Quality of the Source Code Lexicon

Transcription

1 Quality of the Source Code Lexicon Venera Arnaoudova Mining & Modeling Unstructured Data in Software Challenges for the Future NII Shonan Meeting, March 206

2 Psychological Complexity meaningless or incorrect comments can be harmful." variable names of a reasonable length should make programs more understandable [Weissman, 974]

3 Psychological Complexity meaningless or incorrect comments can be harmful." variable names of a reasonable length should make programs more understandable [Weissman, 974] public static int (final String[] els, final String e) { int h = els.length; int l = -; while (h - l > ) { final int m = (l + h) / 2; if (els[m].compareto(e) > 0) { h = m; else { l = m; if (l == - els[l].compareto(e)!= 0) { return -; else { return l;

4 Psychological Complexity meaningless or incorrect comments can be harmful." variable names of a reasonable length should make programs more understandable public static int (final String[] sortedlist, final String element) { int highindex = sortedlist.length; int lowindex = -; while (highindex - lowindex > ) { final int middleindex = (lowindex + highindex) / 2; if (sortedlist[middleindex].compareto(element) > 0) { highindex = middleindex; else { lowindex = middleindex; if (lowindex == - sortedlist[lowindex].compareto(element)!= 0) { return -; else { return lowindex; [Weissman, 974]

5 Psychological Complexity meaningless or incorrect comments can be harmful." variable names of a reasonable length should make programs more understandable [Weissman, 974] public static int (final String[] els, final String e) { int h = els.length; int l = -; while (h - l > ) { final int m = (l + h) / 2; if (els[m].compareto(e) > 0) { h = m; else { l = m; if (l == - els[l].compareto(e)!= 0) { return -; else { return l; public static int (final String[] sortedlist, final String element) { int highindex = sortedlist.length; int lowindex = -; while (highindex - lowindex > ) { final int middleindex = (lowindex + highindex) / 2; if (sortedlist[middleindex].compareto(element) > 0) { highindex = middleindex; else { lowindex = middleindex; if (lowindex == - sortedlist[lowindex].compareto(element)!= 0) { return -; else { return lowindex;

6 Psychological Complexity meaningless or incorrect comments can be harmful." variable names of a reasonable length should make programs more understandable [Weissman, 974] Equal structural complexity: binary search algorithm public static int (final String[] els, final String e) { int h = els.length; int l = -; while (h - l > ) { final int m = (l + h) / 2; if (els[m].compareto(e) > 0) { h = m; else { l = m; if (l == - els[l].compareto(e)!= 0) { return -; else { return l; public static int (final String[] sortedlist, final String element) { int highindex = sortedlist.length; int lowindex = -; while (highindex - lowindex > ) { final int middleindex = (lowindex + highindex) / 2; if (sortedlist[middleindex].compareto(element) > 0) { highindex = middleindex; else { lowindex = middleindex; if (lowindex == - sortedlist[lowindex].compareto(element)!= 0) { return -; else { return lowindex; Extremely short identifiers More descriptive identifiers

7 Measuring the Quality of the Source Code Lexicon Shallow measures, e.g., length of variable names Individual identifiers, e.g., Lexicon Bad Smells [Abebe et al., 2009] Consistency among identifiers, e.g., Linguistic Antipatterns [Arnaoudova et al., 205] Consistency of identifiers with functionality, e.g., Debugging Method Names [Høst et Østvold, 2009]

8 Lexicon Bad Smells (LBS) [Abebe et al., 2009] Terms Structural Meaning Naming Rules Extreme contraction Odd grammatical structure Meaningless terms Useless type indication Misspelled terms Overloaded identifiers Similar names Wrong context Identifier construction Examples: Whole and parts Conceptual Hierarchy No hyponyms Inconsistent term

9 Lexicon Bad Smells (LBS) [Abebe et al., 2009] Terms Structural Meaning Naming Rules Extreme contraction Odd grammatical structure Meaningless terms Useless type indication Misspelled terms Overloaded identifiers Similar names Wrong context Identifier construction Examples: asz Whole and parts Conceptual Hierarchy No hyponyms Inconsistent term

10 Lexicon Bad Smells (LBS) [Abebe et al., 2009] Terms Structural Meaning Naming Rules Extreme contraction Odd grammatical structure Meaningless terms Useless type indication Misspelled terms Overloaded identifiers Similar names Wrong context Identifier construction Examples: asz addition() Whole and parts Conceptual Hierarchy No hyponyms Inconsistent term

