I'm new to ruby ... wondering if the following is possible:
I currently run a test app within irb (irb -r test.rb) and manually execute
various command implemented in test.rb. One of these functions is currently implemented as follows:
def cli(cmd)
ret=$client.Cli(cmd)
print ret, "\n"
end
Where $client.Cli() takes a string. I currently type the following in the IRB prompt
> cli "some command with parameters"
This is sent over socket and results are returned
I would like to be able to do this WITHOUT the quotes. This would be just for this command
Is there a way to do this generally in ruby? if not how would you extend irb to do this?
For those who know 'C' this would be like the following:
#define CLI(CMD) cli(#CMD)
CLI(Quadafi and Sheen walk into a bar...)
where the pre-processed output is:
cli("Quadafi and Sheen walk into a bar...")
Thanks
You could actually monkey patch the gets method of the IRB::StdioInputMethod and IRB::ReadlineInputMethod classes, and perform a rewrite if the cli method is called, by adding the following to your test.rb file:
module IRB
def self.add_quotes(str)
str.gsub(/^cli (..+?)(\\+)?$/, 'cli "\1\2\2"') unless str.nil?
end
class StdioInputMethod
alias :old_gets :gets
def gets
IRB::add_quotes(old_gets)
end
end
class ReadlineInputMethod
alias :old_gets :gets
def gets
IRB::add_quotes(old_gets)
end
end
end
This way, any input line matching cli ... will be replaced with cli "..." before it's evaluated.
I don't think it's possible, because the command that you type to irb has to parse as ruby, and all those bare words will report errors like this:
NameError: undefined local variable or method `hello' for main:Object
(My first attempt, I just called it via cli hello.)
But if you didn't mind a more radical change, you could do something like this:
$ cat /tmp/test_cases
hello world
one
two
three
riding the corpse sled
$ ruby -e 'def f(arg) puts arg end' -ne 'f($_)' < /tmp/test_cases
hello world
one
two
three
riding the corpse sled
I just defined a simple function f() here that would show how it works; you could replace f($_) with $Client.cli($_), and set up the $Client global variable in the first -e argument. And you can leave off the < /tmp/test_cases if you want to type them in interactively:
$ ruby -e 'def f(arg) puts arg end' -ne 'f($_)'
hello
hello
world
world
Of course, if you want it to be any more advanced than this, I'd just write a script to do it all, rather than build something hideous from -pe or -ne commands.
Related
My question is a follow up to this question: No return on command line when running Ruby script because the answer doesn't offer an explanation:
Take the following file, script.rb:
def hello(names)
names.map {|name| "#{name} is awesome!"}
end
hello(["mark", "tony", "scott"])
When executing the file on the command line with ruby script.rb the return value of the following function does not appear. However, testing the method in IRB or by dropping into the code with PRY outputs an explicit return value. Why isn't the return value visible via script execution?
P.S. I am aware that using puts will output code into the terminal but I'm interested in why the return value doesn't output.
Because both IRB or Pry are REPL's
REPL stands for: read, evaluate, print and loop. And that's exactly what both Pry and IRB are doing.
They will first read your input, evaluate your code, print the result of the code execution and then start over.
A Ruby script can't return a value directly like you want it to, the Bash shell works in the same way, a Bash function can't return a string directly. You can return a string (with stdout) and assign it to the variable.
~:$~ cat scr.rb
~:$~ puts "Something here"
~:$~ echo $(ruby ./scr.rb)
Something here
~:$~ myvar=$(echo $(ruby ./scr.rb))
~:$~ echo $myvar
Something here
It's really simple: Bash (or whatever shell you are using) and Ruby are different programming languages. Bash doesn't know anything about Ruby. It doesn't know what a " Ruby return" is, it doesn't know what a "Ruby array" is, it doesn't know what a "Ruby string" is. Therefore, you simply cannot possibly return a Ruby object to Bash.
In fact, the shell usually just uses some operating system function to execute the Ruby script (e.g. the classical fork and exec or something like vfork or clone). If you wanted to return values this way, the operating system kernel would have to know about the semantics of every programming language ever invented plus every programming language that is going to be invented in the future. That is just not feasible.
Note that a command can return a value to the shell, namely an integer between 0 and 255 intended as a status code (with 0 meaning "success" and nonzero meaning "error"), and you can set that return value by calling Kernel#exit.
