| Commit message (Collapse) | Author | Age | Files | Lines |
|\
| |
| | |
SI-6345 Remove scala.runtime.ScalaRunTime.Try
|
| | |
|
|\ \
| |/
|/| |
SI-6326. Makes getClass of a value class work.
|
|/ |
|
|\
| |
| | |
SI-6318 fixes ClassTag.unapply for primitives
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
ClassTag.unapply now has overloads for primitive value classes
so that it can preserve boxiness when performing subtyping tests.
First I wanted to annotate ClassTag.unapply with a ClassTag itself,
i.e. to transform its signature from "def unapply(x: Any): Option[T]"
to "def unapply[U: ClassTag](x: U): Option[T]".
But then virtpatmat_typetag.scala exhibited a nasty problem.
When pattern matching with this unapply, patmat first infers U as something
and then tries to pattern match against this inferred type. And if U gets
inferred as an abstract type itself, bad things happen:
warning: The outer reference in this type test cannot be checked at run time.
That's why I decided to drop the ClassTag idea and go with 9 extra overloads.
Not very beautiful, but definitely robust.
|
|\ \
| | |
| | | |
removes isLocatable from the public API
|
| | |
| | |
| | |
| | |
| | |
| | | |
Hard to come up with a good name, would need extensive documentation
to justify its purpose => no go. If necessary we can always reintroduce
it later in 2.10.1.
|
|\ \ \
| | | |
| | | | |
fixes for class tags and manifests
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Also introduces an important change to Manifest.Nothing and Manifest.Null.
Previously their `erasure` was equal to classOf[Object]. Now it's correctly
set to classOf[scala.runtime.Nothing$] and classOf[scala.runtime.Null$]
correspondingly.
See a discussion here:
https://groups.google.com/forum/#!topic/scala-internals/Y0ALGo7QPqE
|
| |/ /
| | |
| | |
| | |
| | | |
Aliases ClassTag.XXX to Manifest.XXX to reuse already existing
implementations of deprecated APIs.
|
|\ \ \
| |/ /
|/| | |
SI-6258 Reject partial funs with undefined param types
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This regressed with virtpatmat.
With -Xoldpatmat, pattern matching anonymous functions with
an expected type of PartialFunction[A, B] are translated to
a Function tree, and typed by typedFunction, which issues an
error of the parameter types are not fully defined.
This commit adds the same check to MatchFunTyper.
It doesn't plug the hole in RefChecks#validateVariance (which is
reminiscent of SI-3577.) Seems to me that in general one should handle:
a) both BoundedWildcardType and WildcardType
when in a place that can be called during inference, or
b) neither otherwise
|
|\ \ \
| | | |
| | | | |
removes Type.isConcrete
|
| | |/
| |/|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This method was no longer used anywhere, except for a place where it wasn't
really necessary.
Its implementation is non-trivial, and I have doubts about it, so I propose
to remove it altogether instead of living with some dubious code necessary
for some dubious matter.
|
|\ \ \
| |/ /
|/| | |
Issue/5031
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Eliminated InvalidCompanions exception entirely. Anyone's
guess why we unholstered this exception every time someone
calls "isCodefinedWith" rather than when symbols are created.
I moved the check into Namers, where it can be done once and
with sufficient finesse not to crash so much.
With this patch in place, "playbench" can be built with java7.
|
|/ / |
|
|\ \
| | |
| | | |
Removing duplication from Duration.
|
|/ /
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
I don't know what good it is to have code review if we are checking
in code like this. We must raise the bar, people. When the
justification for code being in the standard library is borderline
at best - as it is here - then the code must be of exceptional
quality. This code is not of exceptional quality.
Mostly these are not behavioral changes, but:
- I removed finite_? as it is a gratuitous deviation from
every isXXX method in the world. This isn't ruby.
- I removed all the regexps, which only made things complicated
- I removed all the unnecessary casts, which is to say, all of them
- I made more things final, sealed, and private
- The unapply structure was all wrong; returning Option[Duration]
on the string unapply meant you'd have to say
case Duration(Duration(x, y)) => ...
So I fixed apply and unapply to be symmetric.
- And I removed the "parse" method, since it was doing what
apply is supposed to do.
