refcode and duplicated code in dude

List overview All Threads
Download

newer

older

Mondrian new method...

build notification to the mailing...

Usman Bhatti

7 May 2012 7 May '12

5:02 p.m.

Attachments:

attachment.htm (text/html — 359 bytes)

Show replies by thread

Richard Wettel

7 May 7 May

6:17 p.m.

referenceCode and duplicateCode are the two pieces that are similar. From the perspective of the algorithm, the referenceCode is the fragment of code that gets compared with all the other fragments, while duplicateCode is the one that matches referenceCode. There is no other (semantic) difference between them. The signature of a duplication chain describes the sequence of exact chunks and non-matching gaps. For example: E4.D1.E3 means: - an exact chunk of 4 lines (4E) - followed by 1 deleted line (1D) which means the line exists in the referenceCode but not in the duplicateCode - finally a exact chunk of 3 (3E) If this is confusing, you can read more about it in my bachelor thesis or in the icsm2005 paper from: http://www.inf.usi.ch/phd/wettel/publications.html I am talking about dude here, but I guess smalldude did not change its semantics. Cheers Ricky On May 7, 2012, at 17:02, Usman Bhatti <usman.bhatti(a)gmail.com> wrote:

...

Some questions on dude. I am using an external tool for computing duplications and trying to instantiate appropriate entities in moose: 1. What is referenceCode and duplicateCode in the context of SmallDude and what is the difference between the two? 2. What is a signature? tx Usman _______________________________________________ Moose-dev mailing list Moose-dev(a)iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev

Usman Bhatti

8 May 8 May

9:35 p.m.

Ok then it seems that Dude implementation targets a specific approach whereby only 2 fragments are compared against each other. What I am trying to represent in Moose are CloneClasses (code fragments) that are copied at least 6 times (computed with an external tool). One possible way is to use DudeCodeFragment class to represent a clone class and each member/occurence of the code is represented with DudeDuplication. But in that case, DudeDuplication lacks the notion of an entity that contains a duplicated code. Is this the right way or should I look somewhere else? thanx Usman On Mon, May 7, 2012 at 6:17 PM, Richard Wettel <richard.wettel(a)usi.ch>wrote;wrote:

...

Some questions on dude. I am using an external tool for computing

duplications and trying to instantiate appropriate entities in moose:

1. What is referenceCode and duplicateCode in the context of SmallDude

and what is the difference between the two?

2. What is a signature? tx Usman _______________________________________________ Moose-dev mailing list Moose-dev(a)iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev

_______________________________________________ Moose-dev mailing list Moose-dev(a)iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev

Tudor Girba

10:44 p.m.

Hi, On 8 May 2012, at 21:35, Usman Bhatti wrote:

...

You should use DudeMultiplication, which holds a collection of DudeDuplication entities. Cheers, Doru

...

thanx Usman On Mon, May 7, 2012 at 6:17 PM, Richard Wettel <richard.wettel(a)usi.ch> wrote: referenceCode and duplicateCode are the two pieces that are similar. From the perspective of the algorithm, the referenceCode is the fragment of code that gets compared with all the other fragments, while duplicateCode is the one that matches referenceCode. There is no other (semantic) difference between them. The signature of a duplication chain describes the sequence of exact chunks and non-matching gaps. For example: E4.D1.E3 means: - an exact chunk of 4 lines (4E) - followed by 1 deleted line (1D) which means the line exists in the referenceCode but not in the duplicateCode - finally a exact chunk of 3 (3E) If this is confusing, you can read more about it in my bachelor thesis or in the icsm2005 paper from: http://www.inf.usi.ch/phd/wettel/publications.html I am talking about dude here, but I guess smalldude did not change its semantics. Cheers Ricky On May 7, 2012, at 17:02, Usman Bhatti <usman.bhatti(a)gmail.com> wrote:

_______________________________________________ Moose-dev mailing list Moose-dev(a)iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev _______________________________________________ Moose-dev mailing list Moose-dev(a)iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev

-- www.tudorgirba.com "Sometimes the best solution is not the best solution."

Usman Bhatti

9 May 9 May

11:18 a.m.

On Tue, May 8, 2012 at 10:44 PM, Tudor Girba <tudor(a)tudorgirba.com> wrote:

...

Hi, On 8 May 2012, at 21:35, Usman Bhatti wrote:

Ok then it seems that Dude implementation targets a specific approach

whereby only 2 fragments are compared against each other. What I am trying to represent in Moose are CloneClasses (code fragments) that are copied at least 6 times (computed with an external tool).

One possible way is to use DudeCodeFragment class to represent a clone

class and each member/occurence of the code is represented with DudeDuplication. But in that case, DudeDuplication lacks the notion of an entity that contains a duplicated code.

Is this the right way or should I look somewhere else?

You should use DudeMultiplication, which holds a collection of DudeDuplication entities.

This does not resolve the problem for me. Multiplication is just a place holder for dude duplications and Dude is made for a 1-1 duplication model and I do not have the same model. I'll try to create my own representation.

...

Cheers, Doru

thanx Usman On Mon, May 7, 2012 at 6:17 PM, Richard Wettel <richard.wettel(a)usi.ch>

wrote:

referenceCode and duplicateCode are the two pieces that are similar.

From the perspective of the algorithm, the referenceCode is the fragment of code that gets compared with all the other fragments, while duplicateCode is the one that matches referenceCode. There is no other (semantic) difference between them.

The signature of a duplication chain describes the sequence of exact

chunks and non-matching gaps. For example:

E4.D1.E3 means: - an exact chunk of 4 lines (4E) - followed by 1 deleted line (1D) which means the line exists in the

referenceCode but not in the duplicateCode

- finally a exact chunk of 3 (3E) If this is confusing, you can read more about it in my bachelor thesis

or in the icsm2005 paper from:

http://www.inf.usi.ch/phd/wettel/publications.html I am talking about dude here, but I guess smalldude did not change its

semantics.

Cheers Ricky On May 7, 2012, at 17:02, Usman Bhatti <usman.bhatti(a)gmail.com> wrote: > Some questions on dude. I am using an external tool for computing

duplications and trying to instantiate appropriate entities in moose:

> > 1. What is referenceCode and duplicateCode in the context of SmallDude

and what is the difference between the two?

2. What is a signature? tx Usman _______________________________________________ Moose-dev mailing list Moose-dev(a)iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev

-- www.tudorgirba.com "Sometimes the best solution is not the best solution." _______________________________________________ Moose-dev mailing list Moose-dev(a)iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev

4382

days inactive

4384

days old

moose-dev@list.inf.unibe.ch

Manage subscription

4 comments

3 participants

tags (0)

participants (3)

Richard Wettel
Tudor Girba
Usman Bhatti