ruby find and replace portion of string

Question

I have a large file in a ruby variable, it follows a common pattern like so:

// ...

// comment
$myuser['bla'] = 'bla';

// comment
$myuser['bla2'] = 'bla2';

// ...

I am trying to given a 'key' replace the 'value'

This replaces the entire string how do I fix it? Another method I thought is to do it in two steps, step one would be to find the value within the quotes then to perform a string replace, what's best?

def keyvalr(content, key, value)
    return content.gsub(/\$bla\[\'#{key}\'\]\s+\=\s+\'(.*)\'/) {|m| value }
end

I assume you need to keep the comments and other info, so running php on the file to do the replacements would not be an option. — willoller
– willoller, Commented Nov 8, 2012 at 0:40
The problem with regex, is that their siren call will woo you into believing you should try to do everything inside them. Break it down into smaller chunks to preserve your sanity, and don't be afraid to combine the results of the individual chunks with other code. Nobody ever said we had to do everything with regex. Fer instance, see my answer. — the Tin Man
– the Tin Man, Commented Nov 8, 2012 at 1:12

Martin Ender · Accepted Answer · 2012-11-08 00:44:31Z

1

The .* is greedy and consumes as much as possible (everything until the very last '). Make that . a [^'] then it is impossible for it to go past the first closing '.

/(\$bla\[\'#{key}\'\]\s+\=\s+\')[^']*(\')/

I also added parentheses to capture everything except for the value, which is to be replaced. The first set of parens will correspond to \1 and the second to \2. So that you replace the match of this with:

"\1yournewvaluehere\2"

edited Nov 8, 2012 at 0:44

answered Nov 8, 2012 at 0:34

Martin Ender

44.4k11 gold badges93 silver badges132 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

user391986 Over a year ago

Thanks the problem for keyvalr(content, "bla", "newbla") replaces the entire line "$myuser['bla'] = 'bla';" to "newbla"

Martin Ender Over a year ago

oh right. I don't know the correct ruby-syntax for this, but you can create capturing groups like this: /(\$bla\[\'#{key}\'\]\s+\=\s+\'=)[^']*(\')/ and then replace this with "\1newbla\2". \1 and \2 will contain what was matched in the two sets of parentheses.

the Tin Man · Accepted Answer · 2012-11-08 01:27:17Z

I'd use something like:

text = %q{
// ...

// comment
$myuser['bla'] = 'bla';

// comment
$myuser['bla2'] = 'bla2';

// ...
}

from_to = {
  'bla'  => 'foo',
  'bla2' => 'bar'
}

puts text.gsub(/\['([^']+)'\] = '([^']+)'/) { |t|
  key, val = t.scan(/'([^']+)'/).flatten
  "['%s'] = '%s'" % [ key, from_to[key] ]
}

Which outputs:

// ...

// comment
$myuser['bla'] = 'foo';

// comment
$myuser['bla2'] = 'bar';

// ...

This is how it works:

If I do:

puts text.gsub(/\['([^']+)'\] = '([^']+)'/) { |t|
  puts t
}

I see:

['bla'] = 'bla'
['bla2'] = 'bla2'

Then I tried:

"['bla'] = 'bla'".scan(/'([^']+)'/).flatten
=> ["bla", "bla"]

That gave me a key, "value" pair, so I could use a hash to look-up the replacement value.

Sticking it inside a gsub block meant whatever matched got replaced by my return value for the block, so I created a string to replace the "hit" and let gsub do its "thang".

I'm not a big believer in using long regex. I've had to maintain too much code that tried to use complex patterns, and got something wrong, and failed to accomplish what was intended 100% of the time. They're very powerful, but maintenance of code is a lot harder/worse than developing it, so I try to keep patterns I write in spoon-size pieces, having mercy on those who follow me in maintaining the code.

Collectives™ on Stack Overflow

ruby find and replace portion of string

2 Answers 2

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related