"

 Copyright (c) 2007-2012 Tony Garnock-Jones

 This code is based on Squeak's DiffMerge package

 written by Tony Garnock-Jones. Original project's web site:

 http://www.squeaksource.com/DiffMerge

 Permission is hereby granted, free of charge, to any person

 obtaining a copy of this software and associated documentation

 files (the 'Software'), to deal in the Software without

 restriction, including without limitation the rights to use,

 copy, modify, merge, publish, distribute, sublicense, and/or sell

 copies of the Software, and to permit persons to whom the

 Software is furnished to do so, subject to the following

 conditions:

 The above copyright notice and this permission notice shall be

 included in all copies or substantial portions of the Software.

 THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND,

 EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES

 OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND

 NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT

 HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,

 WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING

 FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR

 OTHER DEALINGS IN THE SOFTWARE.

"

"{ Package: 'stx:libtool' }"

Object subclass:#Diff2

	instanceVariableNames:'file1 file2'

	classVariableNames:''

	poolDictionaries:''

	category:'Collections-Sequenceable-Diff2'

!

Object subclass:#Chunk

	instanceVariableNames:'offset length'

	classVariableNames:''

	poolDictionaries:''

	privateIn:Diff2

!

Diff2 subclass:#HuntMcilroy

	instanceVariableNames:'lcs'

	classVariableNames:''

	poolDictionaries:''

	privateIn:Diff2

!

Object subclass:#Candidate

	instanceVariableNames:'file1index file2index chain'

	classVariableNames:''

	poolDictionaries:''

	privateIn:Diff2::HuntMcilroy

!

Diff2 subclass:#MyersUkkonen

	instanceVariableNames:'lcs'

	classVariableNames:''

	poolDictionaries:''

	privateIn:Diff2

!

Object subclass:#Patch

	instanceVariableNames:'chunks snippets'

	classVariableNames:''

	poolDictionaries:''

	privateIn:Diff2

!

Diff2 comment:'Generic diff/comm utilities. Agnostic as to the longestCommonSubsequence algorithm used.

Instance Variables
	file1:		<SequenceableCollection> One of the two files to compare.
	file2:		<SequenceableCollection> The other of the files to compare.

-- 
Copyright (c) 2008 Tony Garnock-Jones <tonyg@lshift.net>
Copyright (c) 2008 LShift Ltd. <query@lshift.net>

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction,including without limitation the rights to use, copy, modify, merge,publish, distribute, sublicense, and/or sell copies of the Software,and to permit persons to whom the Software is furnished to do so,subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
'

!

!Diff2 class methodsFor:'documentation'!

copyright

"

 Copyright (c) 2007-2012 Tony Garnock-Jones

 This code is based on Squeak's DiffMerge package

 written by Tony Garnock-Jones. Original project's web site:

 http://www.squeaksource.com/DiffMerge

 Permission is hereby granted, free of charge, to any person

 obtaining a copy of this software and associated documentation

 files (the 'Software'), to deal in the Software without

 restriction, including without limitation the rights to use,

 copy, modify, merge, publish, distribute, sublicense, and/or sell

 copies of the Software, and to permit persons to whom the

 Software is furnished to do so, subject to the following

 conditions:

 The above copyright notice and this permission notice shall be

 included in all copies or substantial portions of the Software.

 THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND,

 EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES

 OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND

 NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT

 HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,

 WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING

 FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR

 OTHER DEALINGS IN THE SOFTWARE.

"

!

documentation

"

Generic diff/comm utilities. Agnostic as to the longestCommonSubsequence algorithm used.

Instance Variables

        file1:          <SequenceableCollection> One of the two files to compare.

        file2:          <SequenceableCollection> The other of the files to compare.

    [author:]

        Tony Garnock-Jones <tonyg@lshift.com>

    [instance variables:]

    [class variables:]

    [see also:]

"

! !

!Diff2 class methodsFor:'accessing'!

HuntMcilroy

    ^HuntMcilroy

    "Created: / 16-03-2012 / 18:54:30 / Jan Vrany <jan.vrany@fit.cvut.cz>"

!

