Viewers: in countries Watching now:
Learn how to find and manipulate text quickly and easily using regular expressions. Author Kevin Skoglund covers the basic syntax of regular expressions, shows how to create flexible matching patterns, and demonstrates how the regular expression engine parses text to find matches. The course also covers referring back to previous matches with backreferences and creating complex matching patterns with lookaround assertions, and explores the most common applications of regular expressions.
In this chapter, we'll learn about capturing groups, and how to use backreferences to access those groups. If you remember back in Chapter 5 when we talked about grouping expressions by using parentheses, I mentioned that grouped expressions are captured by the regex engine. What this means is that as the regex engine is going through and finding matches, it stores the matched portion that's in parentheses. It bookmarks it, and remembers it for later. So for example, let's say we have a regex like A, open parentheses, P, and then the two quantifier after it, L, close parentheses, E.
Well, we already know, obviously, that's going to match apple, but at the same time that it's matching it, the regex engine also says, ah! Inside those parentheses, it matched P, P, L. I am going to store that, and remember it for later. Notice it stores the actual data that was matched; not the expression. So the actual match that it made, it stores that data from the string, so we can have it for later, and this happens automatically, and by default. It doesn't matter if you were using those parentheses for repetition, for alternation, or just helping keep your regex organized.
When the regex engine sees those parentheses, it captures the data that's in there for later. Of course, if it stores that data, then we need a way to be able to access it, and the way that we do that is using backreferences. Backreferences are the tool that allows us to access the captured data, and we can refer to those different backreferences with the syntax backslash, followed by a number. So the first backreference would be backslash one. So the metacharacters you would use for the backreferences would be backslash one, through backslash nine, and those would refer to backreference positions one to nine, referring back to the part that was captured, so we can use it again.
Now, there's typically two ways that you would use those backreferences. The first is as you would refer back in the same expression. So in the same expression where you captured the group, later on you refer back to what was captured earlier. Secondly, what you do is refer back to that captured data after the matching is complete. To do that, you really would need to be inside, like, a programming language, so that the matched data can be stored in a variable, and refer to those different matched portions. Or another place that it comes up is inside a text editor.
If you are doing Find and Replace, then the part that's captured during the Find action, can be referred to while you're doing to Replace action. One important note with these backreferences, though, is that they cannot be used inside character classes, and there is no real reason you would need to. They are kind of two fundamentally different concepts. Character classes is defining a set of characters to match, whereas this is talking about literal data that got matched, so the concepts really shouldn't overlap. Now, as far support for these goes, most regex engines support backslash one, through backslash nine.
Some regex engines actually support backslash ten, through backslash ninety-nine. I find that one through nine is usually enough, and I think it's best to try and stay within that limit, and then if you find that you really do need more than that for a special case, check to see if your platform supports it. There are some regex engines that instead of using the backslash followed by a number, they use dollar sign followed by a number. So if you find that the backslash backreference isn't working, you might just try that dollar sign, and see if that works for you instead. So let's see some concrete examples.
Let's see we've got a regex that is apples, being captured in a group, space, T, O, space, and then backslash, one; that is, a backreference to what was just captured, so that is a reference back to apples. So of course it matches apples to apples. It captured it and referred to it all in the same expression. It would not match apples to oranges; it would not match apple to apple. It would only match apples to apples. Of course, you could use multiple backreferences in the same expression. So if you had A, B; C, D; and E, F, each one being captured, then we can refer to those with backreference three, two, and one, in the reverse order, and that would match A, B; C, D; E, F; E, F; C, D; A, B.
Now, both of those examples are just using literal text, but we can put any expression inside that captured group. For example, let's put alternation in there. So let's say we are looking at an HTML document, and we are looking for any tag that is either an I tag, or an E, M tag. So you see there in the captured group in parentheses, I have got I, or E, M. Any text inside the tag, I have made it not greedy, so we don't accidentally skip over to the end of another tag. At the end, the closing tag should match.
