Commit graph

82 commits

Author SHA1 Message Date
Chris Targett
f778419cb9 Refactored parameterizer to use mutimethods
It was briefly mentioned in #honeysql on clojurians.slack.com that this
mechanism of extension uses a register method and an atom to support
extension when multimethods are used elsewhere.
2018-05-15 21:22:32 +01:00
Michael Blume
524887a9cd test SET after JOIN 2018-04-16 00:03:35 -07:00
Michael Blume
43e94bf497
Merge pull request #201 from vincent-dm/master
adds support for JOIN ... USING (...) syntax (#188)
2018-04-12 23:39:59 -07:00
Juvenn Woo
2abe128008
Support register customized parameterizer 2018-03-09 11:54:14 +08:00
Nuttanart Pornprasitsakul
3206b099df Remove nil predicate from :and 2018-02-18 13:27:07 +07:00
Michael Blume
8ad38d28af
Merge pull request #196 from emidln/emidln/insert-into-drop-parens
Avoid wrapping QUERY with parens while formatting INSERT INTO ... QUERY
2018-02-10 23:48:01 -08:00
Vincent
8a6ecd94aa adds support for JOIN ... USING (...) syntax (#188) 2018-01-16 03:13:14 +01:00
Tianxiang Xiong
ccff2d2c23 Limit value context to sequences in value positions
Fix #198
2018-01-09 23:42:53 -08:00
Brandon Adams
c7df208e3e Avoid wrapping QUERY with parens while formatting INSERT INTO ... QUERY
This also adds a test using honeysql.format/format for an existing
format-clause test case for :insert-into which encodes the previously
failing behavior.
2017-12-12 16:09:31 -06:00
Michael Blume
faba02f2b4
Merge pull request #186 from arichiardi/parameterizer-none
Add :parameterizer :none option
2017-11-15 21:23:11 -08:00
Andrea Richiardi
6cf89436d4 Add :parameterizer :none option
Sometimes you want to completely skip using parameters.

Postgres, in particular, complains with cryptic "syntax error at or near "$1"
whenever query parameters cannot be used for the query. An example would be
CREATE TABLE queries where the only parameters are the VARCHAR size.

This patch adds a :none option to :parameterizer which completely skips
parameter filling (and numbering), returning the SQL string only.
2017-11-06 17:13:02 -08:00
Andrea Richiardi
65037939b2 [Fix #167] Add ClojureScript self-host support
This patch uses the wonderful cgrand/macrovich for supporting both JVM and
self-host ClojureScript. The library avoids destructive code changes, no
namespaces were moved.

The patch also adds a runner for self-host tests by adding the lein-tach
plugin. Tests can be launched with:

  lein tach lumo (or planck)

See the lein-tach README for details.
2017-11-01 10:13:47 -07:00
Michael Blume
7256f7afba Revert "Merge pull request #169"
Revert the reversion, keep the behavior from 0.9.0, vectors will be
primarily used for function calls.
2017-08-25 09:12:53 -07:00
Michael Blume
3eaf5edbc0 Merge pull request #169 from MichaelBlume/fix-tuples
Revert #149 and restore tuple behavior
2017-07-18 13:47:56 -07:00
Michael Blume
9ab7d3a2f8 restore function call test using sql/call 2017-07-18 13:45:12 -07:00
Michael Blume
d99efb8606 allow inlining values
Most of the time we want to parameterize numbers, but sometimes we don't

Fixes #171
2017-07-18 12:50:27 -07:00
Michael Blume
a0b73cdcc6 Test that tuples work correctly
Thanks to @jetmind who gave this test in #168
2017-07-17 20:49:35 -07:00
Michael Blume
62c4760c6a Revert "Merge pull request #149"
Should #168
2017-07-17 20:49:35 -07:00
Brandon Adams
f94e343f31 Reprioritize WITH wrt UNION and UNION ALL
WITH needs to come before clauses that are part of a UNION
or UNION ALL.[1][2][3][4]

[1] "A CTE must be followed by a single SELECT, INSERT, UPDATE, or DELETE
statement that references some or all the CTE columns."
https://docs.microsoft.com/en-us/sql/t-sql/queries/with-common-table-expression-transact-sql

[2] "the WITH clause itself is attached to a primary statement that can
also be a SELECT, INSERT, UPDATE, or DELETE"
https://www.postgresql.org/docs/9.6/static/queries-with.html

[3] https://mariadb.com/kb/en/mariadb/with/

[4] "All common table expressions (ordinary and recursive) are created
by prepending a WITH clause in front of a SELECT, INSERT, DELETE, or
UPDATE statement."
https://sqlite.org/lang_with.html