MyersUkkonen

    ^MyersUkkonen

    "Created: / 16-03-2012 / 18:54:40 / Jan Vrany <jan.vrany@fit.cvut.cz>"

! !

!Diff2 methodsFor:'accessing'!

file1

	^ file1

!

file1: anObject

	file1 := anObject.

	self emptyCaches.

!

file2

	^ file2

!

file2: anObject

	file2 := anObject.

	self emptyCaches.

! !

!Diff2 methodsFor:'diffing'!

comm

	"Returns a collection of similarities and differences between the two files. Each entry in the resulting collection is either (#common -> {...}) or (#different -> ({...} -> {...}))."

	| result common p1 p2 |

	result := OrderedCollection new.

	p1 := 0.

	p2 := 0.

	common := OrderedCollection new.

	self longestCommonSubsequence do: [:entry |

		common := self maybeAddCommonBlock: common to: result

						p1: p1 p2: p2 limit1: entry key limit2: entry value.

		common add: (self file1 at: entry key).

		p1 := entry key.

		p2 := entry value.].

	common := self maybeAddCommonBlock: common to: result

					p1: p1 p2: p2 limit1: file1 size + 1 limit2: file2 size + 1.

	self addCommonBlock: common ifNonEmptyTo: result.

	^ result asArray

!

diff

        "Returns a DiffPatch instance that can be used in future to transform file1 into file2."

        | p |

        p := Diff2::Patch new.

        p initChunks: self diffIndices file1: file1 file2: file2.

        ^ p

    "Modified: / 16-03-2012 / 19:16:55 / Jan Vrany <jan.vrany@fit.cvut.cz>"

!

diffIndices

	"Returns a collection of (DiffChunk -> DiffChunk) associations mapping differing regions of file1 and file2 onto each other."

	| result p1 p2 |

	result := OrderedCollection new.

	p1 := 0.

	p2 := 0.

	self longestCommonSubsequence do: [:entry |

		self maybeAddDiffChunkTo: result p1: p1 p2: p2 limit1: entry key limit2: entry value.

		p1 := entry key.

		p2 := entry value.].

	self maybeAddDiffChunkTo: result p1: p1 p2: p2 limit1: file1 size + 1 limit2: file2 size + 1.

	^ result asArray

!

longestCommonSubsequence

	"The longestCommonSubsequence (LCS) algorithm is at the heart of a diff/comm algorithm."

	self subclassResponsibility.

! !

!Diff2 methodsFor:'private'!

addCommonBlock: aSubCollection ifNonEmptyTo: aCollection

	^ aSubCollection isEmpty

		ifFalse: [aCollection add: #common -> aSubCollection asArray. OrderedCollection new]

		ifTrue: [aSubCollection]

!

emptyCaches

	"Subclasses should implement this to clear any cached state they may have built up."

!

maybeAddCommonBlock: common to: result p1: p1 p2: p2 limit1: limit1 limit2: limit2

	"For internal use by comm."

	((p1 + 1 ~= limit1) or: [p2 + 1 ~= limit2])

			ifTrue: [| newCommon |

					newCommon := self addCommonBlock: common ifNonEmptyTo: result.

					result add: #different -> ((file1 copyFrom: p1 + 1 to: limit1 - 1) ->

											(file2 copyFrom: p2 + 1 to: limit2 - 1)).

					^ newCommon]

			ifFalse: [^ common].

!

maybeAddDiffChunkTo: result p1: p1 p2: p2 limit1: limit1 limit2: limit2

        "For internal use by diffIndices."

        ((p1 + 1 ~= limit1) or: [p2 + 1 ~= limit2])

                        ifTrue: [result add: ((Diff2::Chunk offset: p1 + 1 length: limit1 - p1 - 1) ->

                                                                (Diff2::Chunk offset: p2 + 1 length: limit2 - p2 - 1))].

    "Modified: / 16-03-2012 / 19:13:01 / Jan Vrany <jan.vrany@fit.cvut.cz>"

! !

