Bash pattern matching is an indispensable concept that comes in handy when selecting different filenames from a directory and checking if a string matches a given format. Whether you are starting out with bash pattern matching or looking to brush up on your skills, this guide covers the various ways and tips for pattern matching.
Pattern Matching
Bash mainly offers two types of pattern matching: globs and extended globs. The two offer simple ways for users to select the filenames and check the different strings against specific rules. However, Bash also supports using regex, especially when working with scripts to validate various user input and parsing data.
1. Bash Pattern Matching with Globs
Glob pattern matching mainly involves checking for specific filenames in a given directory. They are referred to as wildcards. There are three types of bash glob patterns:
2. Asterisk (*)
It is used when you want to match zero or more filenames.
Let’s have some examples.
Checking Filenames Starting or Ending with a Given Character
For instance, if we needed to search for a file starting with l or li, we could use the following commands:
ls li*
Note how it matches all the filenames that start with li or l in the current directory, including the directories.
To only match with directories, add the / after the wildcard, as shown in the following:
Similarly, if we needed to match the filenames that end with a given character or characters, we could use the wildcard as shown in the following illustration. In this example, we match the filenames that end with e.
Suppose that you are unsure of the filename that you want to search for but you know the given characters. As in the following example, you can still match the filenames by adding the wildcard before and after the characters.
We matched the filenames that include the nux keyword in their name.
Checking Files with a Given Extension
With Bash pattern matching, you can locate the files with a given extension. For instance, if you want to find all the pdfs or text files. In that case, you add the wildcard before the file extension. Let’s see the following example to match all the text files:
All the text files are matched as shown in the following:
3. Question Mark (?)
The question mark wildcard matches the filenames that match a single character. However, you must know the number of characters that a given filename has, such that each “?” represents each character.
For instance, to match a filename with four characters, but we know the last character, we could have the following example:
Similarly, to find a given extension where you know the number of characters of the filename, we could use the following command:
4. Square Brackets []
When you want a more specific search such as matching the filenames that contain the given characters, the characters to match can be added in the square brackets.
For instance, to match the filenames that contain any characters or numbers specified regardless of the extension, we can have the following command:
Here, we match the filenames that contain any character from e to m and has any extension.
If you wanted to match the filenames that start with specific characters, let’s say starting with [f-l] and having a given extension, we could use the following pattern:
5. Bash Pattern Matching with Extended Globs
Extended globs have a huge resemblance to regular expressions. However, they are turned off by default. When you must use them, you have to turn them on using the following command:
The extended globs include:
6. ?(patterns)
It is used when matching with zero or one occurrence of the specified pattern.
For instance, if you want to match all the filenames with a given extension or those that contain specific characters in their naming, we could have the following command:
7. *(patterns)
It is used when matching with zero or more occurrences of the specified pattern.
For instance, we could use the following command to match the filenames that start with zero or more “f” and are text files:
8. +(patterns)
It is used when matching with one or more occurrences of the specified pattern.
If we run the previous command, note that it only matches the filenames that must contain at least one “f” which means that the .txt file won’t be a match.
9. @(patterns)
It is used when matching with one occurrence of the specified pattern.
If we only need to match the filenames that contain only one “f” using the same pattern that is used in the previous example, we could use the @ as shown in the following:
10. !(patterns)
It is used when you want to match everything else that doesn’t match the specified pattern.
Let’s have an example where we want to match all the filenames that don’t have the specified pattern. In this example, we match all the filenames that are not text files or Bash scripts. The command is as follows:
Conclusion
This guide presented the Bash pattern matching and the various examples to ensure that you understand how to work with the various pattern matching options. Using this foundation, you can advance your Bash pattern matching and use the same logic when you want to match the various patterns on the terminal or when creating the Bash scripts.
