A low-level Clojure wrapper for JDBC-based access to databases. This project is "Inactive". It has effectively been superseded by seancorfield/next.jdbc.
For higher level DSLs and migration libraries that are compatible, see the documentation.
Formerly known as clojure.contrib.sql
.
This library is mature and stable. It is widely used and its use is described in many books and tutorials. It will continue to get bug fixes and minor releases. Based on my experience using and maintaining this library, I've created a faster, more modern JDBC wrapper called next.jdbc. I consider it to be the "next generation" of clojure.java.jdbc
but it exposes a different API -- a better API, I think.
- API Reference (Autogenerated)
- Overview
- Manipulating Data with SQL
- How to Reuse Database Connections
- Using DDL and Metadata
- Mailing List
- #sql on Clojurians Slack
Latest stable release: 0.7.12 -- requires Clojure 1.7 or later!
CLI/deps.edn
dependency information:
org.clojure/java.jdbc {:mvn/version "0.7.12"}
Leiningen dependency information:
[org.clojure/java.jdbc "0.7.12"]
Maven dependency information:
<dependency>
<groupId>org.clojure</groupId>
<artifactId>java.jdbc</artifactId>
<version>0.7.12</version>
</dependency>
Note: Earlier versions of Clojure are supported by older versions of clojure.java.jdbc
: e.g., version 0.6.1 supports Clojure 1.4 and later.
You will also need to add dependencies for the JDBC driver you intend to use. Here are links (to Maven Central) for each of the common database drivers that clojure.java.jdbc is known to be used with:
- Apache Derby
- H2
- HSQLDB
- Microsoft SQL Server jTDS
- Microsoft SQL Server -- Official MS Version
- MySQL
- PostgreSQL
- SQLite
Note: different versions of various database drivers have different Java/JVM version requirements. In particular, recent versions of Apache Derby require at least Java 8 and recent versions of H2 require at least Java 7. Clojure's Continuous Integration system uses older versions so tests can be run on Java 6 (see pom.xml
); local testing is done with more recent versions on Java 8.
(require '[clojure.java.jdbc :as j])
;; there are many ways to write a db-spec but the easiest way is to
;; use :dbtype and then provide the :dbname and any of :user, :password,
;; :host, :port, and other options as needed:
(def mysql-db {:dbtype "mysql"
:dbname "clojure_test"
:user "clojure_test"
:password "clojure_test"})
(def pg-db {:dbtype "postgresql"
:dbname "mypgdatabase"
:host "mydb.server.com"
:user "myuser"
:password "secret"
:ssl true
:sslfactory "org.postgresql.ssl.NonValidatingFactory"})
;; if the dbtype is not known to clojure.java.jdbc, or you want to override the
;; default choice of JDBC driver class name, you can provide :classname and the
;; name of the class to use:
(def redshift42 {:dbtype "redshift"
:dbname "myredstore"
:classname "com.amazon.redshift.jdbc42.Driver"
...})
;; you can also specify a full connection string if you'd prefer:
(def pg-uri
{:connection-uri (str "postgresql://myuser:secret@mydb.server.com:5432/mypgdatabase"
"?ssl=true&sslfactory=org.postgresql.ssl.NonValidatingFactory")})
(j/insert-multi! mysql-db :fruit
[{:name "Apple" :appearance "rosy" :cost 24}
{:name "Orange" :appearance "round" :cost 49}])
;; ({:generated_key 1} {:generated_key 2})
(j/query mysql-db
["select * from fruit where appearance = ?" "rosy"]
{:row-fn :cost})
;; (24)
For more detail see the API reference or documentation.
-
Testing:
- Currently by default tests run only against Derby and HSQLDB, the in-process databases.
-
To test against PostgreSQL, first create the user and database:
$ sudo -u postgres createuser clojure_test -P clojure_test $ sudo -u postgres createdb clojure_test -O clojure_test
-
Or similarly with MySQL:
$ mysql -u root mysql> create database clojure_test; mysql> grant all on clojure_test.* to clojure_test identified by "clojure_test";
-
Then run the tests with the TEST_DBS environment variable:
$ TEST_DBS="mysql postgres" mvn test
-
Also see the
run-tests.sh
shell script which uses theclj
CLI anddeps.edn
for multi-version testing!
