dbz - Database routines for InterNetNews

Bugs

       Unlike dbm, dbz will refuse to dbzstore with a key already in the database.  The user is responsible for
       avoiding this.

       The RFC5322 case mapper implements only a first approximation to the hideously-complex RFC5322 case
       rules.

       dbz no longer tries to be call-compatible with dbm in any way.

Description

These functions provide an indexing system for rapid random access to a text file, hereafter named the
base file.

dbz stores offsets into the base file for rapid retrieval. All retrievals are keyed on a hash value that
is generated by the HashMessageID function in libinn(3).

dbzinit opens a database, an index into the base file name, consisting of files name.dir, name.index, and
name.hash which must already exist. (If the database is new, they should be zero-length files.)
Subsequent accesses go to that database until dbzclose is called to close the database. When tagged hash
format is used (if --enable-tagged-hash was given at configure time), a name.pag file is used instead of
.index and .hash.

dbzfetch searches the database for the specified key, assigning the offset of the base file for the
corresponding key to value, if any.

dbzstore stores the key-data pair in the database. It will return "DBZSTORE_EXISTS" for duplicates
(already existing entries), and "DBZSTORE_OK" for success. It will fail with "DBZSTORE_ERROR" if the
database files are not writable or not opened, or if any other error occurs.

dbzexists will verify whether or not the given hash exists or not. dbz is optimized for this operation
and it may be significantly faster than dbzfetch.

dbzfresh is a variant of dbzinit for creating a new database with more control over details. The size
parameter specifies the size of the first hash table within the database, in number of key-value pairs.
Performance will be best if the number of key-value pairs stored in the database does not exceed about
2/3 of size, or 1/2 of size when the tagged hash format is used. (The dbzsize function, given the
expected number of key-value pairs, will suggest a database size that meets these criteria.) Assuming
that an fseek offset is 4 bytes, the .index file will be 4 * size bytes. The .hash file will be
"DBZ_INTERNAL_HASH_SIZE" * size bytes (the .dir file is tiny and roughly constant in size) until the
number of key-value pairs exceeds about 80% of size. (Nothing awful will happen if the database grows
beyond 100% of size, but accesses will slow down quite a bit and the .index and .hash files will grow
somewhat.)

dbz stores up to "DBZ_INTERNAL_HASH_SIZE" bytes (by default, 4 bytes if tagged hash format is used, 6
otherwise) of the Message-ID's hash in the .hash file to confirm a hit. This eliminates the need to read
the base file to handle collisions.

A size of "0" given to dbzfresh is synonymous with the local default; the normal default is suitable for
tables of about 6,000,000 key-value pairs (or 500,000 key-value pairs when the tagged hash format is
used). That default value is used by dbzinit.

When databases are regenerated periodically, as it is the case for the history file, it is simplest to
pick the parameters for a new database based on the old one. This also permits some memory of past sizes
of the old database, so that a new database size can be chosen to cover expected fluctuations. dbzagain
is a variant of dbzinit for creating a new database as a new generation of an old database. The database
files for oldname must exist. dbzagain is equivalent to calling dbzfresh with a size equal to the result
of applying dbzsize to the largest number of entries in the oldname database and its previous 10
generations.

When many accesses are being done by the same program, dbz is massively faster if its first hash table is
in memory. If the pag_incore flag is set to "INCORE_MEM", an attempt is made to read the table in when
the database is opened, and dbzclose writes it out to disk again (if it was read successfully and has
been modified). dbzsetoptions can be used to set the pag_incore and exists_incore flags to different
values which should be "INCORE_NO" (read from disk), "INCORE_MEM" (read from memory) or "INCORE_MMAP"
(read from a mmap'ed file) for the .hash and .index files separately; this does not affect the status of
a database that has already been opened. The default is "INCORE_NO" for the .index file and
"INCORE_MMAP" for the .hash file. The attempt to read the table in may fail due to memory shortage; in
this case dbz fails with an error. Stores to an in-memory database are not (in general) written out to
the file until dbzclose or dbzsync, so if robustness in the presence of crashes or concurrent accesses is
crucial, in-memory databases should probably be avoided or the writethrough option should be set to true
(telling to systematically write to the filesystem in addition to updating the in-memory database).

If the nonblock option is true, then writes to the .hash and .index files will be done using non-blocking
I/O. This can be significantly faster if your platform supports non-blocking I/O with files. It is only
applicable if you're not mmap'ing the database.

dbzsync causes all buffers etc. to be flushed out to the files. It is typically used as a precaution
against crashes or concurrent accesses when a dbz-using process will be running for a long time. It is a
somewhat expensive operation, especially for an in-memory database.

Concurrent reading of databases is fairly safe, but there is no (inter)locking, so concurrent updating is
not.

An open database occupies three stdio streams and two file descriptors; Memory consumption is negligible
except for in-memory databases (and stdio buffers).

Diagnostics

       Functions returning bool values return true for success, false for failure.

       dbzinit attempts to have errno set plausibly on return, but otherwise this is not guaranteed.  An errno
       of "EDOM" from dbzinit indicates that the database did not appear to be in dbz format.

       If "DBZTEST" is defined at compile-time, then a main() function will be included.  This will do
       performance tests and integrity test.

History

       The original dbz was written by Jon Zeeff <zeeff@b-tech.ann-arbor.mi.us>.  Later contributions by David
       Butler and Mark Moraes.  Extensive reworking, including this documentation, by Henry Spencer
       <henry@zoo.toronto.edu> as part of the C News project.  MD5 code borrowed from RSA.  Extensive reworking
       to remove backwards compatibility and to add hashes into dbz files by Clayton O'Neill
       <coneill@oneill.net>.  Rewritten into POD by Julien Elie.

Name

       dbz - Database routines for InterNetNews

Synopsis

           #include <inn/dbz.h>

           #define DBZMAXKEY              ...
           #define DBZ_INTERNAL_HASH_SIZE ...

           typedef enum
           {
               DBZSTORE_OK,
               DBZSTORE_EXISTS,
               DBZSTORE_ERROR
           } DBZSTORE_RESULT;

           typedef enum
           {
               INCORE_NO,
               INCORE_MEM,
               INCORE_MMAP
           } dbz_incore_val;

           typedef struct {
               bool writethrough;
               dbz_incore_val pag_incore;
               dbz_incore_val exists_incore;
               bool nonblock;
           } dbzoptions;

           typedef struct {
               char hash[DBZ_INTERNAL_HASH_SIZE];
           } __attribute__((__packed__)) erec;

           extern bool dbzinit(const char *name);
           extern bool dbzclose(void);

           extern bool dbzfresh(const char *name, off_t size);
           extern bool dbzagain(const char *name, const char *oldname);
           extern bool dbzexists(const HASH key);
           extern bool dbzfetch(const HASH key, off_t *value);
           extern DBZSTORE_RESULT dbzstore(const HASH key, off_t data);
           extern bool dbzsync(void);
           extern long dbzsize(off_t contents);
           extern void dbzsetoptions(const dbzoptions options);
           extern void dbzgetoptions(dbzoptions *options);