fdd1f152a7
improve lambda syntax (allow single expression)
2016-06-16 11:02:06 -04:00
ccad99fb5c
Merge pull request #18911 from rmuir/noncapturing_lambdas
...
non-capturing lambda support
2016-06-16 10:31:54 -04:00
8d3ef742db
Painless: =~ and ==~ operators
...
Adds support for the find operator (=~) and the match operator (==~)
to painless's regexes. Also whitelists most of the Matcher class and
documents regex support in painless.
The find operator (=~) returns a boolean that is the result of building
a matcher on the lhs with the Pattern on the RHS and calling `find` on
it. Use it like this:
```
if (ctx._source.last =~ /b/)
```
The match operator (==~) returns boolean like find but instead of calling
`find` on the Matcher it calls `matches`.
```
if (ctx._source.last ==~ /[^aeiou].*[aeiou]/)
```
Finally, if you want the actual matcher you do:
```
Matcher m = /[aeiou]/.matcher(ctx._source.last)
```
2016-06-16 08:42:33 -04:00
4293030fd9
painless: Remove stale comment and remove declared exception
2016-06-16 11:46:11 +02:00
f2d6219426
painless: remove useless dropArguments and throws statement in megamorphic cache; add tests
2016-06-16 11:08:10 +02:00
18ff051ad5
Simplify ScriptModule and script registration ( #18903 )
...
Registering a script engine or native scripts still uses Guice today
and is much more complicated than needed. This change moves to a pull
based model where script plugins have to implement a dedicated interface
`ScriptPlugin` and defines simple getter returning instances rather than
classes.
2016-06-16 09:35:13 +02:00
ec7215e5f2
get non-capturing lambdas working
2016-06-16 01:25:43 -04:00
60176afdde
clean up a bit more
2016-06-15 20:30:16 -04:00
a4503c2aed
Plugins: Remove name() and description() from api
...
In 2.0 we added plugin descriptors which require defining a name and
description for the plugin. However, we still have name() and
description() which must be overriden from the Plugin class. This still
exists for classpath plugins. But classpath plugins are mainly for
tests, and even then, referring to classpath plugins with their class is
a better idea. This change removes name() and description(), replacing
the name for classpath plugins with the full class name.
2016-06-15 17:12:22 -07:00
54797b7d09
don't let megamorphic cache "capture" arbitrary arguments.
...
pass the shit we need as bootstrap params
2016-06-15 19:33:59 -04:00
ae6ce1c42e
Small refactoring and cleanup to separate the polymorphic and megamorphic case
2016-06-15 22:27:30 +02:00
6e83f41e94
Merge branch 'more_def_cleanup' of https://github.com/rmuir/elasticsearch into more_def_cleanup
2016-06-15 21:47:28 +02:00
5b11b15300
fix typo
2016-06-15 21:46:58 +02:00
786c3b0fa8
speed up compound assignments
2016-06-15 14:38:37 -04:00
a4036b8069
add compound assignment tests
2016-06-15 14:02:19 -04:00
949e99bde3
Merge branch 'more_def_cleanup' of https://github.com/rmuir/elasticsearch into more_def_cleanup
2016-06-15 19:48:53 +02:00
8d3c411298
Remove unneeded type adaption and add error message on violation
2016-06-15 19:48:38 +02:00
e3f34730ae
and compound assignment
2016-06-15 13:37:10 -04:00
27f8b6e6db
get compound assignment working
2016-06-15 13:30:21 -04:00
ca2e0e1660
Some small cleanups around the lookup
2016-06-15 19:07:13 +02:00
767bb49da8
Fix bug with first param (receiver) on the cache lookup
2016-06-15 19:00:22 +02:00
283410b7ac
Merge branch 'more_def_cleanup' of https://github.com/rmuir/elasticsearch into more_def_cleanup
2016-06-15 18:37:07 +02:00
04ab42cdf5
Use rethrow puzzler in the ClassValue
2016-06-15 18:36:45 +02:00
d4db8c9c82
some cleanup
2016-06-15 12:32:38 -04:00
6487940a79
Move type changes up to static part
2016-06-15 18:26:11 +02:00
a473298f44
Add megamorphic cache
2016-06-15 18:21:42 +02:00
b71f42a627
split MIC from PIC
2016-06-15 09:07:11 -04:00
2b1ebc55a9
remove null guards from +, if not needed.