-
Release 0.7.12 on 2021-02-01
- Make the protocols
ISQLValue
,ISQLParameter
, andIResultSetReadColumn
extensible via metadata.
- Make the protocols
-
Release 0.7.11 on 2019-12-24
- Address edge case in transaction rollback failure JDBC-179.
-
Release 0.7.10 on 2019-08-24
- Use a US-locale
lower-case
function to avoid problems in certain locales (e.g., Turkish). A similar issue has been fixed recently in both HoneySQL andnext.jdbc
. - Clean up
db-spec
options that are passed to the JDBC connection manager as properties JDBC-178. - Relax restriction on
create-table-ddl
column specs to allow numbers (as well as keywords and strings) JDBC-177.
- Use a US-locale
-
Release 0.7.9 on 2019-02-21
- Fix behavior of multi-inserts when database does not support generated keys JDBC-176.
- Added highly experimental support for
datafy
/nav
(inclojure.java.jdbc.datafy
namespace). This includes a convention-based approach to foreign keys with some assistance from a:schema
option. This is subject to change and is provided mostly for informational purposes, as an example of the new functionality in Clojure 1.10. This includes a fix for the conventions from JDBC-175. - Add note about rewriting batched operations to
insert-multi!
for some drivers JDBC-174. - Support Oracle SID style URLs (
dbtype
can beoracle:sid
which maps tooracle:thin
and uses:
as the separator before thedbname
value) JDBC-173.
-
Release 0.7.8 on 2018-08-13
-
Release 0.7.7 on 2018-06-23
- Support
:as-arrays?
,:result-set-fn
, and:row-fn
in operations that return generated keys as a result set (execute!
,insert!
, andinsert-multi!
) JDBC-169. get-connection
provides much better feedback if you accidentally call a function that expects adb-spec
but pass ajava.sql.Connection
object instead (which is only required forprepare-statement
).
- Support
-
Release 0.7.6 on 2018-04-24
execute!
now supports:return-keys
as a vector of column names, rather than just a simple Boolean value, for drivers that support that JDBC-166.- Add built-in support for H2 in-memory database (
:dbtype "h2:mem"
). - Add missing spec for
db-spec
being ajava.net.URI
object. - Fix
add-connection
handling of stringdb-spec
(becomes:connection-uri
, not:connection-string
). - Fix specs for
with-db-*
functions, to support options in the binding form JDBC-165. - Update tests so they work properly with string
db-spec
test databases. - Ensure no reflection warnings are present.
- Switched local test infrastructure over to CLI and
deps.edn
(from Leiningen) as an example of multi-version testing without a "build tool".
-
Release 0.7.5 on 2017-12-29
- Add support for
:return-keys
inexecute!
and:multi?
indb-do-prepared-return-keys
JDBC-163.
- Add support for
-
Release 0.7.4 on 2017-12-14
- Improved discoverability of other
java.jdbc
documentation JDBC-160. - Optional specs updated with
:keywordize?
and:connection-uri
changes from 0.7.2 and 0.7.3 releases. - Performance improvements, primarily in
query
andreducible-query
. - Experimental
:raw?
result set handling inreducible-query
. modify-connection
is more robust in the face ofnull
connections and bad option values.
- Improved discoverability of other
-
Release 0.7.3 on 2017-10-05
- Added
:keywordize?
option alongside:identifiers
that defaults totrue
but can be set tofalse
to opt-out of converting identifiers to keywords (so column names etc will only be processed by the function passed as:identifiers
) JDBC-159. - If an exception occurs during a transaction, and then rollback fails with another exception, both exceptions will now be combined into an
ex-info
. Previously the rollback exception obscured the transaction exception JDBC-158.
- Added
-
Release 0.7.2 on 2017-10-02
connection-uri
was incorrectly spec'd as ajava.net.URI
but should bestring?
JDBC-156.- Allow for
:user
and:password
to be passed with:connection-uri
, so credentials can be omitted from the connection string. - Clarified docstring for
get-connection
to show where:user
and:password
can be passed.
-
Release 0.7.1 on 2017-08-30
- Connection strings with empty values were not parsed correctly JDBC-155.
