I am using Azure Devops to build and deploy my git repo to a third party vps. I do this by logging into the server from Azure Devops through SSH, executing a shell script to pull git repo, and build it with ie. vue-cli and Laravel.
When the bash script is executed I receive a lot of errors on nearly all commands although everything is succeeding - can anyone tell me how to get rid of these unless something is really failing (would be nice to fail if npm build exit with code 1 for instance).
See screenshot below.
Screenshots are only really helpful for visual issues. You can use PasteBin or etc to share long logs if necessary.
According to this issue Azure just follows the lead of whatever shell it's running code in. So, in Bash it continues unless explicitly told to stop.
To easily change this behavior you can add set -e (or set -o errexit) at the start of your script. The errexit option causes Bash to exit as soon as a command/etc returns a non-zero exit code.
Another worthy addition is the set -o pipefail option. If you've got any pipes like command1 | command2 this will return the first non-zero exit code from a chain of pipes of any length as the result. So, if command1 fails above but command2 succeeds it would return the failure code from command1 instead of overwriting it.
Finally, set -u (or -o nounset) causes an error when unset variables are encountered during parameter expansion. If running in a non-interactive shell, it will also exit.
Many scripts combine these by running set -euo pipefail at the beginning to stop them from running after the first problem is encountered.
If you want to explicitly force a bash script to exit you can use || and && accordingly. The expression command || exit will exit if the command fails and command && exit will exit if the command succeeds.
This seems to be one bug starting from npm V.3.10.8. You can check this discussion.
As a workaround you can add this script to package.json and run the command with --no-optional switch:
"optionalDependencies": {
"fsevents": "*"
},
Also, there's possibility that your NPM version is too old. You can use Node.js tool installer task with version spec = 12.x to install higher node.js and npm versions.
I get this error:
Can't locate Foo.pm in #INC
Is there an easier way to install it than downloading, untarring, making, etc?
On Unix:
usually you start cpan in your shell:
$ cpan
and type
install Chocolate::Belgian
or in short form:
cpan Chocolate::Belgian
On Windows:
If you're using ActivePerl on Windows, the PPM (Perl Package Manager) has much of the same functionality as CPAN.pm.
Example:
$ ppm
ppm> search net-smtp
ppm> install Net-SMTP-Multipart
see How do I install Perl modules? in the CPAN FAQ
Many distributions ship a lot of perl modules as packages.
Debian/Ubuntu: apt-cache search 'perl$'
Arch Linux: pacman -Ss '^perl-'
Gentoo: category dev-perl
You should always prefer them as you benefit from automatic (security) updates and the ease of removal. This can be pretty tricky with the cpan tool itself.
For Gentoo there's a nice tool called g-cpan which builds/installs the module from CPAN and creates a Gentoo package (ebuild) for you.
Try App::cpanminus:
# cpanm Chocolate::Belgian
It's great for just getting stuff installed. It provides none of the more complex functionality of CPAN or CPANPLUS, so it's easy to use, provided you know which module you want to install. If you haven't already got cpanminus, just type:
# cpan App::cpanminus
to install it.
It is also possible to install it without using cpan at all. The basic bootstrap procedure is,
curl -L http://cpanmin.us | perl - --sudo App::cpanminus
For more information go to the App::cpanminus page and look at the section on installation.
I note some folks suggesting one run cpan under sudo. That used to be necessary to install into the system directory, but modern versions of the CPAN shell allow you to configure it to use sudo just for installing. This is much safer, since it means that tests don't run as root.
If you have an old CPAN shell, simply install the new cpan ("install CPAN") and when you reload the shell, it should prompt you to configure these new directives.
Nowadays, when I'm on a system with an old CPAN, the first thing I do is update the shell and set it up to do this so I can do most of my cpan work as a normal user.
Also, I'd strongly suggest that Windows users investigate strawberry Perl. This is a version of Perl that comes packaged with a pre-configured CPAN shell as well as a compiler. It also includes some hard-to-compile Perl modules with their external C library dependencies, notably XML::Parser. This means that you can do the same thing as every other Perl user when it comes to installing modules, and things tend to "just work" a lot more often.
