I have a hash of hashes that I need to embed in an exec resource command. My thought was to serialize the hash to a string and interpolate it into the exec call. The exec call will be executing ruby code via ruby -e 'ruby code here'.
Using irb, I know that hash.to_s creates a single line parse-able version of the hash. Or I could use json. I doubt you can call to_s in puppet, but am not sure.
The stdlib for Puppet has parseyaml and parsejson to deserialize, but is there a way to serialize to a parse-able string? I can write a custom puppet function to do it, but prefer an already built in solution if there is one.
Update
I am considering defining a puppet function. I have never written one before, so am not sure of the syntax. Here is my first attempt:
Puppet::Parser::Functions.newfunction(
:serialize_hash,
:arity => 2,
:doc => "Serialize a hash to any depth and optionally escape the double quotes.",
:type => :rvalue) do |args|
hash = args[0]
escape_quotes = args[1]
serialized = hash.to_s
if (escape_quotes)
serialized.sub!(/"/, "\\\"")
end
serialized
end
You can always execute ruby code inline with your puppet module:
$my_string = inline_template('<%= #my_hash.to_s %>')
Obviously it is important to not overuse this, but it is particularly useful when a very simple ruby function can achieve what you need.
Related
Coming to Ruby from a PHP background, I'm used to being able to use require, require_once, include, or include_once which all have a similar effect, but the key being they continue to process code in the same scope where the include / require command was invoked.
Example:
sub.php
<?php
echo $foo;
main.php
<?php
$foo = 1234;
include('sub.php'); // outputs '1234'
When I first started using Ruby I tried to include / require / require_relative / load other .rb files, and after becoming a little frustrated with not having it work how I would expect it to I decided that there were better ways to go about breaking up large files and that Ruby didn't need to behave in the same way PHP did.
However, occasionally I feel that for testing purposes it would be nice to to load code from another .rb file in the way PHP does - in the same scope with access to all the same variables - without having to use class / instance variables or constants. Is this possible? Maybe somehow using a proc / binding / or eval command?
Again, I'm not advocating that this should be used during development - but I am curious if it is possible - and if so, how?
Yes, this is possible, although certainly not something I'd recommend doing. This works:
includer.rb:
puts var
include.rb:
var = "Hello!"
eval(File.read("include.rb"), binding)
Running this (Ruby 2.2.1, Ruby 1.9.3) will print Hello!. It works simply: eval takes an optional binding with which to evaluate the code it is passed, and Kernel#binding returns the current binding.
To have code run in same binding, you could simply eval the file contents as follows:
example.rb
class Example
def self.called_by_include
"value for bar"
end
def foo
puts "Called foo"
end
eval( File.read( 'included.rb' ) )
end
Example.new.bar
included.rb
BAR_CONSTANT = called_by_include
def bar
puts BAR_CONSTANT
end
Running ruby example.rb produces output
value for bar
The important thing is the eval( File.read( 'included.rb' ) ) code, which if you really wanted you could define as a class method on Object, to allow arbitrary source to be included with a convenience function*. The use of constants, class variables etc just shows influences working in both directions between the two pieces of source code.
It would be bad practice to use this in any production code. Ruby gives you much better tools for meta-programming, such as ability to use mix-ins, re-open classes, define methods from blocks etc.
* Something like this
class Object
def self.include_source filename
eval( File.read( filename ) )
end
end
And the line in example.rb would become just
include_source 'included.rb'
Again I have to repeat this is not such a great idea . . .
To import external .rb file in your code, I'm not sure but I think it have to be a gem.
Use require followed by the name of the gem you want to import.
Example
require 'foobar'
# do some stuff
Or you can use load to import entire rb file
load 'foobar.rb'
# do some stuff
Good luck and sorry for my english
Background
I'm trying to test the formatting of some HAML templates outside of Rails. The idea is to pass in some instance variables on the command line or via an included Ruby file, rendering the template to standard output. I tried this several different ways without success, as outlined below.
Requiring a Ruby File
For example, given the following two files:
HAML template: "test.haml"
!!!
%h1 Testing HAML CLI
%p= #bar
%p= #baz
Ruby file: "test.rb"
#foo = 'abc'
#bar = '123'
I would expect an invocation like haml -r ./test test.haml to return an interpolated HTML file on standard output, but it doesn't. Instead, I get just the HTML:
<!DOCTYPE html>
<h1>Testing HAML CLI</h1>
<p></p>
<p></p>
Programmatic Attempt
Since this didn't work, I also tried to do this programmatically. For example:
#!/usr/bin/env ruby
require 'haml'
#foo = 'abc'
#bar = '123'
engine = Haml::Engine.new(File.read 'test.haml')
puts engine.render
with exactly the same results, e.g. just the HTML with no variable interpolation.
Restating the Question
Clearly, something else is needed to get HAML to render the template with its associated variables. I would prefer to do this from the command line, either by passing arguments or including a file. How should I be invoking HAML from the command line to make it happen?
If that's not possible for whatever reason, how should I invoke HAML programmatically to perform the interpolation without depending on Rails?
You can supply a scope object and a local variables hash to the render method. In your example case, you would call:
engine = Haml::Engine.new(File.read 'test.haml')
engine.render(Object.new, { :#foo => 'abc', :#bar => '123' })
The reason that both of these examples are not working is that you are attempting to access instance variables from a different class. The simplest solution is to define and use methods instead of attempting to access another classes instance variables as if they were your own.