...
improve DefBootstrapTests
2016-06-14 23:52:21 -04:00
5e4b2f2815
clean up compound assignment tests, fix shift bug
2016-06-14 20:40:16 -04:00
cd4f4594d6
fix perf bug with xor promotion
2016-06-14 19:19:50 -04:00
5cef06fa3e
Add a test.
2016-06-14 14:28:47 -07:00
36f7632232
Correct type checking during casting related to Object.
2016-06-14 14:12:23 -07:00
b69e8aadbd
refactor tests
2016-06-14 15:07:56 -04:00
e4dc469e58
Merge pull request #18849 from rmuir/give_indy_ops_types
...
Remove casts and boxing for dynamic math
2016-06-14 13:01:04 -04:00
79f0ab2fc6
rename this variable
2016-06-14 12:51:29 -04:00
49c09f1527
Test: Fix error message check for Windows
...
Betrayal! CRLF strikes again. Weakened the test from equals to
containsString.
2016-06-14 09:56:49 -04:00
a12937d548
painless: remove casts and boxing for dynamic math
2016-06-14 00:49:13 -04:00
7ecb4ca4e9
Refactor def math
2016-06-13 18:56:28 -04:00
6617b53422
Painless: Add support for /regex/
...
Adds `/regex/` as a regex constructor. A couple of fun points:
1. This makes generic the idea of arbitrary stuff adding a constant.
Both SFunction and LRegex create a statically initialized constant.
Both go through Locals to do this because they LRegex isn't directly
iterable from SScript.
2. Differentiating `/` as-in-division from `/` as-in-start-of-regex
is hard. See:
http://www-archive.mozilla.org/js/language/js20-2002-04/rationale/syntax.html#regular-expressions
The javascript folks have a way, way tougher time of it then we do
because they have semicolon insertion. We have the much simpler
delimiter rules. Even with our simpler life we still have to add
a hack to get lexing `/regex/` to work properly. I chose to add
token-level lookbehind because it seems to be a pretty contained hack.
I considered and rejected lexer modes, a lexer member variable,
having the parser set variables on the lexer (this is a fairly common
solution for js, I believe), and moving regex parsing to the parser
level.
3. I've only added a very small subset of java.util.regex to the
whitelist because it is the subset I needed to test LRegex sanely.
More deserves to be added, and maybe more regex syntax like `=~` and
`==~`. Those can probably be added without too much pain.
2016-06-13 18:11:45 -04:00
c6e144f669
remove stray tab, fix comment
2016-06-13 13:35:11 -04:00
faee2323ab
add array ctor refs
2016-06-12 12:53:37 -04:00
5be9211761
cleanup grammar
2016-06-12 11:25:52 -04:00
60cc152caf
Remove unused field
2016-06-11 22:34:17 +02:00
5c352a063a
need not be public, should not be public
2016-06-11 16:03:45 -04:00
07baefac32
Add comments
2016-06-11 16:01:50 -04:00
fd9b0f9087
Remove some hardcoded strings, fix exception handling (remove rethrow), add some utility methods around the "handle$" fields
2016-06-11 21:53:15 +02:00
454881eeb4
test interface default methods
2016-06-11 15:51:32 -04:00
c08b45a277
More cleanup
2016-06-11 21:20:39 +02:00
65352face2
fix too long line
2016-06-11 15:04:49 -04:00
592042601f
def case working
2016-06-11 15:02:43 -04:00
876f0a613e
Unfuck the MethodWriter nesting. All MethodWriters should only live separately from each other, no nesting
2016-06-11 21:00:06 +02:00
e2f999b025
make this static final
2016-06-11 14:33:48 -04:00
8d5a70200f
write refs for functions
2016-06-11 14:29:15 -04:00
ab457be9d8
remove unnecessary semicolon and return
2016-06-11 13:48:04 -04:00
0bfb166eeb
merge master
2016-06-11 13:23:17 -04:00
4a265d0279
Painless: Add } as a delimiter. Kindof.
