Steven Fackler
cae12218ab
Encapsulate information about unknown types
2015-02-06 09:03:49 -08:00
Steven Fackler
f96664e63e
Rename ResultDescription to Column and encapsulate
2015-02-06 09:03:49 -08:00
Steven Fackler
7cfde75dce
Encapsulate DbError
2015-02-06 09:03:49 -08:00
Steven Fackler
173075bae1
Remove Error::WrongParamCount
2015-02-06 09:03:49 -08:00
Steven Fackler
cd1d9b9e3f
Remove deprecated methods
2015-02-06 09:03:49 -08:00
Steven Fackler
7101cab178
Remove WrongConnection and WrongTransaction Errors
...
These only occur during contract violations, so panicing is recommended
by the erorr handling RFC.
2015-02-06 09:03:48 -08:00
Steven Fackler
d35124e214
Make next_block_for return an option
2015-02-06 09:03:48 -08:00
Steven Fackler
b4aff18501
Remove unneeded feature
2015-02-04 19:29:59 -08:00
Steven Fackler
cb1ae4d115
Fix for stability changes
2015-01-29 08:59:43 -08:00
Steven Fackler
8d704f71ce
Fix for IO changes
2015-01-28 08:52:33 -08:00
Steven Fackler
af6084871f
Add an is_active method
...
Will be needed when Error::WrongTransaction is removed
2015-01-26 21:09:50 -08:00
Steven Fackler
99a3ba22a6
Fix tests
2015-01-23 11:10:04 -08:00
Steven Fackler
5cce51723a
Move lazy_query to Statement
...
Symmetry is good
2015-01-22 22:52:29 -08:00
Steven Fackler
ee7f9ee806
Avoid accidentally closing finished statements
2015-01-22 22:33:35 -08:00
Steven Fackler
247e320275
Add explicit statement caching
...
Also remove transaction depth checks to preparation methods on
Connection since lifetimes of statements are tied to the connection, not
any transaction that may be active.
cc #84
2015-01-22 22:27:14 -08:00
Steven Fackler
b30bd46c2b
Set element_type for range types
...
Closes #95
2015-01-22 00:04:58 -08:00
Steven Fackler
290d572166
Redesign COPY IN execution setup.
...
The old Iterator of Iterators setup basically required all of the rows
to sit in memory at the same time, which defeats the entire purpose of
that configuration. There's probably a library that defines
`StreamIterator`, so this may jump to that.
2015-01-21 20:25:01 -08:00
Steven Fackler
9e9b0ee54a
Reenable test_unix_connection
2015-01-16 22:07:04 -08:00
Steven Fackler
92a3593d7f
Update for upstream changes
...
String vs Show still needs to be resolved, but it compiles now
2015-01-09 10:50:54 -08:00
Steven Fackler
347c79def9
Fix for upstream changes
2015-01-07 08:23:40 -08:00
Steven Fackler
01ea8afce8
Fix up tests
...
PostgreSQL v9.4 dropped SSLv3 support (yay)
2014-12-22 12:14:20 -05:00
Steven Fackler
d5c0cc0140
Fix for upstream changes
2014-12-19 10:43:42 -08:00
Steven Fackler
810a7442e0
Use out of tree serialize
2014-12-18 17:08:58 -08:00
Steven Fackler
cdfcef8e05
Update for upstream changes
2014-12-15 12:39:49 -08:00
Steven Fackler
42eb9fb01e
Move Send bound for NoticeHandler to trait def
...
This is a breaking change
2014-12-08 23:43:01 -08:00
Steven Fackler
39a89de5f0
Fix off by one in index impl
2014-12-06 10:41:49 -08:00
Steven Fackler
587aa653a3
Fix for equality changes
2014-12-05 07:58:11 -08:00
Steven Fackler
d5998d8f2a
Add Notifications::next_block_for method
...
The setup is a little hairy, but seems correct.
cc #19
2014-12-02 20:36:31 -08:00
Steven Fackler
75641e121f
Allow statements to be prepared on any object
...
Statements aren't cleaned up at the end of a transaction, so this is
safe.
2014-12-01 22:45:22 -08:00
Steven Fackler
6334f045cb
Be more conservative when closing things
2014-11-28 15:01:01 -08:00
Steven Fackler
dd9c712ea9
Fix for upstream changes
2014-11-28 12:24:34 -08:00
Steven Fackler
546967f68c
Add a blocking next method to Notifications
...
cc #19
2014-11-25 18:30:28 -08:00
Steven Fackler
a46c524200
Finish variant naming convention transition
2014-11-20 21:47:13 -08:00
Kevin Butler
d9618c9ba0
tests: [T,.. N] no longer coerces to &[T]
2014-11-19 18:20:20 +00:00
Steven Fackler
9a87c6e18a
Make time support optional
2014-11-17 19:11:32 -08:00
Steven Fackler
e5c2edaf97
More enum namespacing preparation
2014-11-17 13:46:33 -08:00
Steven Fackler
a4fe4e2056
Flatten and rename error stuff
2014-11-03 22:24:11 -08:00
Steven Fackler
ad82d28d18
Demangle PostgresType
2014-11-03 21:31:08 -08:00
Steven Fackler
2da41be987
Remove the old connection pool
...
It's been deprecated for a long time
2014-11-01 16:22:41 -07:00
Steven Fackler
c11cbd6b7d
PostgresConnection -> Connection
2014-11-01 16:21:47 -07:00
Steven Fackler
570512725f
PostgresNotification -> Notification
2014-11-01 16:18:09 -07:00
Steven Fackler
c6c7d8f574
PostgresNoticeHandler -> NoticeHandler
2014-11-01 16:16:50 -07:00
Steve Klabnik
7d1be73156
update macro definition too
2014-10-30 01:13:32 -04:00
Steve Klabnik
675534c26f
fail -> panic
2014-10-30 00:26:03 -04:00
Steven Fackler
e16fca84d7
switch to the username encode set for a test
...
This'll also escape `:` just in case someone has a super weird
filesystem
2014-10-26 12:39:53 -07:00
Steven Fackler
eaf61bec7b
Use servo rust-url for test encoding
2014-10-26 12:05:43 -07:00
Steven Fackler
886f24ea9d
Get rid of literal type specifiers
...
Yay trait reform!
2014-10-25 23:43:59 -07:00
Steven Fackler
7929d07f74
Fix a desynchronization issue
2014-10-25 20:26:45 -07:00
Steven Fackler
f62aa26917
Fix for upstream changes
2014-10-09 20:43:14 -07:00
Steven Fackler
557a159a8a
Add a trait abstracting over transactions and connections
...
This is implemented *in addition* to the intrinsic methods because
having to import a trait to call `prepare` sucks.
Closes #63
2014-10-08 20:29:26 -07:00