X-Git-Url: http://mj.ucw.cz/gitweb/?a=blobdiff_plain;f=lib%2Findex.h;h=b17e9a62626440c342d1632086ffa94aa8bc7e1b;hb=8a5c234a1c81e4d199d6c8a20b25548068d843f7;hp=8bfbe88d4d6862f58118de370d52f41d3d483553;hpb=7a84ff60630fb4211e60320843955ff2cf013ac2;p=libucw.git

diff --git a/lib/index.h b/lib/index.h
index 8bfbe88d..b17e9a62 100644
--- a/lib/index.h
+++ b/lib/index.h
@@ -11,9 +11,21 @@
 #include SHERLOCK_CUSTOM
 #include "charset/unistream.h"
 
-/* Words */
+/*
+ *  Words and word complexes
+ *
+ *  MAX_WORD_LEN is the maximum length (measured in UTF-8 characters, excluding
+ *  the terminating zero byte if there's any) of any word which may appear in the
+ *  indices or in the bucket file. Naturally, the same constant also bounds
+ *  the number of UCS-2 characters in a word.
+ *
+ *  Caveat: If you are upcasing/downcasing the word, the UTF-8 encoding can
+ *  expand, although at most twice, so you need to reserve 2*MAX_WORD_LEN bytes.
+ *
+ *  MAX_COMPLEX_LEN is the upper bound on number of words in any word complex.
+ */
 
-#define MAX_WORD_LEN		64
+#define MAX_WORD_LEN		64	/* a multiple of 4 */
 #define MAX_COMPLEX_LEN		10
 
 /* Word and string types are defined in lib/custom.h */