From: Martin Mares <mj@ucw.cz>
Date: Fri, 30 Mar 2001 19:38:45 +0000 (+0000)
Subject: Added indexing of URL words (partially ported from our old alter ego).
X-Git-Tag: holmes-import~1487
X-Git-Url: http://mj.ucw.cz/gitweb/?a=commitdiff_plain;h=981de590927a038ab4b844d16de5a8706daabd1b;p=libucw.git

Added indexing of URL words (partially ported from our old alter ego).

Robert, please ignore word types present in WORD_TYPES_HIDDEN when
searching for contexts -- URL's and other tricky stuff shouldn't show up.
---

diff --git a/lib/index.h b/lib/index.h
index 0fe46f6b..f7ad7c9a 100644
--- a/lib/index.h
+++ b/lib/index.h
@@ -21,13 +21,14 @@ enum word_type {
   WT_KEYWORD,				/* Explicitly marked keyword */
   WT_META,				/* Various meta-information */
   WT_ALT,				/* Alternate texts for graphical elements */
+  WT_URL,				/* Word extracted from document URL */
   WT_MAX
 };
 
 /* Descriptive names used for user output */
 #define WORD_TYPE_USER_NAMES							\
    "reserved", "text", "emph", "small", "title", "hdr1", "hdr2", "keywd",	\
-   "meta", "alt", "type10", "type11", "type12", "type13", "type14", "type15"
+   "meta", "alt", "url", "type11", "type12", "type13", "type14", "type15"
 
 /* Keywords for word type names */
 #define WORD_TYPE_NAMES	       			\
@@ -43,6 +44,9 @@ enum word_type {
 	T(META, 1 << WT_META)			\
 	T(ALT, 1 << WT_ALT)
 
+/* These types are not shown in document contexts */
+#define WORD_TYPES_HIDDEN (1 << WT_URL)
+
 /* String types */
 
 enum string_type {