ICU Footprint (Version 2.6)

The following is a breakdown of the size of ICU, both Java and C/C++ versions. All figures are in kilobytes, rounded to 1 KB.

ICU4J

755

32%

Code

1,596

68%

Data

2,351

 

total

Data breakdown

693

43%

General Locale data

388

24%

Collation data (inc UCA rules)

228

14%

Unicode data (props, names,...

178

11%

Translit data

54

3%

Thai break

55

3%

Thai dict

1,596

 

total

ICU4C

1,364

14%

Code

8,373

86%

Data

9,737

 

total

Code breakdown

586

43%

Common library (Unicode strings and properties, conversion, resource bundles, break iteration)

688

50%

I18N library (collation, transforms/transliteration, formatting/parsing, regex)

90

7%

Layout library (text line layout)

1,364

 

total

Data breakdown

4,515

54%

Conversion (160 files, .cnv and alias table)
1,685 20% Collation Data (55 files, collation data only; UCA collation [223KB, 2 files]; UCA Rules [93KB])*

728

9%

Other Locale data (244 files, includes calendar, formatting, etc; excluding collation data)

588

7%

Script transliteration (34 files)

347

4%

Unicode (4 files - properties, property names, character names, normalization)

250

3%

Text boundaries (7 files)

23

0%

IDNA (1 file)

16

0%

Timezone data (1 file)

8,373

 

total

* Note: in ICU2.8 we are in the process of splitting out service data into separate data trees.