Normalization Demo

This page contains a simple applet that demonstrates the differences among the normalization forms discussed in  UTR #15: Unicode Normalization Forms, and the source code used to produce that applet.

The demo uses a subset of the data in the Unicode Character Database (UCD) to allow faster loading with browsers. It supports Latin-1 characters, plus a-circumflex-grave and a-circumflex-acute to show characters with multiple accents. Since many browsers won't actually display the combining accents, substitutes are used in display and editing. (see caveats).

Operation [This button needs Java support in the browser.]

• Hit this button to see this demo.
• Type the source text in the top text field.
• To insert the accents, hit the appropriate button at the top of the applet window.
• To produce the different normalization forms, hit the appropriate button in the middle of the window. This will show the result in the bottom text field.

Source

The following source files are used in the demo applet. In your browser, you may have to view them with View>Page Source to keep the browser from messing up the line breaks and indentation.

• Normalizer.java (the core conversions)
• NormalizerData.java (container for the tables)
• NormalizerBuilder.java (for building the tables from the UCD)
• IntHashtable.java (utility)
• IntStringHashtable.java (utility)

Caveats

• This applet requires Java 1.1.4 or later.
• It has not been tested on a wide variety of browsers.
• a-circumflex-grave (Ấ,ấ) and a-circumflex-acute (Ầ,ầ) may not be displayed correctly in some browsers, and may show up as boxes or question marks.
• In Internet Explorer, hitting a button to type an accent leaves the caret (insertion point) in front of the accent.