Commit Graph

106 Commits

Author SHA1 Message Date
Brian J. Cardiff 2db12d99f5 fix formatting issues 2016-09-06 13:24:57 -03:00
Brian J. Cardiff 47e7d826e8 Add connection retry logic to connection pool 2016-08-31 17:32:01 -03:00
Brian J. Cardiff f568e4506d add cache of pool statements per query
refactor/reuse connection statement cache
2016-08-30 16:20:18 -03:00
Brian J. Cardiff 75aa821f5f make Database return PoolStatement
create StatementMethods for common interface between Statement and PoolStatment.
2016-08-29 16:56:34 -03:00
Brian J. Cardiff 6a0a450622 make statement_specs run over a single connection 2016-08-29 16:50:37 -03:00
Brian J. Cardiff 8d891a5a4e changes needed due to 3c91978d36
add posibility to inspect availability of a resource in a pool (testing only)
allow access to internal connection pool of db (testing only)
2016-08-29 13:14:47 -03:00
Brian J. Cardiff 3c91978d36 Merged branch master into feature/pool 2016-08-29 12:57:42 -03:00
Brian J. Cardiff 325fa9d4ae ensure the connection is released after exec
avoid releasing connection twice on exec.
avoid releasing connection when QueryMethod#query is blockless called.
2016-08-29 01:23:20 -03:00
Brian J. Cardiff bed461c01b allow connection initialization 2016-08-18 00:55:43 -03:00
Ary Borenszweig b5112d9a48 Add `ResultSet#column_names`. Fixes #11 2016-07-17 21:50:52 -03:00
Ary Borenszweig 1cce4685ff Make DB::Pool accept a block 2016-07-11 15:38:31 -03:00
Brian J. Cardiff a2c22c16cf allow DB to use a connection pool.
allow Driver to parse connection pool options for extensibility.
fix waiting_resource counter after a timeout was generated.
2016-07-07 15:50:09 -03:00
Brian J. Cardiff 421996b952 resource pool implementation with
* max_pool_size
* initial_pool_size
* max_idle_pool_size
* checkout_timeout
configuration options
2016-07-05 15:21:39 -03:00
Santiago Palladino 9ca0b19d9e Support mappable classes in query_one and query_all methods 2016-07-04 12:46:45 -03:00
Santiago Palladino 552b6e12b4 Rebase to latest DB version and upgrade to Crystal 0.18 2016-07-04 12:13:39 -03:00
Santiago Palladino 7fcedc6711 Database mapping macro
Add `from_rs` method to class to load instances from a resultset. Inspired by YAML and JSON mapping macros.
2016-07-04 11:19:16 -03:00
Santiago Palladino 6b065bd6b6 Handle different number of columns in dummy driver
Number of cols is inferred from the number of fields in the first row.
2016-07-04 11:15:04 -03:00
Santiago Palladino 5773faaa5c Enumerate columns in result_set 2016-07-04 11:11:21 -03:00
Ary Borenszweig 9c88f718e8 Require ResultSet to just implement `read`, optionally implementing `read(T.class)`. Fixes #5 2016-06-30 09:46:25 -03:00
Brian J. Cardiff 6838784f7b fix #1. avoid marking as closed if do_close raise exception 2016-06-24 10:54:30 -03:00
Brian J. Cardiff 933014c1aa fix bug related to Slice(T).new(0). It was messing with the data in the stack 2016-06-23 22:36:34 -03:00
Brian J. Cardiff b2da4f3f6e allow scalar to return custom types 2016-06-23 22:19:53 -03:00
Brian J. Cardiff 34ae9d5775 avoid messing with finalize cycle in Disposable.