Added a test case demonstrating the expected behavior for each of
UNION and UNION ALL.
2017-06-27 20:01:52 -07:00
Eli Naeher
00316fc6bd Add helpers for :with and :with-recursive clauses 2017-05-19 22:59:21 -07:00
Michael Blume
e0072efbdc Merge pull request #132, allow namespaced keywords 2017-04-05 21:11:45 -07:00
Tom Connors
bdfee0883e Fix using maps with different key orders in :values 2017-03-24 00:18:04 -07:00
Joost Diepenmaat
5e7e224df4 Show problem with operators in value positions 2017-02-17 10:20:40 +01:00
Richard Newman
a0d0880ac6 Add an additional test for phrasing complex UNION queries. 2016-10-12 09:26:51 -07:00
Richard Newman
9dd87d14a0 Don't parenthesize the subclauses of a UNION, UNION ALL, or INTERSECT clause. Fixes #141. 2016-10-12 08:50:54 -07:00
Richard Newman
0539e33d9a Get tests running with 'lein doo'. 2016-07-18 14:12:01 -07:00
Mike Blume
7c8cdf241d MichaelBlume's first pass at ClojureScript support. 2016-07-18 14:12:01 -07:00
Mike Blume
5603e9df9f Add ClojureScript build, move files to cljc.
ClojureScript build fails.
2016-07-18 14:10:52 -07:00
Jon Doane
5b197f9756 Solve an issue where namespaced keywords and symbols where only taking the name portion of the identifier.
- This now checks if there is a namespace and prepends it to the name if it exists.
2016-07-09 10:37:33 -04:00
Alexander Hudek
7397b3a48f Update tests for parameterized numbers. 2016-05-18 17:21:51 -04:00
Michael Blume
43df955a86 Merge pull request #111 from stuarth/master
allow joins without a pred
2016-03-30 23:10:18 -07:00
Michael Blume
d9567e64e0 Merge pull request #110 from csummers/master
escape quote characters in quoted identifiers
2016-03-12 17:33:09 -08:00
Dave Della Costa
0662a00722 Merge pull request #103 from jakemcc/add-intersect
add intersect support
2016-03-02 12:37:17 -05:00
Stuart Hinson
80e9d2ec44 allow joins without a pred 2016-03-01 09:39:21 -05:00
Curtis Summers
47cefd0371 escape quote characters in quoted identifiers 2016-02-24 06:45:49 -06:00
Michael Blume
06c97376ef prevent records from being treated as maps
Thanks to @djwhitt and @blinsay for reporting and for providing
test cases.

Fixes #106
2016-02-09 14:09:22 -08:00
Michael Blume
f8a2eedf45 demonstrate incorrect raw/call behavior 2016-02-09 14:07:51 -08:00
Jake McCrary
8fa44d22b3 add INTERSECT support 2015-11-29 17:49:45 -06:00
Michael Blume
607ee17cca Merge pull request #95 from emidln/emidln/with_column_names
Support column names in :with clauses
2015-10-18 11:07:03 -07:00
Brandon Adams
10a1565467 Support column names in :with clauses
This binds `honeysql.format/*subquery?*` to `false` which turns
on processing of a cte-name like `[:foo {:columns [:a :b :c]}]`
resulting in a string like `WITH foo (a, b, c)`.
2015-10-12 15:13:16 -05:00
Jon Doane
f38668edc6 Added one more assertion to test dashed names that are split by periods. 2015-10-11 11:38:50 -04:00
Jon Doane
7e78b10ae4 Added simple test to check for preserved dashes when the var is bound to true. 2015-10-11 11:29:55 -04:00
Dave Della Costa
2123c01fb5 adds tests for :union and :union-all 2015-08-26 00:11:29 +09:00
loganlinn
b51611d4d6 Add fn-handler for CASE statement 2015-08-06 11:29:54 -07:00
Donald Ball
713f72cfb6 Fix the parameterizable protocol to include nil
I forgot that Object doesn't include nil on protocols.

I added test coverage demonstrating the regression, only to note that
parameterized and literal queries produce different sql for the nil
case. I believe only the former is actually valid ANSI SQL, though mysql
at least does not care.
2015-05-20 11:23:36 -04:00
Mike Blume
115ac81fee Merge pull request #74 2015-05-08 16:41:11 -07:00
Andy Chambers
203296149f Add support for exists 2015-04-24 08:08:13 -07:00
Donald Ball
4bdaae39af Convert seq param values to literal sql param lists
This allows seqs to be used as param values, e.g. for IN clauses. It
also converts sets to seqs for the same purpose.

It does not expand the parameter name when this occurs, which will lead
to a mismatch between the *param-names* and *params* collections. It is
not clear how *param-names* are intended to be used, so I have no better
fix to suggest other than repeating a seq's param name the corresponding
number of times into that collection.
2015-04-21 14:28:28 -04:00
Donald Ball
2820662f59 Apply seq to sets when converting to sql
This allows them to be used as values, e.g. for in clauses, as
demonstrated in the test.
2015-04-20 23:40:59 -04:00
Mike Blume
72784452b9 Merge remote-tracking branch 'dball/support-locking-selects' into HEAD 2015-04-17 14:25:23 -07:00