summaryrefslogtreecommitdiff
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Fix 27 typos (p-r)Janek Bogucki2015-06-3014-17/+17
|
* Merge pull request #4569 from retronym/ticket/6985Adriaan Moors2015-06-301-2/+3
|\ | | | | SI-6895 Test cases to explain the limitations in tcpoly inference
| * SI-8892 Fix incorrect qualification in error messageJason Zaugg2015-06-231-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Since beadafa2, we've ended up with nutty error messages for type errors involving aliases that target types in `java.lang` and `scala` packages. I think the intent of that change was to force the error messages to qualify types like `String` when needed, but to leave them unqualified by default. However, this led to this flat out wrong message in the enclosed test. found : B required: C.this.java.lang.B (which expands to) String I've changed the heuristic slightly limit this code to aliases that are eponymous with their targets. Still feels pretty kludgy, but we can at least book a little progress.
| * SI-8777 Avoid redundant disambiguation in error messagesJason Zaugg2015-06-231-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When printing types in error messages, we attempt to disambiguate symbol names by qualifying them in various ways. Type paramters symbols are qualified by adding `(in someMethod)`. However, the type errors generated by higher kinded subtyping can contain clones of type parameter symbols, as creater in `isPolySubType`. The disambiguation tries fruitlessly to distinguish them but ended up adding the same suffix to both names repeatedly. ``` found : [F[_]]Foo[[X(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)]Bar[F,X(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)]] required: Foo[[X(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)]Bar[[X]Or[String,X],X(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)(in type L)]] ``` This commit limits the post qualification of type parameter symbols to a single attempt to limit the damage. An alternative might be to mark a clone (we could determine its status by checking whether it is a type parameter of its owner.) But I'm not sure how to present this information in a comphrenensible way, so for now I'm limiting my ambitions to stopping the stutter.
* | Merge pull request #4591 from vsalvis/vsalvis-stringdocSeth Tisue2015-06-302-4/+8
|\ \ | | | | | | SI-8140 Documentation references java.lang.String directly
| * | SI-8140 Documentation references java.lang.String directlyvsalvis2015-06-292-4/+8
| | |
* | | Merge pull request #4588 from retronym/topic/icodishAdriaan Moors2015-06-291-0/+5
|\ \ \ | | | | | | | | GenBCode: fix incrementatal compilation by mimicing GenASM
| * | | GenBCode: fix incrementatal compilation by mimicing GenASMJason Zaugg2015-06-281-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The incremental compiler in SBT uses an implementation detail of the compiler backend to enumerate the classes that are actually written to disk. This commit mimics this in GenBCode by populating `Run#icode` with an `IClass` for each `ClassDef` processed. We should revisit this by creating a dedicated API for this purpose and migrating SBT to use that. We should also revisit this code as we implement closure elimination in the GenBCode; this commit assumes that all `ClassDef`s that enter the backend will generate classfile products. The enclosed test is extracted from the incrementatl compiler. I've also manually integration tested this with SBT: https://gist.github.com/retronym/fabf6f92787ea9c1ce67
* | | | Merge pull request #4586 from som-snytt/issue/missing-argsSeth Tisue2015-06-291-19/+18
|\ \ \ \ | | | | | | | | | | Improved message for missing argument list
| * | | | Improved message for missing argument listSom Snytt2015-06-271-19/+18
| | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Clarifies the language and rules for eta-expansion. A missing argument in a list, as opposed to a missing argument list, results in a different message. The comical expansion in parens does not attempt to show what was already applied, but succeeds in showing at a glance the shape of the method in question. ``` scala> def m(i: Int, j: Int)(x: Int) = ??? m: (i: Int, j: Int)(x: Int)Nothing scala> m <console>:12: error: missing argument list for method m Unapplied methods are only converted to functions when a function type is expected. You can make this conversion explicit by writing `m _` or `m(_,_)(_)` instead of `m`. m ^ ``` The original submission was due to sschaef and the wording due to adriaanm, with a minor tweak.