ensure statements are closed when connection is closed
2016-06-23 22:07:59 -03:00
Brian J. Cardiff 8025ecaa13 update Slice(UInt8) to Bytes 2016-06-22 14:10:09 -03:00
Brian J. Cardiff 11e24e1c65 #5 Make sure queries (insert and query) work with any type 2016-06-22 00:44:16 -03:00
Brian J. Cardiff c5d7778a0d update DB::ExecResult#rows_affected to Int64 2016-06-21 18:31:13 -03:00
Brian J. Cardiff 9f3f0a9836 #5 remove read_object. allow read / read? to receive any type 2016-06-21 12:10:09 -03:00
Brian J. Cardiff 76e3a35d59 allow exec to receive any object arguments 2016-06-21 10:55:44 -03:00
Brian J. Cardiff d4b2046a65 refactor and add specs for query without blocks 2016-06-21 10:55:44 -03:00
Brian J. Cardiff 3464a1d1cc allow query to receive any object arguments 2016-06-21 10:55:41 -03:00
Brian J. Cardiff 5b23114095 add ResultSet#read_object 2016-06-21 10:54:58 -03:00
Brian J. Cardiff 312609b722 allow specs to use multiple drivers 2016-06-21 10:54:56 -03:00
Brian J. Cardiff d69555b372 make Nil a proper column type 2016-06-21 10:53:34 -03:00
Ary Borenszweig f834b0f8a6 Updated to Crystal 0.17.4 2016-06-16 12:16:39 -03:00
Brian J. Cardiff 44559f6f95 move spec and src files to a shardy way 2016-02-25 22:37:10 -03:00
Brian J. Cardiff 4ed7f28fe6 ensure exec and scalar release connection
ensure exec, query can be executed with Enumerable(Any)
update db doc sample
2016-02-25 22:27:42 -03:00
Brian J. Cardiff d45427dfdd expose database in connection
get/return from pool while result set is been used. still single connection pool
2016-02-25 22:27:42 -03:00
Brian J. Cardiff fa111fd698 add statement cache in connection.
hide methods from api
2016-02-25 22:27:42 -03:00
Brian J. Cardiff 562d5076bf use URI as connection string 2016-02-25 22:27:42 -03:00
Brian J. Cardiff b92d08bf74 refactor db disposables 2016-02-25 22:27:42 -03:00
Brian J. Cardiff 6a5b2a8758 assert connection is closed after db is used.
tidy up close methods
2016-02-25 22:27:35 -03:00
Brian J. Cardiff 3598dddb65 split perform_query and perform_exec
add DB::ExecResult to have a concurrency safety exec result
remove connection string from abstract DB::Connection
2016-02-25 22:25:55 -03:00
Brian J. Cardiff 67fe5c9aae leave a single scalar value returns DB::Any 2016-02-25 22:25:55 -03:00
Brian J. Cardiff d01da912f7 code review notes
close / do_close in result_set
avoid closing statements
remove named arguments
refactor positioned arguments query
2016-02-25 22:25:55 -03:00
Brian J. Cardiff fd804dd592 switch to 0-pased positional arguments
add docs, many docs
2016-02-25 22:25:55 -03:00
Brian J. Cardiff a96776e336 use connection_string as database configuration instead of a hash 2016-02-25 22:25:55 -03:00
Brian J. Cardiff 8c0313a306 add Connection#last_inserted_id
extract QueryMethods modules and use it in Database and Connection
"add" finalize methods, but they cause GC warnings
2016-02-25 22:25:55 -03:00
Brian J. Cardiff 1fb0b8a82b add `ResultSet#exec` to ensure the query is performed. By default generates a `move_next`. 2016-02-25 22:25:54 -03:00
Brian J. Cardiff 683e6bdfa7 major db refactor for a better api.
`#query`, `#exec`, `#scalar`, `#scalar?` as main query methods from `Database`
blocks overrides that ensure statements are closed.
2016-02-25 22:25:54 -03:00
Brian J. Cardiff caf2676aad column_count and column_names
close result_set and statements
main entry points for exec_non_query and exec_query closing them when ready
2016-02-25 22:25:54 -03:00
Brian J. Cardiff e50cc4ec73 add specs for statement arguments
add support for blobs (i.e. Slice(UInt8) )
2016-02-25 22:25:54 -03:00
Brian J. Cardiff 8a8b86e31a arguments support 2016-02-25 22:25:54 -03:00
Brian J. Cardiff 1572062501 introduce database as driver wrapper.
expose list of types to support by the drivers implementors.
deal with nilable types with `#read?(T.class) : T?` methods.
change `#has_next` to `#move_next`
2016-02-25 22:25:54 -03:00
Brian J. Cardiff c16dc42e96 allow `result_set#each` 2016-02-25 22:25:54 -03:00
Brian J. Cardiff 4dd0312934 prepare/exec a statement
dummy driver that parses the query
initial type support for result_sets
2016-02-25 22:25:54 -03:00
Brian J. Cardiff cc1545a58e db driver registration 2016-02-25 22:25:54 -03:00