⟦15749c8ce⟧

TextFile

This is the initial release of GNU dbm.  Better documentation will
be written soon.  For now, this file briefly describes the contents
of this release and how to use it.

The files are:

COPYING		- Copying information.
README		- This file
bucket.c, dbminit.c, delete.c, extern.h, falloc.c, fetch.c, findkey.c,
gdbm.h, gdbmclose.c, gdbmdelete.c, gdbmerrno.h, gdbmfetch.c,
gdbmopen.c, gdbmreorg.c, gdbmseq.c, gdbmstore.c, global.c, hash.c
seq.c, store.c, update.c  -  Source for GNU dbm library.
Makefile 	- Makefile, will make gdbm.a 
test.c		- A simple test program.
version.c	- The current version number in a string.


GNU dbm is a set of database routines that use extendible hashing and
works similar to the standard UNIX dbm routines.  The basic unit of
data is the structure

  typedef struct {
             char *dptr;
             int   dsize;
          } datum;

The following is a quick list of the routines.  After this list, a
longer description of each routine will be given.  The routines are:

  char *gdbm_open ( name, block_size, read_write, fatal_func )

  gdbm_close ( dbf )
  
  int gdbm_store ( dbf, key, content )

  datum gdbm_fetch ( dbf, key )

  int gdbm_delete ( dbf, key )

  datum gdbm_firstkey ( dbf )

  datum gdbm_nextkey ( dbf, key )

  int gdbm_reorganize ( dbf )

  gdbm_recovery ( dbf, state )

For compatibility with the standard dbm, the following routines are 
defined.

  int dbminit ( name )

  int store ( key, content )

  datum fetch ( key )

  int delete ( key )

  datum firstkey ()

  datum nextkey ( key )

An additional routine is needed for use with the compatibility routines.

  dbmclose ()




Description of GNU dbm routines.
--------------------------------

GNU dbm allows multiple data files.  A routine that opens a gdbm file
is designated as a "reader" or a "writer".  Only one writer may open a
gdbm file and many readers may open the file.  Readers and writers can
not open the gdbm file at the same time. The procedure for opening a
gdbm file is:

char * dbf;

dbf = gdbm_open ( name, block_size, read_write, fatal_func )

The parameters are:
  char *name - the name of the file (the complete name, gdbm does
	  not append any characters to this name)
  int block_size - the size of a single transfer from disk to memory.
	  This parameter is ignored unless the file is a new file.
	  The minimum size is 512.  If it is less than 512, dbm will
	  use the stat block size for the file system.
  int read_write - 0 => reader, 1 => writer
  void (*fatal_func) () - a function for dbm to call if it detects a
	  fatal error. The only parameter of this function is a string.
	  If the value of 0 is provided, dbm will use a default function.

The return value, dbf, is the pointer needed by all other routines to
access that gdbm file.  If the return is the NULL pointer, gdbm_open
was not successful.  The errors can be found in "gdbm_errno" for gdbm
errors and in "errno" for file system errors.  (For error codes, see
gdbmerrno.h.)

In all of the following calls, the parameter "dbf" refers to the pointer
returned from gdbm_open.

It is important that every file opened is also closed.  This is needed to
update the reader/writer count on the file.  This is done by:

  gdbm_close(dbf);


The database is used by 3 primary routines.  The first stores data in the
database.

  ret = gdbm_store ( dbf, key, content )

  The parameters are:
     char *dbf - the pointer returned by gdbm_open
     datum key - the key data
     datum content - the data to be associated with the key

  If a reader calls store, ret gets -1.  Otherwise, ret is 0.
  NOTICE: If you store data for a key that is already in the data base,
  gdbm replaces the old data with the new data.  You do not get two
  data items for the same key and you do not get an error from gdbm_store.

To search for some data:

  content = gdbm_fetch ( dbf, key )

  The parameters are:
     char *dbf - the pointer returned by gdbm_open
     datum key - the key data

  The "datum" returned in content is a pointer to the data found.  If the
  dptr is NULL, no data was found.  If dptr is not NULL, then it points
  to data allocated by malloc.  gdbm does not automatically free this data.
  The user must free this storage when done using it.  This eliminates the
  need to copy the result to save it for later use. (You just save the 
  pointer.)