11 Lexicon Bad Smells (LBS) [Abebe et al., 2009] Terms Structural Meaning Naming Rules Extreme contraction Odd grammatical structure Meaningless terms Useless type indication Misspelled terms Overloaded identifiers Similar names Wrong context Identifier construction Examples: asz addition() foo() Whole and parts Conceptual Hierarchy No hyponyms Inconsistent term

12 Lexicon Bad Smells (LBS) [Abebe et al., 2009] Terms Structural Meaning Naming Rules Extreme contraction Odd grammatical structure Meaningless terms Useless type indication Misspelled terms Overloaded identifiers Similar names Wrong context Identifier construction Examples: asz addition() foo() namestring Whole and parts Conceptual Hierarchy No hyponyms Inconsistent term

13 Linguistic Antipatterns (LAs) [Arnaoudova et al., 205] Behavior A. Does more than it says B. Says more than it does C. Does the opposite than it says State D. Contains more than it says E. Says more than it contains F. Contains the opposite than it says A. A.2 A.3 A.4 B. B.2 B.3 B.4 B.5 B.6 C. C.2 D. D.2 E. F. F.2 Get - more than an accessor Is returns more than a Boolean Set method returns Expecting but not getting a single instance Not implemented condition Validation method does not confirm Get method does not return Not answered question Transform method does not return Expecting but not getting a collection Method name and return type are opposite Method signature and comment are opposite Says one but contains many Name suggests Boolean but type does not Says many but contains one Attribute name and type are opposite Attribute signature and comment are opposite

14 Linguistic Antipatterns (LAs) [Arnaoudova et al., 205] Behavior A. Does more than it says B. Says more than it does C. Does the opposite than it says State D. Contains more than it says E. Says more than it contains F. Contains the opposite than it says A. A.2 A.3 A.4 B. B.2 B.3 B.4 B.5 B.6 C. C.2 D. D.2 E. F. F.2 Get - more than an accessor Is returns more than a Boolean Set method returns Expecting but not getting a single instance Not implemented condition Validation method does not confirm Get method does not return Not answered question Transform method does not return Expecting but not getting a collection Method name and return type are opposite Method signature and comment are opposite Says one but contains many Name suggests Boolean but type does not Says many but contains one Attribute name and type are opposite Attribute signature and comment are opposite

15 Linguistic Antipatterns (LAs) [Arnaoudova et al., 205] Behavior A. Does more than it says B. Says more than it does C. Does the opposite than it says State D. Contains more than it says E. Says more than it contains F. Contains the opposite than it says A. A.2 A.3 A.4 B. B.2 B.3 B.4 B.5 B.6 C. C.2 D. D.2 E. F. F.2 Get - more than an accessor Is returns more than a Boolean Dimension setbreadth(..) Expecting but not getting a single instance Not implemented condition Validation method does not confirm Get method does not return Not answered question Transform method does not return Expecting but not getting a collection Method name and return type are opposite Method signature and comment are opposite Says one but contains many Name suggests Boolean but type does not Says many but contains one Attribute name and type are opposite Attribute signature and comment are opposite

16 Linguistic Antipatterns (LAs) [Arnaoudova et al., 205] Behavior A. Does more than it says B. Says more than it does C. Does the opposite than it says State D. Contains more than it says E. Says more than it contains F. Contains the opposite than it says A. A.2 A.3 A.4 B. B.2 B.3 B.4 B.5 B.6 C. C.2 D. D.2 E. F. F.2 Get - more than an accessor Is returns more than a Boolean Dimension setbreadth(..) Expecting but not getting a single instance Not implemented condition Validation method does not confirm Get method does not return Not answered question Transform method does not return Expecting but not getting a collection Method name and return type are opposite Method signature and comment are opposite Says one but contains many Name suggests Boolean but type does not Says many but contains one Attribute name and type are opposite Attribute signature and comment are opposite

17 Linguistic Antipatterns (LAs) [Arnaoudova et al., 205] Behavior A. Does more than it says B. Says more than it does C. Does the opposite than it says State D. Contains more than it says E. Says more than it contains F. Contains the opposite than it says A. A.2 A.3 A.4 B. B.2 B.3 B.4 B.5 B.6 C. C.2 D. D.2 E. F. F.2 Get - more than an accessor Is returns more than a Boolean Dimension setbreadth(..) Expecting but not getting a single instance Not implemented condition Validation method does not confirm Get method does not return Not answered question Transform method does not return Expecting but not getting a collection EnterTransport exittransport(..) Method signature and comment are opposite Says one but contains many Name suggests Boolean but type does not Says many but contains one Attribute name and type are opposite Attribute signature and comment are opposite

