I have a problem that I cannot seem to find an answer to.
With Perl I need to use a script across Windows and unix platforms. Te problem is that on Windows we use Win32-pecific modules like Win32::Process, and those modules do not exist on unix.
I need a way to include those Win32 modules only on Windows.
if($^O =~ /win/i)
{
use win32::process qw(CREATE_NEW_CONSOLE);
}
else
{
#unix fork
}
The problem lies in that use statement for windows. No matter what I try this does not compile on unix.
I have tried using dynamic evals, requires, BEGIN, etc.
Is there a good solution to this problem? Any help will be greatly appreciated.
Thanks in advance,
Dan
Update:
A coworker pointed out to me this is the correct way to do it.
require Win32;
require Win32::Process;
my $flag = Win32::Process::CREATE_NEW_CONSOLE();
Win32::Process::Create($process,
$program,
$cmd,
0,
$flag, ".") || die ErrorReport();
print "Child started, pid = " . getPID() . "\n";
Thank you all for your help!
Dan
use is executed at compile time.
Instead do:
BEGIN {
if( $^O eq 'MSWin32' ) {
require Win32::Process;
# import Win32::Process qw(CREATE_NEW_CONSOLE);
Win32::Process->import(qw/ CREATE_NEW_CONSOLE /);
}
else {
#unix fork
}
}
See the perldoc for use.
Also see perlvar on $^O.
Update:
As Sinan Unur points out, it is best to avoid indirect object syntax.
I use direct method calls in every case, except, with calls to import. Probably because import masquerades as a built-in. Since import is really a class method, it should be called as a class method.
Thanks, Sinan.
Also, on Win32 systems, you need to be very careful that you get the capitalization of your module names correct. Incorrect capitalization means that symbols won't be imported properly. It can get ugly.use win32::process may appear to work fine.
Are you sure win32::process can be loaded on OSX? "darwin" matches your /win/i.
You may want to use http://search.cpan.org/dist/Sys-Info-Base/ which tries to do the right thing.
That aside, can you post an example of the code that you actually are using, the failure message you're receiving, and on which unix platform (uname -a) ?
What about a parser that modifies the file on each OS?
You could parse your perl file via a configure script that works on both operating systems to output perl with the proper Use clauses. You could even bury the parse action in the executable script to launch the code.
Originally I was thinking of precompiler directives from C would do the trick, but I don't know perl very well.
Here's an answer to your second set of questions:
Are you using strict and warnings?
Did you define an ErrorReport() subroutine? ErrorReport() is just an example in the synopsis for Win32::Process.
CREATE_NEW_CONSOLE is probably not numeric because it didn't import properly. Check the capitalization in your call to import.
Compare these one-liners:
C:\>perl -Mwin32::process -e "print 'CNC: '. CREATE_NEW_CONSOLE;
CNC: CREATE_NEW_CONSOLE
C:\>perl -Mwin32::process -Mstrict -e "print 'CNC: '. CREATE_NEW_CONSOLE;
Bareword "CREATE_NEW_CONSOLE" not allowed while "strict subs" in use at -e line 1.
Execution of -e aborted due to compilation errors.
C:\>perl -MWin32::Process -e "print 'CNC: '. CREATE_NEW _CONSOLE;
CNC: 16
You could just place your platform specific code inside of an eval{}, and check for an error.
BEGIN{
eval{
require Win32::Process;
Win32::Process->import(qw'CREATE_NEW_CONSOLE');
};
if( $# ){ # $# is $EVAL_ERROR
# Unix code here
}
}
Related
We have two versions of Active perl 5.6 and 5.24. We have web services which has to be executed on Active perl '5.24' versions(to adopt TLS 1.2 version) and this needs to be invoked from Active perl '5.6' version. We are using windows operating system.
Steps followed :
Caller code which is executed in 5.6 version invokes the 5.24 version using system /require command.
Problem:
How to call the 5.24 perl function(example: webservicecall(arg1){return "xyz") from 5.6 perl script through system command, require or etc..?
Also how to get the return value of perl function 5.24?
Note:
Its a temporary work around to have two perl versions and the we have a plan to do upgrade it for higher version.
Here perl version 5.6 installed in "C:\Perl\bin\perl\" and perl version 5.24 installed in "D:\Perl\bin\perl\".
"**p5_6.pl**"
print "Hello Perl5_6\n";
system('D:\Perl\bin\perl D:\sample_program\p5.24.pl');
print $OUTFILE;
$retval = Mul(25, 10);
print ("Return value is $retval\n" );
"**p5_24.pl**"
print "Hello Perl5_24\n";
our $OUTFILE = "Hello test";
sub Mul($$)
{
my($a, $b ) = #_;
my $c = $a * $b;
return($c);
}
I have written sample program for detail information to call perl 5.24 version from perl script 5.6 version. During execution I didn't get the expected output. How to get the "return $c" value & the "our $OUTFILE" value of p5_24.pl in p5_6.pl script?
