Issue on output when using javac or scalac under emacs

Issue on output when using javac or scalac under emacs

[Luc Henninger]

[-- Attachment #1: trace_w32 --]
[-- Type: application/octet-stream, Size: 1401 bytes --]

[-- Attachment #2: trace_nox --]
[-- Type: application/octet-stream, Size: 1363 bytes --]

[-- Attachment #3: trace_nt --]
[-- Type: application/octet-stream, Size: 767 bytes --]

-*- mode: compilation; default-directory: "~/Projects/Java_1.7/src/synthetiseurC/" -*-
Compilation started at Tue Aug 25 23:17:03

make 
Java initialisé avec 'Trace.java MainSynthetiseur.java '
../../bin/synthetiseurC/Trace.class: due to Trace.java ../../bin/kawa/util/Tracer.class
rm -f ../../bin/synthetiseurC/Trace*.class
/cygdrive/c/Program\ Files/Java/jdk7/bin/javac -classpath ../../bin -sourcepath ../../src -g -d "../../bin" -deprecation -Werror -Xlint:all -Xlint:-serial -encoding UTF-8  Trace.java
Trace.java:7: error: cannot find symbol
  A bug;
  ^
  symbol:   class A
  location: class Trace
1 error
make: *** [../../../makeJavaAll.mk:192 : ../../bin/synthetiseurC/Trace.class] Erreur 1

Compilation exited abnormally with code 2 at Tue Aug 25 23:17:05

RE: [cygwin] Issue on output when using javac or scalac under emacs

[Jason Pyeron]

> -----Original Message-----
> From: Luc Henninger
> Sent: Tuesday, August 25, 2020 6:40 PM
> 

I'm thick - what the issue? There was no body to the email.

Re: [cygwin] Issue on output when using javac or scalac under emacs

[Brian Inglis]

On 2020-08-25 17:19, Jason Pyeron wrote:
>> -----Original Message-----
>> From: Luc Henninger
>> Sent: Tuesday, August 25, 2020 6:40 PM
>>
> 
> I'm thick - what the issue? There was no body to the email.

There were three attachments with javac complaining about a missing symbol, but
it is unclear under what conditions which product produced each of the outputs,
what output was expected, and what was the issue with what was actually output?

-- 
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

This email may be disturbing to some readers as it contains
too much technical detail. Reader discretion is advised.
[Data in IEC units and prefixes, physical quantities in SI.]

Re: [cygwin] Issue on output when using javac or scalac under emacs

[Luc Henninger]

Le 26/08/2020 à 01:19, Jason Pyeron a écrit :
>> -----Original Message-----
>> From: Luc Henninger
>> Sent: Tuesday, August 25, 2020 6:40 PM
>>
> I'm thick - what the issue? There was no body to the email.
>
>
Hello,

( My previous message dos not contain the text, only the attachments!!! )

For years, I use emacs-nt (the one provided by gnu.org) under cygwin. I 
just discover that emacs is also directly provided by cygwin (late is 
better than never :-)). So I try to use it successfully up to a 
compilation error in one of my program in java.

To compile java (or scala), I use my own makefile that I start from emacs.

In case of error, I encountered a strange output undex emacs_w32 and 
emacs_nox compared to emacs_nt (see trace_nox, trace_w32 and trace_nt in 
attachment).

I currently use emacs release 27.1 for all emacses, but I verify that I 
have the same strange behaviour with emacs 26.3.

Can anybody explain this? and more gane give a workaround?

Regards

-- 
Luc.Henninger@orange.fr    +33 6 32 96 32 27


-- 
L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast.
https://www.avast.com/antivirus

Re: [cygwin] Issue on output when using javac or scalac under emacs

[Takashi Yano]

On Wed, 26 Aug 2020 08:31:35 +0200
Luc Henninger wrote:
> Le 26/08/2020 à 01:19, Jason Pyeron a écrit :
> >> -----Original Message-----
> >> From: Luc Henninger
> >> Sent: Tuesday, August 25, 2020 6:40 PM
> >>
> > I'm thick - what the issue? There was no body to the email.
> >
> >
> Hello,
> 
> ( My previous message dos not contain the text, only the attachments!!! )
> 
> For years, I use emacs-nt (the one provided by gnu.org) under cygwin. I 
> just discover that emacs is also directly provided by cygwin (late is 
> better than never :-)). So I try to use it successfully up to a 
> compilation error in one of my program in java.
> 
> To compile java (or scala), I use my own makefile that I start from emacs.
> 
> In case of error, I encountered a strange output undex emacs_w32 and 
> emacs_nox compared to emacs_nt (see trace_nox, trace_w32 and trace_nt in 
> attachment).

