String replacement using a dictionary

Here's a perl version. It creates a hash containing pre-compiled regular expressions and then loops through each line of input applying all the regexes to each line. perl's -i is used for "in place editing" of the input file. You can easily add or change any of the regexes or replacement strings.

Pre-compiling the regexes using qr// greatly improves the speed of the script, which will be very noticable if there are a lot of regexes and/or a lot of input lines to process.

#! /usr/bin/perl -i

use strict;

# the dictionary is embedded in the code itself.
# see 2nd version below for how to read dict in
# from a file.
my %regex = (
    qr/yes/      => 'no',
    qr/stop/     => 'go, go, go!',
    qr/wee-ooo/  => 'ooooh nooo!',
    qr/gooodbye/ => 'hello',
    qr/high/     => 'low',
    qr/why\?/    => 'i don\'t know',
);

while (<>) {
      foreach my $key (keys %regex) {
            s/$key/$regex{$key}/g;
      }
}

Here's another version which reads in the dictionary from the first filename on the command line, while still processing the second (and optional subsequent) filenames:

#! /usr/bin/perl -i

use strict;

# the dictionary is read from a file.
#
# file format is "searchpattern replacestring", with any
# number of whitespace characters (space or tab) separating
# the two fields.  You can add comments or comment out dictionary
# entries with a '#' character.
#
# NOTE: if you want to use any regex-special characters as a
# literal in either $searchpattern or $replacestring, you WILL
# need to escape them with `\`.  e.g. for a literal '?', use '\?'.
#
# this is very basic and could be improved.  a lot.

my %regex = ();

my $dictfile = shift ;
open(DICT,'<',$dictfile) || die "couldn't open $dictfile: $!\n";
while(<DICT>) {
    s/#.*// unless (m/\\#/); # remove comments, unless escaped.
                             # easily fooled if there is an escaped 
                             # '#' and a comment on the same line.

    s/^\s*|\s*$//g ;         # remove leading & trailing spaces
    next if (/^$/) ;         # skip empty lines

    my($search, $replace) = split;
    $regex{qr/$search/} = $replace;
};
close(DICT);


# now read in the input file(s) and modify them.
while (<>) {
      foreach my $key (keys %regex) {
            s/$key/$regex{$key}/g;
      }
}

Started writing this as a comment, but it got too complicated, hence a second perl answer. Given your source file, you can use a neat perl trick to build a regex:

#!/usr/bin/env perl

use strict;
use warnings; 
use Data::Dumper;

#build key-value pairs
my %replace = map { /"(.+)"\s*:\s*"(.+)"/ } <DATA>;
print Dumper \%replace; 

#take the keys of your hash, then build into capturing regex
my $search = join ( "|", map {quotemeta} keys %replace ); 
$search = qr/($search)/;

print "Using match regex of: $search\n";

#read stdin or files on command line, line by line
while ( <> ) { 
    #match regex repeatedly, replace with contents of hash. 
    s/$search/$replace{$1}/g;
    print;
}

__DATA__
"yes"      : "no"
"stop"     : "go, go, go!"
"wee-ooo"  : "ooooh nooo!"
"gooodbye" : "hello"

"high"     : "low"
"why?"     : "i don't know"

We generate a hash using a multi-line pattern match and map to create key value pairs.

We build a search regex, and use the values captured in it to replace.

Using <> is perl's magic filehandle - STDIN or files specified on command line. Much how sed does it. (You can use a file and read it 'normally' for the pattern, the use of DATA is purely illustrative).