X-Git-Url: http://mj.ucw.cz/gitweb/?a=blobdiff_plain;f=lib%2Findex.h;h=bab49b952dcaca54ecbcb43058eecfb494f91e09;hb=5bb0345810e5d0097488fffd1216b9ed0b4b9d05;hp=cb23ffd63a7b85b654432b8793cdf433e6613551;hpb=54b504cb60e57a2bc00be2329a3f7ffbf82b9845;p=libucw.git

diff --git a/lib/index.h b/lib/index.h
index cb23ffd6..bab49b95 100644
--- a/lib/index.h
+++ b/lib/index.h
@@ -1,20 +1,18 @@
 /*
  *	Sherlock: Data structures used in indices
  *
- *	(c) 2001--2003 Martin Mares <mj@ucw.cz>
+ *	(c) 2001--2004 Martin Mares <mj@ucw.cz>
  */
 
 #ifndef _SHERLOCK_INDEX_H
 #define _SHERLOCK_INDEX_H
 
-#include "lib/fastbuf.h"
-#include SHERLOCK_CUSTOM
-#include "charset/unistream.h"
+#include "custom/lib/custom.h"
 
-#define INDEX_VERSION (0x32240100+sizeof(struct card_attr))	/* Increase with each incompatible change in index format */
+#define INDEX_VERSION (0x32260200+sizeof(struct card_attr))	/* Increase with each incompatible change in index format */
 
 /*
- *  Words and word complexes
+ *  Words
  *
  *  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
@@ -23,15 +21,39 @@
  *
  *  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	/* a multiple of 4 */
-#define MAX_COMPLEX_LEN		10
 
 /* Word and string types are defined in lib/custom.h */
 
+/* Types used for storing contexts */
+
+#ifdef CONFIG_CONTEXTS
+#if CONFIG_MAX_CONTEXTS == 32768
+typedef u16 context_t;
+#define bget_context bgetw
+#define bput_context bputw
+#define GET_CONTEXT GET_U16
+#define PUT_CONTEXT PUT_U16
+#elif CONFIG_MAX_CONTEXTS == 256
+typedef byte context_t;
+#define bget_context bgetc
+#define bput_context bputc
+#define GET_CONTEXT GET_U8
+#define PUT_CONTEXT PUT_U8
+#else
+#error CONFIG_MAX_CONTEXTS set to an invalid value.
+#endif
+#else
+struct fastbuf;
+typedef struct { } context_t;
+static inline uns bget_context(struct fastbuf *b UNUSED) { return 0; }
+static inline void bput_context(struct fastbuf *b UNUSED, uns context UNUSED) { }
+#define GET_CONTEXT(p) 0
+#define PUT_CONTEXT(p,x) do {} while(0)
+#endif
+
 /* Index card attributes */
 
 struct card_attr {
@@ -57,7 +79,7 @@ enum card_flag {
   CARD_FLAG_MERGED = 8,			/* Destination of a merge [merger] */
   CARD_FLAG_IMAGE = 16,			/* Is an image object [scanner] */
   CARD_FLAG_FRAMESET = 32,		/* Contains a frameset to be ignored [scanner] */
-  CARD_FLAG_GIANT_CLASS = 64,		/* Belongs to a very large class, subject to penalties [merger] */
+  CARD_FLAG_OVERRIDEN = 64,		/* Overriden by another index [sherlockd] */
 };
 
 #define CARD_POS_SHIFT 5		/* Card positions are shifted this # of bits to the right */
@@ -77,8 +99,10 @@ enum card_flag {
 #define CA_GET_FILE_INFO(a) ((a)->type_flags & 0x1f)
 #define CA_GET_FILE_LANG(a) ((a)->type_flags & 0x80 ? 0 : CA_GET_FILE_INFO(a))
 #define FILETYPE_ATTRS SMALL_SET_ATTR(ftype, FILETYPE, CA_GET_FILE_TYPE, ext_ft_parse)
+#define MAX_FILE_TYPES 8
+#define FILETYPE_IS_TEXT(f) ((f) < 4)
 byte *ext_ft_parse(u32 *dest, byte *value, uns intval);
-extern byte *custom_file_type_names[8];
+extern byte *custom_file_type_names[MAX_FILE_TYPES];
 #else
 #define FILETYPE_ATTRS
 #endif
@@ -91,7 +115,7 @@ byte *ext_lang_parse(u32 *dest, byte *value, uns intval);
 #define LANG_ATTRS
 #endif
 
-#define EXTENDED_ATTRS CUSTOM_ATTRS FILETYPE_ATTRS LANG_ATTRS
+#define EXTENDED_ATTRS CUSTOM_ATTRS LANG_ATTRS		/* Beware, FILETYPE_ATTRS are handled separately */
 
 /* String fingerprints */
 
@@ -107,72 +131,19 @@ fp_hash(struct fingerprint *fp)
   return (fp->hash[0] << 24) | (fp->hash[1] << 16) | (fp->hash[2] << 8) | fp->hash[3];
 }
 
-/* Reading of tagged text (Unicode values, tags mapped to 0x80000000 and higher) */
-
-#define GET_TAGGED_CHAR(p,u) do {				\
-  u = *p;							\
-  if (u >= 0xc0)						\
-    GET_UTF8_CHAR(p,u);						\
-  else if (u >= 0x80)						\
-    {								\
-      p++;							\
-      if (u >= 0xb0)						\
-        {							\
-	  ASSERT(u == 0xb0);					\
-	  u += 0x80020000;					\
-        }							\
-      else if (u >= 0xa0)					\
-        {							\
-	  ASSERT(*p >= 0x80 && *p <= 0xbf);			\
-	  u = 0x80010000 + ((u & 0x0f) << 6) + (*p++ & 0x3f);	\
-        }							\
-      else							\
-	u += 0x80000000;					\
-    }								\
-  else								\
-    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)
-{
-  uns u = bgetc(f);
-  if ((int)u < 0x80)
-    ;
-  else if (u < 0xc0)
-    {
-      if (u >= 0xb0)
-	{
-	  ASSERT(u == 0xb0);
-	  u += 0x80020000;
-	}
-      else if (u >= 0xa0)
-	{
-	  uns v = bgetc(f);
-	  ASSERT(v >= 0x80 && v <= 0xbf);
-	  u = 0x80010000 + ((u & 0x0f) << 6) + (v & 0x3f);
-	}
-      else
-	u += 0x80000000;
-    }
-  else
-    {
-      bungetc(f);
-      u = bget_utf8(f);
-    }
-  return u;
-}
+/* The card fingerprints */
+
+struct card_print {
+  struct fingerprint fp;
+  u32 cardid;
+};
+
+/* URL keys */
+
+#define URL_KEY_BUF_SIZE (3*MAX_URL_SIZE)
+byte *url_key(byte *url, byte *buf);
+void url_fingerprint(byte *url, struct fingerprint *fp);
+void url_key_init(void);
 
 /* Conversion of document age from seconds to our internal units */