...
Add `}` is statement delimiter but only in places where it is
otherwise a valid part of the syntax, specificall the end of a block.
We do this by matching but not consuming it. Antlr 4 doesn't have
syntax for this so we have to kind of hack it together by actually
matching the `}` and then seeking backwards in the token stream to
"unmatch" it. This looks reasonably efficient. Not perfect, but way
better than the alternatives.
I tried and rejected a few options:
1. Actually consuming the `}` and piping a boolean all through the
grammar from the last statement in a block to the delimiter. This
ended up being a rather large change and made the grammar way more
complicated.
2. Adding a semantic predicate to delimiter that just does the
lookahead. This doesn't work out well because it doesn't work (I
never figured out why) and because it generates an *amazing*
`adaptivePredict` which makes a super huge DFA. It looks super
inefficient.
Closes #18821
2016-06-11 12:51:25 -04:00
e293000d8d
Merge branch 'master' into stubby
2016-06-10 16:22:17 -07:00
bd94c62adf
Added a stub node for lambda expressions.
2016-06-10 16:18:36 -07:00
16d17351f7
don't use loadLocal, add new node to package-info
2016-06-10 17:37:57 -04:00
9e0a70c4b2
Merge branch 'master' into capturingReferences
2016-06-10 17:10:56 -04:00
f98be6fa06
Add functions to Painless.
...
Closes #18810
2016-06-10 13:59:59 -07:00
edb0fd3b88
painless: add capturing method references
2016-06-10 14:24:58 -04:00
9826b64e1f
Exempt LocalDate from bridge method checks on java 9
...
The painless whitelist has a lot of self-checking, in this case, it checks
for missing covariant overrides. It fails on java 9, because LocalDate.getEra()
now returns IsoEra instead of Era: https://bugs.openjdk.java.net/browse/JDK-8072746
To our checker, it thinks we were lazy with whitelisting :)
This means painless works on java 9 again
2016-06-08 15:47:09 -04:00
be168f528b
Merge pull request #18771 from jdconrad/methods
...
Add Method to Get New MethodWriters
2016-06-07 11:47:29 -07:00
f9b45107c4
Added method to allow creation of new methods on-the-fly.
2016-06-07 11:24:33 -07:00
f23e8c3722
more cleanups
2016-06-07 12:42:19 -04:00
89224dbc3c
switch FunctionRef over to methodhandles
2016-06-07 11:56:44 -04:00
32965f043a
convert FunctionRef/Def usage to methodhandles.
2016-06-07 10:23:49 -04:00
3108c399f0
use painless types in FunctionRef
2016-06-07 09:49:54 -04:00
39550354a7
compute functional interface stuff in Definition
2016-06-07 09:41:03 -04:00
3238868cc4
Make checkstyle happy (add Lookup import, line length)
2016-06-07 13:06:23 +02:00
cac111ff73
Don't hide LambdaConversionException and behave like real javac compiled code when a conversion fails. This works anyways, because fallback is allowed to throw any Throwable
2016-06-07 13:00:36 +02:00
6913be2ec2
Pass through the lookup given by invokedynamic to the LambdaMetaFactory. Without it real lambdas won't work, as their implementations are private to script class
2016-06-07 12:32:59 +02:00
950b093fc7
Merge branch 'master' into painless_method_references
2016-06-06 22:39:28 -04:00
b3804c47f7
Cleaned up SEach node.