Note: The above is the sample program based on this I will modify the actual program using serialized data.
Place the code for the function that needs v5.24 in a wrapper script, written just so that it runs that function (and prints its result). Actually, I'd recommend writing a module with that function and then loading that module in the wrapper script.
Then run that script under the wanted (5.24) interpreter, by invoking it via its full path. (You may need to be careful to make sure that all libraries and environment are right.) Do this in a way that allows you to pick up its output. That can be anything from backticks (qx) to pipe-open or, better, to good modules. There is a range of modules for this, like IPC::System::Simple, Capture::Tiny, IPC::Run3, or IPC::Run. Which to use would depend on how much you need out of that call.
You can't call a function in a running program but to have it somehow run under another program.
Also, variables (like $OUTFILE) defined in one program cannot be seen in another one. You can print them from the v5.24 program, along with that function result, and then parse that whole output in the v5.6 program. Then the two programs would need a little "protocol" -- to either obey an order in which things are printed, or to have prints labeled in some way.
Much better, write a module with functions and variables that need be shared. Then the v5.24 program can load the module and import the function it needs and run it, while the v5.6 program can load the same module but only to pick up that variable (and also run the v5.24 program).
Here is a sketch of all this. The package file SharedBetweenPerls.pm
package SharedBetweenPerls;
use warnings;
use strict;
use Exporter qw(import);
our #EXPORT_OK = qw(Mul export_vars);
my $OUTFILE = 'test_filename';
sub Mul { return $_[0] * $_[1] }
sub export_vars { return $OUTFILE }
1;
and then the v5.24 program (used below as program_for_5.24.pl) can do
use warnings;
use strict;
# Require this to be run by at least v5.24.0
use v5.24;
# Add path to where the module is, relative to where this script is
# In our demo it's the script's directory ($RealBin)
use FindBin qw($RealBin);
use lib $RealBin;
use SharedBetweenPerls qw(Mul);
my ($v1, $v2) = #ARGV;
print Mul($v1, $v2);
while the v5.6 program can do
use warnings;
use strict;
use feature 'say';
use FindBin qw($RealBin);
use lib $RealBin;
use SharedBetweenPerls qw(export_vars);
my $outfile = export_vars(); #--> 'test_filename'
# Replace "path-to-perl..." with an actual path to a perl
my $from_5.24 = qx(path-to-perl-5.24 program_for_5.24.pl 25 10); #--> 250
say "Got variable: $outfile, and return from function: $from_5.24";
where $outfile has the string test_filename while $from_5.24 variable is 250.†
This is tested to work as it stands if both programs, and the module, are in the same directory, with names as in this example. (And with path-to-perl-5.24 replaced with the actual path to your v5.24 executable.) If they are at different places you need to adjust paths, probably the package name and the use lib line. See lib pragma.
Please note that there are better ways to run an external program --- see the recommended modules above. All this is a crude demo since many details depend on what exactly you do.
Finally, the programs can also connect via a socket and exchange all they need but that is a bit more complex and may not be needed.
† The question's been edited, and we now have D:\Perl\bin\perl for path-to-perl-5.24 and D:\sample_program\p5.24.pl for program_for_5.24.
Note that with such a location of the p5.24.pl program you'd have to come up with a suitable location for the module and then its name would need to have (a part of) that path in it and to be loaded with such name. See for example this post.
A crude demo without a module (originally posted)
As a very crude sketch, in your program that runs under v5.6 you could do
my $from_5.24 = qx(path-to-perl-5.24 program_for_5.24.pl 25 10);
where the program_for_5.24.pl then could be something like
use warnings;
use strict;
sub Mul { return $_[0] * $_[1] }
my ($v1, $v2) = #ARGV;
print Mul($v1, $v2);
and the variable $from_5.24 ends up being 250 in my test.
You cannot directly call a Perl function running with another Perl version. You would need to create a program which explicitly invokes the function. The input and output need to be explicitly serialized in order to be transported between these two programs.
Serializing could be done with Data::Dumper, Storable or similar. If lower performance is needed you could invoke the program which provides the function with system and share the serialized data with temporary files or pipes. Or you could create some client-server architecture and share the serialized data with sockets. The latter is faster since it skips the repeated start and teardown of the other process but instead keeps it running.
My team has all our Golang code in a monorepo.
Various package subdirectories with library code.
Binaries/services/tools under cmd
We've had it for a while and are doing some cleanup. Are there any tools or techniques that can find functions not used by the binaries under cmd?
I know go vet can find private functions that are unused in a package. However I suspect we also have exported library functions that aren't used either.