There's a test case to exercise accessing it from java,
which also reveals what I hope are bugs.
Thanks to viktor klang for DurationConversions.
|
|\ \
| | |
| | | |
Fixed positions in de-aliased special symbols..
|
| | |
| | |
| | |
| | |
| | | |
`apply` methods.
Fixed #5064, thanks to @paulp who showed the right direction (and how to test it).
|
|\ \ \
| | | |
| | | | |
More worksheet nstrumentation changes
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Made instrumenter more robust by taking into account the positions of the tokens in the source. This allows us to reliably put instrumentation code at the end of the previous token, which tends to put it at end of lines. Furthermore, we now skip left parents. Previously, the instrumenter got confused for a statement like
( x + 1 )
because it thought that the statement started at the `x` and not at the `(`.
Another small change is that we now use decoded names in the worksheet. So ??? will show as ??? not $qmark$qmark$qmark.
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
(1) Handle empty worksheets
(2) Handle for expressions
Review by @dragos
|
|\ \ \ \
| | | | |
| | | | | |
Fixes typos in the ScalaDoc of StringContext
|
| | | | | |
|
|\ \ \ \ \
| | | | | |
| | | | | | |
Allow nested calls to `askForResponse` in the presentation compiler.
|
| | |_|/ /
| |/| | |
| | | | |
| | | | |
| | | | | |
Fix #6312.
review by @odersky,@lrytz.
|
|\ \ \ \ \
| | | | | |
| | | | | | |
Fixes SI-6285 - ParIterableLike no longer says sequential foreach.
|
| | | | | | |
|
|\ \ \ \ \ \
| |/ / / / /
|/| | | | | |
SI-6274 Fix owners when eta-expanding function with byName param
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
When eta-expanding a function that takes a by-name param the local definition
for the corresponding argument is a function-0
val eta$1 = () => { argument-to-by-name }
If there are any definitinos in the `argument-to-by-name`, the symbol
owner needs to be changed to the anonymous function's symbol.
To know the function symbol in eta expand, we need to type-check the
function, and therefore pass the `Typer` instance to `etaExpand`.
|
|\ \ \ \ \ \
| | | | | | |
| | | | | | | |
Fix SI-6294.
|
| | | | | | | |
|
|\ \ \ \ \ \ \
| |_|/ / / / /
|/| | | | | | |
Deprecate unmaintained/old classes for removal in 2.11
|
| | | | | | | |
|
|\ \ \ \ \ \ \
| |_|_|/ / / /
|/| | | | | | |
Made Dynamic extend Any.
|
|/ / / / / /
| | | | | |
| | | | | |
| | | | | | |
So it can be used in value classes.
|
|\ \ \ \ \ \
| | | | | | |
| | | | | | | |
Formatting cleanup in def typed.
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
As seen on scala-internals.
|
|\ \ \ \ \ \ \
| | | | | | | |
| | | | | | | | |
further polishing of reflection
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
I think `isVal` and `isVar` are the right names, because they
exactly map on Scala syntax.
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
As discussed, this is not the best API to expose, because this is an
implementation detail that might change.
However the ship has sailed. We're already imposing
the moduleClass <-> sourceModule quirk to the users of our API:
Evidence: http://stackoverflow.com/questions/12128783.
There are reflection tasks that cannot be pulled without the knowledge
of this implementation detail, so we shouldn't pretend that we can
change it on a whim and not break anything.
Hence I propose to add sourceModule to the public contract
and bear the potential consequences.
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
Symbol.pos is moved to macros.Universe, because it's empty at runtime.
Also added Symbol.associatedFile, because it has valid usages during runtime.
When reflecting one might want to get to a classfile of the symbol and then
analyze it to e.g. decompile the bytecode or find out the associated src file.
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
no longer necessary, since scala.tools.nsc.io.AbstractFile
has been long moved to scala-reflect.jar
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
By giving ToolBox.mirror the correct type and constructing the underlying
mirror from a proper classloader, we improve user experience w.r.t
runtime code generation.
Direct motivator: http://stackoverflow.com/questions/12122939, in which
instances of a toolbox-generated class are accessed with Scala reflection.
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
Doesn't seem to be inferrable from the API we expose right now
|