-
Release 0.7.0 on 2017-07-16
:conditional?
option forcreate-table-ddl
anddrop-table-ddl
to provide for existence check (or a function to manipulate the generated DDL).- Add better support for Oracle connections (default port to
1521
, support:dbtype "oracle"
-- as"oracle:thin"
-- and:dbtype "oracle:oci"
, with@
instead of//
before host).
-
Release 0.7.0-beta5 on 2017-07-05
get-connection
now accepts anopts
map with:auto-commit?
and:read-only?
options. If present, the appropriate methods will be called on the connection obtained. These options are valid in any function call that may callget-connection
under the hood. This should allow for streaming results in a query for most databases JDBC-153.- Additional validation of options is performed in
prepared-statement
to avoid silently ignoring invalid combinations of:concurrency
,:cursors
,:result-type
, and:return-keys
.
-
Release 0.7.0-beta4 on 2017-07-04
opts
are now correctly passed fromreducible-query
todb-query-with-resultset
.- Updated the
::query-options
spec to make it clear that::prepare-options
are also acceptable there.
-
Release 0.7.0-beta3 on 2017-07-04
- Reflection warnings removed in
reducible-result-set
JDBC-152.
- Reflection warnings removed in
-
Release 0.7.0-beta2 on 2017-06-30 (a.k.a The Reducible Saga, Part 2)
- Support for Clojure 1.5 and 1.6 has been dropped -- breaking change.
- Or, put another way,
clojure.java.jdbc
now requires Clojure 1.7 or later! - All public functions now have specs in the optional
clojure.java.jdbc.spec
namespace (requiresclojure.spec.alpha
). reducible-query
andreducible-result-set
useIReduce
and correctly support the no-init
arity ofreduce
by using the first row of theResultSet
, if present, as the (missing)init
value, and only callingf
with no arguments if theResultSet
is empty. Theinit
arity ofreduce
only ever callsf
with two arguments.
-
Release 0.7.0-beta1 on 2017-06-29
- Support for Clojure 1.4.0 has been dropped -- breaking change.
- Optional spec support now uses
clojure.spec.alpha
. reducible-query
accepts adb-spec
and a SQL/parameters vector and returns a reducible (IReduce
on Clojure 1.7 or later;CollReduce
on Clojure 1.5/1.6): when reduced, it runs the query, obtains a reducible result set, and then reduces that. A reducible query will run the query each time it is reduced. The helper functionreducible-result-set
is public: it accepts aResultSet
and produces a reducible that offers a single pass reduce over the rows. Both functions honorreduced
values to short-circuit the process JDBC-99.
-
Release 0.7.0-alpha3 on 2017-03-23
classname
is now accepted withdbtype
/dbname
so you can easily specify a JDBC driver class name for a database type that is not known JDBC-151.redshift
has been added as adbtype
withcom.amazon.redshift.jdbc.Driver
as the driver name.
-
Release 0.7.0-alpha2 on 2017-03-01
pgsql
and the Impossibl PostgresSQL 'NG' driver are now supported (note that:max-rows
does not work with this driver!); also, providing unknowndbtype
orsubprotocol
in adb-spec
should now throw a better exception JDBC-150.quoted
now accepts keywords for database / dialect (:ansi
(including PostgresSQL),:mysql
,:oracle
,:sqlserver
-- these match the keywords used in HoneySQL which is the recommended third party SQL DSL for java.jdbc) JDBC-149.- Reorder
get-connection
clauses to make it easier to combine keys in adb-spec
JDBC-148. - Force load
DriverManager
beforeclassForName
call on drivers to avoid potential race condition on initialization JDBC-145.
-
Release 0.7.0-alpha1 on 2016-11-12 -- potentially breaking changes
- The signatures of
as-sql-name
andquoted
have changed slightly: the former no longer has the curried (single argument) version, and the latter no longer has the two argument version. This change came out of a discussion on Slack which indicated curried functions are non-idiomatic. If you relied on the curried version ofas-sql-name
, you will not need to usepartial
. If you relied on the two argument version ofquoted
, you will need to add an extra( )
for the one argument call. I'd be fairly surprised if anyone is usingas-sql-name
at all since it is really an implementation detail. I'd also be surprised if anyone was using the two argument version ofquoted
since the natural usage is:entities (quoted [\[ \]])
to create a naming strategy (that provides SQL entity quoting). - Clarified that
insert-multi!
with a sequence of row maps may be substantially slower than with a sequence of row value vectors (the former performs an insert for each row, the latter performs a single insert for all the data together) JDBC-147. - All options are passed through all function calls, expanding the range of options you can pass into high-level functions such as
insert!
andupdate!