* | | | Merge pull request #4589 from janekdb/2.11.x-method-names-m-oJason Zaugg2015-06-291-2/+2
|\ \ \ \ | | | | | | | | | | Improve method names (m-o)
| * | | | Improve method names (m-o)Janek Bogucki2015-06-281-2/+2
| |/ / /
* / / / Fix 23 typos (m-o)Janek Bogucki2015-06-288-10/+10
|/ / /
* | | Merge pull request #4577 from janekdb/2.11.x-typos-j-lSeth Tisue2015-06-244-7/+7
|\ \ \ | |/ / |/| | Fix 8 typos (j-l)
| * | Fix 8 typos (j-l)Janek Bogucki2015-06-234-7/+7
| | |
* | | Merge pull request #4552 from lrytz/opt/closureInliningJason Zaugg2015-06-2413-113/+1005
|\ \ \ | |/ / |/| | Closure elimination for new backend
| * | Cast arguments where necessary before rewriting closure invocationsLukas Rytz2015-06-231-15/+70
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The parameter types of a closure invocation can be supertypes of the parameter types in the implementation method. The closure automatically casts arguments to the right type. We need to do the same when rewriting closure invocations. Example: val fun: String => String = l => l val l = List("") fun(l.head) The closure object calls `apply(Object)Object`. The body method takes an argument of type `String`.
| * | Rewrite closure invocations to the lambda body methodLukas Rytz2015-06-2210-108/+489
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When an indylambda closure is allocated and invoked within the same method, rewrite the invocation to the implementation method. This works for any indylambda / SAM type, not only Scala functions. However, the Scala compiler (under -Xexperimental) currently desugars function literals for non-FunctionN types to an anonymous class during typer. No testing yet, waiting for FunctionN to become SAMs first. The feature requires scala-java8-compat to be on the classpath and a number of compiler flags: -Ydelambdafy:method -Ybackend:GenBCode -Yopt:closure-elimination -target:jvm-1.8 ➜ scala git:(opt/closureInlining) ant -Dscala-java8-compat.package=1 -Dlocker.skip=1 ➜ scala git:(opt/closureInlining) cd sandbox ➜ sandbox git:(opt/closureInlining) cat Fun.java public interface Fun<T> { T apply(T x); } ➜ sandbox git:(opt/closureInlining) javac Fun.java ➜ sandbox git:(opt/closureInlining) cat Test.scala class C { val z = "too" def f = { val kap = "me! me!" val f: Tuple2[String, String] => String = (o => z + kap + o.toString) f(("a", "b")) } def g = { val f: Int => String = x => x.toString f(10) } def h = { val f: Fun[Int] = x => x + 100 // Java SAM, requires -Xexperimental, will create an anonymous class in typer f(10) } def i = { val l = 10l val f: (Long, String) => String = (x, s) => s + l + z + x f(20l, "n") } def j = { val f: Int => Int = x => x + 101 // specialized f(33) } } ➜ sandbox git:(opt/closureInlining) ../build/quick/bin/scalac -target:jvm-1.8 -Yopt:closure-elimination -Ydelambdafy:method -Ybackend:GenBCode -Xexperimental -cp ../build/quick/scala-java8-compat:. Test.scala ➜ sandbox git:(opt/closureInlining) asm -a C.class ➜ sandbox git:(opt/closureInlining) cat C.asm [...] public g()Ljava/lang/String; L0 INVOKEDYNAMIC apply()Lscala/compat/java8/JFunction1; [ // handle kind 0x6 : INVOKESTATIC java/lang/invoke/LambdaMetafactory.altMetafactory(Ljava/lang/invoke/MethodHandles$Lookup;Ljava/lang/String;Ljava/lang/invoke/MethodType;[Ljava/lang/Object;)Ljava/lang/invoke/CallSite; // arguments: (Ljava/lang/Object;)Ljava/lang/Object;, // handle kind 0x6 : INVOKESTATIC C.C$$$anonfun$2$adapted(Ljava/lang/Object;)Ljava/lang/String;, (Ljava/lang/Object;)Ljava/lang/String;, 3, 1, Lscala/Serializable;.class, 0 ] CHECKCAST scala/Function1 L1 ASTORE 1 L2 ALOAD 1 BIPUSH 10 INVOKESTATIC scala/runtime/BoxesRunTime.boxToInteger (I)Ljava/lang/Integer; ASTORE 2 POP ALOAD 2 INVOKESTATIC C.C$$$anonfun$2$adapted (Ljava/lang/Object;)Ljava/lang/String; CHECKCAST java/lang/String L3 ARETURN [...]