I'm not sure what you meant by emacs_w32, but if you start make by
M-x compile, I guess
env CYGWIN=disable_pcon emacs
will solve the issue.

> I currently use emacs release 27.1 for all emacses, but I verify that I 
> have the same strange behaviour with emacs 26.3.
> 
> Can anybody explain this? and more gane give a workaround?

Cygwin 3.1.x and later enables pseudo console for non-cygwin apps.
Pseudo console generates some ANSI escape sequences if non-cygwin
apps are executed. However, emacs provides dumb terminal for M-x
compile, M-x shell or M-x eshell. As a result, the problem occurs.

CYGWIN=disable_pcon disables pseudo console, so it will resolve
the issue.

-- 
Takashi Yano <takashi.yano@nifty.ne.jp>

Re: [cygwin] Issue on output when using javac or scalac under emacs solved

[Luc Henninger]

Very thanks to Takashi Yano

CYGWIN=disable_pcon solved my problem

--

Luc.Henninger@orange.fr    +33 6 32 96 32 27


--
L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast.
https://www.avast.com/antivirus

Re: [cygwin] Issue on output when using javac or scalac under emacs solved

[Eliot Moss]

On 8/26/2020 6:09 AM, Luc Henninger wrote:
> Very thanks to Takashi Yano
> 
> CYGWIN=disable_pcon solved my problem

Aha!  As I thought - something about the terminal!

Best wishes - EM

Re: [cygwin] Issue on output when using javac or scalac under emacs

[Eliot Moss]

On 8/26/2020 2:31 AM, Luc Henninger wrote:
> Le 26/08/2020 à 01:19, Jason Pyeron a écrit :
>>> -----Original Message-----
>>> From: Luc Henninger
>>> Sent: Tuesday, August 25, 2020 6:40 PM
>>>
>> I'm thick - what the issue? There was no body to the email.
>>
>>
> Hello,
> 
> ( My previous message dos not contain the text, only the attachments!!! )
> 
> For years, I use emacs-nt (the one provided by gnu.org) under cygwin. I just discover that emacs is 
> also directly provided by cygwin (late is better than never :-)). So I try to use it successfully up 
> to a compilation error in one of my program in java.
> 
> To compile java (or scala), I use my own makefile that I start from emacs.
> 
> In case of error, I encountered a strange output undex emacs_w32 and emacs_nox compared to emacs_nt 
> (see trace_nox, trace_w32 and trace_nt in attachment).
> 
> I currently use emacs release 27.1 for all emacses, but I verify that I have the same strange 
> behaviour with emacs 26.3.
> 
> Can anybody explain this? and more gane give a workaround?

Here's a thought.  Emacs NT would be a native Windows build of Emacs, so is definitely organized to 
interact properly with Windows subprocesses.  The Cygwin versions are designed to interact most 
easily with Cygwin subprocesses.  They can probably be made to work with Windows subprocesses, but 
as you found, there is at least one rough edge here.  My first guess would be something to do with 
line endings as they are conveyed back to the parent Emacs - the Windows "extra" CRs may confuse a 
Cygwin Emacs if it is not expecting them.  We sometimes see this with bash scripts brought over from 
Linux to Cygwin.  Anyway, this, or other character set related things, is where I would start to 
diagnose the issue and try different settings.  I would ask questions around what the terminal 
settings are that the subprocesses see, what exact characters are they outputting, etc.

Other are more expert than I in these details and may have specific settings to suggest that you 
look at :-) ...

Regards - Eliot Moss

Re: Issue on output when using javac or scalac under emacs

[Eliot Moss]

Dear Luc -- You should be aware that while Cygwin can invoke Java, the Windows Java still expects 
_Windows-style_ arguments, including Windows paths, etc.

For example, not a/b/c:d/e but 'a\b\c;d\e" for a path.

Just because you're using Cygwin, the installed Java is not "magically" transported to the Linux world!

You may also run into issues around line endings (CR/LF vs LF only, and so forth).

Regards - Eliot Moss

Re: Issue on output when using javac or scalac under emacs

[Brian Inglis]

On 2020-08-25 18:36, Eliot Moss wrote:
> Dear Luc 
> You should be aware that while Cygwin can invoke Java, the Windows Java still
> expects _Windows-style_ arguments, including Windows paths, etc.
> > For example, not a/b/c:d/e but 'a\b\c;d\e" for a path.
> 
> Just because you're using Cygwin, the installed Java is not "magically"
> transported to the Linux world!
> 
> You may also run into issues around line endings (CR/LF vs LF only, and so forth).

