From e4ebdbec894a0edf237c76737218a309db967794 Mon Sep 17 00:00:00 2001
From: Francois Fleuret <francois@fleuret.org>
Date: Tue, 17 Mar 2009 17:26:37 +0100
Subject: [PATCH] Changed the removal of duplicates so that the most recent one
 is kept instead of the older one.

---
 selector.cc | 32 ++++++++++++++++++++++++++++----
 1 file changed, 28 insertions(+), 4 deletions(-)

diff --git a/selector.cc b/selector.cc
index 2725c2b..7422fd6 100644
--- a/selector.cc
+++ b/selector.cc
@@ -114,13 +114,17 @@ int test_and_add(char *new_string, int new_index,
   code = code % hash_table_size;
 
   while(hash_table[code] >= 0) {
-    if(strcmp(new_string, strings[hash_table[code]]) == 0) return 1;
+    if(strcmp(new_string, strings[hash_table[code]]) == 0) {
+      int result = hash_table[code];
+      hash_table[code] = new_index;
+      return result;
+    }
     code = (code + 1) % hash_table_size;
   }
 
   hash_table[code] = new_index;
 
-  return 0;
+  return -1;
 }
 
 //////////////////////////////////////////////////////////////////////
@@ -546,16 +550,36 @@ int main(int argc, char **argv) {
         while(*s == ' ' || (*s >= '0' && *s <= '9')) s++;
       }
 
-      if(!hash_table || !test_and_add(s, nb_lines, lines, hash_table, hash_table_size)) {
+      int dup;
+
+      if(hash_table) {
+        dup = test_and_add(s, nb_lines, lines, hash_table, hash_table_size);
+      } else {
+        dup = -1;
+      }
+
+      if(dup < 0) {
         lines[nb_lines] = new char[strlen(s) + 1];
         strcpy(lines[nb_lines], s);
-        nb_lines++;
+      } else {
+        lines[nb_lines] = lines[dup];
+        lines[dup] = 0;
       }
+
+      nb_lines++;
     }
   }
 
   delete[] hash_table;
 
+  int n = 0;
+  for(int k = 0; k < nb_lines; k++) {
+    if(lines[k]) {
+      lines[n++] = lines[k];
+    }
+  }
+  nb_lines = n;
+
   if(inverse_order) {
     for(int i = 0; i < nb_lines/2; i++) {
       char *s = lines[nb_lines - 1 - i];
-- 
2.39.5