| * | Producers / Consumers AnalysisLukas Rytz2015-06-223-5/+461
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ASM has a built-in `SourceValue` analysis which computes for each value a set of instructions that may possibly have constructed it. The ProdConsAnalyzer class provides additional queries over the result of the SourceValue analysis: - consumers of values - tracking producers / consumers through copying operations (load, store, etc) A fix to (and therefore a new version of) ASM was required. See here: https://github.com/scala/scala-asm/commit/94106a5472
* | | Merge pull request #4574 from janekdb/2.11.x-typos-g-iJason Zaugg2015-06-2310-10/+10
|\ \ \ | | | | | | | | Fix 25 typos (g-i)
| * | | Fix 25 typos (g-i)Janek Bogucki2015-06-2210-10/+10
| |/ /
* | | Merge pull request #4564 from som-snytt/issue/promptv2.11.7Adriaan Moors2015-06-228-101/+109
|\ \ \ | | | | | | | | SI-9206 Fix REPL code indentation
| * | | SI-9206 Local refactor to save eyesightSom Snytt2015-06-212-42/+39
| | | | | | | | | | | | | | | | | | | | | | | | We talk about bit rot but not about how dust accumulates on code that hasn't been swept since the last time the furniture was moved around.
| * | | SI-9206 Accept paste with custom promptSom Snytt2015-06-213-24/+28
| | | | | | | | | | | | | | | | But sans test.
| * | | SI-9206 Verbose REPL prompt for info modeSom Snytt2015-06-191-6/+7
| | | | | | | | | | | | | | | | | | | | | | | | Only for exactly `-Dscala.repl.info`, include the complete version number string in the REPL prompt. One could imagine this is the mode for posting snippets to stackoverflow.
| * | | SI-9206 Fix REPL code indentationSom Snytt2015-06-195-33/+30
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | To make code in error messages line up with the original line of code, templated code is indented by the width of the prompt. Use the raw prompt (without ANSI escapes or newlines) to determine the indentation. Also, indent only once per line.
| * | | SI-9206 REPL prompt is more easily configuredSom Snytt2015-06-192-12/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The scala shell prompt can be provided as either a system property or in compiler.properties. The prompt string is taken as a format string with one argument that is the version string. ``` $ scala -Dscala.repl.prompt="%nScala %s> " Welcome to Scala version 2.11.7-20150616-093756-43a56fb5a1 (Java HotSpot(TM) 64-Bit Server VM, Java 1.8.0_45). Type in expressions to have them evaluated. Type :help for more information. Scala 2.11.7-20150616-093756-43a56fb5a1> 42 res0: Int = 42 Scala 2.11.7-20150616-093756-43a56fb5a1> :quit ```
| * | | SI-9206 BooleanProp if set and not untrueSom Snytt2015-06-181-3/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously, handy `sys.BooleanProp.keyExists` ignored the property value. While trying not to make any real estate puns, this commit will let it go false if a value is supplied that is not true in the usual Java sense. But what is truth? Allows `scala -Dscala.color=off`, for example.