JDBC-144. - Added
get-isolation-level
to return the current transaction's isolation level, if any JDBC-141. - Added support for
read-columns
option to allow more flexible customization of reading column values from a result set (particularly in a multi-database application). Also expandsset-parameters
support to options (previously it was just part of the db-spec) JDBC-137. - Expanded optional
clojure.spec
coverage to almost the whole library API.
- The signatures of
-
Release 0.6.2-alpha3 on 2016-08-25
- Fixed bad interaction between
:qualifier
and existing:identifiers
functionality JDBC-140. - Updated the README and docstrings to reflect that
:dbtype
is the easiest / preferred way to writedb-spec
maps JDBC-139. - Fixed postgres / postgresql alias support JDBC-138. This also adds aliases for mssql (sqlserver), jtds (jtds:sqlserver), oracle (oracle:thin), and hsql (hsqldb).
- Fixed bad interaction between
-
Release 0.6.2-alpha2 on 2016-07-21
- Update
clojure.spec
support to work with Clojure 1.9.0 Alpha 10.
- Update
-
Release 0.6.2-alpha1 on 2016-07-05
- Experimental support for
clojure.spec
via the newclojure.java.jdbc.spec
namespace. Requires Clojure 1.9.0 Alpha 8 (or later). - All options to all functions can now have defaults within the
db-spec
itself JDBC-136. query
(and by extensionfind-by-keys
andget-by-id
) now support:explain?
and:explain-fn
options to help support basic performance analysis JDBC-135.insert!
andinsert-multi!
now respect:identifiers
and:qualifier
because inserting rows on PostgreSQL returns full rows, not just the newly inserted keys JDBC-134.- In addition to the
:identifiers
option, you can now use:qualifier
to specify a namespace qualifier (string) to be used when constructing keywords from SQL column names JDBC-133.
- Experimental support for
-
Release 0.6.1 on 2016-05-12 -- IMPORTANT BUG FIX!
insert!
andinsert-multi!
now default:transaction?
totrue
(as they should have done in 0.6.0!) JDBC-128. These two functions also have improved docstrings to clarify the difference in behavior between inserting rows as maps compared to inserting rows as a series of column values.- PostgreSQL support has been improved: java.jdbc is now tested against PostgreSQL locally (as well as SQLite, Apache Derby, HSQLDB, H2, MySQL, MS SQL Server (both MS Type 4 driver and jTDS driver). JDBC-127 and JDBC-129.
-
Release 0.6.0 on 2016-05-11 -- BREAKING RELEASE! DEPRECATED FUNCTIONALITY REMOVED!
find-by-keys
now correctly handlesnil
values JDBC-126. 0.6.0 / 2016-05-11.find-by-keys
callsseq
on:order-by
to treat[]
as noORDER BY
clause. 0.6.0 / 2016-05-11.db-query-with-resultset
now accepts an options map and passes it toprepare-statement
JDBC-125. 0.6.0-rc2 / 2016-05-07.- Passing the
prepare-statement
options map as the first element of the[sql & params]
vector is no longer supported and will throw anIllegalArgumentException
. It was always very poorly documented and almost never used, as far as I can tell.
- Passing the
db-query-with-resultset
no longer requires thesql-params
argument to be a vector: a sequence is acceptable. This is in line with other functions that accept a sequence. 0.6.0-rc2 / 2016-05-07.db-query-with-resultset
now accepts a bare SQL string orPreparedStatement
as thesql-params
argument, when there are no parameters needed. This is in line with other functions that accept SQL or aPreparedStatement
. 0.6.0-rc2 / 2016-05-07.query
's options map now is passed todb-query-with-resultset
and thus can contain options to be used to construct thePreparedStatement
JDBC-125. 0.6.0-rc2 / 2016-05-07.- Adds
get-by-id
andfind-by-keys
convenience functions (these were easy to add after the API changes in 0.6.0 and we rely very heavily on them at World Singles so putting them in the core for everyone seemed reasonable). 0.6.0-rc1 / 2016-05-04.find-by-keys
accepts an:order-by
option that expects a sequence of orderings; an ordering is a column name (keyword) or a map from column name (keyword) to direction (:asc
or:desc
). 0.6.0-rc2 / 2016-05-07.
