| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
| |
Deprecated scala.reflect.generic, except for ByteCodecs. After we move
some non-internal things out of scala.reflect.internal we should go back
and improve some deprecation messages (i.e. point somewhere useful.) No
review.
|
| |
|
|
|
|
| |
Suppressing the ability to suppress stack trace suppression in
NonLocalReturnControl for performance reasons, no review.
|
| |
|
|
|
| |
Fix for that slightly not compiling previous patch, no review.
|
| | |
|
| |
|
|
|
| |
Eliminating accumulated dead ends from the pattern matcher. No review.
|
| |
|
|
|
|
|
| |
Triumph over various unsolved mysteries of the pattern matcher. More
elimination of redundant ways of doing things. More allowing the typer
to do its typing thing. No review.
|
| |
|
|
|
|
|
| |
Bytecode inspection reveals the pattern matcher leaving a lot more
redundancy in generated trees than is necessary. Added more inteligence
to the process. Review by dragos.
|
| |
|
|
|
|
|
|
|
|
|
| |
closes #4692: unification in type constructor inference now widens *and*
dealiases when necessary
in 2.8.1 implicit conversion search started with a widened type, so that
combo never came up
no review
|
| |
|
|
|
|
| |
got rid of grotesque isCompat closure in exprTypeArgs and friends. no
review
|
| | |
|
| |
|
|
|
|
|
|
| |
Update manually created tree which was calling Predef.error to call
scala.sys.error instead. Created convenience functions in Definitions
for getting package objects akin to those for getting classes and
objects. No review.
|
| |
|
|
|
|
|
|
|
| |
More batched performance improvements for io.{ File, Classpath } and
others in the neighborhood. Avoids calling the expensive
getCanonicalPath in favor of getAbsolutePath: I note that because
it has the potential to change compiler behavior at the borders.
No review.
|
| |
|
|
|
| |
review by rompf -- odersky may want to take a quick look and update the spec
|
| |
|
|
|
|
|
| |
Documented what I could figure out about late flags and anti-flags.
Eliminated some indirection in flag usage. Made -Xshow-phases print out
more info about flags if -Ydebug is also given. No review.
|
| | |
|
| |
|
|
|
|
|
| |
Test case for implicits which unwrap typeclasses, something which must
really live on the edge given the multiple ways we've busted it lately.
Also some Array/signature and repl tests. No review.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Views using methods implemented in terms of isEmpty (in particular,
headOption and lastOption) were traversing the collection twice up
to the nonEmpty element, because "if (isEmpty) None else Some(head)"
means calling isEmpty separately from head. I overrode those methods in
TraversableViewLike to avoid the second traversal.
This leaves at least init and tail still in that boat, but they were
getting too involved.
How do I say "review by pool of reviewers", who can help set that up? In
the meantime no review.
|
| |
|
|
|
|
|
| |
Marked some more by-name-argument-using methods with final @inline to
make possible and then encourage their inlining. Knocked almost 300 more
classfiles and 200K off the compiler jar. No review.
|
| |
|
|
|
|
| |
In the interests of unbreaking the ide, changing the reporter type in
the repl to ConsoleReporter. No review.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Revert r24960, "Change Types#narrow to create an existential rather than
a refinement type."
For many days I have been haunted by the knowledge that the 20 seconds
I took off quick.comp between r24893 and r24920 all came roaring back
around r24960 for no reason which I could discern. To verify r24960 was
not at fault I had compared it against r24959 more than once, and the
difference was negligible.
It was negligible, that is, until I compared under -optimise, and then
the wheels came off. In fact the differences there are so spectacular
(quick.lib under -optimise goes from 8:50 to 4:26 with this patch,
quick.comp from 8:34 to 6:30) that there must be some major disconnect
between what the benchmark charts are measuring and what I am measuring.
In any case, the charts and I do agree on the direction of the arrow.
For reasons which remain to be determined, this commit was bad on the
nanos. Review by odersky.
|
| |
|
|
|
| |
We migrated the Scala wiki from Trac to Confluence. Review by rytz.
|
| |
|
|
|
| |
We rename Trac and start using JIRA as the issue tracking system.
|
| |
|
|
|
|
| |
The '#' sign may easily to lead to problems, So we use '_' instead of
'#'. Closes #4641.
|
| |
|
|
|
|
|
| |
Disabled the addition to isImpossibleSubType which prevents scalaz from
building with trunk. Not sure what should happen with it; review by
rompf.
|
| |
|
|
|
|
| |
Don't issue unchecked warnings on higher-kinded types. Closes #1439,
review by moors.