2016-06-06 18:21:08 -07:00
231268c89d
remove outdated TODO
2016-06-06 21:04:59 -04:00
6dace47c1f
Reverted S-node design change.
2016-06-06 17:59:39 -07:00
8db9a971e5
enable test
2016-06-06 20:30:42 -04:00
2852e82ab9
add def iteration over arrays
2016-06-06 20:07:32 -04:00
d331f8768e
implement dynamic case for iterables
2016-06-06 19:41:37 -04:00
aeaf39f36d
Added a for each stub node for a target type of def.
2016-06-06 16:07:12 -07:00
a088d367f7
More comments.
2016-06-06 15:58:56 -07:00
f897342296
Updated package info documentation.
2016-06-06 15:34:51 -07:00
9a78f6955b
Added foreach for array types.
2016-06-06 15:25:09 -07:00
23d7cdd043
ban slow methods in MethodHandle
2016-06-06 17:58:00 -04:00
b32f6abb26
remove unnecessary caching
2016-06-06 17:29:03 -04:00
a4ffaa6e7a
Working iterable foreach with tests.
2016-06-06 13:53:31 -07:00
b2e648dfde
Implementation of iterable.
2016-06-06 12:33:25 -07:00
260b0fd40f
Stubbed foreach node.
2016-06-06 10:49:40 -07:00
0c9174decc
Merge branch 'master' into each
2016-06-06 09:23:01 -07:00
881e4058ee
don't do a no-op filter, that was just for testing
2016-06-06 09:39:39 -04:00
6dbf7ab1ea
initial messy impl of painless method references
2016-06-06 09:27:19 -04:00
c0a3a20071
painless: Add support for the new Java 9 MethodHandles#arrayLength() factory (see https://bugs.openjdk.java.net/browse/JDK-8156915 )
2016-06-03 23:35:11 +02:00
9a0d0d7cff
Fixed a grammar mistake in a comment.
2016-06-02 15:30:11 -07:00
b5393ce55e
Improve painless compile-time exceptions
2016-06-02 18:23:47 -04:00
b6e1429760
Merge branch 'master' into each
2016-05-31 09:36:04 -07:00
2d1eb89aef
improve date api for expressions/painless fields
2016-05-31 09:32:33 -04:00
d1f4e2ef04
Merge branch 'master' into each
2016-05-30 16:22:14 -07:00
f5be0982d9
Added initial infrastructure to allow for each to be made.
2016-05-27 17:41:31 -07:00
2d8030f09f
Added foreach in grammar.
2016-05-27 13:50:38 -07:00
aa53285448
add java.time packages to painless whitelist
2016-05-27 15:07:27 -04:00
3f06d9f3b8
Merge pull request #18600 from rmuir/new_script_exception
...
replace ScriptException with a better one
2016-05-26 17:51:34 -04:00
cde83e363a
Add an update script equivalent test.
2016-05-26 13:16:22 -07:00
f037807117
replace ScriptException with a better one
2016-05-26 11:43:29 -04:00
eefcafd831
Add method reference stub to Painless.
2016-05-25 10:34:21 -07:00
82ae65bbb5
Tests: Make score tests unit tests
...
This change simply removes the need for single node tests from score
tests.
2016-05-24 19:53:50 -07:00
a5d0ccf50b
painless: add trap tests
2016-05-23 23:36:40 -04:00
0a9c036c3e
Merge branch 'master' into painless_more_whitelisting
2016-05-23 20:35:44 -04:00
1424f09c02
Remove Painless grammar ambiguities.
...
Closes #18531
2016-05-23 16:39:52 -07:00
a56a9cb818
improve painless whitelist coverage of java api
2016-05-23 19:35:28 -04:00
fdfd2a2f18
Remove ScriptMode class in favor of boolean true/false
...
