[Tutor] gdbm / python howto or tutorial

[Justin Heath]

Danny Yoo wrote:

>On Mon, 18 Aug 2003, Justin Heath wrote:
>
>  
>
>>I am looking for information on using gdbm with python. I have not used
>>gdbm in the past but am interested in using it in cases where I cannot
>>use mysql. Therefore, I do need something pretty introductory. I have
>>looked for documentation but mostly what I have found is C code, and
>>don't want to "learn" C just to get started. Any pointers would be
>>appreciated.
>>    
>>
>
>Hi Justin,
>
>
>The documenation to gdbm:
>
>    http://www.python.org/doc/lib/module-gdbm.html
>
>doesn't itself have an example, but it say that 'gdbm' objects behave a
>lot like dictionaries (except that keys and values are forced to be
>strings).
>
>
>Here's a quicky example: Let's say that we'd like to create a new DBM
>file:
>
>###
>  
>
>>>>import gdbm
>>>>definitions = gdbm.open('test_gdbm.dbm', 'c')
>>>>definitions['pair programming'] = '''
>>>>        
>>>>
>... An XP practice requiring that each piece of source code
>... to be integrated into the software product should be created
>... by two programmers jointly at one computer.'''.strip()
>  
>
>>>>definitions['code smell'] = '''
>>>>        
>>>>
>... The unpleasant feeling you get when you see bad code'''.strip()
>  
>
>>>>definitions.close()
>>>>        
>>>>
>###
>
>
>This saves a 'dbm' file onto disk.  Another program can then open this dbm
>file up:
>
>###
>  
>
>>>>import gdbm
>>>>definitions = gdbm.open('test_gdbm.dbm')
>>>>        
>>>>
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>gdbm.error: Flag ' ' is not supported.
>###
>
>
>Oh.  Unlike regular file opening, this one doesn't default to 'read' mode,
>so we'll have to directly specify it:
>
>###
>  
>
>>>>definitions = gdbm.open('test_gdbm.dbm', 'r')
>>>>definitions.keys()
>>>>        
>>>>
>['code smell', 'pair programming']
>  
>
>>>>definitions['code smell']
>>>>        
>>>>
>'The unpleasant feeling you get when you see bad code'
>###
>
>
>Ok, better.  *grin*
>
>
>
>
>By the way, Bob Gailer recently mentioned Sqlite,
>
>    http://www.hwaci.com/sw/sqlite/
>
>
>which does appear to be an alternative way of getting very lightweight SQL
>behavior.  The equivalent code, using Sqlite, looks like:
>
>###
>  
>
>>>>import sqlite
>>>>conn = sqlite.connect(db='definitions')
>>>>cursor = conn.cursor()
>>>>cursor.execute("""create table definitions
>>>>        
>>>>
>...  (word varchar not null, defn varchar not null)""")
>  
>
>>>>cursor.execute("insert into definitions values ('hello', 'world')")
>>>>cursor.execute("select * from definitions")
>>>>cursor.fetchall()
>>>>        
>>>>
>[('hello', 'world')]
>  
>
>>>>conn.commit()
>>>>        
>>>>
>###
>
>One warning: when using any SQL database, don't forget to commit.  I made
>this mistake a couple of times since MySQL has autocommit behavior, and I
>still bungle it sometimes... *grin* But the changes to a Sqlite database
>won't persist and will be rolled back when we quit the program, unless we
>explicitely do the commit().
>
>
>Hope this helps!
>
>  
>
I have since found the refered document. I can now create and re-open db 
files I have created. However, I cannot seem to open redhat rpm database 
files or freebsd pkgdb files. I keep getting an invalid argument error 
when using the syntax:

 >>> import anydbm
 >>> dbf = anydbm.open('pkgdb', 'r')
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/anydbm.py", line 86, in open
    return mod.open(file, flag, mode)
  File "/usr/lib/python2.2/dbhash.py", line 16, in open
    return bsddb.hashopen(file, flag, mode)
bsddb.error: (22, 'Invalid argument')

**Note I have also tried the above using bsdb abd dbhash with the same 
result.

This appears that it may be a versioning issue between versions of 
berkley db but I am unsure since I have tried using the db_upgrade on 
(copies of) the files with the same results. If anyone has gotten this 
to work or has any additional info please let me know.

Thanks,
Justin