If you're on Ubuntu and you want to install the pre-packaged perl module (for example, geo::ipfree) try this:
$ apt-cache search perl geo::ipfree
libgeo-ipfree-perl - A look up country of ip address Perl module
$ sudo apt-get install libgeo-ipfree-perl
A couple of people mentioned the cpan utility, but it's more than just starting a shell. Just give it the modules that you want to install and let it do it's work.
$prompt> cpan Foo::Bar
If you don't give it any arguments it starts the CPAN.pm shell. This works on Unix, Mac, and should be just fine on Windows (especially Strawberry Perl).
There are several other things that you can do with the cpan tool as well. Here's a summary of the current features (which might be newer than the one that comes with CPAN.pm and perl):
-a
Creates the CPAN.pm autobundle with CPAN::Shell->autobundle.
-A module [ module ... ]
Shows the primary maintainers for the specified modules
-C module [ module ... ]
Show the Changes files for the specified modules
-D module [ module ... ]
Show the module details. This prints one line for each out-of-date module (meaning,
modules locally installed but have newer versions on CPAN). Each line has three columns:
module name, local version, and CPAN version.
-L author [ author ... ]
List the modules by the specified authors.
-h
Prints a help message.
-O
Show the out-of-date modules.
-r
Recompiles dynamically loaded modules with CPAN::Shell->recompile.
-v
Print the script version and CPAN.pm version.
sudo perl -MCPAN -e 'install Foo'
Also see Yes, even you can use CPAN. It shows how you can use CPAN without having root or sudo access.
Otto made a good suggestion. This works for Debian too, as well as any other Debian derivative. The missing piece is what to do when apt-cache search doesn't find something.
$ sudo apt-get install dh-make-perl build-essential apt-file
$ sudo apt-file update
Then whenever you have a random module you wish to install:
$ cd ~/some/path
$ dh-make-perl --build --cpan Some::Random::Module
$ sudo dpkg -i libsome-random-module-perl-0.01-1_i386.deb
This will give you a deb package that you can install to get Some::Random::Module. One of the big benefits here is man pages and sample scripts in addition to the module itself will be placed in your distro's location of choice. If the distro ever comes out with an official package for a newer version of Some::Random::Module, it will automatically be installed when you apt-get upgrade.
Already answered and accepted answer - but anyway:
IMHO the easiest way installing CPAN modules (on unix like systems, and have no idea about the wondows) is:
curl -L http://cpanmin.us | perl - --sudo App::cpanminus
The above is installing the "zero configuration CPAN modules installer" called cpanm. (Can take several minutes to install - don't break the process)
and after - simply:
cpanm Foo
cpanm Module::One
cpanm Another::Module
Many times it does happen that cpan install command fails with the message like
"make test had returned bad status, won't install without force"
In that case following is the way to install the module:
perl -MCPAN -e "CPAN::Shell->force(qw(install Foo::Bar));"
Lots of recommendation for CPAN.pm, which is great, but if you're using Perl 5.10 then you've also got access to CPANPLUS.pm which is like CPAN.pm but better.
And, of course, it's available on CPAN for people still using older versions of Perl. Why not try:
$ cpan CPANPLUS
Use cpan command as cpan Modulename
$ cpan HTML::Parser
To install dependencies automatically follow the below
$ perl -MCPAN -e shell
cpan[1]> o conf prerequisites_policy follow
cpan[2]> o conf commit
exit
I prefer App::cpanminus, it installs dependencies automatically. Just do
$ cpanm HTML::Parser
On ubuntu most perl modules are already packaged, so installing is much faster than most other systems which have to compile.
To install Foo::Bar at a commmand prompt for example usually you just do:
sudo apt-get install libfoo-bar-perl
Sadly not all modules follow that naming convention.