This removes the ScriptMode class entirely, which was an enum with two
options (ON and OFF) which essentially boiled down to true and false.
Now the boolean values are used instead.
2016-05-20 15:01:30 -06:00
b156438957
Proxy box method to use valueOf.
2016-05-20 11:22:18 -07:00
0e24ed49e6
Use the build-in valueOf method instead of the custom one.
2016-05-20 11:13:43 -07:00
d836194095
Fixed tests and added a comment to the box method.
2016-05-20 11:04:09 -07:00
2b793c1e06
Fix boxing.
2016-05-20 10:38:06 -07:00
c0eb813e10
Merge branch 'clean' into whitelist_cleanup
2016-05-20 10:59:13 -04:00
ce9b36590a
simplify tree node use of compiler settings
2016-05-20 09:36:09 -04:00
b3c71c1e3a
improve style for these constants
2016-05-20 08:42:29 -04:00
7bc91c00d9
More elegant way to force class initialization
2016-05-20 14:12:15 +02:00
6205f79554
Remove Definitions's copy-ctor; fix RuntimeClass to be unmodifiable
...
Please note: The maps inside the pirvate singleton instance of Defininition are no longer unmodifiable, but nothing from the outside can modify it! All private :-)
2016-05-20 14:03:46 +02:00
4ffa92c7c1
Remove the Definition instance passed around everywhere
2016-05-20 12:35:08 +02:00
91f4bba042
Make Definition's public API completely static.
...
TODO: Remove Definition arguments everywhere and hide INSTANCE field!
2016-05-20 11:22:03 +02:00
f5fc60ac67
Add utility type constant
2016-05-20 11:05:40 +02:00
4e454439bb
Fixed bugs in comparison with Def. Fixed may tests.n
2016-05-20 02:04:38 -07:00
87ab39613d
more constants
2016-05-20 10:56:50 +02:00
82aed08301
Restore constants as static final (first step)
2016-05-20 10:48:41 +02:00
64e2ef5807
Removed transforms. No user-facing boxing.
2016-05-19 23:38:14 -07:00
3962553551
Removed user-facing boxing.
2016-05-19 18:51:55 -07:00
cfe555d216
move unused @SuppressWarnings one level up
2016-05-20 00:46:05 +02:00
afa9ebad28
dce
2016-05-19 16:03:11 -04:00
0ee726ed0d
nuke eclipse warnings
2016-05-19 15:35:26 -04:00
a569130368
fix typo
2016-05-19 15:28:59 -04:00
b9c7dbcfbd
nuke overflow detection as we cannot guarantee for def. simplify life :)
2016-05-19 15:25:31 -04:00
915ab16176
First steps at whitelist cleanup
2016-05-19 09:29:48 -04:00
1022123f59
Merge pull request #18400 from uschindler/painless_java9StringConcats
...
Use Java 9 Indy String Concats, if available
2016-05-18 08:00:20 -04:00
ea9026e6d8
painless: fix checkstyle
2016-05-18 00:31:12 +02:00
b418d01117
painless: Improve string concat tests; add state machine check after MethodWriter
2016-05-18 00:26:47 +02:00
f055ad641a
Made def variable casting consistent with invokedynamic rules.
2016-05-17 13:16:49 -07:00
b27b0483d5
Merge branch 'master' into painless_java9StringConcats
...
# Conflicts:
# modules/lang-painless/src/test/java/org/elasticsearch/painless/StringTests.java
2016-05-17 18:47:33 +02:00
5622e9c9d3
painless: use Java 9 Indy String Concats, if available
2016-05-17 17:44:46 +02:00
fa5f099486
make semicolons only optional at the end
2016-05-17 10:55:40 -04:00
92339c4763
painless: add method overloading based on arity
...