UPD 2020: The unused tool has been incorporated into staticcheck.
Unfortunately, v0.0.1-2020.1.4 will probably be the last to support this
feature. Dominik explains that it is because the check consumes a lot of
resources and is hard to get right.
To get that version:
env GO111MODULE=on go get honnef.co/go/tools/cmd/staticcheck#v0.0.1-2020.1.4
To use it:
$ staticcheck --unused.whole-program=true -- ./...
./internal/pkg/a.go:5:6: type A is unused (U1001)
Original answer below.
Dominik Honnef's unused tool might be what you're looking for:
Optionally via the -exported flag, unused can analyse all arguments as a
single program and report unused exported identifiers. This can be useful for
checking "internal" packages, or large software projects that do not export
an API to the public, but use exported methods between components.
Try running go build -gcflags -live. This passes the -live flag to the compiler (go tool compile), instructing it to output debugging messages about liveness analysis. Unfortunately, it only prints when it's found live code, not dead code, but you could in theory look to see what doesn't show up in the output.
Here's an example from compiling the following program stored in dead.go:
package main
import "fmt"
func main() {
if true {
fmt.Println(true)
} else {
fmt.Println(false)
}
}
Output of go build -gcflags -live:
# _/tmp/dead
./dead.go:7: live at call to convT2E: autotmp_5
./dead.go:7: live at call to Println: autotmp_5
If I'm reading this correctly, the second line states that the implicit call to convT2E (which converts non-interface types to interface types, since fmt.Println takes arguments of type interface{}) is live, and the third line states that the call to fmt.Println is live. Note that it doesn't say that the fmt.Println(false) call is live, so we can deduce that it must be dead.
I know that's not a perfect answer, but I hope it helps.
It is a bit dirty, but it works for me.
I had a lot of structs which I did not want to test manually, so I wrote a script that renames the struct then runs all the tests (ci/test.sh) and renames it back if any test failed:
#!/bin/sh
set -e
git grep 'struct {' | grep type | while read line; do
file=$(echo $line | awk -F ':' '{print $1}')
struct=$(echo $line | awk '{print $2}')
sed "s/$struct struct/_$struct struct/g" -i $file
echo "testing for struct $struct changed in file $file"
if ! ./ci/test.sh; then
sed "s/_$struct struct/$struct struct/g" -i $file
fi
done
It's not an open source solution, but it works.
If you guys are using Goland, you should consider using its code-inspections feature, includes useful features:
Reports unused constant
Reports unused exported functions
Reports unused exported types in the main package and in tests
Reports unused unexported functions
Reports global variables that are defined but are never used in code
Reports unused function parameters
Reports unused types
(It looks like the implementation of this feature is black box, jetbrains does not open source this feature)
Go-related detection tools seem to place more emphasis on accuracy, and they want to do their best to minimize error reporting. And using Goland's code-inspections feature may require more self-judgment. :)
Interests: Paid users only, not working for Jetbrains, simply think this feature works well.
A reliable but inelegant method I've used is to rename or comment out functions you suspect might not be used and then recompile everything -- no errors means you didn't need them.
If they are needed, it shows you where these functions are called so it's good for getting familiar with a code base and seeing how things connect.
I am trying to build Win32::Daemon by myself. The reason I not use CPAN is because I want to dig deeper into the working of Perl modules. In the end I hope to come up with a solution for another problem by seeing this working (not of importance here).
I would see 3 options to build the module: cygwin, mingw, microsoft compiler (cl)
On MinGW it reports that it is not supported (simple if in the Makefile.PL) which expands to more errors once I modify the check to match MinGW
On Cygwin it complains about tchar.h which, as I found out, is a Windows header (MinGW does have it).
But my real goal anyway is building it with the MS compiler, so while any compilation that does not require any special libs (like it would do with cygwin I suppose) will more.
So now here goes my nmake output from running just name /f Makefile:
NMAKE : fatal error U1073: "C:/Program" could not be created.
Stop.
I roughly translated the error message from german, but the statement is simple.