* | | | Merge pull request #4566 from lrytz/t9359Adriaan Moors2015-06-227-56/+85
|\ \ \ \ | | | | | | | | | | SI-9359 Fix InnerClass entry flags for nested Java enums
| * | | | SI-9359 Fix InnerClass entry flags for nested Java enumsLukas Rytz2015-06-197-56/+85
| |/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The access flags in InnerClass entries for nested Java enums were basically completely off. A first step is to use the recently introduced backend method `javaClassfileFlags`, which is now moved to BCodeAsmCommon. See its doc for an explanation. Then the flags of the enum class symbol were off. An enum is - final if none of its values has a class body - abstract if it has an abstract method (https://docs.oracle.com/javase/specs/jls/se7/html/jls-8.html#jls-8.9) When using the ClassfileParser: - ENUM was never added. I guess that's just an oversight. - ABSTRACT (together with SEALED) was always added. This is to enable exhaustiveness checking, see 3f7b8b5. This is a hack and we have to go through the class members in the backend to find out if the enum actually has the `ACC_ABSTRACT` flag or not. When using the JavaParser: - FINAL was never added. - ABSTRACT was never added. This commit fixes all of the above and tests cases (Java enum read from the classfile and from source).
* | / / SI-9253 avoid IndexOutOfBoundsException in TypeMaps.correspondingTypeArgumentBruno Bieth2015-06-221-1/+1
| |/ / |/| |
* | | Fix 36 typos (d-f)Janek Bogucki2015-06-2116-18/+18
| |/ |/|
* | Merge pull request #4563 from adriaanm/jline-shadeAdriaan Moors2015-06-1915-273/+334
|\ \ | | | | | | Corral, shade & embed jline.
| * | SI-9339 Support classpaths with no single compatible jlineAdriaan Moors2015-06-185-11/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As usual, the repl will use whatever jline 2 jar on the classpath, if there is one. Failing that, there's a fallback and an override. If instantiating the standard `jline.InteractiveReader` fails, we fall back to an embedded, shaded, version of jline, provided by `jline_embedded.InteractiveReader`. (Assume `import scala.tools.nsc.interpreter._` for this message.) The instantiation of `InteractiveReader` eagerly exercises jline, so that a linkage error will result if jline is missing or if the provided one is not binary compatible. The property `scala.repl.reader` overrides this behavior, if set to the FQN of a class that looks like `YourInteractiveReader` below. ``` class YourInteractiveReader(completer: () => Completion) extends InteractiveReader ``` The repl logs which classes it tried to instantiate under `-Ydebug`. # Changes to source & build The core of the repl (`src/repl`) no longer depends on jline. The jline interface is now in `src/repl-jline`. The embedded jline + our interface to it are generated by the `quick.repl` target. The build now also enforces that only `src/repl-jline` depends on jline. The sources in `src/repl` are now sure to be independent of it, though they do use reflection to instantiate a suitable subclass of `InteractiveReader`, as explained above. The `quick.repl` target builds the sources in `src/repl` and `src/repl-jline`, producing a jar for the `repl-jline` classes, which is then transformed using jarjar to obtain a shaded copy of the `scala.tools.nsc.interpreter.jline` package. Jarjar is used to combine the `jline` jar and the `repl-jline` into a new jar, rewriting package names as follows: - `org.fusesource` -> `scala.tools.fusesource_embedded` - `jline` -> `scala.tools.jline_embedded` - `scala.tools.nsc.interpreter.jline` -> `scala.tools.nsc.interpreter.jline_embedded` Classes not reachable from `scala.tools.**` are pruned, as well as empty dirs. The classes in the `repl-jline` jar as well as those in the rewritten one are copied to the repl's output directory. PS: The sbt build is not updated, sorry. PPS: A more recent fork of jarjar: https://github.com/shevek/jarjar.