Closes #18385
Squashed commit of the following:
commit b2819df4d392d69b86e5c96d358eb03424e67e02
Author: Robert Muir <rmuir@apache.org>
Date: Tue May 17 09:15:47 2016 -0400
add note about tuple
commit 85fcac6a0d0674da24535121eab23e2c407d683f
Author: Robert Muir <rmuir@apache.org>
Date: Mon May 16 23:39:25 2016 -0400
painless: add method overloading based on arity
2016-05-17 09:17:43 -04:00
75f27df376
painless: Refactor WriterUtils to extends ASM GeneratorAdapter as "MethodWriter" for easy usage (and more coming later)
2016-05-17 02:25:05 +02:00
3fd3d367ef
Moved ADefLink to be a marker interface instead.
2016-05-16 15:48:29 -07:00
3f23186712
Merge branch 'master' into types
2016-05-16 14:22:59 -07:00
95a1b5b2a3
Fixed a casting bug in Painless.
2016-05-16 14:18:05 -07:00
e69305ae21
Merge pull request #18372 from rmuir/painless_whitelist_expansion
...
painless whitelist expansion
2016-05-16 17:08:49 -04:00
8d4c1befe5
Merge pull request #18364 from rmuir/nukeRunAsFloat
...
Remove LeafSearchScript.runAsFloat(): Nothing calls it.
2016-05-16 17:08:25 -04:00
1d394868d4
Merge branch 'master' into types
2016-05-16 12:43:24 -07:00
43344eee82
Add more def tests including checking bytecode for specific
...
optimizations.
2016-05-16 12:42:25 -07:00
f664fa5f33
Merge pull request #18350 from uschindler/painless_isAssignableFrom
...
Use isAssignableFrom instead of relying on ClassCastException
2016-05-16 12:41:09 -07:00
a2c2628a78
Merge pull request #18359 from uschindler/painless_removeStoreBoxing
...
Remove boxing when loading and storing values in "def" fields/arrays, remove boxing onsimple method calls of "def" methods
2016-05-16 12:11:37 -07:00
d6cbbdeb98
painless: Some reformatting in EChain to make it clear if "this" or inner "expression" is affected
2016-05-16 21:04:03 +02:00
b05ac87a5d
painless: Add Jack's test
2016-05-16 20:37:51 +02:00
3a5ef68ad0
painless: Fix issue with dup and cast
2016-05-16 20:31:27 +02:00
2fc7ca46e1
expand whitelist for basic types
2016-05-16 09:56:29 -04:00
07e4f416e6
painless: Small cleanup of Debugger class to use StringWriter
2016-05-16 15:52:14 +02:00
8edf213492
Remove LeafSearchScript.runAsFloat(): Nothing calls it.
2016-05-15 22:59:28 -04:00
65aca4f71c
painless: make compound statement like a[1]++ work with def. There was also a bug with the size of LDefArray: fixed
2016-05-15 23:00:32 +02:00
d221cd14d2
painless: Also remove boxing for reads and method calls
2016-05-15 19:49:25 +02:00
604bcd9320
painless: Make field stores not box; use GeneratorAdapter.invokeDynmaic for consistency with other method calls
2016-05-15 18:31:53 +02:00
8195ef9e7a
painless: Transform more try-catch-ClassCastException to isAssignableFrom
2016-05-15 01:06:29 +02:00
7211017813
painless: Some cleanups
2016-05-14 17:23:29 +02:00
c78bf8c867
painless: Use isAssignableFrom instead of relying on ClassCastException
2016-05-14 16:04:18 +02:00
c8bd521ed3
painless: restore accidentally removed test
2016-05-14 12:00:30 +02:00
ba9dae820a
Add test for wrong array index
2016-05-13 23:22:41 +02:00
1a76329d89
Take return type from "after" field.
