add --enable-camelcase option to configure

src/common/autoconfig.h.in

@@ -44,5 +44,8 @@
 /* Use file extended attributes */
 #undef RCL_USE_XATTR
 
+/* Split camelCase words */
+#undef RCL_SPLIT_CAMELCASE
+
 /* No X11 */
 #undef WITHOUT_X11

src/common/textsplit.cpp

@@ -18,6 +18,7 @@ static char rcsid[] = "@(#$Id: textsplit.cpp,v 1.38 2008-12-12 11:53:45 dockes E
  *   59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
  */
 #ifndef TEST_TEXTSPLIT
+#include "autoconfig.h"
 
 #include <assert.h>
 
@@ -450,9 +451,14 @@ bool TextSplit::text_to_words(const string &in)
 	    }
 	    break;
 
+#ifdef RCL_SPLIT_CAMELCASE
             // Camelcase handling. 
             // If we get uppercase ascii after lowercase ascii, emit word.
             // This emits "camel" when hitting the 'C' of camelCase
+            // Not enabled by defaults as this makes phrase searches quite 
+            // confusing. 
+            // ie "MySQL manual" is matched by "MySQL manual" and 
+            // "my sql manual" but not "mysql manual"
 	case A_ULETTER:
 	    if (m_span.length() && 
                 charclasses[(unsigned char)m_span[m_span.length() - 1]] == 
@@ -483,6 +489,7 @@ bool TextSplit::text_to_words(const string &in)
                 m_wordLen++;
             }
             goto NORMALCHAR;
+#endif /* CAMELCASE */
 
 
 	default:

src/configure.ac

@@ -154,6 +154,25 @@ if test X$xattrEnabled = Xyes ; then
   AC_DEFINE(RCL_USE_XATTR, 1, [Use file extended attributes])
 fi
 
+# Enable CamelCase word splitting. This is optional because it causes 
+# problems with phrases: with camelcase enabled, "MySQL manual"
+# will be matched by "MySQL manual" and "my sql manual" but not 
+# "mysql manual" (which would need increased slack as manual is now at pos
+# 2 instead of 1
+AC_ARG_ENABLE(camelcase,
+    AC_HELP_STRING([--enable-camelcase],
+   [Enable splitting camelCase words. This is not enabled by default as
+   this makes phrase matches more difficult: you need to use matching
+   case in the phrase query to get a match. Ie querying for 
+   "MySQL manual" and "my sql manual" are the same, but not the same as
+   "mysql manual" (in phrases only and you could raise the phrase slack to
+   get a match).]),
+        camelcaseEnabled=$enableval, camelcaseEnabled=no)
+
+if test X$camelcaseEnabled = Xyes ; then
+  AC_DEFINE(RCL_SPLIT_CAMELCASE, 1, [Split camelCase words])
+fi
+
 
 AC_CHECK_FUNCS(mkdtemp)