Meta

In a previous post about string parsing, I noted that regular expressions were available, but did no dive into an example.

Parsing a String in D

Jesse Phillips ・ Oct 20 '19

#dlang #tutorial #string

Intro

Regular Expressions (regex) is a way of defining a search pattern. It is a very common operation in programming and comes with an age old joke.

I had a problem which I solved with regular expressions

Now you have two programs

What does that mean

For learning how and what regular expressions look like I suggest this help site regularexpressions.info. But you can keep reading to learn about them and how to take advantage of them using D.

Format Phone Number

I wanted to start with something simple, but too simple seemed not useful. In this example I decided to show changing from one US phone format to another.

import std.regex;
void main()
{
    auto data = "555.876.9846";

    auto re = regex(r"(\d{3})\.(\d{3})\.(\d{4})");

    auto ans = replaceAll(data, re, "($1) $2-$3");

    assert(ans == "(555) 876-9846");
}

There are a number of things in this along with the regular expression.

• Regex
  • groupings
  • digits
  • repeat
  • escape
• regex string literal
• replacement of all matches
• match group referencing

String Literal

string literal

One thing that can be problematic is that the programming language provides special codes string literals can use to represent unprintables. D provides the regex string literal r"" otherwise the above would look like

"(\\d{3})\\.(\\d{3})\\.(\\d{4})"
r"(\d{3})\.(\d{3})\.(\d{4})"

This string is still problematic if you need to use quotes. D also has a wysiwyg literal with back ticks, which is problematic for if that is needed. And D provides... Alright that is enough you can follow the link above.

Regex

std.regex

Digits

\d

The backslash enters into a special "code" in this case d which stands for digit.

Other common codes

• w - word character (alphanumeric)
• s - whitespace

Repeat

\d{3}

Here I'm using an explicit number of instances. It will only match if three digits are written in a row. Other common repeat codes:

• \d* - zero or more
• \d+ - one or more

Grouping

Regular Expressions allows matched content to be stored in groupings. This is done with the parentheses.

r"(\d{3})"

This creates a match group 1. This is used in the replacement string as $1.

Escape Sequence

\.

In regular expressions the dot . has a special meaning, match any character. I wanted to match on the literal character. Thus I escape the special processing using backslash, just like when I escape into into special meaning with \d.