Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Re: [perl #33801] [PATCH] GDBMHash - a dynamic PMC that binds to GNU dbm

8 views
Skip to first unread message

Leopold Toetsch

unread,
Jan 17, 2005, 5:31:53 AM1/17/05
to perl6-i...@perl.org
Bernhard Schmalhofer <parrotbug...@parrotcode.org> wrote:

> Hi,

> I have been working on a dynamic PMC that calls into 'gdbm'. 'gdbm' is a
> file based database, that provides dictionary lookup.

> There are no fancy features yet. INVAL, FLOATVAL, PMC and STRINGS are
> converted to char arrays and are written to the database file.

Shouldn't PMC use freeze/thaw?

> The availability of 'libgdbm.so' is checked during the Parrot configuration,
> which is not nice.

Why?

> CU, Bernhard

leo

Bernhard Schmalhofer

unread,
Jan 17, 2005, 6:29:12 AM1/17/05
to l...@toetsch.at, perl6-i...@perl.org
Leopold Toetsch wrote:

> Bernhard Schmalhofer <parrotbug...@parrotcode.org> wrote:
>
>>I have been working on a dynamic PMC that calls into 'gdbm'. 'gdbm' is a
>>file based database, that provides dictionary lookup.
>
>>There are no fancy features yet. INVAL, FLOATVAL, PMC and STRINGS are
>>converted to char arrays and are written to the database file.
>
> Shouldn't PMC use freeze/thaw?
That depends on what GDBMHash is used for. I was thinking primarily
about data exchange with non-Parrot programs. In this case one wouldn't
expect a frozen PMC as content.

>>The availability of 'libgdbm.so' is checked during the Parrot configuration,
>>which is not nice.
>
> Why?

As an extension writer, or language implementator, my prefered starting
point is a completely built Parrot core. So when adding things I
wouldn't want to rebuild or reconfigure Parrot itself.

CU, Bernhard


--
**************************************************
Dipl.-Physiker Bernhard Schmalhofer
Senior Developer
Biomax Informatics AG
Lochhamer Str. 11
82152 Martinsried, Germany
Tel: +49 89 895574-839
Fax: +49 89 895574-825
eMail: Bernhard.S...@biomax.com
Website: www.biomax.com
**************************************************

Leopold Toetsch

unread,
Jan 17, 2005, 8:46:15 AM1/17/05
to Bernhard Schmalhofer, perl6-i...@perl.org
Bernhard Schmalhofer <Bernhard.S...@biomax.de> wrote:
> Leopold Toetsch wrote:
>> Bernhard Schmalhofer <parrotbug...@parrotcode.org> wrote:
>>
>>>There are no fancy features yet. INVAL, FLOATVAL, PMC and STRINGS are
>>>converted to char arrays and are written to the database file.
>>
>> Shouldn't PMC use freeze/thaw?
> That depends on what GDBMHash is used for. I was thinking primarily
> about data exchange with non-Parrot programs. In this case one wouldn't
> expect a frozen PMC as content.

Ah, ok.

>>>The availability of 'libgdbm.so' is checked during the Parrot configuration,
>>>which is not nice.
>>
>> Why?
> As an extension writer, or language implementator, my prefered starting
> point is a completely built Parrot core. So when adding things I
> wouldn't want to rebuild or reconfigure Parrot itself.

Ok. In the long run I expect a multi-stage config system. First is to
get miniparrot built and running, then Parrot core. Extensions like gdbm
could be configured at that point or even in a third step.

> CU, Bernhard

leo

Leopold Toetsch

unread,
Jan 17, 2005, 9:02:06 AM1/17/05
to perl6-i...@perl.org
Bernhard Schmalhofer <parrotbug...@parrotcode.org> wrote:

> I have been working on a dynamic PMC that calls into 'gdbm'. 'gdbm' is a
> file based database, that provides dictionary lookup.

Thanks, applied.
leo

0 new messages