To remove some data from the database:

  ret = gdbm_delete ( dbf, key )

  The parameters are:
     char *dbf - the pointer returned by gdbm_open
     datum key - the key data

  The ret value is -1 if the item is present or the requester is a reader.
  The ret value is 0 if there was a successful delete.


The next two routines allow for accessing all items in the database.  This 
access is not key sequential, but it is guaranteed to visit every key in
the database once.  (The order has to do with the hash values.)

  key = gdbm_firstkey ( dbf )

  nextkey = gdbm_nextkey ( dbf, key )

  The parameters are:
     char *dbf - the pointer returned by gdbm_open
     datum key - the key data

  The return values are both datum.  If key.dptr or nextkey.dptr is NULL,
  there is no first key or next key.  Again notice that dptr points to
  data allocated by malloc and gdbm will not free it for you. 

The following routine should be used very seldom.  
  
  ret = gdbm_reorganize ( dbf )

  If you have had a lot of deletions and would like to shrink the space
  used by the gdbm file, the this routine will reorganize the database.
  gdbm will not shorten the length of a gdbm file except by using this
  reorganization.  (Deleted file space will be reused.)

Finally, gdbm does crash detection and recovery.  This causes a little
overhead in processing of the gdbm_store and gdbm_delete routines.  If
crash detection and recovery are unimportant or costs too much, you can
turn off crash detection and recovery.  This is done on an individual
file basis.  It remains off until a request is made to turn it on again.
The following routine changes the state of crash detection and recovery:

  gdbm_recovery ( dbf, state )
 
  The parameters are:
     char *dbf - the pointer returned by gdbm_open
     int  state - 0 to turn it off, anything else to turn on.


The following two are variables that may need to be used:

  gdbm_error gdbm_errno   -   the variable that contains more information
                              about gdbm errors.

  char * gdbm_version     -   the string containing the version information.


There are a few more things of interest.  First, gdbm files are not
"sparse".  You can copy them with the UNIX cp command and they will
not expand in the copying process.  Also, there is a compatibility
mode for use with programs that already use UNIX dbm.  In this
compatibility mode, no gdbm file pointer is required by the user.
Only one file may be opened at a time.  All users in compatibility
mode are assumed to be writers.  Also, default crash detection and
recovery is turned off for compatibility mode.  All returned pointers
in datum structures point to data that gdbm WILL free.  They should be
treated as static pointers (as standard UNIX dbm does).  The
compatibility routine names are the same as the UNIX dbm routine
names.  Their definitions follow:

  int dbminit ( name )

  int store ( key, content )

  datum fetch ( key )

  int delete ( key )

  datum firstkey ()

  datum nextkey ( key )

An additional routine is needed for use with the compatibility routines.  This
closes the database file.  Adding this routine should be the only change
to programs using the standard UNIX dbm.

  dbmclose ()


WARNING: standard UNIX dbm and GNU dbm do not have the same data format in the
file.  You cannot access a standard UNIX dbm file with GNU dbm!




Notes on making GNU dbm.
------------------------

The "Makefile" will make both "gdbm.a", the collection of gdbm routines and
a simple test program that uses the gdbm routines.  There are three make
commands:
   make gdbm.a		makes the gdbm.a archive.
   make test		makes both gdbm.a and the test program.
   make install		installs gdbm.a as /usr/lib/libgdbm.a.

The makefile is distributed with RANDOM_HASH defined.  This causes the
hash function in gdbm to use the UNIX random function to produce a well
distributed and random hash value.  If this computation is too costly
for you, you can remove the definition of RANDOM_HASH from the makefile.
WARNING:  GNU dbm compiled with RANDOM_HASH will not read databases
created by GNU dbm compiled without RANDOM_HASH!

Also, similar to UNIX dbm, GNU dbm will not create a gdbm file.  If it
finds a file with zero length, ti will initialize the file.  To make
a new gdbm database file, "touch file.name" and then run the program
that opens "file.name" as a gdbm file.  The first execution will
initialize the file.

Please send bug reports to phil@wwu.edu.

Thank you.
DataMuseum.dk

DKUUG/EUUG Conference tapes

⟦15749c8ce⟧ TextFile

Derivation

TextFile
