keysym: fix locale dependence in xkb_keysym_from_name()

We currently use strcasecmp, which is locale-dependent. In particular,
one well-known surprise even if restricted just ASCII input is found in
the tr_TR (Turkish) locale, see e.g.
https://msdn.microsoft.com/en-us/library/ms973919.aspx#stringsinnet20_topic5

We have known to avoid locale-dependent functions before, but in this
case, we forgot.

Fix it by implementing our own simple ASCII-only strcasecmp/strncasecmp.
Might have been possible to use strcasecmp_l() with the C locale, but
went the easy route.

Side advantage is that even this non-optimized version is faster than
the optimized libc one (__strcasecmp_l_sse42) since it doesn't need to
do the locale stuff. xkb_keysym_from_name(), which uses strcasecmp
heavily, becomes faster, and so for example Compose file parsing, which
uses xkb_keysym_from_name() heavily, becomes ~20% faster.

Resolves https://github.com/xkbcommon/libxkbcommon/issues/42
Signed-off-by: Ran Benita <ran234@gmail.com>

1 files changed, 11 insertions, 3 deletions

diff --git a/src/utils.h b/src/utils.h
index d63d23a..cb98e8e 100644
--- a/src/utils.h
+++ b/src/utils.h
@@ -29,7 +29,6 @@
 #include <stdbool.h>
 #include <stdio.h>
 #include <string.h>
-#include <strings.h>
 
 #include "darray.h"
 
@@ -44,6 +43,15 @@
     switch (0) { case 0: case (expr): ; } \
 } while (0)
 
+char
+to_lower(char c);
+
+int
+istrcmp(const char *a, const char *b);
+
+int
+istrncmp(const char *a, const char *b, size_t n);
+
 static inline bool
 streq(const char *s1, const char *s2)
 {
@@ -61,13 +69,13 @@ streq_not_null(const char *s1, const char *s2)
 static inline bool
 istreq(const char *s1, const char *s2)
 {
-    return strcasecmp(s1, s2) == 0;
+    return istrcmp(s1, s2) == 0;
 }
 
 static inline bool
 istreq_prefix(const char *s1, const char *s2)
 {
-    return strncasecmp(s1, s2, strlen(s1)) == 0;
+    return istrncmp(s1, s2, strlen(s1)) == 0;
 }
 
 static inline char *