X-Git-Url: http://mj.ucw.cz/gitweb/?a=blobdiff_plain;f=lib%2Findex.h;h=8bfbe88d4d6862f58118de370d52f41d3d483553;hb=3f29125f27f7ee82281e49fef16ad32811cc3dcc;hp=696a92191c3996011a5cf3774e47cdd9b8ad2ec7;hpb=3fa8e5e40f25d94a220b272a42937336c2400493;p=libucw.git

diff --git a/lib/index.h b/lib/index.h
index 696a9219..8bfbe88d 100644
--- a/lib/index.h
+++ b/lib/index.h
@@ -8,11 +8,13 @@
 #define _SHERLOCK_INDEX_H
 
 #include "lib/fastbuf.h"
+#include SHERLOCK_CUSTOM
 #include "charset/unistream.h"
 
 /* Words */
 
 #define MAX_WORD_LEN		64
+#define MAX_COMPLEX_LEN		10
 
 /* Word and string types are defined in lib/custom.h */
 
@@ -26,10 +28,10 @@ struct index_params {
 
 struct card_attr {
   u32 card;				/* Reference to card description (either oid or filepos) */
+#ifdef CONFIG_SITES
   u32 site_id;
-#define INT_ATTR(t,i,o,k,g,p) t i;
-  CUSTOM_ATTRS				/* Include all custom attributes */
-#undef INT_ATTR
+#endif
+  CUSTOM_CARD_ATTRS			/* Include all custom attributes */
   byte weight;
   byte flags;
   byte age;				/* Document age in pseudo-logarithmic units wrt. reference time */
@@ -86,6 +88,17 @@ fp_hash(struct fingerprint *fp)
     p++;							\
 } while (0)
 
+#define SKIP_TAGGED_CHAR(p) do {				\
+  if (*p >= 0x80 && *p < 0xc0)					\
+    {								\
+      uns u = *p++;						\
+      if (u >= 0xa0 && u < 0xb0 && *p >= 0x80 && *p < 0xc0)	\
+	p++;							\
+    }								\
+  else								\
+    UTF8_SKIP(p);						\
+} while (0)
+
 static inline uns
 bget_tagged_char(struct fastbuf *f)
 {