A simple solution would be to use replace() with a replacer function.

I use the regex /#(\D+)(\d+)/g. Matching #, followed by one or more non-digits (placed in capture group 1), followed by one or more digits (placed in capture group 2).

^Regexper

All capture groups are passed as arguments of the replacer function. The full match is passed as the first argument, capture group 1 as the second, etc.

Within the replacer function you can then access the value based on the captured attribute name and index.

var data = {
  items: ["item", "HI", "item2"],
  moreitems: ["moreitem", "moreitem1", "HELLO"]
}
var str = "hey #items1 hey #moreitems2";

const result = str.replace(
  /#(\D+)(\d+)/g,
  (_match, attr, index) => data[attr][index]
);
console.log(result);

you can use

• string.replace to replace the substr in the end of the process
• calculate the index in string array by recovered the last character of match string with match[match.length - 1]
• recover category of data to use by removing first and last character category = match.slice(0, -1).substring(1);

var data = {
  items: ["item", "HI", "item2"],
  moreitems: ["moreitem", "moreitem1", "HELLO"]
}

var str = "hey #items1 hey #moreitems2";

var newStr = str;
var match = str.match(/#(.*?)\d+/g);
var index, category;
console.log(match);

match.forEach(match => {
  index = match[match.length - 1];
  category = match.slice(0, -1).substring(1);
  newStr = newStr.replace(match, data[category][index]);
});

console.log(newStr)

const dic       = {
  xxx: 'hello',
  yyy: 'world',
  zzz: 'peace'
}
const string    = 'hey! #xxx, #yyy! #zzz on you'
const newString = string.replace(/#\w+/g, item => dic[item.substring(1)])

console.log(string)
console.log(newString)