- Ensures SQL / params are actually vectors prior to destructuring (this addresses an interop edge case from other languages) JDBC-124. 0.6.0-alpha2 / 2016-04-18.
- Fix typo in
insert-multi!
argument validation exception JDBC-123. 0.6.0-alpha2 / 2016-04-18. - ALL DEPRECATED FUNCTIONALITY HAS BEEN REMOVED! JDBC-118. 0.6.0-alpha1 / 2016-04-13
- See changes described in versions 0.5.5 through 0.5.8 for what was deprecated
- Use version 0.5.8 as a bridge to identify any deprecated API calls on which your code relies!
db-transaction
(deprecated in version 0.3.0) has been removed- The
java.jdbc.deprecated
namespace has been removed
-
Release 0.5.8 on 2016-04-12
db-do-commands
now expects multiple commands to be be wrapped in a vector JDBC-122. The single command form is unchanged (but may be wrapped in a vector). Callingdb-do-commands
with multiple commands (not wrapped in a single vector) will produce a "DEPRECATED" warning printed to the console.db-do-prepared
anddb-do-prepared-return-keys
now expect to receive adb-spec
, an optionaltransaction?
boolean, asql-params
argument, and an optional options map.sql-params
is a vector containing a SQL string orPreparedStatement
followed by parameters -- like other APIs in this library. In addition, like the:multi? true
version ofexecute!
,db-do-prepared
can accept a vector that has parameter groups: multiple vectors containing groups of parameter values JDBC-122. Callingdb-do-prepared
with unrolled arguments -- the SQL string / statement followed by parameter groups -- is deprecated and will produce "DEPRECATED" warnings printed to the console.
-
Release 0.5.7 on 2016-04-10
(insert! db table [:col] ["val"] {})
syntax, introduced in 0.5.6, threw an exception JDBC-121.
-
Release 0.5.6 on 2016-04-10
create-table-ddl
now expects the column specs to be wrapped in a single vector and no longer needs the:options
delimiter to specify the options map JDBC-120.- If column specs are not wrapped in a vector, you will get a "DEPRECATED" warning printed to the console.
insert!
now supports only single row insertion; multi-row insertion is deprecated.insert-multi!
has been added for multi-row insertion.:options
is no longer needed as a delimiter for the options map JDBC-119.- If
insert!
is called with multiple rows, or:options
is specified, you will get a "DEPRECATED" warning printed to the console.
- If
- NOTE: all deprecated functionality will go away in version 0.6.0!
-
Release 0.5.5 on 2016-04-09
- Allow options map in all calls that previously took optional keyword arguments JDBC-117.
- The unrolled keyword argument forms of call are deprecated -- and print a "DEPRECATED" message to the console! -- and will go away in 0.6.0.
- Allow options map in all calls that previously took optional keyword arguments JDBC-117.
-
Release 0.5.0 on 2016-03-27
-
Release 0.4.2 on 2015-09-15
- Remove redundant type hints JDBC-113 - Michael Blume.
- Avoid reflection on
.prepareStatement
JDBC-112 - Michael Blume. - Add
metadata-query
macro to make metadata query / results easier to work with for JDBC-107. prepare-statement
:return-keys
may now be a vector of (auto-generated) column names to return, in addition to just being truthy or falsey. This allows keys to be returned for more databases. JDBC-104.- Officially support H2 (and test against it) to support JDBC-91 and clarify docstrings to improve debugging driver-specific restrictions on SQL.
-
Release 0.4.0 / 0.4.1 on 2015-07-26
db-do-prepared
now allowstransaction?
to be omitted when aPreparedStatement
is passed as the second argument JDBC-111 - Stefan Kamphausen.- Nested transaction checks isolation level is the same JDBC-110 - Donald Ball.
- Default PostgreSQL port; Support more dbtype/dbname variants JDBC-109.
- Drop Clojure 1.2 compatibility.
