String search logic - not language specific - string

Question

I'm having to check data entry on an address field. The client does not want users to use terms like Rd. or Rd for road, ave or ave. for avenue etc. I have no problem with most of the terms. Where I have issues is with 'Ave' lets say. If I look for ' AVE ', that's fine but it will not pick up on ' AVE' at the end of the string and if I look for ' AVE' it will get a false positive on ' Avenue' since it will find ' Ave' within that string. Anyone have an idea of how I can go about this?
Thank you for any help.
Norst
Although the Q: is not language specific, here is how I'm going about this in JS:
//function to check address for rd rd. ave ave. st st. etc
function checkaddy() {
//array of items to look for
var watchfor = Array();
watchfor[0] = " RD";
watchfor[1] = " RD.";
watchfor[2] = " AVE ";
watchfor[3] = " AVE.";
watchfor[4] = " ST ";
watchfor[5] = " ST.";
watchfor[6] = " BLVD.";
watchfor[7] = " CRT ";
watchfor[8] = " CRT.";
watchfor[9] = " CRES ";
watchfor[10] = " CRES.";
watchfor[11] = " E ";
watchfor[12] = " E.";
watchfor[13] = " W ";
watchfor[14] = " W.";
watchfor[15] = " N ";
watchfor[16] = " N.";
watchfor[17] = " S ";
watchfor[18] = " S.";
watchfor[19] = " PKWY ";
watchfor[20] = " PKWY.";
watchfor[21] = " DR ";
watchfor[22] = " DR.";
//get what the user has in the address box
var addcheck = $("#address").val();
//upper case the address to check
addcheck = addcheck.toUpperCase();
//check to see if any of these terms are in our address
watchfor.forEach(function(i) {
if (addcheck.search(watchfor[i]) > 0 ) {
alert("Found One!");
}
});
}

Answer 1

Perhaps you need to look for word boundary character \b. Here are some Ruby examples:
irb(main):002:0> " Avenue" =~ / AVE\b/i
=> nil
irb(main):003:0> " Ave" =~ / AVE\b/i
=> 0
irb(main):005:0> " Ave" =~ /\bAVE\b/i
=> 1
irb(main):006:0> " Ave" =~ /\bAVE\b/i
Notice how " Avenue" doesn't match while " AVE" does match. Also notice how the '\b' behaves and we get 0 and 1 respectively.
There are other characters classes as well in regular expressions. So all you need to do is formulate correct REs for your problem set.
I hope that helps.

Answer 2

Why would your client insist on spelling out names? The United States Postal Service actually encourages abbreviations. Not only that, they prefer addresses to be in all uppercase letters and no more than 5 lines. Such specifications are what their automated sorters were built for. But I digress.
To actually answer your question, you may consider the following code. There was a mistake in your forEach declaration. You were using i as an index, but, in fact, the forEach function uses the whole entry of the array. I modified it below. Also, because we're using a string expression in the constructor for the RegExp, the \ in the \b has to be escaped, so we add two \'s inside the string. Because we using the \b construct for word boundaries, we don't need to add extra periods into the test array. I hope you find this helpful.
//array of items to look for
var watchfor = ['RD','AVE','ST','BLVD','CRT','CRES','E','W','N','S','PKWY','DR'];
//function to check address for rd rd. ave ave. st st. etc
function checkaddy(address) {
//check to see if any of these terms are in our address
watchfor.forEach(function(entry) {
var patt1 = RegExp('.*\\b' + entry + '\\b.*','gim');
if (patt1.test(address)) {
document.write("Found " + entry);
}
});
}