"

 COPYRIGHT (c) 1994 by Claus Gittinger

	      All Rights Reserved

 This software is furnished under a license and may be used

 only in accordance with the terms of that license and with the

 inclusion of the above copyright notice.   This software may not

 be provided or otherwise made available to, or used by, any

 other person.  No title to or ownership of the software is

 hereby transferred.

"

"{ Package: 'stx:libbasic' }"

ByteArray variableByteSubclass:#CharacterArray

	instanceVariableNames:''

	classVariableNames:'PreviousMatch DecoderTables EncoderTables DecodingFailedSignal

		EncodingFailedSignal'

	poolDictionaries:''

	category:'Collections-Text'

!

!CharacterArray class methodsFor:'documentation'!

copyright

"

 COPYRIGHT (c) 1994 by Claus Gittinger

	      All Rights Reserved

 This software is furnished under a license and may be used

 only in accordance with the terms of that license and with the

 inclusion of the above copyright notice.   This software may not

 be provided or otherwise made available to, or used by, any

 other person.  No title to or ownership of the software is

 hereby transferred.

"

!

documentation

"

    CharacterArray is a superclass for all kinds of Strings (i.e.

    (singleByte-)Strings, TwoByteStrings and whatever comes in the future.

    All this class does is provide common protocol for concrete subclasses.

    [author:]

    [see also:]

"

! !

!CharacterArray class methodsFor:'initialization'!

initialize

    ]

    "

     CharacterArray initialize

    "

    "Modified: 3.8.1997 / 18:15:59 / cg"

! !

!CharacterArray class methodsFor:'instance creation'!

basicNew

    "return a new empty string"

    ^ self basicNew:0

!

fromBytes:aByteCollection

    "return an instance of the receiver class,

     taking untranslated bytes from the argument, aByteCollection.

     Only useful, when reading twoByteStrings from external sources."

    |mySize nBytes newString dstIdx|

    nBytes := aByteCollection size.

    mySize := self basicNew bitsPerCharacter.

    mySize == 16 ifTrue:[

    ].

    ^ (self basicNew:nBytes) replaceFrom:1 with:aByteCollection

    "

     TwoByteString fromBytes:#[16r21 16r21]

    "

    "Modified: 30.6.1997 / 20:08:37 / cg"

!

fromString:aString

    "return a copy of the argument, aString"

    ^ (self basicNew:(aString size)) replaceFrom:1 with:aString

    "TwoByteString fromString:'hello'"

!

fromStringCollection:aCollectionOfStrings

    "return new string formed by concatenating a copy of the argument, aString"

    ^ self fromStringCollection:aCollectionOfStrings separatedBy:''

    "

     String fromStringCollection:#('hello' 'world' 'how' 'about' 'this')

    "

    "Created: 20.11.1995 / 15:26:59 / cg"

!

fromStringCollection:aCollectionOfStrings separatedBy:aSeparatorString

    "return new string formed by concatenating a copy of the argument, aString"

    |newString first|

    newString := ''.

    first := true.

	first ifFalse:[

	    newString := newString , aSeparatorString

	] ifTrue:[

	    first := false

	].

	newString := newString , s

    ].

    ^ newString

    "

     String fromStringCollection:#('hello' 'world' 'how' 'about' 'this') separatedBy:' '

    "

    "Created: 20.11.1995 / 15:32:17 / cg"

!

fromUTF8Bytes:aByteCollection

    "return a new string which represents the characters as decoded

     from the utf8 encoded bytes, aByteCollection.

     Returns either a normal String, or a TwoByteString instance.

     Only useful, when reading twoByteStrings from external sources.

     This only handles up-to 16bit characters"

    ^ self decodeFromUTF8:aByteCollection.

    "

     CharacterArray fromUTF8Bytes:#[ 16rEF 16rBF 16rBF ]

   rfc2279 examples:

   invalid:

    "

!

new

    "return a new empty string"

    ^ self basicNew:0

! !

!CharacterArray class methodsFor:'Signal constants'!

decodingFailedSignal

    "return the signal, raised when decoding of a string is not possible

     due to invalid characters contained in the source.

     This may happen for example, if a non EUC coded 8-bit string

     is attempted to be decoded into a JIS string."

    ^ DecodingFailedSignal

    "Created: 28.6.1997 / 20:09:55 / cg"

    "Modified: 3.8.1997 / 18:16:47 / cg"

!

encodingFailedSignal

    "return the (query-) signal, raised when encoding of a string is not possible

     due to invalid characters contained in the source."

    ^ EncodingFailedSignal

    "Modified: 28.6.1997 / 20:09:35 / cg"

    "Created: 3.8.1997 / 18:16:40 / cg"

! !

!CharacterArray class methodsFor:'cleanup'!

lowSpaceCleanup

    "cleanup in low-memory situations"

    DecoderTables := EncoderTables := nil

    "

     CharacterArray lowSpaceCleanup

    "

    "Created: 22.2.1996 / 16:30:30 / cg"

    "Modified: 22.2.1996 / 17:58:05 / cg"

! !

!CharacterArray class methodsFor:'encoding & decoding'!

decodeFromUTF8:aStringOrByteCollection

    "return a string which represents the characters as decoded

     from the utf8 encoded bytes, aByteCollection.

     Returns either a normal String, or a TwoByteString instance.

     Only useful, when reading twoByteStrings from external sources.

     This only handles up-to 16bit characters."

    ^ CharacterEncoderImplementations::ISO10646_to_UTF8 decodeString:aStringOrByteCollection

    "

     CharacterArray fromUTF8Bytes:#[ 16rEF 16rBF 16rBF ]

   rfc2279 examples:

   invalid:

    "

! !

!CharacterArray class methodsFor:'pattern matching'!

matchEscapeCharacter

    "return the character used to escape a matchCharacter

     (i.e. make it a regular character in a matchPattern)"

    ^ $\

!

matchScan:matchScanArray from:matchStart to:matchStop with:aString from:start to:stop ignoreCase:ignoreCase

    "helper for match; return true if the characters from start to stop in

     aString are matching the scan in matchScan from matchStart to matchStop.

     The matchScan is as created by asMatchScanArray.

     pattern matching package should be added."

     mStart "{ Class: SmallInteger }"

     mStop  "{ Class: SmallInteger }"

     sStart "{ Class: SmallInteger }"

     sStop  "{ Class: SmallInteger }"

     mSize  "{ Class: SmallInteger }"

     sSize  "{ Class: SmallInteger }"

     index  "{ Class: SmallInteger }"

     quickCheck matchLast nextMatchEntry

     checkChar included|

    mStart := matchStart.

    mStop := matchStop.

    sStart := start.

    sStop := stop.

    [true] whileTrue:[

"/                    ''' against:' , (matchScanArray copyFrom:mStart to:mStop) printString).

    ].

    "

     |scanArray s|

     scanArray := self matchScanArrayFrom:'*hello'.

     s := 'foo bar hello world'.

     CharacterArray

    "

    "

     |scanArray s|

     scanArray := self matchScanArrayFrom:'*hello*'.

     s := 'foo bar hello world'.

     CharacterArray

    "

    "Modified: / 15.10.1998 / 13:39:25 / cg"

!

matchScanArrayFrom:aString

    "scan a pattern string and decompose it into a scanArray.

     This is processed faster (especially with character ranges), and

     can also be reused later. (if the same pattern is to be searched again)"

     idx "{ Class: SmallInteger }"

     end c1 c2 matchSet previous escape|

    previous := nil.

    escape := self matchEscapeCharacter.