X-Git-Url: http://jsfdemo.indexdata.com/?a=blobdiff_plain;f=src%2Ftermlists.c;h=fdfc7f0ea81067e27c5defeb3b658afe3f293c93;hb=0e0ddf5be474a96ae8838fe9845858b7ca96b9d8;hp=892e6d9a5229bf6dd3c95f1fcd28f47a885a78b5;hpb=85a4a8d3f48c3d2c95efb6437b25f0804d966161;p=pazpar2-moved-to-github.git

diff --git a/src/termlists.c b/src/termlists.c
index 892e6d9..fdfc7f0 100644
--- a/src/termlists.c
+++ b/src/termlists.c
@@ -1,5 +1,5 @@
 /* This file is part of Pazpar2.
-   Copyright (C) 2006-2009 Index Data
+   Copyright (C) 2006-2010 Index Data
 
 Pazpar2 is free software; you can redistribute it and/or modify it under
 the terms of the GNU General Public License as published by the Free
@@ -42,8 +42,7 @@ struct termlist_bucket
 struct termlist
 {
     struct termlist_bucket **hashtable;
-    int hashtable_size;
-    int hashmask;
+    unsigned hash_size;
 
     struct termlist_score **highscore;
     int highscore_size;
@@ -53,24 +52,14 @@ struct termlist
     NMEM nmem;
 };
 
-struct termlist *termlist_create(NMEM nmem, int numterms, int highscore_size)
+struct termlist *termlist_create(NMEM nmem, int highscore_size)
 {
-    int hashsize = 1;
-    int halfnumterms;
-    struct termlist *res;
-
-    // Calculate a hash size smallest power of 2 larger than 50% of expected numterms
-    halfnumterms = numterms >> 1;
-    if (halfnumterms < 0)
-        halfnumterms = 1;
-    while (hashsize < halfnumterms)
-        hashsize <<= 1;
-    res = nmem_malloc(nmem, sizeof(struct termlist));
-    res->hashtable = nmem_malloc(nmem, hashsize * sizeof(struct termlist_bucket*));
-    memset(res->hashtable, 0, hashsize * sizeof(struct termlist_bucket*));
-    res->hashtable_size = hashsize;
+    struct termlist *res = nmem_malloc(nmem, sizeof(struct termlist));
+    res->hash_size = 399;
+    res->hashtable =
+        nmem_malloc(nmem, res->hash_size * sizeof(struct termlist_bucket*));
+    memset(res->hashtable, 0, res->hash_size * sizeof(struct termlist_bucket*));
     res->nmem = nmem;
-    res->hashmask = hashsize - 1; // Creates a bitmask
 
     res->highscore = nmem_malloc(nmem, highscore_size * sizeof(struct termlist_score *));
     res->highscore_size = highscore_size;
@@ -118,7 +107,7 @@ static void update_highscore(struct termlist *tl, struct termlist_score *t)
     }
 }
 
-void termlist_insert(struct termlist *tl, const char *term)
+void termlist_insert(struct termlist *tl, const char *term, int freq)
 {
     unsigned int bucket;
     struct termlist_bucket **p;
@@ -131,12 +120,12 @@ void termlist_insert(struct termlist *tl, const char *term)
     for (cp = buf + strlen(buf); cp != buf && strchr(",. -", cp[-1]); cp--)
         cp[-1] = '\0';
     
-    bucket = jenkins_hash((unsigned char *)buf) & tl->hashmask;
+    bucket = jenkins_hash((unsigned char *)buf) % tl->hash_size;
     for (p = &tl->hashtable[bucket]; *p; p = &(*p)->next)
     {
         if (!strcmp(buf, (*p)->term.term))
         {
-            (*p)->term.frequency++;
+            (*p)->term.frequency += freq;
             update_highscore(tl, &((*p)->term));
             break;
         }
@@ -146,7 +135,7 @@ void termlist_insert(struct termlist *tl, const char *term)
         struct termlist_bucket *new = nmem_malloc(tl->nmem,
                 sizeof(struct termlist_bucket));
         new->term.term = nmem_strdup(tl->nmem, buf);
-        new->term.frequency = 1;
+        new->term.frequency = freq;
         new->next = 0;
         *p = new;
         update_highscore(tl, &new->term);