Hi Masood

Firstly, in response to your original question…

When working with regular expressions, you can be far more efficient with your code if you use quantifiers. For example, instead of writing (\d\d\d) you can just write \d{3}. Since you’re not sure how many digits there will be in a row, you can just search for one or more with a plus character: \d+. Each group of digits may or may not have a comma after it—in other words, the comma is optional. We can match an optional character with a question mark: \d+,?. The complete number has one or more of these groups, so our regex now looks like this: (\d+,?)+. Now you only want to match currencies (not phone numbers, years, etc), so lets’s only match strings that have one of the currency symbols before it. We can achieve this with a ‘positive lookbehind’: (?<=\$|£|€). We’re not really interested in everything after the decimal point, so we’ll just ignore it. So… here’s our complete regex:

(?<=\$|£|€)(\d+,?)+

(For more help with regex basics, check out my article: [Getting started with regular expressions](https://inkwire.app/articles/getting-started-with-regular-expressions.html).)

So how do we combine this approach with capture groups, so you can reconstruct each number with $1, $2, etc? Unfortunately we can’t. Quantifiers don’t dynamically create groups as they go. This kind of problem isn’t one we can easily solve with regular expressions alone, so yeah… we need to write a script.

I’ve modified and simplified Ariel’s script to your purposes here. You can define a thousands separator (set to a comma by default) and whether you want to ignore 4-digit numbers (set to true by default). The script will search your selected text, or the whole story if you place your cursor anywhere in the text or select a text frame…

app.doScript (main, undefined, undefined, UndoModes.ENTIRE_SCRIPT, "Format numbers");

function main(){
    // Script settings
    var delimiter = ","; // Thousands separator
    var ignore4 = true // Ignore 4-digit numbers?
    
    // Get selection
    var mySelection = textSelection();
    if (!mySelection) {
        alert("No text selected.");
        return;
    }
    
    var mySourceText;
    var mySelectionType = mySelection.constructor.name;
    if (mySelectionType == "InsertionPoint" || mySelectionType == "TextFrame") {
        // Select whole story
        mySourceText = mySelection.hasOwnProperty("storyOffset") ? mySelection.storyOffset.parentStory : mySelection.insertionPoints[0].parentStory;
    } else {
        mySourceText = mySelection;
    }

    var mySearch = /(?<=\$|£|€)(\d+,?)+/;

    app.findGrepPreferences = null;
    app.findGrepPreferences.findWhat = mySearch.source;
    var myFinds = mySourceText.findGrep();
    var f, n, t;
    while (f = myFinds.pop()){
        n = f.contents.toString().replace(/,/g,''); // Remove existing commas from string
        // Add commas to thousands
        if (!ignore4 || (n && n.length > 4)){
            t = n.split('').reverse().join('').match(/.{1,3}/g).join(delimiter).split('').reverse().join('');
        }
        t && f.contents = t;
    }
    
    function textSelection() {  // Returns text-based object
        if ( app.selection.length == 1 && app.selection[0].hasOwnProperty("findGrep") ) {
            return (app.selection[0]);
        }
        // Invalid selection
        return null;
    }
}

Let me know if you want me to explain anything about the script. Cheers!

Hi Khaled. It’s been a good while since I threw that script together, but looking back at it now, I can see that it only works if all the hyperlinks in the document have URL Destinations. If you have other types of destinations (like Text Destinations for example), the script will fail at the line that tries to get the destinationURL. So let’s fix it…

I think the easiest way is to just catch the error when it happens. Since we’re only interested in finding actual URLs, anything that isn’t a URL can safely be ignored. So, here’s an updated version of the script that does just that…

// Get all our document hyperlinks
var myDocument = app.documents[0];
var myDocumentHyperlinks = myDocument.hyperlinks.everyItem().getElements();
// Create an empty object to store all the URLs (along with linked text)
var indexOfURLs = {};
// Iterate over all the document hyperlinks and add each one to the object
for (var i = 0; i < myDocumentHyperlinks.length; i++) {
    var hyperlink = myDocumentHyperlinks[i];
    // Get linked text (title) and URL
    var title = hyperlink.source.sourceText.contents;
    var url;
    try {
        url = hyperlink.destination.destinationURL;
    } catch (err) {
        // The destination is not a URL. Ignore it.
        continue;
    }
    // Add title and URL to the list
    indexOfURLs[title] = url;
}
// Print out all the URLs
for (var key in indexOfURLs) {
  $.writeln(key + ": " + indexOfURLs[key]);
}

Please let us know how you go. :-) (I find that it’s not uncommon to give up precious time on forums like this one to try and help others only to never hear back from people, as was the case with the OP on this thread. We’re all human and a little ‘thanks for your help’ goes a long way to keeping any community alive and helpful.)

The printer has agreed to print 1 copy x 500 unique artwork files?? That sounds like a crazy way to do it, even if you do work out how to automate the numbering. I’d be choosing a printer who can do the automatic numbering for you, with just one artwork file.

Yep, I was about to say that a positive lookahead is what you need… This means that the letter needs to be there, but it won’t be included in the matched text.

Ah I see. No worries.

Well yes, but Adobe also calls it a ‘shortcut’. :-) I always give my styles these shortcuts, and have done for so many years, I forget that not everyone works with an extended keyboard.

Okay, reading Michel’s comment again, and actually checking the behaviour in InDesign, I think I get it. If you select multiple cells, say in column 1 of a two column table, then shift-click on ANY cell in column 2, all the rows get selected. (It doesn’t matter which row you target.) So I can see how SelectionOptions.ADD_TO just mimics this behaviour.

And I guess app.selection[0].rows.everyItem().select() probably fails because there’s no guarantee of everyItem() delivering a contiguous selection (even though it does in this case). Since InDesign doesn’t support multiple selections, it makes sense that select() would require a single object. Jeremy, this may begin to explain why your selection (going back to your original question) was reported as a single object (Cell) and not a collection?? That seems like a bit of a hack on Adobe’s part though. Perhaps a genuine ExtendScript guru will stumble onto this thread and explain it to us in a way that makes sense.