!Diff2::Chunk class methodsFor:'as yet unclassified'!

negativeSize: s

	"Returns a pseudo-chunk with *negative* length, useful as a kind of zero for destructiveMergeWith: operations intended to build up coverage over some set of chunks."

	^ self new offset: s + 1; length: s negated

!

offset: o length: l

	^ self new offset: o; length: l

! !

!Diff2::Chunk class methodsFor:'documentation'!

copyright

"

 Copyright (c) 2007-2012 Tony Garnock-Jones

 This code is based on Squeak's DiffMerge package

 written by Tony Garnock-Jones. Original project's web site:

 http://www.squeaksource.com/DiffMerge

 Permission is hereby granted, free of charge, to any person

 obtaining a copy of this software and associated documentation

 files (the 'Software'), to deal in the Software without

 restriction, including without limitation the rights to use,

 copy, modify, merge, publish, distribute, sublicense, and/or sell

 copies of the Software, and to permit persons to whom the

 Software is furnished to do so, subject to the following

 conditions:

 The above copyright notice and this permission notice shall be

 included in all copies or substantial portions of the Software.

 THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND,

 EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES

 OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND

 NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT

 HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,

 WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING

 FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR

 OTHER DEALINGS IN THE SOFTWARE.

"

!

documentation

"

A DiffChunk represents a span of items within a collection (e.g. a collection of lines representing a text file).

Instance Variables

        length:                 <Integer> Count of lines within the chunk; 0 is permitted

        offset:                 <Integer> Index of first line within the chunk; 1-based

    [author:]

        Tony Garnock-Jones <tonyg@lshift.com>

    [instance variables:]

    [class variables:]

    [see also:]

"

! !

!Diff2::Chunk methodsFor:'accessing'!

correctForSkewFrom: smallerChunk to: biggerChunk

	"Given a biggerChunk that definitely contains smallerChunk but might have an extra head or tail, updates the receiver to include such an extra head or tail."

	| headSize tailSize |

	headSize := smallerChunk offset - biggerChunk offset.

	tailSize := biggerChunk lastIndex - smallerChunk lastIndex.

	offset := offset - headSize.

	length := length + headSize + tailSize.

!

destructiveMergeWith: aChunk

	| newLastIndex |

	newLastIndex := self lastIndex max: aChunk lastIndex.

	offset := offset min: aChunk offset.

	length := newLastIndex - offset + 1.

!

lastIndex

	"Returns the rightmost index contained in my range. (Offset is the leftmost index.) If my length is zero, will return an index lower than my offset."

	^ offset + length - 1

!

length

	^ length

!

length: anObject

	length := anObject

!

offset

	^ offset

!

offset: anObject

	offset := anObject

!

printOn: aStream

	aStream

		nextPut: $(;

		nextPutAll: self class name;

		nextPutAll: ' offset: ';

		nextPutAll: self offset asString;

		nextPutAll: ' length: ';

		nextPutAll: self length asString;

		nextPut: $).

! !

!Diff2::Chunk methodsFor:'as yet unclassified'!

extractFrom: aCollection

	"Extracts a subcollection from aCollection corresponding to my offset and length."

	^ aCollection copyFrom: offset to: offset + length - 1.

! !

!Diff2::Chunk methodsFor:'comparing'!

< aDiffChunk

	"Used to sort changed chunks during three-way merge; see Diff3"

	^ self offset < aDiffChunk offset

!

= otherChunk

        ^ (otherChunk isKindOf: Diff2::Chunk) and:

        [(self offset = otherChunk offset) and:

        [(self length = otherChunk length)]]

    "Modified: / 16-03-2012 / 19:13:07 / Jan Vrany <jan.vrany@fit.cvut.cz>"

! !

!Diff2::HuntMcilroy class methodsFor:'documentation'!

copyright

"

 Copyright (c) 2007-2012 Tony Garnock-Jones

 This code is based on Squeak's DiffMerge package

 written by Tony Garnock-Jones. Original project's web site:

 http://www.squeaksource.com/DiffMerge

 Permission is hereby granted, free of charge, to any person

 obtaining a copy of this software and associated documentation

 files (the 'Software'), to deal in the Software without

 restriction, including without limitation the rights to use,

 copy, modify, merge, publish, distribute, sublicense, and/or sell

 copies of the Software, and to permit persons to whom the

 Software is furnished to do so, subject to the following

 conditions:

 The above copyright notice and this permission notice shall be

 included in all copies or substantial portions of the Software.

 THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND,

 EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES

 OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND

 NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT

 HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,

 WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING

 FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR

 OTHER DEALINGS IN THE SOFTWARE.

"

!

documentation

"

A HuntMcilroyDiff provides a longestCommonSubsequence algorithm following Hunt and McIlroy 1976 for use by the methods on GenericDiff.

J. W. Hunt and M. D. McIlroy, An algorithm for differential file comparison, Bell Telephone Laboratories CSTR #41 (1976).

http://www.cs.dartmouth.edu/~doug/

Instance Variables

        lcs:            cached longest common subsequence

    [author:]

        Tony Garnock-Jones <tonyg@lshift.com>

    [instance variables:]

    [class variables:]

    [see also:]

"

! !

!Diff2::HuntMcilroy methodsFor:'accessing'!

emptyCaches

	lcs := nil.

! !

!Diff2::HuntMcilroy methodsFor:'diffing'!

longestCommonSubsequence

        | equivalenceClasses candidates |

        lcs ifNotNil: [ ^ lcs ].

        equivalenceClasses := self computeEquivalenceClasses.

        candidates := OrderedCollection with: (Candidate new 

                        file1index: 0

                        file2index: 0

                        chain: nil).

        file1 withIndexDo: 

                [ :line :file1index | 

                self 

                        mergeCandidates: candidates

                        file1index: file1index

                        file2indices: (equivalenceClasses 

                                        at: line

                                        ifAbsent: #()) ].

        lcs := self postprocessCandidateChain: candidates.

        ^ lcs

    "Modified: / 16-03-2012 / 19:14:04 / Jan Vrany <jan.vrany@fit.cvut.cz>"

! !

!Diff2::HuntMcilroy methodsFor:'private'!

computeEquivalenceClasses

	| result |

	result := Dictionary new.

	file2 withIndexDo: 

		[ :line :index | 

		(result 

			at: line

			ifAbsentPut: [ OrderedCollection new ]) add: index ].

	^ result

!

findCandidateFrom: candidates forLine: file2index startingAt: lowIndex

	"Find the index k in the given subrange of candidates where file2index falls strictly between the file2indexes of the kth and k+1th candidates. If no such k exists, return 0."

	(lowIndex to: candidates size) do: [ :k |

		(candidates at: k) file2index >= file2index ifTrue: [^ 0].

		(k = candidates size or: [ (candidates at: k + 1) file2index > file2index ])

			ifTrue: [^ k] ].

	^ 0

!

mergeCandidates: candidates file1index: file1index file2indices: file2indices 

        | r c s newCandidate |

        r := 1.

        c := candidates at: r.

        file2indices do: 

                [ :file2index | 

                s := self findCandidateFrom: candidates forLine: file2index startingAt: r.

                s > 0 ifTrue: 

                        [ newCandidate := Candidate new   

                                file1index: file1index

                                file2index: file2index

                                chain: (candidates at: s).

                        self storeCandidate: c at: r in: candidates.

                        c := newCandidate.

                        r := s + 1.

                        "optimise by leaving early if s was the end of the candidates list, since none of the subsequent file2indices will have a place to go"

                        s = candidates size ifTrue: 

                                [ self storeCandidate: c at: r in: candidates. ^ self ] ] ].

        self storeCandidate: c at: r in: candidates.

    "Modified (format): / 16-03-2012 / 19:08:50 / Jan Vrany <jan.vrany@fit.cvut.cz>"

!

postprocessCandidateChain: candidates 

	| result c |

	result := OrderedCollection new.

	c := candidates at: candidates size.

	[ c chain notNil ] whileTrue: 

		[ result add: c file1index -> c file2index.

		c := c chain ].

	^ result reversed.

!