I used to have the same question myself when I started coding. If you have a closer look at your code you can see why it doesn't print anything. You are actually no asking it in your code. Imagine having a huge script of thousands of lines and you want to execute it. You would have millions of pointless outputs if ruby myscript.rb worked the same way as the REPLs.
In addition, if you do want it to work that way, you can just do require the script inside the REPL session ( require_relative 'your_script' ) and then if you call your function hello it will work the way you describe.
I can use the ruval gem. It evaluates each statement and returns its value.
$ ruval names.rb
def hello(names)
names.map {|name| "#{name} is awesome!"}
end
=> hello
hello(["mark", "tony", "scott"])
=> ["mark is awesome!", "tony is awesome!", "scott is awesome!"]
I'm new to ruby. I understand that, when I see a ruby script, it usually contains lines similar to this:
#!/usr/bin/env ruby
require 'rubyfunction1'
require 'rubyfunction2'
I understand that the require lines are basically (to put it in simple basic terms), calling other scripts. That is really all there is to it. These other scripts are functions.
Now, suppose, I put the content of the rubyfunction1 and rubyfunction2 scripts into two different variables. How do I require the content of a variable?
Or, suppose I want to be able to do something like this:
require '`/home/swenson/rubyfunction1.rb`'
I understand this is a roundabout way of requiring gems/ruby functions, but I'm curious to know if it is at all possible in this manner.
Basically, if I were to run the /home/swenson/rubyfunction1.rb script by itself on the command line, it will basically output to you the content of the script. It would be equivalent to doing "cat /home/swenson/rubyfunction1.rb".
I want to be able to do something like this:
require '`/home/swenson/rubyfunction1.rb`'
require '`/home/swenson/rubyfunction2.rb`'
or
specvar1 = `/home/swenson/rubyfunction1.rb`
specvar2 = `/home/swenson/rubyfunction2.rb`
require specvar1
require specvar2
Is this possible? Any suggestions I can apply to get it to work?
UPDATE:
So here's what I ended up doing.
Main Script called example.rb:
#!/usr/bin/env ruby
add = `./add.rb` # for my purposes, this will serve as require
subtract = `./subtract.rb` # for my purposes, this will serve as require
eval add
puts "I can add: #{add(3, 2)}"
eval subtract
puts "I can now subtract #{subtract(3, 2)}"
Content of add.rb:
#!/usr/bin/env ruby
puts <<-function
#!/usr/bin/env ruby
def add(a, b)
a + b
end
function
Content of subtract.rb:
#!/usr/bin/env ruby
puts <<-function
#!/usr/bin/env ruby
def subtract(a, b)
a - b
end
function
When run from the command line, I get no errors:
# ./example.rb
I can add: 5
I can now subtract 1
Basically, what I want done is precisely this. However, I know there's probably a optimized way of doing this (without having to directly require the relative file). So please, feel free to help me update or optimize this.
I understand that the require lines are basically (to put it in simple basic terms), calling other scripts. That is really all there is to it.
Yes. load, require, and require_relative simply run a Ruby file. That's it.
These other scripts are functions.
No. They are scripts. There is no such thing as a function in Ruby.
Now, suppose, I put the content of the rubyfunction1 and rubyfunction2 scripts into two different variables. How do I require the content of a variable?
You can't. require runs a file. It takes the name of a file (more precisely, a relative path) as an argument. Ruby code is not the name of a file.
Or, suppose i want to be able to do something like this:
require '`/home/swenson/rubyfunction1.rb`'
I understand this is a roundabout way of requiring gems/ruby functions, but im curious to know if it is at all possible in this manner.
This is possible. There's nothing special about this. It will simply run a file at the path `/home/swenson/rubyfunction1.rb`. That is a slightly unusual path, but there is nothing special about it. It's just a path like any other, with some funny characters in it.
so to iterate what im trying to do, i want to be able to do something like this:
require '`/home/swenson/rubyfunction1.rb`'
require '`/home/swenson/rubyfunction2.rb`'
or
specvar1 = `/home/swenson/rubyfunction1.rb`
specvar2 = `/home/swenson/rubyfunction2.rb`
require specvar1
require specvar2
Is this possible? Any suggestions I can apply to get it to work?
It's not quite clear what you want here. Those two code snippets are in no way equivalent, they do completely different things!