So I have got the forward slash, and then followed by a backreference to whatever was matched before. So it matches hello inside I tags, and hello inside E, M tags. But it does not match hello inside some combination of I and E, M. It has to be the same tag both times. It's the literal text that got matched, not the expression. Let's try these out for ourselves. So let's start with apples to apples, and then for our regular expression, we are going to put in apples, we are going to capture that to, backslash one.
So it matched the whole thing. It doesn't match apple to apples, or any other combination, but it does match exactly this phrase. Let's try our second example: A, B; C, D; E, F; and then backslash 3, backslash 2, backslash 1. And then down here, A, B; C, D; E, F; E, F; C, D; A, B. We can put in more here. Let's say we had G, H, and I, J, and then make the reference here out of sequence.
After the E, F, it's expecting G, H, and I, J, and after the A, B, it's expecting G, H, and I, J. Let's try our example with the HTML tags. I am just going to paste in a couple here, and let's try crafting that real quick. So, we know that we need tags, and inside that tag, we have either I, or E, M. So it matched each one of those. And then after that, we don't care what it is, we'll put in our wildcard; we'll make it not greedy. If we take away the greediness, you will see how it stretches all the way to the end. Make it not greedy; let's put our first part of the closing tag there. See? We are still in good shape.
We know we need the forward slash, and then let's make our backreference 1. Now, notice what it matches here. Italics matched, emphasis matched, this one appears to have matched, but it didn't actually. Notice here that what it's doing is it's being greedy in order to get the match. It's saying, ah! I see an I tag here, and I see an I tag here. If we take either one of those out, then you will see that it doesn't match either of those. So it has to be the same text both times. And you could then come back, and you can continue to enhance this and say alright, I am looking for B tags, and strong tags.
Then it would respond to those as well. B for bold, and let's do strong; my forward slash there. Alright, I think that's a good practical example working with HTML tags. Let's try another one. Let's say that we had a list of people, and we wanted to find people whose names were repeated in their last name. So, for example, John Johnson. John occurs twice in it. So what we are going to do is we are going to say, anytime we have a word boundary, and then we know that the first thing is going to be A to Z for a capital letter, followed by a to z lowercase letter, repeated, and then let's capture that whole thing.
That's their first name, followed by backslash b, for the word boundary. But what we are interested in is not just finding each of the names. We are interested in anything that has a space, followed by a word boundary, and then backslash one, and then son at the end. So now we have John Johnson, Evan Evanson, and you can put in other combinations here in place of son if you wanted other variations, but you can see now we have names that are repeated, like John Johnson, and Evan Evanson. But it did not pick up Eric Erikson, because of course, it's spelled differently.
Let's try one more useful example. Let's say we have a phrase like, Paris in the the spring. Maybe you've seen this before; this is an exercise in how your brain works. You don't notice the double words right away. Well, with regex, we can find those double words. Let's say we are looking for anything that starts of word boundary, and that is any word character, and then after that is a space that is repeated, and then that same word again, followed by another word boundary. We're capturing the word, so it's anything between a boundary, and some spaces -- we don't care how many spaces there are, the new line return doesn't matter -- and then, whatever that word is again gets repeated the second time.
This can be a very helpful regular expression that you can run on a document, and make sure that you didn't accidentally duplicate two words. So now I think you have a good understanding of the fundamentals of how captured groups, and backreferences to those captured groups, work. In the next movie, let's talk about how backreferences work with optional groups.
There are currently no FAQs about Using Regular Expressions.
Access exercise files from a button right under the course name.
Search within course videos and transcripts, and jump right to the results.
Remove icons showing you already watched videos if you want to start over.
Make the video wide, narrow, full-screen, or pop the player out of the page into its own window.
Click on text in the transcript to jump to that spot in the video. As the video plays, the relevant spot in the transcript will be highlighted.