-
Release 0.3.7 on 2015-05-18
-
Release 0.3.6 on 2014-10-28
-
Release 0.3.5 on 2014-08-01
-
Release 0.3.4 on 2014-06-30
-
Release 0.3.3 on 2014-01-30
-
Release 0.3.2 on 2013-12-30
- Add nil protocol implementation to ISQLParameter
-
Release 0.3.1 on 2013-12-29 (broken; use 0.3.2 instead)
- Improve docstrings and add :arglists for better auto-generated documentation.
- Make insert-sql private - technically a breaking change but it should never have been public: sorry folks!
- Provide better protocol for setting parameters in prepared statements JDBC-86.
- Fix parens in two deprecated tests JDBC-85.
- Made create-table-ddl less aggressive about applying as-sql-name so only first name in a column spec is affected.
-
Release 0.3.0 on 2013-12-16
- Ensure canonical Boolean to workaround strange behavior in some JDBC drivers JDBC-84.
- Rename recently introduced test to ensure unique names JDBC-83.
- Rename unused arguments in protocol implementation to support Android JDBC-82.
- Correctly handle empty param group sequence in execute! (which only seemed to affect SQLite) JDBC-65.
-
Release 0.3.0-rc1 on 2013-12-12
- Deprecate db-transaction (new in 0.3.0) in favor of with-db-transaction JDBC-81.
- Add with-db-metadata macro and metadata-result function to make it easier to work with SQL metadata JDBC-80.
- Add with-db-connection macro to make it easier to run groups of operations against a single open connection JDBC-79.
- Add ISQLValue protocol to make it easier to support custom SQL types for parameters in SQL statements JDBC-77.
- Add support for :isolation in with-db-transaction JDBC-75.
- Add :user as an alias for :username for DataSource connections JDBC-74.
-
Release 0.3.0-beta2 on 2013-11-24
- BREAKING CHANGES!
- The DSL namespaces introduced in 0.3.0-alpha1 have been retired - see java-jdbc/dsl for a migration path if you wish to continue using the DSL (although it is recommended you switch to another, more expressive DSL).
- The older API (0.2.3) which was deprecated in earlier 0.3.0 builds has moved to
clojure.java.jdbc.deprecated
to help streamline the API for 0.3.0 and clean up the documentation.
-
Release 0.3.0-beta1 on 2013-11-03
- query as-arrays? now allows you to leverage lazy result fetching JDBC-72.
- "h2" is recognized as a protocol shorthand for org.h2.Driver
- Tests no longer use :1 literal JDBC-71.
- Conditional use of javax.naming.InitialContext so it can be compiled on Android JDBC-69.
- New db-query-with-resultset function replaces private
db-with-query-results*
and processes a raw ResultSet object JDBC-63. - Allow :set-parameters in db-spec to override set-parameters internal function to allow per-DB special handling of SQL parameters values (such as null for Teradata) JDBC-40.
-
Release 0.3.0-alpha5 on 2013-09-15
- DDL now supports entities naming strategy JDBC-53.
- Attempt to address potential memory leaks due to closures - see Christophe Grand's blog post on Macros, closures and unexpected object retention.
- Documentation has moved to Using java.jdbc on Clojure-Doc.org
- Added Leiningen support for easier development/testing (Maven is still the primary build tool).
- Added create-index / drop-index DDL JDBC-62 - moquist
- Make transaction? boolean optional in various
db-do-*
functions - Create clojure.java.jdbc.ddl namespace
- Add create-table, drop-table, create-index and drop-index
- Deprecate create-table, create-table-ddl and drop-table in main namespace
- Update README to clarify PostgreSQL instructions.
- Fix test suite for PostgreSQL JDBC-59
- Improve hooks for Oracle data type handling JDBC-57
- Fix reflection warnings JDBC-55
-
Release 0.3.0-alpha4 on 2013-05-11
- Fix connection leaks JDBC-54
- Allow order-by to accept empty sequence (and return empty string)
-
Release 0.3.0-alpha3 on 2013-05-04
- Fix macro / import interaction by fully qualifying Connection type.
-
Release 0.3.0-alpha2 on 2013-05-03
- Address JDBC-51 by declaring get-connection returns java.sql.Connection
- Add IResultSetReadColumn protocol extension point for custom read conversions JDBC-46
- Add :multi? to execute! so it can be used for repeated operations JDBC-52
- Reverted specialized handling of NULL values (reopens JDBC-40)
- Rename :as-arrays to :as-arrays? since it is boolean
- Add curried version of clojure.java.jdbc.sql/as-quoted-str
- Officially deprecate resultset-seq
-
Release 0.3.0-alpha1 on 2013-04-07
- MAJOR API OVERHAUL!
