| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
|
| |
closes #2750: cooking java raw types in info of type parameters of Java
classes review by: odersky
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
[This patch submitted by ismael juma - commit message his words,
but condensed.]
Fix ticket #1600: Serialization and deserialization of hash-based
collections should not re-use hashCode. The collection is rebuilt
on deserialization - note that this is not compatible with the previous
serialization format. All @SerialVersionUIDs have been reset to 1.
WeakHashMap is not Serializable and should not be so. TreeHashMap has
not been reintegrated yet. OpenHashMap has not been updated. (I think
this collection is flawed and should be removed or reimplemented.)
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
tests in pending for see #2705 -- note that partest does not work for jvm tests under pending, move to files and it works, see #2711
|
| |
|
|
|
|
|
|
| |
Fixed test cases to the new method names (indexOfSeq -> indexOfSlice).
See r19892
|
| |
|
| |
|
|
|
|
|
|
| |
quick fix for #2673: relaxed zipped on Tuple2/Tuple3 (back to using A =>
B for constraint instead of A <%< B)
|
|
|
|
|
|
|
|
|
| |
closed #2624: instantiate type params after eta expansion of polymorphic
method does not blow the stack on pos/t0674 (previous private fix did
typed(tree); instantiate(tree) instead of instantiate(typed(tree)))
fixed check file for neg/bug608 -- got better error message
|
|
|
|
|
| |
Disabled docgenerator test, which isn't compatible with Scaladoc 2.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
close #2665 and close #2667: use weak conformance in polymorphic case of
isApplicable reviewed by: odersky
exprTypeArgs now takes a comparison function: isWeaklyCompatible is )
passed in isApplicable's typesCompatible (to mimic what happens in the )
monomorphic case Martin: please review as this is different from my )
original proposal (that one broke type inference, this one passes all )
tests and does not slow down quick.comp )
|
| |
|
| |
|
|
|
|
|
|
| |
Priority queue reverse is undefined - overriden to throw an exception.
Reverse iterator seems to have sense - it is overriden and is defined, and some methods in SeqLike are implemented in terms of it.
|
| |
|
| |
|
|
|
|
|
| |
Applied performance patch and test case from ijuma; closes #2526.
|
| |
|
| |
|
|
|
|
|
|
|
| |
closes #2585: generate more precise Java generic signatures for classes
nested in parametric outer classes fix based on review by Martin
baseType is your friend
|
| |
|
|
|
|
|
|
| |
Commented out anonymous function invocation for which Hudson was
complaining it couldn't find the classdef.
|
|
|
|
|
| |
Changes made in the clean up phase - now the symbols get interned during classload for each symbol literal - references to them reside in static fields. These static fields get initialized in static constructors - the java backend will now identify ctors with static flags and generate a static initializer containing the necessary code.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
Simplifiations in collections libraries, enabled by introduction of Self
type in TraversableLike.
|
| |
|
|
|
|
|
|
| |
Finally completed the incredibly tedious task of removing the lower case
primitive aliases from Predef. Had to rebuild msil.jar along the way.
|
|
|
|
|
|
| |
since 2.7.2 (still except for lower case primitive type aliases) and
removes every deprecated method which has never shipped in a release.
|
|
|
|
|
|
| |
Allow implicit modifier on single-parameter function literals. Fixes and
closes #1492.
|
|
|
|
|
|
| |
Contents of scala.Math moved into scala.math package object, and
scala.Math deprecated. Also a couple janitorial cleanups.
|
| |
|
| |
|
| |
|
|
|
|
|
| |
review requests can now be created outside a checkout
|
|
|
|
|
|
| |
Tweaked a test which has been regularly failing due to heap exhaustion,
although what it's supposed to be testing is stack utilization.
|
| |
|
|
|
|
|
| |
Fixed #2848 and #2630; Improvements in equality speed
|
|
|
|
|
|
|
| |
fixing a long-standing bug in fjbg and recompiling fjbg.jar, which had
the side effect of revealing that the current fjbg jar had never been
recompiled with target 1.5, so now it's smaller and (I imagine) faster.
|