| * | Centralize dependencies on jlineAdriaan Moors2015-06-1715-273/+314
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Code that depends on jline is now in package `scala.tools.nsc.interpreter.jline`. To make this possible, remove the `entries` functionality from `History`, and add the `historicize` method. Also provide an overload for `asStrings`. Clean up a little along the way in `JLineHistory.scala` and `JLineReader.scala`. Next step: fall back to an embedded jline when the expected jline jar is not on the classpath. The gist of the refactor: https://gist.github.com/adriaanm/02e110d4da0a585480c1
* | | Merge pull request #4331 from Ichoran/issue/8930Adriaan Moors2015-06-191-5/+12
|\ \ \ | |_|/ |/| | SI-8930 - Vector updated, +:, and :+ slow when typed as Seq[A]
| * | SI-8930 - Vector updated, +:, and :+ slow when typed as Seq[A]Rex Kerr2015-06-181-5/+12
| | | | | | | | | | | | | | | | | | Vector was intercepting only the IndexedSeq CanBuildFrom to quickly generate new vectors. Now it intercepts immutable.Seq and collection.Seq as well. There are other possibilities (collection.IndexedSeq), but they will probably arise rarely, and to avoid an absurdly long set of checks we would need a marker trait (that is not binary compatible).
* | | Merge pull request #4559 from janekdb/2.11.x-scaladoc-2Seth Tisue2015-06-1811-17/+17
|\ \ \ | | | | | | | | Fix some typos (a-c)
| * | | Fix another several typosMichał Pociecha2015-06-185-7/+7
| | | | | | | | | | | | | | | | | | | | | | | | I just used text search to check whether there are no more typos like these corrected by janekdb, and by the way fixed also some other ones which I saw.
| * | | Fix some typos (a-c)Janek Bogucki2015-06-189-10/+10
| | | |
* | | | Merge pull request #4529 from lrytz/inlineAccessibilityJason Zaugg2015-06-192-18/+74
|\ \ \ \ | | | | | | | | | | Fix illegal inlining of instructions accessing protected members
| * | | | Fix illegal inlining of instructions accessing protected membersLukas Rytz2015-05-282-18/+74
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There were two issues in the new inliner that would cause a VerifyError and an IllegalAccessError. First, an access to a public member of package protected class C can only be inlined if the destination class can access C. This is tested by t7582b. Second, an access to a protected member requires the receiver object to be a subtype of the class where the instruction is located. So when inlining such an access, we need to know the type of the receiver object - which we don't have. Therefore we don't inline in this case for now. This can be fixed once we have a type propagation analyis. https://github.com/scala-opt/scala/issues/13. This case is tested by t2106. Force kmpSliceSearch test to delambdafy:inline See discussion on https://github.com/scala/scala/pull/4505. The issue will go away when moving to indy-lambda.
* | | | | Merge pull request #4543 from som-snytt/issue/7773Seth Tisue2015-06-181-10/+12
|\ \ \ \ \ | | | | | | | | | | | | SI-7773 Restore phase id to icode filename
| * | | | | SI-7773 Restore phase id to icode filenameSom Snytt2015-06-071-10/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Having icode output files of the form `X-24.icode` went in and out of style using the long-form phase name because it broke the windows nightly build somehow. Here's hoping using just the phase id works on this year's infrastructure. As previously, the long name is still available under `-Ydebug`, because why not debug.
* | | | | | Merge pull request #4539 from vsalvis/vsalvis-docfixesSeth Tisue2015-06-186-31/+12
|\ \ \ \ \ \ | | | | | | | | | | | | | | Doc fixes
| * | | | | | SI-6131 doc: remove broken show member link from all scaladoc pagesvsalvis2015-06-171-1/+0
| | | | | | |
| * | | | | | SI-8543 doc: Move TODO out of NumericRange's scaladocvsalvis2015-06-171-4/+4
| | | | | | |
| * | | | | | SI-8858 doc: fix note about PartialFunction in Function0, F1 and F2vsalvis2015-06-174-26/+8
| |/ / / / /
* | | | | | Merge pull request #4553 from som-snytt/issue/implicit-docSeth Tisue2015-06-182-2/+12
|\ \ \ \ \ \ | |_|_|/ / / |/| | | | | SI-9354 ScalaDoc members added via by-name view
| * | | | | SI-9354 ScalaDoc members added via by-name viewSom Snytt2015-06-142-2/+12
| |/ / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Eligible views were looked up by exact from type without including the by-name dodge. By-name views are now included without consideration whether ScalaDoc processes possible duplicates correctly.
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-23 14:39:04 +0000