From 8c56e563826cf599bfb75a0dfc67c3796037c6e6 Mon Sep 17 00:00:00 2001
From: George Rhoten <grhoten@users.noreply.github.com>
Date: Sat, 16 Feb 2008 11:26:53 +0000
Subject: [PATCH] ICU-6076 Make it a little easier to detect initialization
 problems, and improve initialization time.

X-SVN-Rev: 23442
---
 icu4c/source/common/unicode/uniset.h |  2 ++
 icu4c/source/common/uniset.cpp       | 10 ++++++++--
 icu4c/source/common/uniset_props.cpp | 19 ++++++++++++++-----
 3 files changed, 24 insertions(+), 7 deletions(-)

diff --git a/icu4c/source/common/unicode/uniset.h b/icu4c/source/common/unicode/uniset.h
index 09e25a3ce39..40e644be4fc 100644
--- a/icu4c/source/common/unicode/uniset.h
+++ b/icu4c/source/common/unicode/uniset.h
@@ -1503,6 +1503,8 @@ private:
                               UnicodeString& rebuiltPat,
                               UErrorCode& ec);
 
+    static const UnicodeSet* getInclusions(int32_t src, UErrorCode &status);
+
     /**
      * A filter that returns TRUE if the given code point should be
      * included in the UnicodeSet being constructed.
diff --git a/icu4c/source/common/uniset.cpp b/icu4c/source/common/uniset.cpp
index 9720d1c3c0d..215bdb34193 100644
--- a/icu4c/source/common/uniset.cpp
+++ b/icu4c/source/common/uniset.cpp
@@ -1446,8 +1446,14 @@ UnicodeSet& UnicodeSet::compact() {
     if (len < capacity) {
         // Make the capacity equal to len or 1.
         // We don't want to realloc of 0 size.
-        capacity = len + (len == 0);
-        list = (UChar32*) uprv_realloc(list, sizeof(UChar32) * capacity);
+        int32_t newCapacity = len + (len == 0);
+        UChar32* temp = (UChar32*) uprv_realloc(list, sizeof(UChar32) * newCapacity);
+        if (temp) {
+            list = temp;
+            capacity = newCapacity;
+        }
+        // else what the heck happened?! We allocated less memory!
+        // Oh well. We'll keep our original array.
     }
     return *this;
 }
diff --git a/icu4c/source/common/uniset_props.cpp b/icu4c/source/common/uniset_props.cpp
index e1a40c51366..cce3a28da8a 100644
--- a/icu4c/source/common/uniset_props.cpp
+++ b/icu4c/source/common/uniset_props.cpp
@@ -136,7 +136,14 @@ U_CDECL_END
 
 U_NAMESPACE_BEGIN
 
-static const UnicodeSet* getInclusions(int32_t src, UErrorCode &status) {
+/*
+Reduce excessive reallocation, and make it easier to detect initialization
+problems.
+Usually you don't see smaller sets than this for Unicode 5.0.
+*/
+#define DEFAULT_INCLUSION_CAPACITY 3072
+
+const UnicodeSet* UnicodeSet::getInclusions(int32_t src, UErrorCode &status) {
     UBool needInit;
     UMTX_CHECK(NULL, (INCLUSIONS[src] == NULL), needInit);
     if (needInit) {
@@ -146,9 +153,10 @@ static const UnicodeSet* getInclusions(int32_t src, UErrorCode &status) {
             _set_add,
             _set_addRange,
             _set_addString,
-            NULL // don't need remove()
+            NULL, // don't need remove()
+            NULL // don't need removeRange()
         };
-
+        incl->ensureCapacity(DEFAULT_INCLUSION_CAPACITY, status);
         if (incl != NULL) {
             switch(src) {
             case UPROPS_SRC_CHAR:
@@ -894,7 +902,7 @@ void UnicodeSet::applyFilter(UnicodeSet::Filter filter,
     clear();
 
     UChar32 startHasProperty = -1;
-    int limitRange = inclusions->getRangeCount();
+    int32_t limitRange = inclusions->getRangeCount();
 
     for (int j=0; j<limitRange; ++j) {
         // get current range
@@ -1322,7 +1330,8 @@ UnicodeSet& UnicodeSet::closeOver(int32_t attribute) {
                 _set_add,
                 _set_addRange,
                 _set_addString,
-                NULL // don't need remove()
+                NULL, // don't need remove()
+                NULL // don't need removeRange()
             };
 
             // start with input set to guarantee inclusion