2016-05-13 23:17:22 +02:00
1b10a14e8d
painless: build descriptor of array and field load/store in code; fix array index to adapt type not DEF
2016-05-13 23:04:04 +02:00
2028691e66
painless: improve exception stacktraces
...
closes #18319
2016-05-13 15:40:45 -04:00
9731d5d314
painless: Rename the dynamic call site factory to DefBootstrap and make the inner class very short (PIC = Polymorphic Inline Cache)
2016-05-13 19:58:05 +02:00
9bcdafedda
Allow only a single extension for a scripting engine
...
Previously multiple extensions could be provided, however, this can lead
to confusion with on-disk scripts (ie, "foo.js" and "foo.javascript")
having different content. Only a single extension is now supported.
The only language currently supporting multiple extensions was the
Javascript engine ("js" and "javascript"). It now only supports the
`.js` extension.
Relates to #10598
2016-05-13 09:54:31 -06:00
efff3918d8
Remove support for mulitple languages per scripting engine
2016-05-13 09:24:31 -06:00
a4060f7436
Remove vestiges of script engine sandboxing
...
This removes all the mentions of the sandbox from the script engine
services and permissions model. This means that the following settings
are no longer supported:
```yaml
script.inline: sandbox
script.stored: sandbox
```
Instead, only a `true` or `false` value can be specified.
Since this would otherwise break the default-allow parameter for
languages like expressions, painless, and mustache, all script engines
have been updated to have individual settings, for instance:
```yaml
script.engine.groovy.inline: true
```
Would enable all inline scripts for groovy. (they can still be
overridden on a per-operation basis).
Expressions, Painless, and Mustache all default to `true` for inline,
file, and stored scripts to preserve the old scripting behavior.
Resolves #17114
2016-05-13 09:24:31 -06:00
a1d0fb5008
painless: cleanup of DynamicCallSite to have all implementation encapsulated in inner class, outer class only has bootstrap method
2016-05-13 15:50:55 +02:00
2e25837deb
Made line number available in painless nodes for debugging.
2016-05-12 10:32:06 -07:00
28164ccf2e
Decouple the ANTLR AST from Painless.
...
Closes #18286
2016-05-12 00:46:48 -07:00
113e239e0f
Merge pull request #18288 from rmuir/painless_debug
...
painless: add debugging tool to tests
2016-05-12 01:16:32 -04:00
8871fddba7
Merge pull request #18282 from jdconrad/longs
...
Long priority over Float
2016-05-11 21:08:28 -07:00
930a215b5c
add debugger
2016-05-11 23:56:19 -04:00
25dd64250b
painless: remove input, support params instead
2016-05-11 21:32:10 -04:00
c5532d3df0
add a rest test for this that seems to work, fix the documentation. thanks @s1monw
2016-05-11 16:07:08 -04:00
6b4e47bf96
this makes aggregations per-document _value fast (bypass hash put, hash get, etc) for painless.
...
but i have no clue how to test it, it seems this feature never worked via REST?
Should we drop the feature instead?
2016-05-11 15:39:00 -04:00
20af482dd7
Removed extra branches. Only check for long first where we already know the
...
second type.
2016-05-11 10:22:40 -07:00
2764b47d22
Give long priority over float in Def math calcs. Promotion of types
...
still works correclty.
2016-05-11 09:32:30 -07:00
d790c76636
painless: _score as double, not float
2016-05-11 11:19:18 -04:00
7689a1af28
Add 'ctx' keyword to painless.
2016-05-11 02:31:19 -04:00
d8b373cac6
Fixed an issue related to bracket shortcuts.
2016-05-10 23:11:13 -07:00
632bb4c720
add reserved word tests for 'doc', too
2016-05-11 00:17:19 -04:00
389ef462c4
special access to doc
2016-05-10 23:43:57 -04:00
abb015bcb9
add tests for reserved words. ban stores against _score
2016-05-10 22:55:06 -04:00
a63285aaf9
painless: retrieve _score directly from Scorer
2016-05-10 20:53:08 -04:00
1d80542199
Merge pull request #18247 from rmuir/needsScore
...
painless: implement needsScore() correctly.