18 Linguistic Antipatterns (LAs) [Arnaoudova et al., 205] Behavior A. Does more than it says B. Says more than it does C. Does the opposite than it says State D. Contains more than it says E. Says more than it contains F. Contains the opposite than it says A. A.2 A.3 A.4 B. B.2 B.3 B.4 B.5 B.6 C. C.2 D. D.2 E. F. F.2 Get - more than an accessor Is returns more than a Boolean Dimension setbreadth(..) Expecting but not getting a single instance Not implemented condition Validation method does not confirm Get method does not return Not answered question Transform method does not return Expecting but not getting a collection EnterTransport exittransport(..) Method signature and comment are opposite Says one but contains many Name suggests Boolean but type does not Says many but contains one Attribute name and type are opposite Attribute signature and comment are opposite

19 Linguistic Antipatterns (LAs) [Arnaoudova et al., 205] Behavior A. Does more than it says B. Says more than it does C. Does the opposite than it says State D. Contains more than it says E. Says more than it contains F. Contains the opposite than it says A. A.2 A.3 A.4 B. B.2 B.3 B.4 B.5 B.6 C. C.2 D. D.2 E. F. F.2 Get - more than an accessor Is returns more than a Boolean Dimension setbreadth(..) Expecting but not getting a single instance Not implemented condition Validation method does not confirm Get method does not return Not answered question Transform method does not return Expecting but not getting a collection EnterTransport exittransport(..) Method signature and comment are opposite Says one but contains many Name suggests Boolean but type does not boolean _statistics Attribute name and type are opposite Attribute signature and comment are opposite

20 Linguistic Anti-Patterns Detector (LAPD) C++ Java srcml xml XPath identifiers + structural information 6

21 C++ Linguistic Anti-Patterns Detector (LAPD) Submit Clear srcml xml XPath Java identifiers + structural information / JavadocNotVisibleReference Javadoc Stanford Not CoreNLP Visible Reference Stanford javadoc not CoreNLP visible reference Noun Adverb Adjective Noun Please NNenter your RB text JJhere: NN javadoc not visible reference amod Submit Clear neg NN RB JJ NN Part-of-Speech: javadoc not visible reference hidden JJ Extracting identifiers Splitting identifiers Stanford CoreNLP WordNet Part-of-Speech: / 8/4/204 Stanford CoreNLP javadoc hidden reference Visualisation provided using the brat visualisation/annotation software. Copyright 20, Stanford University, All Rights Reserved. amod Collapsed CC-processed endencies: javadoc hidden reference amod Collapsed endencies: javadoc hidden reference amod Basic endencies: javadoc hidden reference Coreference: javadoc hidden reference Named Entity Recognition: javadoc hidden reference Part-of-Speech: JavadocHiddenReference javadoc hidden reference 8/5/204 8/4/204 Stanford CoreNLP Transform terms to lowercase Output format: Output format: Visualise Please enter your text here: Output format: Visualise javadoc not visible reference javadoc not visible reference Submit Clear NN RB JJ NN Part-of-Speech: javadoc not visible reference Named Entity Recognition: javadoc not visible reference javadoc not visible reference Named Entity Recognition: Coreference: javadoc not visible reference javadoc not visible reference Coreference: Basic endencies: NN RB JJ NN javadoc not visible reference amod neg NN RB JJ NN Basic endencies: javadoc not amod visible reference 8/5/204 Named Javadoc Stanford Entity Hidden CoreNLP Recognition: Reference javadoc hidden reference javadoc Stanford hidden CoreNLP reference Please enter your text here: Coreference: javadoc hidden reference Visualisation provided using the brat visualisation/annotation software. Copyright 20, Stanford University, All Rights Reserved. antonyms Visualise javadoc hidden reference amod javadoc Noun hidden Adjective reference Noun Visualisation provided using the brat visualisation/annotation software. Copyright 20, Stanford University, All Rights Reserved. synonyms Output format: Please enter your text here: Collapsed NN CC-processed JJ NN endencies: javadoc hidden reference Basic endencies: javadoc hidden hidden reference reference amod javadoc hidden reference Submit Clear amod amod Collapsed CC-processed endencies: Part-of-Speech: javadoc hidden reference Collapsed endencies: javadoc hidden reference Collapsed endencies: amod Part-of-Speech: javadoc hidden reference amod Named NN NNEntity JJ JJRecognition: NN NN javadoc hidden reference javadoc hidden reference javadoc hidden reference amod Collapsed endencies: Basic endencies: javadoc hidden reference Collapsed Named Entity Coreference: CC-processed Recognition: endenci amod Submit Clear Visualise javadoc hidden reference javadoc hidden reference javadoc hidden reference amod Coreference: Basic endencies: Coreference: Basic endencies: javadoc hidden reference javadoc hidden reference Visualisation provided using the brat visualis Copyright NN 20, JJStanford amod Basic endencies: NNUniversity, All Ri javadoc hidden reference javadoc javadoc hidden hidden reference reference Named Coreference: Entity Recognition: javadoc javadoc hidden hidden reference reference Collapsed endencies: amod Named Entity Recognition: 6

