Changeset 33563 in vbox for trunk/include/iprt/cpp

Timestamp:

Oct 28, 2010 2:46:26 PM (15 years ago)

Author:

vboxsync

svn:sync-xref-src-repo-rev:

67169

Message:

iprt:ministring: Added the java-style equals() and equalsIgnoreCase() as equals() can optimize the comparison by first checking if the length is the same (compare() cannot as it needs to determin the ordering). Added appendCodePoint() for UTF-8. Fixed the incorrect assumption in toUpper and toLower that the string length remained unchanged - the string might shrink as the folded code points may have a shorter encoding. Added testcase that verifies that a code point will not grow during folding and that have a stable encoding length after it has been changed in a folding.

File:

• trunk/include/iprt/cpp/ministring.h (modified) (16 diffs)

trunk/include/iprt/cpp/ministring.h

@@ -65 +65 @@
     MiniString()
         : m_psz(NULL),
-          m_cbLength(0),
+          m_cch(0),
           m_cbAllocated(0)
     {
@@ -118 +118 @@
     size_t length() const
     {
-        return m_cbLength;
+        return m_cch;
     }
 
@@ -149 +149 @@
     {
         if (    cb != m_cbAllocated
-             && cb > m_cbLength + 1
+             && cb > m_cch + 1
            )
         {
@@ -235 +235 @@
      * Appends the given character to "this".
      *
-     * @param   c               The character to append.
+     * @param   ch              The character to append.
      *
      * @throws  std::bad_alloc  On allocation error.  The object is left unchanged.
@@ -241 +241 @@
      * @returns Reference to the object.
      */
-    MiniString &append(char c);
+    MiniString &append(char ch);
+
+    /**
+     * Appends the given unicode code point to "this".
+     *
+     * @param   uc              The unicode code point to append.
+     *
+     * @throws  std::bad_alloc  On allocation error.  The object is left unchanged.
+     *
+     * @returns Reference to the object.
+     */
+    MiniString &appendCodePoint(RTUNICP uc);
 
     /**
@@ -287 +298 @@
     {
         if (length())
+        {
+            /* Folding an UTF-8 string may result in a shorter encoding (see
+               testcase), so recalculate the length afterwars. */
             ::RTStrToUpper(m_psz);
+            size_t cchNew = strlen(m_psz);
+            Assert(cchNew <= m_cch);
+            m_cch = cchNew;
+        }
         return *this;
     }
@@ -299 +317 @@
     {
         if (length())
+        {
+            /* Folding an UTF-8 string may result in a shorter encoding (see
+               testcase), so recalculate the length afterwars. */
             ::RTStrToLower(m_psz);
+            size_t cchNew = strlen(m_psz);
+            Assert(cchNew <= m_cch);
+            m_cch = cchNew;
+        }
         return *this;
     }
@@ -361 +386 @@
         if (m_psz)
         {
-            m_cbLength = strlen(m_psz);
-            m_cbAllocated = m_cbLength + 1; /* (Required for the Utf8Str::asOutParam case) */
+            m_cch = strlen(m_psz);
+            m_cbAllocated = m_cch + 1; /* (Required for the Utf8Str::asOutParam case) */
         }
         else
         {
-            m_cbLength = 0;
+            m_cch = 0;
             m_cbAllocated = 0;
         }
@@ -434 +459 @@
     }
 