Many Windows APIs and programs accept either path separator and in such cases
there are seldom issues as long as the paths are not too long e.g. try:

	$ /proc/cygdrive/c/Windows/System32/icacls path/with/slash/delimiters

and many programs, including editors and terminals, do a good job of adapting to
line endings, including mintty and less.

-- 
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

This email may be disturbing to some readers as it contains
too much technical detail. Reader discretion is advised.
[Data in IEC units and prefixes, physical quantities in SI.]

Re: Issue on output when using javac or scalac under emacs

[Eliot Moss]

Agreed that / vs \ is sometimes no problem. I do know that Windows Java really wants ; and not : when giving multiple path names ... EM

Sent from my iPhone

> On Aug 25, 2020, at 10:50 PM, Brian Inglis <Brian.Inglis@systematicsw.ab.ca> wrote:
> 
> On 2020-08-25 18:36, Eliot Moss wrote:
>> Dear Luc 
>> You should be aware that while Cygwin can invoke Java, the Windows Java still
>> expects _Windows-style_ arguments, including Windows paths, etc.
>>> For example, not a/b/c:d/e but 'a\b\c;d\e" for a path.
>> 
>> Just because you're using Cygwin, the installed Java is not "magically"
>> transported to the Linux world!
>> 
>> You may also run into issues around line endings (CR/LF vs LF only, and so forth).
> 
> Many Windows APIs and programs accept either path separator and in such cases
> there are seldom issues as long as the paths are not too long e.g. try:
> 
>    $ /proc/cygdrive/c/Windows/System32/icacls path/with/slash/delimiters
> 
> and many programs, including editors and terminals, do a good job of adapting to
> line endings, including mintty and less.
> 
> -- 
> Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada
> 
> This email may be disturbing to some readers as it contains
> too much technical detail. Reader discretion is advised.
> [Data in IEC units and prefixes, physical quantities in SI.]
> --
> Problem reports:      https://cygwin.com/problems.html
> FAQ:                  https://cygwin.com/faq/
> Documentation:        https://cygwin.com/docs.html
> Unsubscribe info:     https://cygwin.com/ml/#unsubscribe-simple

Re: Issue on output when using javac or scalac under emacs

[Eliot Moss]

On 8/25/2020 11:51 PM, Eliot Moss wrote:
> Agreed that / vs \ is sometimes no problem. I do know that Windows Java really wants ; and not : when giving multiple path names ... EM
> 
> Sent from my iPhone
> 
>> On Aug 25, 2020, at 10:50 PM, Brian Inglis <Brian.Inglis@systematicsw.ab.ca> wrote:
>>
>> On 2020-08-25 18:36, Eliot Moss wrote:
>>> Dear Luc
>>> You should be aware that while Cygwin can invoke Java, the Windows Java still
>>> expects _Windows-style_ arguments, including Windows paths, etc.
>>>> For example, not a/b/c:d/e but 'a\b\c;d\e" for a path.
>>>
>>> Just because you're using Cygwin, the installed Java is not "magically"
>>> transported to the Linux world!
>>>
>>> You may also run into issues around line endings (CR/LF vs LF only, and so forth).
>>
>> Many Windows APIs and programs accept either path separator and in such cases
>> there are seldom issues as long as the paths are not too long e.g. try:
>>
>>     $ /proc/cygdrive/c/Windows/System32/icacls path/with/slash/delimiters
>>
>> and many programs, including editors and terminals, do a good job of adapting to
>> line endings, including mintty and less.

Apologies for the top-posting.  Not sure my phone supports any other way, but if I end up wanting to 
post from my phone, I will see what I can do!   Eliot Moss

Re: Issue on output when using javac or scalac under emacs

[Luc Henninger]

Le 26/08/2020 à 02:36, Eliot Moss a écrit :
> Dear Luc -- You should be aware that while Cygwin can invoke Java, the 
> Windows Java still expects _Windows-style_ arguments, including 
> Windows paths, etc.
>
> For example, not a/b/c:d/e but 'a\b\c;d\e" for a path.
>
> Just because you're using Cygwin, the installed Java is not 
> "magically" transported to the Linux world!
>
> You may also run into issues around line endings (CR/LF vs LF only, 
> and so forth).
>
> Regards - Eliot Moss

Thanks, but I already use "cygpath" within my makefiles in case of 
"java" to convert file paths.

Regards

-- 
Luc.Henninger@orange.fr    +33 6 32 96 32 27


-- 
L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast.
https://www.avast.com/antivirus