The first one passes the literal strings '`/home/swenson/rubyfunction1.rb`' and '`/home/swenson/rubyfunction2.rb`' as arguments to require. The second one executes two files named /home/swenson/rubyfunction1.rb and /home/swenson/rubyfunction2.rb using the default system shell (CMD.EXE on Windows, /bin/sh on POSIX), gets the standard output as String and passes those strings to require.
Note that in the first case, the backticks ` are part of the filename, whereas in the second case, they are Ruby syntax for calling the Kernel#` method.
So, I think I understand your question correctly, let's say we have 3 files
add.rb
#!/usr/bin/env ruby
def add(a, b)
a + b
end
subtract.rb
#!/usr/bin/env ruby
puts "def subtract(a, b)"
puts " a - b"
puts "end"
example.rb
require './add.rb'
subtract = `./subtract.rb`
puts "I can add: #{add(3, 2)}"
# can't do `subtract`, yet, as we haven't `eval`ed the code even though we've run executed the file
eval subtract
puts "I can now subtract #{subtract(3, 2)}"
And the output of running ruby example.rb on the command line is:
$ ruby example.rb
I can add: 5
I can now subtract 1
So, add.rb just defines a function add. When we require that file, it gets loaded in so we can use that function in our code with no problems.
But, subtract.rb doesn't define a function...it just outputs some code, so running it on the command line looks like:
$ ./subtract.rb
def subtract(a, b)
a - b
end
So now, in our third file example.rb, we require the add.rb and then we can start using add in our code as is, but then we want to execute the subtract.rb (using back ticks here) and capture the output of it. At this point, we can't subtract 2 numbers, because we haven't done anything with the output. Then we use eval to evaluate the output of the subtract method, which will define a method for us, then we can subtract the 2 numbers without a problem.
Note that eval is generally frowned upon because it allows arbitrary code to be executed. Never eval untrusted code unless you know how to tame it. In this case, as #JörgWMittag has pointed out in the comments, this code should be trusted, otherwise you just executed an un-trusted file to get this code. Be careful with user input, though, as that's not trusted.
I'd like to create a Ruby script that prefixes the console output. For example:
I want to implement an interface like this:
puts 'MainLogger: Saying hello'
prefix_output_with('MainLogger') do
system 'echo hello'
end
So this shows up in the console:
MainLogger: Saying hello
MainLogger: hello
What would be a good approach to prefix all of the syscall output to have some logger?
Note: I don't care if we echo what the system call is or not
The important point here is that there's no way to know if system will actually produce output. I'm assuming you don't want a blank MainLogger: whenever a system call doesn't print anything, so you'll need to do the prefixes in the shell:
def prefix_system_calls pre
sys = Kernel.instance_method(:system)
# redefine to add prefix
Kernel.send(:define_method, :system) do |cmd|
sys.bind(self)["#{cmd} | sed -e 's/^/#{pre}: /'"]
end
yield
# redefine to call original method
Kernel.send(:define_method, :system) do |*args|
sys.bind(self)[*args]
end
end
system "echo foo"
prefix_system_calls("prefix") do
system "echo bar"
end
system "echo baz"
# foo
# prefix: bar
# baz
This implementation is pretty fragile, though. It doesn't handle all the different ways you can call system, and prefixes containing special shell characters could cause an error.
I have a task of writing a simple Ruby script which would do the following.
Upon execution from the UNIX command line, it would present the user with a prompt at which he should be able to run certain commands, like "dir", "help" or "exit". Upon "exit" the user should return to the Unix shell.
I'm not asking for the solution; I would just like to know how this "shell" functionality can be implemented in Ruby. How do you present the user with a prompt and interpret commands.
I do not need a CLI script that takes arguments. I need something that creates a shell interface.
The type of program you require can easily be made with just a few simple constructs.
I know you're not asking for a solution, but I'll just give you a skeleton to start off and play around with:
#!/usr/bin/env ruby
def prnthelp
puts "Hello sir, what would you like to do?"
puts "1: dir"
puts "2: exit"
end
def loop
prnthelp
case gets.chomp.to_i
when 1 then puts "you chose dir!"
when 2 then puts "you chose exit!"
exit
end
loop
end
loop
Anyways, this is a simplistic example on how you could do it, but probably the book recommended in the comments is better. But this is just to get you off.