What I see here seems to be a path problem (probably the spaces). I also notice the forward slash. The Makefile was created by the Makefile.PL script (I am using Active Perl v5.12.1):
use strict;
use warnings;
use Config qw(%Config);
use ExtUtils::MakeMaker;
unless ($^O eq "MSWin32" || $^O eq "cygwin") {
die "OS unsupported\n";
}
require Win32;
my %param = (
NAME => 'Win32::Daemon',
VERSION_FROM => 'Daemon.pm',
DEFINE => '-DPERL_NO_GET_CONTEXT',
OBJECT => 'CCallbackList$(OBJ_EXT) CCallbackTimer$(OBJ_EXT) Constant$(OBJ_EXT) CWinStation$(OBJ_EXT) Daemon$(OBJ_EXT) ServiceThread$(OBJ_EXT)',
XS => { 'Daemon.xs' => 'Daemon.cpp' },
);
$param{INC} .= ' -EHsc' if $Config{'cc'} =~ /^cl/i;
$param{NO_META} = 1 if eval "$ExtUtils::MakeMaker::VERSION" >= 6.10_03;
WriteMakefile(%param);
sub MY::xs_c {
'
.xs.cpp:
$(PERL) -I$(PERL_ARCHLIB) -I$(PERL_LIB) $(XSUBPP) $(XSPROTOARG) $(XSUBPPARGS) $*.xs >xstmp.c && $(MV) xstmp.c $*.cpp
';
}
I don't know much about the MakeMaker but I don't see anything here that I could fix and would expect that it boils down to fixing the Makefile itself by hand. I tried a couple of things like quoting but nothing helped.
The thing is, I am used to problems like this when building on Windows, but normally this is for tools that were created for Unix. This one is explicitly ONLY Windows and so I would expect it to work out of the box. So I figure that I am doing something wrong.
Any help on where to find the solution?
Thanks in advance.
Edit/Addition: I tried this on another Win7 machine with Active Perl 5.16.x and it worked like a charm. I looked at the different output from this machine and the current one which fails when running perl Makefile.PL and I recieve the following output:
... Detected uninstalled Perl. Trying to continue.
Unable to find a perl 5 (by these names: C:\Program Files\Perl64\bin\perl.exe perl.exe perl5.exe perl5.12.1.exe miniperl.exe, in these dirs: . [...] C:\Program Files\Perl64\site\bin C:\Program Files\Perl64\bin [...])
Have \progra~1\perl64\lib
Want \temp\perl---please-run-the-install-script---\lib
Writing Makefile for Win32::Daemon
I truncated the output. Now please someone explain to me: Why can I run perl Makefile.PL or perl -v but it does not find my Perl in the exact directory it is in? I reinstalled it but it did not work...
Okay I finally seem to have solved this after hours of searching. The problem lies within multiple issues.
The first command of "uninstalled perl" does not make any sense to be, but you can fix it, by supplying perl Makefile.PL PERL_SRC="C\:Program Files\Perl64". Warning: This did not work in a command shell for me, I had to use powershell, because he would not treat the path correctly. You maybe need to juggle with this a bit. Note: In the end I fixed it by installing the original Active Perl, not the one provided by my installer (company software distribution)
Now to the issue of not finding perl: This is a problem with spaces in the path. I fixed this (seemingly) by creating a symlink without spaces. Now perl Makefile.PL does not throw any errors, but nmake -f "Makefile" failed. So the solution really was: Do not have spaces in your perl-path! This sucks, and quite frankfly in 2012 this shouldn't be a problem any more but here you go.
Thanks for all the effort everyone put in, this was a tough one to solve.
I need to get the current encoding according to the system local settings. I'm looking for such function working this way:
my $sysEncoding = getSystemEncoding();
#and now $sysEncoding equals e.g. 'windows-1250'
I looked everywhere on the internet. I've found just the module PerlIO::locale. But I thing that the system encoding should be recognized easier without additional modules.
Encode::Locale provides the means to handle this.
use Win32::API;
if (Win32::API->Import('kernel32', 'int GetACP()')) {
$enc = GetACP();
print "Current local encoding is '$enc'\n";
}
Thanks for hint to Ikegami.
In Perl (on Windows) how do I determine the last modified time of a directory?
Note:
opendir my($dirHandle), "$path";
my $modtime = (stat($dirHandle))[9];
results in the following error:
The dirfd function is unimplemented at scriptName.pl line lineNumber.
Apparently the real answer is just call stat on a path to the directory (not on a directory handle as many examples would have you believe) (at least for windows).
example:
my $directory = "C:\\windows";
my #stats = stat $directory;
my $modifiedTime = $stats[9];
if you want to convert it to localtime you can do:
my $modifiedTime = localtime $stats[9];
if you want to do it all in one line you can do:
my $modifiedTime = localtime((stat("C:\\Windows"))[9]);
On a side note, the Win32 UTCFileTime perl module has a syntax error which prevents the perl module from being interpreted/compiled properly. Which means when it's included in a perl script, that script also won't work properly. When I merge over all the actual code that does anything into my script and retry it, Perl eventually runs out of memory and execution halts. Either way there's the answer above.
my $dir_path = "path_of_your_directory";
my $mod_time = ( stat ( $dir_path ) )[9];
Use the Win32::UTCFileTime module on CPAN, which mirrors the built-in stat function's interface:
use Win32::UTCFileTime qw(:DEFAULT $ErrStr);
#stats = stat $file or die "stat() failed: $ErrStr\n";