Fix #12318: Support options to repl

[liufengyun]

Fix #12318: Support options to repl

[philwalk]

Some things I found while manually verifying each option in the -help message:

The -d <destination-for-generated-files> option is broken, due to -d being treated as -debug in dist/bin/scala line 101.

scala -encoding utf8
Unrecognized option: -encoding

If a script name is provided, the error disappears, so maybe a difference between the repl.Main and scripting.Main.

Also, scripts should immediately capture all remaining arguments as soon as the target_script is initialized (dist\bin\scala near line 138). Otherwise, scripts are unable to have arguments such as -run or -verbose, etc.
This is similar to the a problem recently fixed in dist/bin/scalac (#11776)

[griggt]

The -d <destination-for-generated-files> option is broken, due to -d being treated as -debug in dist/bin/scala line 101.

I believe this has been "broken" for quite a long time, I have been using scalac -d /path/to/output -repl as a workaround for as long as I can remember.

It would be nice if this were fixed such that -d has the same meaning for both scalac and scala.

[michelou]

Fix #12318: Support options to repl

@liufengyun Got your request; I'll have a look.

[philwalk]

Another existing issue with dist/bin scripts that hasn't yet been logged as a bug (AFAIK), is that there's an apparent conflict between jline3 and windows shell environments (cygwin, mingw, msys, etc.). The fix is simple, it would be nice to get it fixed here, if that's acceptable.

The problem can be illustrated with a Windows batch file for launching a REPL session:

@echo off
set SCALA_HOME="c:/opt/scala3"
set CP="%SCALA_HOME%/lib/scala-library-2.13.5.jar;%SCALA_HOME%/lib/scala3-library_3-3.0.1-RC1-bin-SNAPSHOT.jar;%SCALA_HOME%/lib/scala-asm-9.1.0-scala-1.jar;%SCALA_HOME%/lib/compiler-interface-1.3.5.jar;%SCALA_HOME%/lib/scala3-interfaces-3.0.1-RC1-bin-SNAPSHOT.jar;%SCALA_HOME%/lib/scala3-compiler_3-3.0.1-RC1-bin-SNAPSHOT.jar;%SCALA_HOME%/lib/tasty-core_3-3.0.1-RC1-bin-SNAPSHOT.jar;%SCALA_HOME%/lib/scala3-staging_3-3.0.1-RC1-bin-SNAPSHOT.jar;%SCALA_HOME%/lib/scala3-tasty-inspector_3-3.0.1-RC1-bin-SNAPSHOT.jar;%SCALA_HOME%/lib/jline-reader-3.19.0.jar;%SCALA_HOME%/lib/jline-terminal-3.19.0.jar;%SCALA_HOME%/lib/jline-terminal-jna-3.19.0.jar;%SCALA_HOME%/lib/jna-5.3.1.jar"
C:/opt/jdk/bin/java.exe -Xmx768m -Xms768m -classpath %CP% -Dscala.usejavacp=true dotty.tools.repl.Main %*

When launched from a CMD.EXE session, this launches a REPL session, as expected.

But if PWD is set, then the session output is garbled:

C:\opt\ue>s0.bat
←[90m~←[0m←[K                                                                                                     ←[?2004h←[34mscala> ←[0m:q
:q
←[?2004lC:\opt\ue>

The fix is to unset PWD before calls to "$JAVACMD", in bin/scala (see #12405 for a more complete explanation).

This garbled prompt appears in bash sessions as well, with the same fix.

[liufengyun]

Thanks for all your input @philwalk @griggt @michelou .

I think we can make -d to mean the output directory, and remove the debug capability to be consistent with bin/scalac and align with Scala 2 behavior.

For the Windows issue, could you push a fix in another PR @philwalk ?

BTW, it seems it will help a lot if we add more bash tests, which can help us from regressions and save manual testing efforts. Do you have any suggestions for testing the bash REPL (jline3)?

[philwalk]

@liufengyun

Do you have any suggestions for testing the bash REPL (jline3)?

I will give it some thought.

I will start a PR as soon as I can. Which of the problems I described should be included? I notice that scripting_args is referenced in dist/bin/scalac, but is not initialized anywhere. Probably it should be spelled script_args, and it would be relevant if someone calls bin/scalac to compile a script with -compile-only and -script <path> args.

Should I include the fix for the potential misdirected script arguments problem?

scripts should immediately capture all remaining arguments as soon as the target_script is initialized (dist\bin\scala near line 138).

[liufengyun]

Which of the problems I described should be included? I notice that scripting_args is referenced in dist/bin/scalac, but is not initialized anywhere. Should I include the fix for this problem:

scripts should immediately capture all remaining arguments as soon as the target_script is initialized (dist\bin\scala near line 138).

@philwalk Yes, it will be better if the PR includes fix for the problem above. Your help is much appreciated.

[liufengyun]

@philwalk If there is a potential to conflict with the current PR, then you can make a PR on top of this PR. And you can rebase after this is merged.

[michelou]

LGTM. Debug was undocumented anyway.

[liufengyun]

Thank you @michelou !

[philwalk]

PR for console problems is #12411