find.c

UNIX编程高级环境（第二版）的网络套接字部分的程序源码

#include    "db.h"

/* Find the specified record.
 * Called by db_delete(), db_fetch(), and db_store(). */

int
_db_find(DB *db, const char *key, int writelock)
{
    off_t    offset, nextoffset;

        /* Calculate hash value for this key, then calculate byte
           offset of corresponding chain ptr in hash table.
           This is where our search starts. */

            /* calc offset in hash table for this key */
    db->chainoff = (_db_hash(db, key) * PTR_SZ) + db->hashoff;
    db->ptroff = db->chainoff;

            /* Here's where we lock this hash chain.  It's the
               caller's responsibility to unlock it when done.
               Note we lock and unlock only the first byte. */
    if (writelock) {
        if (writew_lock(db->idxfd, db->chainoff, SEEK_SET, 1) < 0)
            err_dump("writew_lock error");
    } else {
        if (readw_lock(db->idxfd, db->chainoff, SEEK_SET, 1) < 0)
            err_dump("readw_lock error");
    }
            /* Get the offset in the index file of first record
               on the hash chain (can be 0) */
    offset = _db_readptr(db, db->ptroff);

    while (offset != 0) {
        nextoffset = _db_readidx(db, offset);
        if (strcmp(db->idxbuf, key) == 0)
            break;        /* found a match */

        db->ptroff = offset;    /* offset of this (unequal) record */
        offset = nextoffset;    /* next one to compare */
    }
    if (offset == 0)
        return(-1);        /* error, record not found */

        /* We have a match.  We're guaranteed that db->ptroff contains
           the offset of the chain ptr that points to this matching
           index record.  _db_dodelete() uses this fact.  (The chain
           ptr that points to this matching record could be in an
           index record or in the hash table.) */
    return(0);
}