2016-05-10 19:21:09 -04:00
ddc2c1f682
Merge pull request #18234 from rmuir/more_indy_type_data
...
painless: use better typing for dynamic method calls
2016-05-10 19:17:13 -04:00
1c2d3b10ef
some cleanups suggested by @uschindler and break some long lines
2016-05-10 12:31:02 -04:00
eb1b2cf111
add test and cleanup
2016-05-10 11:34:33 -04:00
7a47d33a9b
Merge branch 'master' into needsScore
2016-05-10 11:23:50 -04:00
546aed8390
hack
2016-05-10 11:23:24 -04:00
243c9e77fa
painless: Add synthetic length property as alias to Lists, so they can really be used like arrays
2016-05-10 15:01:49 +02:00
2e613f49b3
painless: Array load/store and length with invokedynamic
2016-05-10 12:41:27 +02:00
3aeb66885b
painless: use better typing for dynamic method calls
2016-05-10 05:44:16 -04:00
ba2fe156e8
Switch over dynamic method calls, loads and stores to invokedynamic.
...
Remove performance hack for accessing a document's fields, its not needed.
Add support for accessing is-getter methods like List.isEmpty() as .empty
Closes #18201
2016-05-09 21:44:32 -04:00
e3ce6c9048
Painless: add fielddata accessors (.value/.values/.distance()/etc)
...
This gives better coverage and consistency with the scripting APIs, by
whitelisting the primary search scripting API classes and using them instead
of only Map and List methods.
For example, accessing fields can now be done with `.value` instead of `.0`
because `getValue()` is whitelisted. For now, access to a document's fields in
this way (loads) are fast-pathed in the code, to avoid dynamic overhead.
Access to geo fields and geo distance functions is now supported.
TODO: date support (e.g. whitelist ReadableDateTime methods as a start)
TODO: improve docs (like expressions and groovy have for document's fields)
TODO: remove fast-path hack
Closes #18169
Squashed commit of the following:
commit ec9f24b2424891a7429bb4c0a03f9868cba0a213
Author: Robert Muir <rmuir@apache.org>
Date: Thu May 5 17:59:37 2016 -0400
cutover to <Def> instead of <Object> here
commit 9edb1550438acd209733bc36f0d2e0aecf190ecb
Author: Robert Muir <rmuir@apache.org>
Date: Thu May 5 17:03:02 2016 -0400
add fast-path for docvalues field loads
commit f8e38c0932fccc0cfa217516130ad61522e59fe5
Author: Robert Muir <rmuir@apache.org>
Date: Thu May 5 16:47:31 2016 -0400
Painless: add fielddata accessors (.value/.values/.distance()/etc)
2016-05-05 18:38:41 -04:00
2cae575f53
Added single-quoted strings.
...
Closes #18150
2016-05-05 09:26:02 -07:00
59c135b58d
make internal Def methods private and add basic javadocs
2016-05-05 03:47:56 -04:00
928e2b904d
painless: optimize/simplify dynamic field and method access
2016-05-05 03:42:14 -04:00
a37e53c50f
Painless clean up including fixing _score issues and improving type
...
error messages.
Closes #17428
2016-03-30 16:40:17 -07:00
068c788ec8
Disable fielddata on text fields by defaults. #17386
...
`text` fields will have fielddata disabled by default. Fielddata can still be
enabled on an existing index by setting `fielddata=true` in the mappings.
2016-03-30 14:35:32 +02:00
800c844ebd
Split up Analyzer and Writer into multiple pieces.
...
Closes #17158
2016-03-17 10:15:13 -07:00
7986770e5f
Moved Painless from a plugin to a module.
...
Closes #16755
2016-02-21 16:50:54 -08:00
Robert Muir
Robert Muir
Nik Everett
Uwe Schindler
Simon Willnauer
Ryan Ernst
Jack Conradson
Lee Hinman
Adrien Grand