|
| | |
|
| |
|
|
|
|
| |
Reverts r25051 due to witnessed movement of scalaz in the noncompiling
direction. Postmortem help I hope in the review by moors.
|
| |
|
|
|
|
|
| |
Informs the bootstrap code which preferences certain source files about
EmbeddedControls.scala, necessary so that starrs from trunk can be used
to build scala-virtualized. No review.
|
| |
|
|
|
| |
A third round of optimizations from Tiark, no review.
|
| |
|
|
|
|
| |
A few additional (but less dramatic) optimizations to implicit search,
courtesy of Tiark. No review.
|
| |
|
|
|
| |
Tiark's main batch of optimizations to implicit search. No review.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Making some recent changes less breaky so sbt can be used with trunk
again. Mainly, renamed scala.reflect.internal's AbstractFile to
AbstractFileType. I can't see a way around renaming it and it's becoming
quite a problem not to have sbt. I inspected scala-ide and I don't think
this is going to break it, but in general I don't know how plausible it
is that we can avoid periodic breakage given that we have at least three
source bases (scala, sbt, scala-ide) each with multiple branches with
interdependencies which are not being continuously integrated. Review by
odersky.
|
| |
|
|
|
|
| |
A test case demonstrating some of the issues with DelayedInit.
References #4680. Review by odersky.
|
| |
|
|
|
|
|
| |
Compilation of spec-List enters an infinite loop under -optimise,
disabling in the hopes of seeing a new build before I die. I'll put it
back. No review.
|
| |
|
|
|
|
|
|
|
| |
Tweaks repl tests to enforce -Yrepl-sync and other setup on all clients.
Also includes some more why-do-tests-hang bosons for the particle
accelerator in case that doesn't wrap it up. I think it will though,
because now constrained-types is the only one which failed and it's also
the only one which overrides Settings. No review.
|
| |
|
|
|
|
| |
Reverts r25036, "Altered the positioning of XML literal trees" because
the IDE did not like it. No review.
|
| |
|
|
|
|
|
|
| |
A somewhat more realistic attempt to fix the build, no review. This
introduces a repl command line option -Yrepl-sync to inhibit the
asynchronous path which makes repl startup seem so snappy. And then it
uses it in the repl tests.
|
| |
|
|
|
| |
First of what might be several attempts to fix the build, no review.
|
| |
|
|
|
|
|
|
|
|
|
| |
Modified erasure not to generate instance tests for statically known
types. It appears the production of types like "Foo with Bar" in the
pattern matcher (where the scrutinee is known to be Foo) has been a
major contributor of suboptimal pattern matches. I will also fix it in
the matcher, but it makes sense to catch it in erasure as both a check
on the matcher and because they may come from elsewhere too. Review by
odersky.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Proliferating the number of debugging modes because it's still way
too hard to see what's going on in there. Until we get hubert's type
debugger with its whiz-bang whizbanginess, we'll have to struggle along
with somewhat prettier ascii. This introduces:
-Yinfer-debug
which tries to print in readable fashion what is happening in the
worlds of inference and implicit search. It should be made a bit more
complementary and less overlappy with -Ytyper-debug. No review.
|
| |
|
|
|
|
| |
Realized most of the remaining methods in Option warrant @inline and
final treatment and modified them accordingly. Review by dragos.
|
| |
|
|
|
|
|
|
| |
Working on the inliner, discovered that a few key methods are not
inlined. With this commit there are 326 fewer classfiles generated
under -optimise. Use getOrElse with even wilder abandon than you were
previously. No review.
|
| |
|
|
|
|
| |
Carved out access exception for java protected statics, which otherwise
cannot be accessed from scala. Changes close status of #1806, no review.
|
| |
|
|
|
|
| |
A few more items so that all may enjoy power mode in the way nature
intended. No review.
|
| | |
|
| |
|
|
|
| |
down into more flexible pieces. No review.
|
| |
|
|
|
|
| |
More adjustments to repl parsing to accomodate inaccurately positioned
parse trees. No review.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The eagerly awaited companion commit to r23622, the surprise hit of last
winter. Singletons now have the singleton type inferred in monomorphic
contexts as well. In english:
scala> object X
defined module X
scala> def f = X
f: X.type
Wait, that's not english. English review by moors.
|
| |
|
|
|
|
|
|
| |
Fixed a repl regression with parentheses handling, reminding me we
really need that honest parser phase which doesn't betray us with
parentheses abandonment and dramatic desugarings. I'll promote it from
page 14 to page 11. Closes #4661, no review.
|
