Globalisation & Computer systems Week 5/6 Character representation ACII and code pages UNICODE

Representation bits and bytes characters code points glyphs fonts standardization

Representation What is a bit? ‘a binary digit’, i.e either 0 or 1 What is a byte? ‘the fixed no. of bits that can be treated as a unit by the computer hardware’ A byte can be used to express a character such as “A”

Representation ASCII: American standard code for information interchange A standard character encoding system The bytes were originally 7-bits Given this, how many bit patterns? Each pattern maps onto a decimal code point, and that maps onto a character

Representation ASCII: The problem with 7-bit bytes… What about French la tête What about Greek κεφαλη Extend ASCII to 8-bit bytes ISO (International organization for standardization) Now 256 bit-patterns

Representation Extended ASCII: With 8-bit bytes you get 256 bit-patterns For consistency, the first 128 code-points remain the same from ISO-7 The next 128 used for a range of languages For each language, you need an interpretation of these 128 code points The encoding is handled by a code page

Representation What about Chinese? Thousands of characters – 256 bit-patterns clearly not enough Make the bytes bigger… Bytes have 16-bits, which gives bit- patterns UNICODE

Representation Glyphs the pictures used to represent a given pictures used to represent a given character; many to one: The character “A” -> A A AA A A A A A

Representation Glyphs the pictures used to represent a given pictures used to represent a given character; many to one: The character “A” -> A A AA A A A A A Fonts the collection, or ‘picture gallery’ of glyphs

Representation Extended ASCII: With 8-bit bytes you get 256 bit-patterns For consistency, the first 128 code-points remain the same from ISO-7 The next 128 used for a range of languages For each language, you need an interpretation of these 128 code points The encoding is handled by a code page

Representation Extended ASCII: For code point 154: CP_EASTEUROPE (code page 1250): š CP_RUSSIAN (code page 1251): љ What about code point 65 for these two code pages? Now represent your names with your own orthographies in mind, using the code pages

Representation Code pages in VB Public Enum ValidCharsets ANSI_CHARSET = 0 GREEK_CHARSET = 161 THAI_CHARSET = 222 End Enum Private Sub Form_Load() Dim X As New StdFont X.Charset = 161 X.Bold = True X.Size = 8 X.Name = "Times New Roman" Set frmTest.Font = X Set frmTest.Label1.Font = X Set frmTest.Text1.Font = X frmTest.Label1.Caption = Chr(181) + Chr(225) + Chr(226) frmTest.Text1.Text = Chr(181) + Chr(225) + Chr(226) End Sub

Representation and UNICODE What about Chinese? Thousands of characters – 256 bit-patterns clearly not enough

Representation and UNICODE What about Chinese? Thousands of characters – 256 bit-patterns clearly not enough Make the bytes bigger… Bytes have 16-bits, which gives bit- patterns UNICODE

UNICODE – design principles Reference: The Unicode Standard, Version Online:

UNICODE – design principles Principle 1: 16-bit bytes For code pages, characters share 8-bit byte code points – determined by interpretation For UNICODE each character assigned a unique code point code values available Byte 1: 256 values X Byte 2: 256 values for character representation; remaining 2048 reserved for extended 32-bit codes This gives 1, 048, 544 code values to cover all languages

UNICODE – design principles Principle 2: allocation of code space General scripts area: alphabetic CJK Ideographs – ideographs Hangul syllables – Korean Hangul syllables 1 st 128 code points for Latin Punctuation symbols grouped together

UNICODE – design principles Principle 3: efficiency All characters have equal status, i.e. no escape characters Characters of a common script grouped together as far as is possible Common punctuation shared

Design principles Principle 4: logical and display order Logical order: how the code is ordered in memory: follows time sequence of input …and ‘logically’ that is L-R Dynamically composed characters: base character ordered ‘before’, i.e. left wrt to the modifying character

Design principles Principle 5: plain text and rich text Unicode encodes unformatted plain text, where rendering aim is legibility only Formatting: extra data, give rich text To preserve plain text requirements? Have layers of plain text representing characters and how they are formatted Use mark-up languages: content + tags

Design principles Principle 6: unification Share characters where you can: Mixed writing systems Ideographs common to CJK Punctuation

Character semantics Character name Representative glyph Properties

Property 1: Case A letter in the alphabet has several variants UPPERCASE variant lowercase variant Five scripts which have case: Latin, Greek, Cyrillic, Armenian, archaic Georgian

Property 2: Decomposition A character which is equivalent to one or more other characters Š = S + ˇ 0160 (Latin Ext.-A)= C (Basic Latin)

Property 3: Combining class Base character i.e. no special graphical combining behaviour when following another character Combining character Some characters have shape-change or position behaviour when combing with other characters Non-spacing combining character Does not take up space, e.g. diacritics Spacing combining character Takes up space as though a base character

Property 3: Combining class Sequence is a convention: Base character + combining character Symbol: dotted circle, representing the space of the base character, and combining character positioned relative to the circle Stacking of diacritics follows the convention: Move from the base character outwards

Property 4: Directionality Two directionality types: Left to Right Right to Left (Arabic, Hebrew, Syriac, Thaana) Logical sequence: Left to Right

Property 5: General Category The full character space is partitioned into several major categories: Letters Punctuation Symbols Numbers Examples of general category codes: Lu: letter, uppercase; Ll: letter, lowercase Nd: number, decimal digit; No: number, other

Property 6: Numeric value For characters that represent numbers Decimal digits Fractions Subscripts and superscripts Currency numerators Portion of the CJK ideographs: e.g. U+4E94

Property 7: Mirrored property For characters that have equivalent mirror image characters, e.g. ‘(‘ Important for directionality

Character properties Summary 1. Case 2. Decomposition 3. Combining class 4. Directionality 5. General category 6. Numeric value 7. Mirrored property