This commit makes psycopg2 responsible for sending the status update
(feedback) messages to the server regardless of whether a synchronous or
asynchronous connection is used.
Feedback is sent every *status_update* (default value is 10) seconds,
which could be configured by passing a corresponding parameter to the
`start_replication()` or `start_replication_expert()` methods.
The actual feedback message is sent by the
`pq_read_replication_message()` when the *status_update* timeout is
reached.
The default behavior of the `send_feedback()` method is changed.
It doesn't send a feedback message on every call anymore but just
updates internal structures. There is still a way to *force* sending
a message if *force* or *reply* parameters are set.
The new approach has certain advantages:
1. The client can simply call the `send_feedback()` for every
processed message and the library will take care of not overwhelming
the server. Actually, in the synchronous mode it is even mandatory
to confirm every processed message.
2. The library tracks internally the pointer of the last received
message which is not keepalive. If the client confirmed the last
message and after that server sends only keepalives with increasing
*wal_end*, the library can safely move forward *flush* position to
the *wal_end* and later automatically report it to the server.
Reporting of the *wal_end* received from keepalive messages is very
important. Not doing so casing:
1. Excessive disk usage, because the replication slot prevents from
WAL being cleaned up.
2. The smart and fast shutdown of the server could last indefinitely
because walsender waits until the client report *flush* position
equal to the *wal_end*.
This implementation is only extending the existing API and therefore
should not break any of the existing code.
This class was deprecated in
27cd6c4880 (Dec 2, 2012), which was first
included in release 2.5. Enough time has passed for library uses to find
an alternative solution.
This class was untested.
I'm still fought whether docs should be in the C module or in the .rst.
I'd prefer the first because DRY, but writing multiline strings in C
really sucks.
For library end users, there is no need to install tests alongside the
package itself. This keeps the tests available for development without
adding extra packages to user's site-packages directory. Reduces the
size of the installed package. Avoids accidental execution of test code
by an installed package.
Deprecated in commit b263fbf274 on
2010-01-13. The deprecation warning was first released in version 2.2.2.
The function used to register an alternate type caster for TIMESTAMP
WITH TIME ZONE to deal with historical time zones with seconds in the
UTC offset. These are now correctly handled by the default type caster,
so currently the function doesn't do anything.
The docs don't need to describe what will happen on Python versions
before 2.6 as they are unsupported by psycopg2.
Should have been included in commit
d58844e548, but was missed.
namedtuple is available on all Python versions supported by psycopg2. It
was first introduced in Python 2.6. Can remove all workarounds and
special documentation.
Recent Sphinx versions seem overly aggressive in autodetecting python,
or I just didn't notice the errors, so be explicit in what language to
use with code examples.
Every point has an example and all the example show wrong/correct. Nice
rhythm.
Among the improvements, added point saying explicitly "thou shall not
quote placeholders". Quoted placeholders will just fail except in the
most contrived cases (a statement raising an exception with all the
strings except with the attack ones...), and an example in the following
section explicitly notes "no quotes", but apparenty someone still thinks
this is not documented enough? (see issue #611) so let's just write it
plain and clear into the list of commandments.
The service is deprecated in favour of RTD. There is also no more way to
upload redirects there, so #581 cannot be fixed.
Official docs are on http://initd.org/psycopg/docs/
