::matches(const Replaceable& text,
         // Match in the reverse direction
         for (i=pattern.length()-1; i>=0; --i) {
             UChar keyChar = pattern.charAt(i);
-            UnicodeMatcher* subm = data.lookupMatcher(keyChar);
+            UnicodeMatcher* subm = data->lookupMatcher(keyChar);
             if (subm == 0) {
                 if (cursor > limit &&
                     keyChar == text.charAt(cursor)) {
@@ -110,7 +110,7 @@ UMatchDegree StringMatcher::matches(const Replaceable& text,
                 return U_PARTIAL_MATCH;
             }
             UChar keyChar = pattern.charAt(i);
-            UnicodeMatcher* subm = data.lookupMatcher(keyChar);
+            UnicodeMatcher* subm = data->lookupMatcher(keyChar);
             if (subm == 0) {
                 // Don't need the cursor < limit check if
                 // incremental is TRUE (because it's done above); do need
@@ -151,7 +151,7 @@ UnicodeString& StringMatcher::toPattern(UnicodeString& result,
     }
     for (int32_t i=0; i<pattern.length(); ++i) {
         UChar keyChar = pattern.charAt(i);
-        const UnicodeMatcher* m = data.lookupMatcher(keyChar);
+        const UnicodeMatcher* m = data->lookupMatcher(keyChar);
         if (m == 0) {
             ICU_Utility::appendToRule(result, keyChar, FALSE, escapeUnprintable, quoteBuf);
         } else {
@@ -176,7 +176,7 @@ UBool StringMatcher::matchesIndexValue(uint8_t v) const {
         return TRUE;
     }
     UChar32 c = pattern.char32At(0);
-    const UnicodeMatcher *m = data.lookupMatcher(c);
+    const UnicodeMatcher *m = data->lookupMatcher(c);
     return (m == 0) ? ((c & 0xFF) == v) : m->matchesIndexValue(v);
 }
 
@@ -219,14 +219,29 @@ UnicodeString& StringMatcher::toReplacerPattern(UnicodeString& rule,
 }
 
 /**
- * Remove any match data.  This must be called before performing a
+ * Remove any match info.  This must be called before performing a
  * set of matches with this segment.
  */
  void StringMatcher::resetMatch() {
     matchStart = matchLimit = -1;
 }
 
+/**
+ * Implement UnicodeFunctor
+ */
+void StringMatcher::setData(const TransliterationRuleData* d) {
+    data = d;
+    int32_t i = 0;
+    while (i<pattern.length()) {
+        UChar32 c = pattern.char32At(i);
+        UnicodeFunctor* f = data->lookup(c);
+        if (f != NULL) {
+            f->setData(data);
+        }
+        i += UTF_CHAR_LENGTH(c);
+    }    
+}
+
 U_NAMESPACE_END
 
 //eof
-

icu4c/source/i18n/strmatch.h

@@ -96,6 +96,11 @@ class StringMatcher : public UnicodeFunctor, public UnicodeMatcher, public Unico
      */
     virtual UBool matchesIndexValue(uint8_t v) const;
 
+    /**
+     * Implement UnicodeFunctor
+     */
+    virtual void setData(const TransliterationRuleData*);
+
     /**
      * Replace characters in 'text' from 'start' to 'limit' with the
      * output text of this object.  Update the 'cursor' parameter to
@@ -151,7 +156,7 @@ class StringMatcher : public UnicodeFunctor, public UnicodeMatcher, public Unico
      * Context object that maps stand-ins to matcher and replacer
      * objects.
      */
-    const TransliterationRuleData& data;
+    const TransliterationRuleData* data;
 
     /**
      * The segment number, 1-based, or 0 if not a segment.

icu4c/source/i18n/strrepl.cpp

@@ -257,6 +257,22 @@ UnicodeString& StringReplacer::toReplacerPattern(UnicodeString& rule,
     return rule;
 }
 
+/**
+ * UnicodeFunctor API
+ */
+void StringReplacer::setData(const TransliterationRuleData* d) {
+    data = d;
+    int32_t i = 0;
+    while (i<output.length()) {
+        UChar32 c = output.char32At(i);
+        UnicodeFunctor* f = data->lookup(c);
+        if (f != NULL) {
+            f->setData(data);
+        }
+        i += UTF_CHAR_LENGTH(c);
+    }
+}
+
 U_NAMESPACE_END
 
 //eof

icu4c/source/i18n/strrepl.h

@@ -126,6 +126,11 @@ class StringReplacer : public UnicodeFunctor, public UnicodeReplacer {
      */
     virtual UnicodeString& toReplacerPattern(UnicodeString& result,
                                              UBool escapeUnprintable) const;
+
+    /**
+     * UnicodeFunctor API
+     */
+    virtual void setData(const TransliterationRuleData*);
 };
 
 U_NAMESPACE_END

icu4c/source/i18n/unicode/unifilt.h

@@ -81,6 +81,11 @@ public:
                                  int32_t limit,
                                  UBool incremental);
 
+    /**
+     * UnicodeFunctor API.  Nothing to do.
+     */
+    virtual void setData(const TransliterationRuleData*) {}
+
 protected:
 
     UnicodeFilter();

icu4c/source/i18n/unicode/unifunct.h

@@ -16,6 +16,7 @@ U_NAMESPACE_BEGIN
 
 class UnicodeMatcher;
 class UnicodeReplacer;
+class TransliterationRuleData;
 
 /**
  * <code>UnicodeFunctor</code> is an abstract base class for objects
@@ -97,6 +98,16 @@ class U_I18N_API UnicodeFunctor {
      */
     virtual UClassID getDynamicClassID(void) const { return getStaticClassID(); };
 
+    /**
+     * Set the data object associated with this functor.  The data
+     * object provides context for functor-to-standin mapping.  This
+     * method is required when assigning a functor to a different data
+     * object.  This function MAY GO AWAY later if the architecture is
+     * changed to pass data object pointers through the API.
+     * @draft
+     */
+    virtual void setData(const TransliterationRuleData*) = 0;
+
  protected:
 
     UnicodeFunctor();