Regex find/replace over several files...

[roguerum]

Hoping I might be able to get some help on writing a regular expression search/replace function that modifies a bunch of files. I've got the pseudocode, but need help with syntax. Here's an example showing the pattern:
Changing this

Class my.foo.bar {

}

to this

Package my.foo{
	Class bar {

	}
}

Here is the pseudo code I came up with
For each file:
1) Find the line that starts with "class[space]"
2) from that file, strip out everything after [space] to the last period on that line, store in myVar without the final period (in this case: my.foo)
3) add tab before every line in file
4) insert "Package[space][myVar]{" at beginning of file
5) insert "}" at the very end of file
6) save file and close

It's my first crack at Ruby, so even if no one can provide me with actual code, any insight would really help.

thanks in advance!

roguerum. . .

[Fenrir2]

You can convert a dir with this code:

def convert_code(a)
  name = a.match(/Class ([a-zA-Z.]+)/)[1]  # "foo.bar.baz"
  code = a.split("\n")[1..-2].join("\n  ") # the code inside { ... }
  dot = name.rindex('.')
  package, klass = name[0...dot], name[(dot+1)..-1] # "foo.bar", "baz"

  <<CODE
Package #{package} {
  Class #{klass} {
  #{code}
  }
}
CODE
end

def convert_file(src)
  r = convert_code(File.open(src, 'r').read)
  File.open(src, 'w').write(r)
end

def convert_dir(d)
  for item in Dir[d + '/**']
    convert_file(item) if File.file?(item)
  end
end

convert_dir('testdir')

If the code doesn't work, or if you need explanation, ask!

[Mandibal]

A minor suggestion for Fen's regexps. It might be benificial to add a check for 1 or more spaces whenever you expect one. Most languages don't enforce having just one space inbetween keywords so for instance you could have 2 spaces inbetween the class declaration and the name of the class so adding \s+ would cover this case as well.

/Class\s+([a-zA-Z.]+)/

You never know when an extra space character could get inserted when typing something out and it won't be reported as a syntax error because its not.

[Fenrir2]

A good suggestion

Do you have more suggestions? I don't like the first method at all, but I couldn't think of a better way.

[Mandibal]

A good suggestion

Do you have more suggestions? I don't like the first method at all, but I couldn't think of a better way.

Unfortunately no. I just noticed the regular expression because it's something thats bitten me quite often just trying to parse HTML on occasion. I'm still a beginner with Ruby so it looks good to me.