I.E. in test.rb
def foo
'abc'
end
test.haml
!!!
%h1 Testing HAML CLI
%p= foo
I can't seem to call any of Java's string methods from within JRuby. The same style of syntax work for the Math class though. What am I doing wrong?
#! /usr/bin/env jruby
require 'rubygems'
require 'java'
puts java.lang.Math::max(1000,200)
puts java.lang.Math::PI
# this doesn't work
puts java.lang.String::toUpperCase("we, the people")
# this doesn't work either
JString = java.lang.String
puts JString.toUpperCase('We, the people')
#toUpperCase exists though, see below
puts java.lang.String.java_class.declarSed_instance_methods
I think this is what you are trying to do:
java.lang.String.new("we, the people").toUpperCase
As mentioned by #Jesper, toUpperCase is an instance method with the String class. Using it as a static method will not work.
Also note, the class returned is of native Ruby type.
I'm using an external API that takes a key string, and would like to pass this key string to the test suite. Something like:
rake test [key=api_key]
The code together with the tests will be open sourced, but I'm not allowed to distribute my key string to other users, so I cannot put it in the test file. Can I pass it as a parameter?
You have two options. Pass it as an environment variable:
API_KEY='key' rake test
You can then access this through the ENV object in your test:
key = ENV['API_KEY']
Second option is to put this key in a file (e.g. key.txt) and you read it from that. To ensure that you don't distribute that file with your code, add it to your .gitignore file (or whatever is the ignore file used by your SCM)
Thank you very much!
I actually was thinking of putting it into a file and gitignoring it, but ended up passing a parameter to rake. May be, I will combine both (it's a long key).
Modify the Rakefile code for the :test task, such as adding a
parameter to it.
task :test, :key do |t, k|
result = system("ruby -Ilib -Itest -e 'ARGV.each { |f| load(f) if File.exists?(f)}' test/unit/* '#{k[:key]}'")
exit(result ? 0 : 1)
end
Call rake test['blah-blah']
It may take more then one key if needed.
How can I directly call a ruby function from the command-line?
Imagine, I would have this script test.rb:
class TestClass
def self.test_function(some_var)
puts "I got the following variable: #{some_var}"
end
end
If this script is run from the command-line (ruby test.rb), nothing happens (as intended).
Is there something like ruby test.rb TestClass.test_function('someTextString')?
I want to get the following output: I got the following variable: someTextString.
First the name of the class needs to start with a capital letter, and since you really want to use a static method, the function name definition needs to start with self..
class TestClass
def self.test_function(someVar)
puts "I got the following variable: " + someVar
end
end
Then to invoke that from the command line you can do:
ruby -r "./test.rb" -e "TestClass.test_function 'hi'"
If you instead had test_function as an instance method, you'd have:
class TestClass
def test_function(someVar)
puts "I got the following variable: " + someVar
end
end
then you'd invoke it with:
ruby -r "./test.rb" -e "TestClass.new.test_function 'hi'"
Here's another variation, if you find that typing ruby syntax at the command line is awkward and you really just want to pass args to ruby. Here's test.rb:
#!/usr/bin/env ruby
class TestClass
def self.test_function(some_var)
puts "I got the following variable: #{some_var}"
end
end
TestClass.test_function(ARGV[0])
Make test.rb executable and run it like this:
./test.rb "Some Value"
Or run it like this:
ruby test.rb "Some Value"
This works because ruby automatically sets the ARGV array to the arguments passed to the script. You could use ARGV[0] or ARGV.first to get the first argument, or you could combine the args into a single string, separated by spaces, using ARGV.join(' ').
If you're doing lots of command-line stuff, you may eventually have a use for Shellwords, which is in the standard ruby lib.
If you have multiple arguments to call in a example like this:
class TestClass
def self.test_function(some_var1, some_var2)
puts "I got the following variables: #{some_var1}, #{some_var2}"
end
end
run it like this (the arguments need to be comma separated in this case)
ruby -r "./test.rb" -e "TestClass.new.test_function 'hi','Mike'"
#!/usr/bin/env ruby
class A
def run
p :Hello_world
end
self
end.new.run
The usual way to script Ruby is to just use the top level execution environment called main. You can just start defining methods and code you write outside of a class, and these will be executed directly. (BTW, code inside a class but outside any method will run "by itself" also.)
Anyway, I'm with you ... I like writing all code in a named class and instantiating that class, so you can combine the techniques .. have the class return its own object .. and then use just a little of that top level code to allocate, initialize, and then dot into the class.
With this, you can just type $ ruby test.rb and it will do what you want. Or you can chmod +x test.rb; ./test.rb since we did add a shebang.
If you are working on a command line interface, then I would suggest to have a look at thor.
Thor directly maps your commands to methods within the defined class, see the thor wiki for an example.
Just an extension to Ingenu's answer for the case that the function does not print something out, but does return something.
We would have the following test.rb
class TestClass
def self.test_function(some_var)
return "I got the following variable: " + some_var
end
end
Then to invoke that from the command line and get the return value:
ruby -r "./test.rb" -e "puts TestClass.test_function('hi')"
If you know that how to call an rb file from commandline
ruby yourfile.rb
This can do the whole trick for you.
What you have done is, just defined your methods in the class. Now you can call it below the definition. If you still want to read, wide open your eyes
class TestClass
def self.test_function(some_var)
puts "I got the following variable: #{some_var}"
end
test_function(var)
end