| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
| | | | | |
|
| | | | | |
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | | |
experimental for now. destined to be automated via scala-jenkins-infra
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
bash's -x flag was flowing down into these scripts, resulting in extra
stuff was getting printed to stderr, resulting in a corrupted
build\pack\META-INF\MANIFEST.MF
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
this was failing because the expected output was:
res1: List[String] = List(shello, world.)
but the actual output was:
res1: List[String] = List(shello, world., Picked up _JAVA_OPTIONS: -Duser.home=y:\jenkins)
but the "Picked up..." part caused partest's filters feature to
ignore the entire line (it doesn't anchor matches to start of line.)
This was a tough one to track down.
|
| |\ \ \ \
| | |_|/ /
| |/| | | |
Remove two lingering `println`s in non-debug methods of the library.
|
| | |/ / |
|
| |\ \ \
| | | | |
| | | | | |
misc upgrades to README
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
* be clearer about required JDK versions
* eliminate redundancy with existing src/eclipse and src/intellij readmes
* give IntelliJ readme a .md extension
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
* link to new scala/contributors room on Gitter
* remove link to old disused CI
* replace list of mailing lists with single link to community page
* link to 2.12 nightlies too, not just 2.11
* update required JDK version info
|
|\ \ \ \ \
| |_|_|_|/
|/| | | | |
SI-9363 fix
|
| | | | | |
|
|/ / / / |
|
|\ \ \ \
| | | | |
| | | | | |
Fix some small warnings that pop up on every compile
|
|/ / / / |
|
|\ \ \ \
| | | | |
| | | | | |
Set the scene for phase 1 of the new trait encoding.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Comment about my poor naming choice in Types.
NullaryMethodType sounds like the method has one empty argument list,
whereas it really has no argument lists at all.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
sbt's [API extraction phase](https://github.com/sbt/sbt/blob/0.13/compile/interface/src/main/scala/xsbt/API.scala#L25)
extends `scala.reflect.internal.Phase`, which implements a bunch of methods,
such as `erasedTypes` as `false`, which are then overridden by scalac
in `GlobalPhase` (nested in scala.tools.nsc.Global).
(`erasedTypes` in particular is again overridden in the back-end -- for performance?)
However, since sbt's compiler phases extend `reflect.internal.Phase`,
the logic for detecting the current phase does not work,
as the default implementation is called (simply returning `false`),
when chasing the `prev` pointers hits an sbt-injected phase,
as its implementation is `reflect.internal`'s constant `false`.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This logic was scattered all over the hierarchy,
even though it's only needed in one spot, and is unlikely to evolve.
|
| | | | |
| | | | |
| | | | |
| | | | | |
A tiny bit more duplicated code, but so much more intelligible.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Keep owner for module (symbol of the tree) and module class (holds the members)
in synch while moving trees between owners (e.g., while duplicating them in specialization)
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Discovered while refactoring the way fields are emitted in traits.
TODO: a lot more cleanup
|
| | | | | |
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The goal is to reduce noise by sorting fields & methods,
suppressing debug & frame info.
This class can be used as a git textconf for classfiles with a script like:
```
scala scala.tools.nsc.backend.jvm.AsmUtils "$1"
```
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Triggered by sjrd's review, it turns out that we
forgot to unlink the symbol for omittable ValDefs.
I left the fix in a comment in the margin, since
this commit is meant to be bytecode-compatible.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Move conditions and template derivation back to the
call sites of `staticConstructor` (formerly `addStaticInits`).
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Codify the scope of mutability for various buffers,
inline one-time methods, further reduce spooky action at a distance.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
When we change the encoding of fields in traits,
this will come in handy...
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Limit communication between traits using ConstructorTransformerBase interface.
Make Constructors a bit less of a test case for itself,
by making a bunch of fields local variables of the `transformed` method.
|
|\ \ \ \ \
| | | | | |
| | | | | | |
SI-9178 Don't eta expand to an Function0-like SAM expected type
|
| | | | | | |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Otherwise, we can end up with a subtle source incompatibility with
the pre-SAM regime. Arguably we should phase out eta expansion to
Function0 as well, but I'll leave that for another day.
|
|\ \ \ \ \ \
| | | | | | |
| | | | | | | |
SI-6162 Make @deprecated{Inheritance,Overriding} public
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
It was kept private first with the expectation that further modifier
changes might warrant a more extensible design.
Until now, there doesn't seem to have appeared other interesting
use-cases that warranted a design overhaul.
This commit makes the existing annotations public to allow all Scala
developers to benefit from it.
|
|\ \ \ \ \ \ \
| | | | | | | |
| | | | | | | | |
Remove ThreadPoolExecutor fallback in ExecutionContextImpl
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
The method createDefaultExecutorService had a fallback if the creation
of a ForkJoinPool didn't succeed. This was necessary, because Scala
shipped its own version of FJP, and the dependency on sun.misc.Unsafe
(which is not an "offical" official API) made portability slightly
questionable.
Now that we can assume that FJP is supplied by the JDK, this concern
goes away.
|
|\ \ \ \ \ \ \ \
| | | | | | | | |
| | | | | | | | | |
SI-9545 SyncVar doc and deboxing
|
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | | |
Don't store the wrapped value double-wrapped in Option.
Also burnish some more docs.
|
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | | |
It used to have javish semantics, but now is what it is.
|
|\ \ \ \ \ \ \ \ \
| | | | | | | | | |
| | | | | | | | | | |
SI-7566 Remove some private, deprecated remnants from scala-actors
|
| | |/ / / / / / /
| |/| | | | | | | |
|
|\ \ \ \ \ \ \ \ \
| | | | | | | | | |
| | | | | | | | | | |
WrapAsScala: Fix documentation
|
| | |/ / / / / / /
| |/| | | | | | | |
|
|\ \ \ \ \ \ \ \ \
| | | | | | | | | |
| | | | | | | | | | |
SI-9555 Refactor SystemProperties to avoid mutable map
|
| | |_|_|_|/ / / /
| |/| | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | | |
A mutable map with help text was lazy populated if and when
lazy vals containing `BooleanProp` for a particular system was
initialized.
Reviewing this code, I noticed two problems: the class doesn't appear
to be designed for multithreaded access (e.g. someone could call
`help` while the map is being updated.), and the help method
won't return anything until the vals below are accessed for
the first time.
We don't use the help method anywhere else, but kept it around
in the unlikely event it was being used by some other tool.
I've done away with the mutable map by just moving the help messages
into a match in the help method itself.
|
|\ \ \ \ \ \ \ \ \
| |/ / / / / / / /
|/| | | | | | | | |
for 2.12, bump JDK version (6 -> 8) in IntelliJ sample config
|
| | | | | | | | | |
|