From 36b266f894f168ff64b66bfd05799539e458f092 Mon Sep 17 00:00:00 2001 From: hniksic Date: Fri, 19 Dec 2003 17:40:19 -0800 Subject: [PATCH] [svn] Trivial doc fix. --- src/hash.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-) diff --git a/src/hash.c b/src/hash.c index af5cf7a2..f86c6726 100644 --- a/src/hash.c +++ b/src/hash.c @@ -106,14 +106,13 @@ so, delete this exception statement from your version. */ The hash table is implemented as an open-addressed table with linear probing collision resolution. - In regular language, it means that all the hash entries (pairs of - pointers key and value) are stored in a contiguous array. The - position of each mapping is determined by the hash value of its key - and the size of the table: location := hash(key) % size. If two - different keys end up on the same position (collide), the one that - came second is placed at the next empty position following the - occupied place. This collision resolution technique is called - "linear probing". + The above means that all the hash entries (pairs of pointers, key + and value) are stored in a contiguous array. The position of each + mapping is determined by the hash value of its key and the size of + the table: location := hash(key) % size. If two different keys end + up on the same position (collide), the one that came second is + placed at the next empty position following the occupied place. + This collision resolution technique is called "linear probing". There are more advanced collision resolution methods (quadratic probing, double hashing), but we don't use them because they incur