+    /**
+     * Compares the two strings.
+     *
+     * @returns true if equal, false if not.
+     * @param   that    The string to compare with.
+     */
+    bool equals(const MiniString &that) const
+    {
+        return that.length() == length()
+            && memcmp(that.m_psz, m_psz, length()) == 0;
+    }
+
+    /**
+     * Compares the two strings.
+     *
+     * @returns true if equal, false if not.
+     * @param   pszThat The string to compare with.
+     */
+    bool equals(const char *pszThat) const
+    {
+        return RTStrCmp(pszThat, m_psz) == 0;
+    }
+
+    /**
+     * Compares the two strings ignoring differences in case.
+     *
+     * @returns true if equal, false if not.
+     * @param   that    The string to compare with.
+     */
+    bool equalsIgnoreCase(const MiniString &that) const
+    {
+        /* Unfolded upper and lower case characters may require different
+           amount of encoding space, so the length optimization doesn't work. */
+        return RTStrICmp(that.m_psz, m_psz) == 0;
+    }
+
+    /**
+     * Compares the two strings ignoring differences in case.
+     *
+     * @returns true if equal, false if not.
+     * @param   pszThat The string to compare with.
+     */
+    bool equalsIgnoreCase(const char *pszThat) const
+    {
+        return RTStrICmp(pszThat, m_psz) == 0;
+    }
+
     /** @name Comparison operators.
      * @{  */
-    bool operator==(const MiniString &that) const { return !compare(that); }
-    bool operator!=(const MiniString &that) const { return !!compare(that); }
+    bool operator==(const MiniString &that) const { return equals(that); }
+    bool operator!=(const MiniString &that) const { return !equals(that); }
     bool operator<( const MiniString &that) const { return compare(that) < 0; }
     bool operator>( const MiniString &that) const { return compare(that) > 0; }
 
-    bool operator==(const char *that) const       { return !compare(that); }
-    bool operator!=(const char *that) const       { return !!compare(that); }
-    bool operator<( const char *that) const       { return compare(that) < 0; }
-    bool operator>( const char *that) const       { return compare(that) > 0; }
+    bool operator==(const char *pszThat) const    { return equals(pszThat); }
+    bool operator!=(const char *pszThat) const    { return !equals(pszThat); }
+    bool operator<( const char *pszThat) const    { return compare(pszThat) < 0; }
+    bool operator>( const char *pszThat) const    { return compare(pszThat) > 0; }
     /** @} */
 
@@ -567 +639 @@
             RTStrFree(m_psz);
             m_psz = NULL;
-            m_cbLength = 0;
+            m_cch = 0;
             m_cbAllocated = 0;
         }
@@ -591 +663 @@
     void copyFrom(const MiniString &s)
     {
-        if ((m_cbLength = s.m_cbLength))
-        {
-            m_cbAllocated = m_cbLength + 1;
+        if ((m_cch = s.m_cch))
+        {
+            m_cbAllocated = m_cch + 1;
             m_psz = (char *)RTStrAlloc(m_cbAllocated);
             if (RT_LIKELY(m_psz))
@@ -599 +671 @@
             else
             {
-                m_cbLength = 0;
+                m_cch = 0;
                 m_cbAllocated = 0;
 #ifdef RT_EXCEPTIONS_ENABLED
@@ -631 +703 @@
         if (pcsz && *pcsz)
         {
-            m_cbLength = strlen(pcsz);
-            m_cbAllocated = m_cbLength + 1;
+            m_cch = strlen(pcsz);
+            m_cbAllocated = m_cch + 1;
             m_psz = (char *)RTStrAlloc(m_cbAllocated);
             if (RT_LIKELY(m_psz))
@@ -638 +710 @@
             else
             {
-                m_cbLength = 0;
+                m_cch = 0;
                 m_cbAllocated = 0;
 #ifdef RT_EXCEPTIONS_ENABLED
@@ -647 +719 @@
         else
         {
-            m_cbLength = 0;
+            m_cch = 0;
             m_cbAllocated = 0;
             m_psz = NULL;
@@ -654 +726 @@
 
     char    *m_psz;                     /**< The string buffer. */
-    size_t  m_cbLength;                 /**< strlen(m_psz) - i.e. no terminator included. */
+    size_t  m_cch;                      /**< strlen(m_psz) - i.e. no terminator included. */
     size_t  m_cbAllocated;              /**< Size of buffer that m_psz points to; at least m_cbLength + 1. */
 };