Re: Reading Sendmail .db files

[waswas-fng]

Try makemap...

makemap -u hash access.db

Will dump a HASH type sendmail db to plaintext, you could also use btree instead. And then you can recreate the map from the text source file:

makemap hash access.db < access.txt

[Maddingue]

Oh thanks, that works. Although I'd still like to know why the Perl scripts (and the short C program) I wrote didn't work. Are the .db files Sendmail create so special that another program can't read them? But, thank you very much for the command, I didn't know the -u option of makemap. At least I can see the content of its db and check what is inside.

[tachyon]

With regard to BerkelyDB => RTFM. You can't just make syntax up and expect it to work. This works fine.....

[root@www mail]# cat reader.pl
#!/usr/bin/perl
use BerkeleyDB;
tie %hash, 'BerkeleyDB::Hash', -Filename => $ARGV[0]
    or die "can't read file '$ARGV[0]': $!";
print map { "  $_ => $hash{$_}\n" } sort keys %hash;

[root@www mail]# ./reader.pl virtusertable.db
  @blech.com.au => blech.com.au
  [snip]

With regard to your second example, once again you need to RTFM. You have a syntax error in that DB_HASH is a $PERL_SCALAR not a constant as you assume. I would agree with your C style brain that it should logically be a constant but I did not design the interface.....

tie %hash, 'DB_File', $file, O_RDONLY, 0666, $DB_HASH

[Maddingue]

Well, I did RTFM of both modules, but missed some parts apparently.. Like $DB_HASH instead of DB_HASH. I'll blame DB_File API for mixing up C-style constants and Perl-style constants. I can also blame myself for not copy&paste the code like I did for the BerkeleyDB example. It seems that it was the -Property => DB_DUP | DB_DUPSORT that prevent the code to run. When I comment this option, the script works. Well, thanks anyway.