On Fedora Linux or Enterprise Linux, yum also tracks perl library dependencies. So, if the perl module is available, and some rpm package exports that dependency, it will install the right package for you.
yum install 'perl(Chocolate::Belgian)'
(most likely perl-Chocolate-Belgian package, or even ChocolateFactory package)
Even it should work:
cpan -i module_name
2 ways that I know of :
USING PPM :
With Windows (ActivePerl) I've used ppm
from the command line type ppm. At the ppm prompt ...
ppm> install foo
or
ppm> search foo
to get a list of foo modules available. Type help for all the commands
USING CPAN :
you can also use CPAN like this (*nix systems) :
perl -MCPAN -e 'shell'
gets you a prompt
cpan>
at the prompt ...
cpan> install foo (again to install the foo module)
type h to get a list of commands for cpan
On Fedora you can use
# yum install foo
as long as Fedora has an existing package for the module.
Easiest way for me is this:
PERL_MM_USE_DEFAULT=1 perl -MCPAN -e 'install DateTime::TimeZone'
a) automatic recursive dependency detection/resolving/installing
b) it's a shell onliner, good for setup-scripts
If you want to put the new module into a custom location that your cpan shell isn't configured to use, then perhaps, the following will be handy.
#wget <URL to the module.tgz>
##unpack
perl Build.PL
./Build destdir=$HOME install_base=$HOME
./Build destdir=$HOME install_base=$HOME install
Sometimes you can use the yum search foo to search the relative perl module, then use yum install xxx to install.
Secure solution
Many answers mention the use of the cpan utility (which uses CPAN.pm) without a word on security. By default, CPAN 2.27 and earlier configures urllist to use a http URL (namely, http://www.cpan.org/), which allows MITM attacks, thus is insecure. This is what is used to download the CHECKSUMS files, so that it needs to be changed to a secure URL (e.g. https://www.cpan.org/).
So, after running cpan and accepting the default configuration, you need to modify the generated MyConfig.pm file (the full path is output) in the following way. Replace
'urllist' => [q[http://www.cpan.org/]],
by
'urllist' => [q[https://www.cpan.org/]],
Note: https is not sufficient; you also need a web site you can trust. So, be careful if you want to choose some arbitrary mirror.
Then you can use cpan in the usual way.
My bug report on rt.cpan.org about the insecure URL.
Simply executing cpan Foo::Bar on shell would serve the purpose.
Seems like you've already got your answer but I figured I'd chime in. This is what I do in some scripts on an Ubuntu (or debian server)
#!/usr/bin/perl
use warnings;
use strict;
#I've gotten into the habit of setting this on all my scripts, prevents weird path issues if the script is not being run by root
$ENV{'PATH'} = '/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin';
#Fill this with the perl modules required for your project
my #perl = qw(LWP::Simple XML::LibXML MIME::Lite DBI DateTime Config::Tiny Proc::ProcessTable);
chomp(my $curl = `which curl`);
if(!$curl){ system('apt-get install curl -y > /dev/null'); }
chomp(my $cpanm = system('/bin/bash', '-c', 'which cpanm &>/dev/null'));
#installs cpanm if missing
if($cpanm){ system('curl -s -L http://cpanmin.us | perl - --sudo App::cpanminus'); }
#loops through required modules and installs them if missing
foreach my $x (#perl){
eval "use $x";
if($#){
system("cpanm $x");
eval "use $x";
}
}
This works well for me, maybe there is something here you can use.
On Windows with the ActiveState distribution of Perl, use the ppm command.
When i am trying to execute the following tcl code i got this
can't find package des
while executing
"package require des"
(file "encdec.tcl" line 1)
Tcl code
package require des
set key "12345678"; # Must be 8 bytes long
set msg "abcde"
##### ENCRYPTION
set encryptedMsg [DES::des -dir encrypt -key $key $msg]
# $encryptedMsg is a bunch of bytes; you'll want to send this around...
##### DECRYPTION
set decryptedMsg [DES::des -dir decrypt -key $key $encryptedMsg]
puts "I got '$decryptedMsg'"
how to get that package ?