- Most of the old 0.2.x API has been deprecated and a new, more idiomatic API introduced, along with a minimal DSL to generate basic SQL
- Specifics:
- Add insert!, query, update!, delete! and execute! high-level API JDBC-20
- Add optional SQL-generating DSL in clojure.java.jdbc.sql (implied by JDBC-20)
- Add db- prefixed versions of low-level API
- Add db-transaction macro
- Add result-set-seq as replacement for resultset-seq (which will be deprecated)
- Transaction now correctly rollback on non-Exception Throwables JDBC-43
- Rewrite old API functions in terms of new API, and deprecate old API JDBC-43
- Add :as-arrays to query / result-set-seq JDBC-41
- Better handling of NULL values JDBC-40 and JDBC-18 Note: JDBC-40 is being reverted in 0.3.0-alpha2 because it introduces regressions in PostgreSQL
- db-do-commands allows you to execute SQL without a transaction wrapping it JDBC-38
- Remove reflection warning from execute-batch
- Add notes to README about 3rd party database driver dependencies
- Add optional :identifiers argument to resultset-seq so you can explicitly pass in the naming strategy
-
Release 0.2.3 on 2012-06-18
- as-str now treats a.b as two identifiers separated by . so quoting produces [a].[b] instead of [a.b]
- Add :connection-uri option JDBC-34
-
Release 0.2.2 on 2012-06-10
-
Release 0.2.1 on 2012-05-10
-
Release 0.2.0 on 2012-04-23
- Merge internal namespace into main jdbc namespace JDBC-19
-
Release 0.1.4 on 2012-04-15
- Unwrap RTE for nested transaction exceptions (we already unwrapped top-level transaction RTEs).
- Remove reflection warning unwrapping RunTimeException (Alan Malloy)
-
Release 0.1.3 on 2012-02-29
- Fix generated keys inside transactions for SQLite3 JDBC-26
-
Release 0.1.2 on 2012-02-29
- Handle prepared statement params correctly JDBC-23
- Add support for SQLite3 JDBC-26
- Replace replicate (deprecated) with repeat JDBC-27
- Ensure MS SQL Server passes tests with both Microsoft and jTDS drivers
- Build server now tests derby, hsqldb and sqlite by default
- Update README per Stuart Sierra's outline for contrib projects
-
Release 0.1.1 on 2011-11-02
-
Release 0.1.0 on 2011-10-16
- Remove dependence on deprecated structmap JDBC-15
-
Release 0.0.7 on 2011-10-11
-
Release 0.0.6 on 2011-08-04
- Improve exception handling (unwrap RTE)
- Don't use batch for update (causes exceptions on Apache Derby) JDBC-12
- Add test suite
-
Release 0.0.5 on 2011-07-18
- Expose prepare-statement API
- Allow with-query-results to accept a PreparedStatement or options for creating one, instead of SQL query string and parameters
- Support databases that cannot return generated keys
-
Release 0.0.4 on 2011-07-17
-
Release 0.0.3 on 2011-07-01
- Key generation compatibility with MS SQL Server, PostgreSQL JDBC-10
-
Release 0.0.2 on 2011-06-07
- Clojure 1.2 compatibility JDBC-7
-
Release 0.0.1 on 2011-05-07
- Initial release
-
Changes from clojure.contrib.sql:
- Expose print-... functions; no longer write exceptions to
\*out\*
- Define resultset-seq to replace clojure.core/resultset-seq
- Add naming / quoting strategies (see name mapping documentation
- Return generated keys from insert operations, where possible
- Add insert-record function
- Clojure 1.3 compatibility
- Expose print-... functions; no longer write exceptions to
Copyright (c) Sean Corfield, Stephen Gilardi, 2011-2023. All rights reserved. The use and distribution terms for this software are covered by the Eclipse Public License 1.0 (https://opensource.org/license/epl-1-0/) which can be found in the file epl-v10.html at the root of this distribution. By using this software in any fashion, you are agreeing to be bound by the terms of this license. You must not remove this notice, or any other, from this software.