Learn about regex (regular expressions) in Java.

Regular Expressions: Lesson

showing the regular expression matches in an email address

Have you ever wondered how search engines like Google Search can interpret the text you type and return results accurately to your inquiry?

Aside from the higher-level steps of crawling, indexing, and ranking, and also ignoring the layers of machine learning behind the information Google has at its disposal, Google, and any other search engine needs a way to interpret and use what you type. It needs to functionally accept whatever you put into that search bar and return results. This process poses the following question:

"How do I generalize all possible inputs but still retain accuracy and precision in the results?"

The answer is [*regular expressions*](https://www.codecademy.com/resources/docs/general/regular-expressions?page_ref=catalog), or *regex* for short.

A regex is a pattern represented by a sequence of characters. They’re used in search engines, word processors, text editors, and many other places due to their general use cases and base functionality. Many programming languages provide regex capabilities either from libraries or built straight into the language. Java is one such language that has an entire package dedicated to regex.

In the simplest sense, regex can be used to perform text search and text replace operations. It’s always first represented as a *search pattern*, which is used to define what exactly is being searched for when scanning text. Many unique ways were developed to add as much utility and specification as possible when creating search patterns. In all regards, search patterns put an emphasis on precision despite the "catch-all" genericness it has to consider.

Regex can be broken into the following categories:
 - Character Classes and Capture Groups
 - Escaped Characters
 - Quantifiers
 - Alternation

We’re going to learn about these categories and how to read, write, and manipulate regex search patterns in Java. Java doesn’t have a built-in regex class, so we import a `java.util.regex` package to work with regular expressions.

Let’s start with talking about this `java.util.regex` package and how it implements regex!

Introduction to Regex

Character classes are one of the primary means of "grouping" characters in a regex pattern without necessarily enforcing an order to those characters. For instance, a character class marks the difference between searching for a specific word and searching for any occurrence of any characters found in that word.

A **character class** matches a character from a specific set, including predefined character classes and custom user-defined sets. The following are examples of how character classes are used (note that any references to "abc" are just examples. These are user-defined, so anything can be used):

| Character Class | Description |
| --- | --- |
| `[abc]` | Includes any of the contained characters |
| `[^abc]` | Excludes all of the contained characters |
| `[a-z]` | Includes any of the contained characters across a range |
| `.` | Any character except newline |
| `\w` | Any word character (alphanumeric and underscore) |
| `\d` | Any digit [0-9] |
| `\s` | Whitespace (spaces, tabs, line breaks) |

These can all be defined as or included in a regex search pattern. Let’s go over a quick example using a character class. Suppose the following is your search pattern:

```java
Pattern pattern = Pattern.compile("[ABCDQ]");
```

The above pattern first creates a character class that checks for any letters from the set {A, B, C, D, Q}.

To see this in action, consider the example of counting the total number of occurrences of characters "B", "C" and "N" in the string "A BANANA CAN FLY BETTER THAN A BIRD":

```java
import java.util.regex.*;
public class Main {
  public static void main(String args[]){
    Pattern pattern = Pattern.compile("[BCN]");
    Matcher matcher = pattern.matcher("A BANANA CAN FLY BETTER THAN A BIRD");                                                                                                        
    int count = 0;
    while(matcher.find())count++;
    System.out.println(count); 
  }
}
```
The output of this program will be 8.

Let's practice using character classes.


Character Classes

Congratulations! You completed the lesson on [Regex](https://www.codecademy.com/resources/docs/general/regular-expressions?page_ref=catalog)! 🙌

Key concepts covered in this lesson include:
 - The Pattern Class
 - The Matcher Class
 - The PatternSyntaxException Class
 - Character Classes and Capture Groups
 - Escaped Characters
 - Quantifiers
 - Alternation


Review

Remember that `+` character used in the character class example? That’s what regex calls a *quantifier*.

Quantifiers look at preceding tokens and indicate how many matches to look for, defaulting to trying to match as many characters as possible unless otherwise defined. The following are examples of quantifiers (note that "a" is user-defined and can be any character or group of characters. Any numerical digits used are also user-defined):

| Quantifier | Description |
| --- | --- |
| `a*` | 0 or more |
| `a+` | 1 or more |
| `a?` | 0 or 1 |
| `a{5}` | Exactly five |
| `a{3,}` | Three or more |
| `a{1,3}` | Between one and three |
| `a+?` | Match as few as possible |
| `a{2,}?` | Match as few as possible with a lower bound |

Patterns with quantifiers are generally straightforward. All they do is grab and group the quantified amount of characters together when found. Let’s cover a quick example:

```java
Pattern pattern = Pattern.compile("a*b");
```

This pattern searches for text that includes 0 or more "a"s followed by a "b." It’ll find any "ab," "aab," "aaaaaaab," but also, because it can match any text that has 0 "a"s followed by a "b," it will also find "b."

Let's practice this concept.


Quantifiers

The `Matcher` class is Java’s "search" feature, allowing it to take a regular expression defined by the `Pattern` class and perform a search based on a compiled pattern. The syntax for the `Matcher` class is as follows:

```java
Pattern pattern = Pattern.compile("Codecademy"); 

// String to match: "Welcome to Codecademy!"
Matcher matcher = pattern.matcher("Welcome to Codecademy!"); 
```
Note that the `Pattern` object is called when using the `matcher()` method. 

