在Perl中，如何从缩小的JavaScript源文件中提取某些字符串？

Question

I have this ugly file.

{message:"What this does is, every time the mouse moves in the canvas area, it sets mouseX and mouseY to the location of the mouse.",},{message:"Then, when each ball is updated, it figures out how far away from the mouse it is, and accelerates toward it.",},{message:"The acceleration is the square root of the distance, so it pulls harder when it is really far away. Imagine all the balls being connected to the mouse by little rubber bands or springs. It's a little like that.",},{message:"Try making the balls smaller! And add more of them! I like it with about 40 small balls chasing the mouse.",},{message:"Great job! Like what you learned? Was it fun?",code:"",hiddenCode:"var c = document.getElementById('pane').getContext('2d');\\nfunction rgba(r,g,b,a) {return 'rgba('+[r,g,b,a].join(',')+')';}\\nfunction rgb(r,g,b,a) {return 'rgb('+[r,g,b].join(',')+')';}\\n\\n",lessonSection:"The End",},{message:"Wow, you did everything! Congratulations, nice work! A lot of these are really hard. I'm impre ssed you finished! I hope you enjoyed it!",code:'var pane = document.getElementById(\\'pane\\');\\nvar s = 3;\\n\\npane.onmousemove = function(evt) {\\n c.fillStyle = randomRGBA();\\n var x = evt.clientX;\\n var y = evt.clientY;\\n c.fillRect(x - s / 2, y - s / 2, s, s);};\\n\\nfunction randomRGBA() {\\n var r = randInt(255);\\n var g = randInt(255);\\n var b = randInt(255);\\n var a = Math.random();\\n var rgba = [r,g,b,a].join(",");\\n return "rgba(" + rgba + ")";\\n}\\nfunction randInt(limit) {\\n var x =

I am trying to use Perl regex to extract the body of the message

I trying two 3 hours working on it, but I can not seems to extract it.

My point is to translate the message from English to other languages, so I wanted the string of the message on a clean file instead of working on this ugly file that combine both messages and code.

I was trying to use this code:

use strict;
use warnings;

my $filename = 'test.txt';
my $row = '';

if (open(my $fh, '<:encoding(UTF-8)', $filename)) {
  while ($row = <$fh>) {
    if ($row =~/message:(.*)/)
    {
        print $1 . "\n";
    }
  }
} 
else {
  warn "Could not open file '$filename' $!";
}

It give me results basically the entire file as an output. I tried \\W+ or \\s+ which gave me the first word only.

Any ideas?

Answer 1

Your problem is that the .* that you use in your regex is "greedy". It grabs as much of the input data as possible - which does right to the end of the file.

You need to change that to .*? so that it grabs as little as possible. But you also need to define better markers for the beginning and end of the regex. Looks to me like your message is always in double-quotes. So let's use that.

#!/usr/bin/perl

use strict;
use warnings;
use 5.010;

my $input = do { local $/; <> };

# Look for 'message:', then capture the following " and
# the minimal amount of test until you get the next ". Also
# check for a following comma - to be safe.
while ($input =~ /message:(".*?"),/) {
  say $1;
}

This will work unless your messages have embedded double-quote marks (which will presumably be escaped as \\" ). If that's the case, you'll need something more complex.

Answer 2

The problem is that there are no newlines in the data so your .* matches the whole of the rest of the file. Try /message:"([^"]*)/ which matches only characters that aren't double quotes

I wrote this

use strict;
use warnings;
use 5.010;

my $data = do {
    local $/;
    <DATA>;
};

say "$1: $2" while $data =~ /[{,](\w+):"([^"]*)/g;

__DATA__
{message:"What this does is, every time the mouse moves in the canvas area, it sets mouseX and mouseY to the location of the mouse.",},{message:"Then, when each ball is updated, it figures out how far away from the mouse it is, and accelerates toward it.",},{message:"The acceleration is the square root of the distance, so it pulls harder when it is really far away. Imagine all the balls being connected to the mouse by little rubber bands or springs. It's a little like that.",},{message:"Try making the balls smaller! And add more of them! I like it with about 40 small balls chasing the mouse.",},{message:"Great job! Like what you learned? Was it fun?",code:"",hiddenCode:"var c = document.getElementById('pane').getContext('2d');\nfunction rgba(r,g,b,a) {return 'rgba('+[r,g,b,a].join(',')+')';}\nfunction rgb(r,g,b,a) {return 'rgb('+[r,g,b].join(',')+')';}\n\n",lessonSection:"The End",},{message:"Wow, you did everything! Congratulations, nice work! A lot of these are really hard. I'm impressed you finished! I hope you enjoyed it!",code:'var pane = document.getElementById(\'pane\');\nvar s = 3;\n\npane.onmousemove = function(evt) {\n c.fillStyle = randomRGBA();\n var x = evt.clientX;\n var y = evt.clientY;\n c.fillRect(x - s / 2, y - s / 2, s, s);};\n\nfunction randomRGBA() {\n var r = randInt(255);\n var g = randInt(255);\n var b = randInt(255);\n var a = Math.random();\n var rgba = [r,g,b,a].join(",");\n return "rgba(" + rgba + ")";\n}\nfunction randInt(limit) {\n var x =

which produced this output

message: What this does is, every time the mouse moves in the canvas area, it sets mouseX and mouseY to the location of the mouse.
message: Then, when each ball is updated, it figures out how far away from the mouse it is, and accelerates toward it.
message: The acceleration is the square root of the distance, so it pulls harder when it is really far away. Imagine all the balls being connected to the mouse by little rubber bands or springs. It's a little like that.
message: Try making the balls smaller! And add more of them! I like it with about 40 small balls chasing the mouse.
message: Great job! Like what you learned? Was it fun?
code: 
hiddenCode: var c = document.getElementById('pane').getContext('2d');\nfunction rgba(r,g,b,a) {return 'rgba('+[r,g,b,a].join(',')+')';}\nfunction rgb(r,g,b,a) {return 'rgb('+[r,g,b].join(',')+')';}\n\n
lessonSection: The End
message: Wow, you did everything! Congratulations, nice work! A lot of these are really hard. I'm impressed you finished! I hope you enjoyed it!

No doubt the syntax, whatever it is, allows for embedding double quotes within each string, but there is no example of it in this fragment

Answer 3

I do not know why you need to do this with the minified and concatenated source code, but, you can reverse that:

#!/usr/bin/env perl

use strict;
use warnings;

use Path::Class;
use JavaScript::Beautifier qw/js_beautify/;

my $js = file('combined.min.js')->slurp('<:encoding(UTF-8)');

my $pretty_js = js_beautify($js);

my @messages = ($pretty_js =~ /message: (.+?)\n/g);

print "$_\n" for @messages;

Answer 4

You already have some perl answers, but you may also be interested in the xgettext tool which is designed specifically to extract strings for internationalisation . Run it like this:

xgettext -a --from-code UTF-8 combined.min.js -o - 

It gives you output on each string like this:

#: combined.min.js:36
msgid ""
"Here is a ball that sticks to the mouse.  Every time the mouse moves, the "
"ball redraws on top of the mouse."
msgstr ""

It is in the gnu gettext package. Look at gnu gettext