Regex Guru

Regex Support Added to The RTL in Delphi XE and C++Builder XE

Jan Goyvaerts — Tue, 24 Aug 2010 08:18:48 +0000

Delphi XE and C++Builder XE will ship with regular expression support built right into the runtime library. The RegularExpressions unit is based on code originally written by Vincent Parrett. It implements a set of classes (records, actually) that mimics the regex classes in the .NET framework. That includes the whole set of convenient static methods as well as automatic cleanup when the record goes out of scope.

The RegularExpressions unit uses the RegularExpressionsCore unit to do the heavy lifting. The RegularExpressionsCore unit is a modified version of the PerlRegEx unit developed by yours truly. This unit defines a single Delphi class called TPerlRegEx that wraps the functionality of the PCRE open source library, with added methods for replacing and splitting. You can use the RegularExpressionsCore unit directly, just as you can already use the PerlRegEx unit in older versions of Delphi.

Vincent and I donated copies of our code to Embarcadero. We keep the full rights to our code, while Embarcadero gets full rights to their copy. That means I can continue to distribute TPerlRegEx for older versions of Delphi. I’ll release a new version of TPerlRegEx that includes some changes I made prior to donating the code to Embarcadero.

I’ll blog about these two units with more technical details once RAD Studio XE ships. The next free minor updates of RegexBuddy and RegexMagic will support Delphi XE and X++Builder XE using these new units, while retaining support for TPerlRegEx for older versions of Delphi.

Regulární výrazy—Kuchařka programátora

Jan Goyvaerts — Mon, 21 Jun 2010 05:07:44 +0000

I received my author copy of “Regulární výrazy—Kuchařka programátora” last week. This is the Czech translation of Regular Expressions Cookbook. You can buy Regulární výrazy Kuchařka programátora from the publisher cpress.cz or any bookstore that sells Czech language books. Ask for ISBN 978-80-251-1935-8.

While the Russian, German, and Japanese translations are full translations, the Czech translation is an abridged translation. It has only 7 chapters compared to 8 with the original English edition. The last chapter, “Markup and Data Interchange”, was omitted. This chapter has examples on using regular expressions on HTML and XML files and demonstrates some more complex regex techniques. All the other chapters were translated completely.

Replacement Text Syntax for JavaScript’s String.replace()

Jan Goyvaerts — Tue, 08 Jun 2010 10:16:09 +0000

A RegexBuddy user told me that he couldn’t easily find a detailed explanation of the replacement text syntax supported by the String.replace() function in JavaScript. I had to admin that my own web page about JavaScript’s regular expression support was also lacking. I’ve now added a new Replacement Syntax section that has all the details. I’ll summarize it here:

$1: Text matched by the first capturing group or the literal text $1 if the regex has no capturing groups.
$99: Text matched by the 99th capturing group if the regex has 99 or more groups. Text matched by the 9th capturing group followed by a literal 9 if the regex has 9 or more but less than 99 groups. The literal text $99 if the regex has fewer than 9 groups.
$+: Text matched by the highest-numbered capturing group. Replaced with nothing if the highest-numbered group didn’t participate in the match.
$&: Text matched by the entire regex. You cannot use $0 for this.
$` (backtick): Text to the left of the regex match.
$' (single quote): Text to the right of the regex match.
$_: The entire subject string.

正規表現クックブック

Jan Goyvaerts — Fri, 02 Apr 2010 10:36:51 +0000

I’ve just been informed that O’Reilly will publish the Japanese translation of Regular Expressions Cookbook on 14 April. The Japanese title is 正規表現クックブック. You can order 正規表現クックブック from O’Reilly, Amazon.co.jp (free shipping in Japan), or wherever Japanese language technical books are sold.