The `matcher()` method used by a `Pattern` object returns a `Matcher` object with information about the search. We can then use this `Matcher` object to sift through that information in various ways, with many methods within the `Matcher` class at our disposal. We’ll start by talking about the two most useful methods:

```java
boolean matchFind = matcher.find();
boolean matchExact = matcher.matches();
```

In the above two methods, `find()` searches the string for the next match based on the pattern and returns `true` if it finds an occurrence of it. For example, if the string is "Codecademy" and the pattern is "cad", `find()` will return `true`. The `find()` function retains a pointer to indicate where it should start searching in the text. The pointer initially starts at the beginning of the text and stops at the next character after matching. If `Matcher` is not reset, the next time `find()` is called it will start searching where it left off. Example:

```java
Pattern pattern = Pattern.compile("ABC"); 
Matcher matcher = pattern.matcher("ABCdefgABCdefAB");

matcher.find() // Starting at index 0 - return true
matcher.find() // Starting at index 4 - return true 
matcher.find() // Starting at index 11 - return false
```
This is useful if we are trying to count the number of occurrences of a substring.

The `matches()` function returns `true` if the string matches exactly the pattern with no other characters in the string. Note: this does not necessarily mean that the strings must be the same. We will see why in later exercises.

The `Matcher` class also includes methods to find and replace text, create search regions, append text, display how the matcher is interpreting the pattern, and many more functional and very useful utilities.

The Matcher Class

Java’s `java.util.regex` package contains two vital classes:
 - `Pattern` Class
 - `Matcher` Class

We’ll start with the `Pattern` class which is what we use to define a pattern we want to find in a string. We create a `Pattern` like so:

```java
Pattern pattern = Pattern.compile("Codecademy");
Pattern pattern = Pattern.compile("codecademy", Pattern.CASE_INSENSITIVE);
```

This method compiles a pattern object in such a way as to designate a given string as the pattern to be used to perform regex operations.

Note the second parameter. It’s optional but contains some useful flags we can set that affect how the regex search is performed. Some examples include:
 - `Pattern.CASE_INSENSITIVE` - The case of letters will be ignored when performing a search.
 - `Pattern.LITERAL` - Special characters in the pattern will not have any special meaning and will be treated as ordinary characters when performing a search.
 - `Pattern.UNICODE_CASE` - Use it together with the `CASE_INSENSITIVE` flag to also ignore the case of letters outside of the English alphabet.

To match a string against a pattern, we need to use the `Matcher` class. We will discuss this in the next exercise.


The Pattern Class

We discussed character classes as one of the means to group characters in [regex](https://www.codecademy.com/resources/docs/general/regular-expressions?page_ref=catalog). The other is known as a **capture group**. Capture groups maintain the order but allow you to section the pattern and allow for substring searching.

The following is known as a capture group (note the "aab" here is user-defined, so anything can be used):

| Capture Group | Description |
| --- | --- |
| (aab) | Groups tokens together and creates a capture group |

Grouped tokens can be used for extracting a substring. Let’s go over an example.

```java
Pattern pattern = Pattern.compile("(co)+");
```
Here, the "+" is known as a quantifier (these are covered in a later exercise). The "+" means "occurring 1 or more times".

The above regex pattern matches any text that contains the character combination "co" one or more times. What this will actually match could look like "co" but it would also match with the following: "coco", "cococo", "cocococo", etc.

Character classes and capture groups may look a little similar but just remember: `[]` denotes character class and `()` denotes capture group.

Now let’s get into some unique character specifications in regex.

The `\` is called an **escape character**, which when used in this manner, helps define special characters with special functions. For example, "\n" represents a new line, and "\s" represents a white space. Here are some others:

| Escaped Character | Description |
| --- | --- |
| `\.` | Escaped period |
| `\*` | Escaped asterisk |
| `\\` | Escaped backslash |
| `\t` | Tab |
| `\n` | Newline or linefeed |
| `\r` | Carriage return |

For example, to search for the `.` to mark the end of a sentence without regex trying to interpret it as "any character except newline", we use "\\." in the pattern.

Escaped characters insert reserved, special, and Unicode characters into the regex pattern. All escaped characters begin with the `\` character, which also allows the pattern to search for special characters without enacting their special properties by treating the special character as plaintext.

Note: Some escape character examples could also be found in the character class examples listed in the previous exercise.


Capture Groups and Escaped Characters

*Alternation* is just the [regex](https://www.codecademy.com/resources/docs/general/regular-expressions?page_ref=catalog) way of saying OR. In regex, this is denoted by the following (note the "ab" and "cd" are user-defined):

| Alternation | Description |
| --- | --- |
| `ab\|cd` | Match `ab` OR `cd` |

The use of alternation works just like the boolean OR. The pattern will search for any text that matches either alternated pattern and only return nothing if neither is found. Otherwise, this pattern will group either match in the search results regardless of which pattern matched.

Alternation is also found implicitly in character classes. For instance, a character class like `[A-Z]` grabs a range of characters. This is the equivalent of grabbing `A` OR `B` OR `C` and so on.

This just allows you to get a bit more complicated, especially if you want to look for slight differences between words that share a pattern. For instance, when looking at code, someone might type the following in two possible ways:

```java
i < 10
10 > i
```

Both ways are technically correct and produce the same result. So to make a regex pattern accept both when looking for the same technical thing, you can write a pattern for each and just alternate it with an `|` to ensure you’re finding both feasible ways to write the same code.
