09.01.2025 23:44, Wietse Venema via Postfix-devel wrote:
dict_cdb.c: In function 'dict_cdbq_sequence':
dict_cdb.c:202:20: error: 'DICT_CDBQ' has no member named 'seq_cpos'; did you
mean 'seq_cptr'?
202 | dict_cdbq->seq_cpos = 0;
|^~~~
|
dict_cdb.c: In function 'dict_cdbq_sequence':
dict_cdb.c:202:20: error: 'DICT_CDBQ' has no member named 'seq_cpos'; did you
mean 'seq_cptr'?
202 | dict_cdbq->seq_cpos = 0;
|^~~~
|seq_cptr
Assuming that seq_cptr was intended, I adde
Michael Tokarev via Postfix-devel:
> Move common code to get a piece of data from cdb into
> its own function.
>
> Signed-off-by: Michael Tokarev
> ---
> On top of the previous patch.
I'll draft some text for CDB_README.html.
Wietse
___
Postfi
Michael Tokarev via Postfix-devel:
> Hi!
>
> I noticed that dict_cdb uses static buffer to read the value,
> while all other maps use per-instance val_buf vstring or similar,
> and even the fold_buf is per-instance. I wonder if something like
> the below is actually necessary.
It does make sense
Move common code to get a piece of data from cdb into
its own function.
Signed-off-by: Michael Tokarev
---
On top of the previous patch.
src/util/dict_cdb.c | 89 +
1 file changed, 74 insertions(+), 15 deletions(-)
diff --git a/src/util/dict_cdb.c b/
use per-instance val_buf vstring buffer instead of
single static buf for all cdb instances, which might
be unsafe if more than one result is being in use at
the same time.
Signed-off-by: Michael Tokarev
---
previous patch has been garbled by cut-n-paste.
Also move val_buf init a bit earlier.
sr
Hi!
I noticed that dict_cdb uses static buffer to read the value,
while all other maps use per-instance val_buf vstring or similar,
and even the fold_buf is per-instance. I wonder if something like
the below is actually necessary.
diff --git a/src/util/dict_cdb.c b/src/util/dict_cdb.c
index a91