This package is a part of "the standard Tcl library", tcllib.
If you're using a "batteries included" Tcl distribution, like ActiveTcl, you should use its means to get this package; for instance, with ActiveTcl, the command to install tcllib would be
teacup install tcllib
(you should run it in your Windows console prompt).
If you're using Tcl installed from a package of your operating system, tcllib is typically available in the form of a package as well. For instance, on a Debian (or its derivative) the command to install tcllib would be
apt-get install tcllib
Next time please tell us about your platform up front, without forcing anyone here to guess.
When i test my debian package with lintian i get the following warning:
executable-not-elf-or-script usr/share/cw1-6005-pp6g11/rss_reader.php
Does anyone know why lintian show this warning.
That file, rss_reader.php, has the +x permissions bit set, making it executable, but it isn't an ELF binary and it isn't a script with #! at the start. It therefore can't be executed by a standard linux kernel (there are modules to allow other binary formats to be executed).
In this particular case, although PHP scripts can be given a #! line and made executable, if they're part of a web application this probably isn't what you want. You should probably just remove the +x bit.
chmod -x /usr/share/cw1-6005-pp6g11/rss_reader.php
this will remove the EXECUTION bit from it and Lintian will not print the warning.
I've made a NSIS script to create an installer for a Rails application.
The application uses JRuby and Java.
In the first section of the installer I set the environment variables: I add jruby\bin to PATH and I create JAVA_HOME variable, which points to Java\jre6.
!define JRubyBinPath "$PROGRAMFILES\${PRODUCT_PUBLISHER}\JRuby\jruby-1.6.3\bin"
Function SetEnvVars
# set JAVA_HOME
${EnvVarUpdate} $0 "JAVA_HOME" "A" "HKCU" "C:\Program Files\Java\jre6"
# add jruby to Path
${EnvVarUpdate} $0 "Path" "A" "HKLM" "${JRubyBinPath}"
FunctionEnd
Section "Pre" SEC01
Call SetEnvVars
...
SectionEnd
** All the paths are correct. **
In the second section of the installer, after packing all the application files and JRuby files, I initialize the database using the following commands:
Section "Installer" SEC02
...
SetOutPath $INSTDIR
nsExec::ExecToLog "jruby -S bundle exec rake db:create RAILS_ENV=production"
nsExec::ExecToLog "jruby -S bundle exec rake db:migrate RAILS_ENV=production"
nsExec::ExecToLog "jruby -S bundle exec rake db:seed RAILS_ENV=production"
...
SectionEnd
The problem is that when I run the installer on a clean Windows system, all the code executes correctly, except for the lines that contain commands using "jruby". I get no error in the installer window, it just won't execute those lines.
Anyway, if I manually run those commands in a console right after the installer finishes, everything works as expected.
From what I have seen so far, the problem is that the installer cannot access the environment variables until it finishes.
* What I have done so far to solve this problem is creating the main installer Installer.exe (here I pack all the files and I set the environment variables), and another executable Init.exe which initializes the database.
If I manually run these executables - first Installer.exe then Init.exe - the application installs correctly. But if I try to launch Init.exe from Installer.exe, it won't work.
However, if there would be a way to execute the "jruby" commands in a new/different thread than the one that sets the environment variables, I think the problem would be solved. But I still couldn't find the way to do that.
Thanks.
NSIS doesn't have any threading. Apparently not quite true—but in your own code, you don't get to use threads.
You may need to set PATH for the current process; I'm not familiar with ${EnvVarUpdate}, but looking at its script it doesn't seem to updates the variable in the current process. Try this:
${EnvVarUpdate} $0 PATH A HKLM "${JRubyBinPath}" # The line you already have
System::Call Kernel32::SetEnvironmentVariable(t"PATH",tr0)
Another note: you are hard-coding JRubyBinPath; are you sure you really want to do that?