22 Debugging Method Names [Høst et Østvold, 2009] Building semantic profiles from the program bytecode E.g., Semantic profile of an iterator method These methods often call other methods with the same name and create objects. They never return void, write parameter values to fields or call themselves recursively, and very rarely write to fields or return parameter values, and rarely have parameters, contain loops, use local variables, do runtime type-checking or casting, return field values, have branches or have multiple return points. public Iterator iterator() throws DomainRegistryException{

23 Debugging Method Names [Høst et Østvold, 2009] E.g., Semantic profile of an is-<adjective> method public void iscaching(boolean value) { this.caching = value; Name: is-<adjective> Implementation: set-<adjective>: returns void, writes field, parameter to field. iscaching => setcaching

24 Relation with Software Quality and Program Comprehension: the good news LBS improve bug prediction [Abebe et al., 202] LAs are perceived as poor practices by the majority of the surveyed developers (69% and 5% of the external and internal developers, respectively) [Arnaoudova et al., 205]

25 Relation with Software Quality and Program Comprehension: the bad news Developers rename towards abbreviations [Arnaoudova et al., 204] Only 0% of the LAs were fixed by developers [Arnaoudova et al., 205]

26 How did we identify poor lexicon? LBS: smart people got together, discussed, and identified a list of poor practices LAs: sampled source code from open-source projects and looked for inconsistencies among name, type, and comments of programming entities Debugging method names: a corpus of Java projects to identify common semantics for methods with similar names

27 Guidelines for writing good quality lexicon? 9 Naming Conventions Naming conventions make programs more understandable by making them easier to read. They can also give information about the function of the identifier for example, whether it's a constant, package, or class which can be helpful in understanding the code. Identifier Type Rules for Naming Examples Packages The prefix of a unique package name is always written in all lowercase ASCII letters and should be one of the top level domain names, currently com, edu, gov, mil, net, org, or one of the English twoletter codes identifying countries as specified in ISO Standard 366, 98. Subsequent components of the package name vary according to an organization's own internal naming conventions. Such conventions might specify that certain directory name components be division, artment, project, machine, or login names. com.sun.eng com.apple.quicktime.v2 edu.cmu.cs.bovik.cheese Classes Class names should be nouns, in mixed case with the first letter of each internal word capitalized. Try to keep your class names simple and descriptive. Use whole words avoid acronyms and abbreviations (unless the abbreviation is much more widely used than the long form, such as URL or HTML). class Raster; class ImageSprite; Interfaces Interface names should be capitalized like class names. interface RasterDelegate; interface Storing; Methods Methods should be verbs, in mixed case with the first letter lowercase, with the first letter of each internal word capitalized. run(); runfast(); getbackground(); Variables Except for variables, all instance, class, and class constants are in mixed case with a lowercase first letter. Internal words start with capital letters. Variable names should not start with underscore _ or dollar sign $ characters, even though both are allowed. Variable names should be short yet meaningful. The choice of a variable name should be mnemonicthat is, designed to indicate to the casual observer the intent of its use. One character variable names should be avoided except for temporary "throwaway" variables. Common names for temporary variables are i, j, k, m, and n for integers; c, d, and e for characters. int i; char c; float mywidth; static final int MIN_WIDTH = 4; Constants The names of variables declared class constants and of ANSI constants should be all uppercase with words separated by underscores ("_"). (ANSI constants should be avoided, for ease of debugging.) static final int MAX_WIDTH = 999; static final int GET_THE_CPU = ;

28 Challenges for the future How should we identify poor lexicon? prioritize the reporting of poor lexicon? define guidelines for writing good quality lexicon?