Some commands to get you started are:
somevar = gets
This gets user input. Maybe learn about some string methods to manipulate this input can do you some good. http://ruby-doc.org/core-2.0/String.html
chomp will chop off any whitespace, and to_i converts it to an integer.
Some commands to do Unix stuff:
system('ls -la') #=> outputs the output of that command
exit #=> exits the program
Anyways, if you want this kind of stuff, I think it's not a bad idea to look into http://www.codecademy.com/ basically they teach you Ruby by writing small scripts such as these. However, they maybe not be completely adapted to Unix commands, but user input and the likes are certainly handled.
Edit:
As pointed out do use this at the top of your script:
#!/usr/bin/env ruby
Edit:
Example of chomp vs. chop:
full_name = "My Name is Ravikanth\r\n"
full_name.chop! # => "My Name is Ravikanth"
Now if you run chop and there are no newline characters:
puts full_name #=> "My Name is Ravikanth"
full_name.chop! #=> "My Name is Ravikant"
versus:
puts full_name #=> "My Name is Ravikanth\r\n"
full_name.chomp! #=> "My Name is Ravikanth"
full_name.chomp! #=> "My Name is Ravikanth"
See: "Ruby Chop vs Chomp"
Here's a really basic loop:
#!/user/bin/ruby
#
while true do
print "$ "
$stdout.flush
inputs = gets.strip
puts "got your input: #{inputs}"
# Check for termination, like if they type in 'exit' or whatever...
# Run "system" on inputs like 'dir' or whatever...
end
As Stefan mentioned in a comment, this is a huge topic and there are scenarios that will make this complicated. This is, as I say, a very basic example.
Adding to the two other (valid) answers posted so far be wary of using #!/usr/bin/ruby, because ruby isn't always installed there. You can use this instead:
#!/usr/bin/env ruby
Or if you want warnings:
#!/usr/bin/env ruby -w
That way, your script will work irrespective of differences where ruby might be installed on your server and your laptop.
Edit: also, be sure to look into Thor and Rake.
http://whatisthor.com
http://rake.rubyforge.org
Use irb.
I was looking into an alternative to bash and was thinking along the same lines... but ended up choosing fish: http://fishshell.com/
Nonetheless, I was thinking of using irb and going along the lines of irbtools: https://github.com/janlelis/irbtools
Example:
> irb
Welcome to IRB. You are using ruby 1.9.3p0 (2011-10-30 revision 33570) [x86_64-linux]. Have fun ;)
>> ls #=> ["bin", "share", "opt", "lib", "var", "etc", "src"]
>>
In any case, irb is the ruby shell.
Take a look at cliqr which comes with inbuilt support for build a custom shell https://github.com/anshulverma/cliqr/
I'm working on implementing Project Euler solutions as semantic Ruby one-liners. It would be extremely useful if I could coerce Ruby to automatically puts the value of the last expression. Is there a way to do this? For example:
#!/usr/bin/env ruby -Ilib -rrubygems -reuler
1.upto(100).into {|n| (n.sum.squared - n.map(&:squared).sum)
I realize I can simply puts the line, but for other reasons (I plan to eval the file in tests, to compare against the expected output) I would like to avoid an explicit puts. Also, it allots me an extra four characters for the solution. :)
Is there anything I can do?
You might try running it under irb instead of directly under a Ruby interpreter.
It seems like the options -f --noprompt --noverbose might be suitable (.
#!/usr/bin/env irb -f --noprompt --noverbose -Ilib -rrubygems -reuler
'put your one-liner here'
The options have these meanings:
-f: do not use .irbrc (or IRBRC)
--noverbose: do not display the source lines
--noprompt: do not prefix the output (e.g. with =>)
result = calculate_result
puts result if File.exist?(__FILE__)
result of eval is last executed operation just like any other code block in ruby
is doing
puts eval(file_contents)
an option for you?
EDIT
you can make use of eval's second parameter which is variables binding
try the following:
do_not_puts = true
eval(file_contents, binding)
and in the file:
....
result = final_result
if defined?(do_not_puts)
result
else
puts(result)
end
Is it an option to change the way you run scripts?
script.rb:
$_= 1.upto(100).into {|n| (n.sum.squared - n.map(&:squared).sum)
invoke with
echo nil.txt | /usr/bin/env/ruby -Ilib -rrubygems -reuler -p script.rb, where nil.txt is a file with a single newline.