| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
Reverting a couple replacements from that last patch which don't look so
safe on re-examination. No review.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Leveraged -Xmigration to burn off some warts which arose in the new
collections. Warnings put in place for behavioral changes, allowing the
following. 1) Buffers: create new collections on
++ and -- like all the other collections. 2) Maps: eliminated
never-shipped redundant method valuesIterable and supplied these
return types:
def keys: Iterable[A]
def keysIterator: Iterator[A]
def values: Iterable[B]
def valuesIterator: Iterator[B]
def keySet: Set[A]
I concluded that keys should return Iterable because keySet also exists
on Map, and is not solely in the province of Maps even if we wanted to
change it: it's defined on Sorted and also appears in some Sets. So it
seems sensible to have keySet return a Set and keys return the more
general type.
Closes #3089, #3145. Review by odersky.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
More support code for the big partest patch I'm working on to finally
finish classpaths for good. No review.
|
|
|
|
|
| |
Improved version where bumping and zero-encoding are rolled into one.
|
| |
|
|
|
|
|
|
| |
Added an object to mangle byte arrays into Java classfile's version of
UTF8.
|
| |
|
|
|
|
|
|
|
| |
do not set the type of the implementation method to be the type of the
original one as this is done properly in cloneSymbol. no review (already
done by Martin)
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
implemented handling of 32-bit collisions in immutable.HashMap. review
by community.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
Small syntactic adjustment so that last patch from adriaan will build.
(Big thumsb up to the aesthetics though.) No review.
|
|
|
|
|
| |
slight (syntactic) cleanup of patch for see #3152 -- sorry, only realised when looking over my patch again
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
closes #3152: refactored adjustTypeArgs and methTypeArgs so that tparams
are correctly split into ones that were inferred successfully, and
that thus have a corresponding type argument, and those that weren't
determined
I didn't investigate the exact cause of the final error message in the
bug report, but Jason Zaugg's observations seems correct and I never
liked that uninstantiated buffer in the first place.
review by odersky
|
| |
|
|
|
|
|
|
| |
...and a line from partest I didn't notice the absence of which would
break the build. No review.
|
|
|
|
|
|
| |
Some minor compiler support bits for my upcoming partest patch. No
review.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
tests you have placed in the test/debug directories - critically
for those of us stuck debugging ant, this lets one run a small
selection of tests by way of ant instead of the console runner.
(Sorry about the empty .gitignore files, but one of git's quirks is that
it won't acknowledge the existence of an empty directory.) No review.
|
|
|
|
|
|
| |
Changed partest ant task not to use reflection, instead using the path
with which scala was invoked. No review.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
make normalize slightly more aggressive in loading symbol info, while )
tolerating the righteous cycle (use sym.info.typeParameters instead )
of unsafeParams this is needed to make sure higher-kinded types have )
their type parameters (otherwise we'd get a PolyType with NoSymbol for )
typeParams )
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
Recursively transform 'new' arguments in specialized programs. Closes
#3149, no review.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Created directory for code which is most likely dead but we want to keep
around a while in case someone else is using it. It's called src/attic
and now it holds four files. Motivations: such files cloud my attempts
to figure out what code in the compiler is really being used, they
require effort to maintain across changes, and they slow down every
build a fraction. Revew by community.
|
|
|
|
|
|
| |
More progress toward immutable Settings, and various cleanups
encountered along the way. No review.
|
|
|
|
|
|
|
|
|
| |
- new immutable HashMap implementation based on a hash trie. this is the
first iteration, more optimizations will be added later. - updated test
cases to reflect new ordering of elements - made Map.empty and Set.empty
singletons, deprecating classes Map.EmptyMap and Set.EmptySet Review by
extempore, odersky.
|
|
|
|
|
|
| |
Removed unnecessary DebugSetting, folding the small extra functionality
back into ChoiceSetting. No review.
|
|
|
|
|
|
| |
Still working my way through all the classpath manipulations in partest.
No review.
|
|
|
|
|
|
|
|
|
| |
One minute too many trying to figure out where some partest classpath
mutation was disappearing on me, and I snapped and started the process
of creating an immutable Settings. This commit is for the most part
infrastructure to enable its smooth and uneventful entrance. Review by
community.
|