Modify the behaviour of mixed C/Fortran string comparisons, so that it

is consistent with the routines converting Fortran strings to C
strings.

Fortran strings have a fixed length, and trailing white space is
usually ignored.  The Cactus routine converting Fortran to C strings
did already ignore trailing white space, but the routine comparing
Fortran and C strings did not.  This is now consistent.


git-svn-id: http://svn.cactuscode.org/flesh/trunk@4590 17b73243-c579-4c4c-a9d2-2d5706c11dac

1 files changed, 10 insertions, 2 deletions

diff --git a/src/util/String.c b/src/util/String.c
index e152efd9..23be0987 100644
--- a/src/util/String.c
+++ b/src/util/String.c
@@ -477,12 +477,20 @@ int Util_StrCmpi (const char *string1, const char *string2)
 int Util_StrMemCmpi (const char *string1, const char *string2, size_t length2)
 {
   int retval;
+  size_t last2;
 
 
+  /* Fortran speciality: ignore trailing blanks */
+  last2 = length2;
+  while (last2 > 0 && string2[last2-1] == ' ')
+  {
+    last2--;
+  }
+
   do
   {
-    retval = tolower (*string1) - (length2 ? tolower (*string2) : '\0');
-  } while (! retval && *string1++ && (string2++, length2--));
+    retval = tolower (*string1) - (last2 ? tolower (*string2) : '\0');
+  } while (! retval && *string1++ && (string2++, last2--));
 
   return (retval);
 }