Submodel Pattern Extraction for Simulink Models James R. Cordy Queen’s University NECSIS Automotive Partnership Canada
MPE near-miss clones near-miss clone detection Simone analysis of GM models pattern extraction pattern evolution
MPE near-miss clones near-miss clone detection Simone analysis of GM models pattern extraction pattern evolution
Model Pattern Engineering discover, catalogue and formalize submodel patterns emergent domain-specific, client-specific
discovery analysis, identification methodology, techniques classification characterization, formalization notation, tooling, catalogues application deployment, analysis organization, documentation, use cases
why? reuse in model development standards/consistency analysis/enforcement failure/change propagation in model maintenance verification/test optimization deployment variation/optimization model product lines
MPE near-miss clones near-miss clone detection Simone analysis of GM models pattern extraction pattern evolution
code clones copy-paste programming efficient, widely used problematic
bool ConfNextToken (char **p) { // skip white space while (1) switch (**p) { case '\t' : // ignore case ' ' : (*p)++; break; case '\0' : return FALSE; default : return TRUE; }; code clones } type 1 - exact bool ConfNextToken (char **p) { // skip white space while (1) switch (**p) { case '\t' : // ignore case ' ' : (*p)++; break; case '\0' : return FALSE; default : return TRUE; }; } [Roy, Cordy, Koschke SCP 2009]
bool ConfNextToken (char **p) { // skip white space while (1) switch (**p) { case '\t' : // ignore case ' ' : (*p)++; break; case '\0' : return FALSE; default : return TRUE; }; code clones } type 1 - exact bool ConfNextToken (char **p) { while (1) switch (**p) { case '\t': case ' ': // just skip (*p)++; break; case '\0': // eof return FALSE; default: // something we want return TRUE; }; } [Roy, Cordy, Koschke SCP 2009]
bool ConfNextToken (char **p) { // skip white space while (1) switch (**p) { case '\t' : // ignore case ' ' : (*p)++; break; case '\0' : return FALSE; default : return TRUE; }; code clones } type 2 - renamed bool NextToken (char **bp) { while (1) // not really switch (**bp) { case '\t': case ' ': // next (*bp)++; break; case '\0': return 0; default: return 1; }; } [Roy, Cordy, Koschke SCP 2009]
bool ConfNextToken (char **p) { // skip white space while (1) switch (**p) { case '\t' : // ignore case ' ' : (*p)++; break; case '\0' : return FALSE; default : return TRUE; }; code clones } bool NextToken (char **bp) type 3 - near miss { while (1) // not really switch (**bp) { case '\t': (*bp)++ case ' ': break; case '\0': return 0; default: return 1; } } [Roy, Cordy, Koschke SCP 2009]
1 2/3*R*muk 1 Fn Tfmaxk Torque Conversion mus/muk 2 Tfmaxs Ratio of static model clones to kinetic type 1 - exact 1 2/3*R*muk 1 Fn Tfmaxk Torque Conversion mus/muk 2 Tfmaxs Ratio of static to kinetic [Alalfi, Cordy, Dean, Stephan, Stevenson ICSM 2012] [Störrle SSM 2013]
model clones type 2 - renamed [Alalfi, Cordy, Dean, Stephan, Stevenson ICSM 2012] [Störrle SSM 2013]
model clones type 3 - near miss [Alalfi, Cordy, Dean, Stephan, Stevenson ICSM 2012] [Störrle SSM 2013]
MPE near-miss clones near-miss clone detection Simone analysis of GM models pattern extraction pattern evolution
ConQAT graph-based model clone detection [Deissenboeck et al. IWSC 2010]
graph flattening ignores hierarchical structure problems with near-miss
graph flattening ignores hierarchical structure problems with near-miss
code-based near-miss works well NiCad, iClones, others mature, accurate, efficient handles unexpected differences threshold-based, tunable scalable
Renaming, Parsing & Potential Clone Filtering, 4 4 3 3 Extraction Normalization 2 2 1 1 Original Code Base Pretty-printed Normalized Potential Clones Potential Clones 1. Parse / Extract 2. Rename / Filter / Normalize Choose Next (Repeat) Potential Clone as Exemplar 5.pc 12.pc 23.pc Pairwise Cluster 17.pc 15.pc 67.pc 4 18.pc 21.pc . . . 22.pc 3 Comparison 78.pc 63.pc 37.pc . . . Comparable 39.pc 2 . . . 97.pc 44.pc 1 . . . with Exemplar Size PCs . . . Comparable Size Normalized Clone Potential Clone Cluster Potential Clones Classes NiCad 3. Clone Analysis parse - extract - normalize - diff threshold [Roy, Cordy ICPC 2008]
crazy idea: can we use near-miss text code methods on graphical models? “Models are source code too” Mark Harman, keynote at SCAM 2010 [Harman SCAM 2010]
MPE near-miss clones near-miss clone detection Simone analysis of GM models pattern extraction pattern evolution
Simone Simulink near-miss clone detection experiment adapt NiCad near-miss code clone detector to graphical models validate vs. ConQAT for types 1 & 2 hand validate type 3 (near-miss) [Alalfi, Cordy,Dean, Stephan, Stevenson ICSM 2012]
Simulink hybrid hardware/software models widespread in industry - automotive, aerospace, embedded systems mature and interesting at GM
Simulink hierarchical models [Alalfi, Cordy,Dean, Stephan, Stevenson ICSM 2012]
Challenge #1 ... � System { � Name "onoff" � Location [168, 385, 668, 686] � code methods Open on � ModelBrowserVisibility off � ModelBrowserWidth 200 � require text ScreenColor "automatic" � PaperOrientationi "landscape" � PaperPositionMode "auto" � PaperType "usletter" � NiCad requires PaperUnits "inches" � ZoomFactor "100" � AutoZoom on � a parser ReportName "simulink-default.rpt" � Block { � BlockType DiscretePulseGenerator � Name "Discrete Pulse\nGenerator" � Solution: Position [45, 25, 75, 55] � Amplitude "1" � Period "2" � grammar inference PulseWidth "1" � PhaseDelay "0" � SampleTime "1" � } � on Simulink’s Block { � BlockType Product � Name "Product" � internal form Ports [2, 1, 0, 0, 0] � Position [145, 67, 175, 98] � Inputs "2" � SaturateOnIntegerOverflow on � } � ... � } � ... �
Challenge #2 ... � System { � Name "onoff" � Location [168, 385, 668, 686] � what granularity? Open on � ModelBrowserVisibility off � ModelBrowserWidth 200 � ScreenColor "automatic" � NiCad requires PaperOrientationi "landscape" � PaperPositionMode "auto" � PaperType "usletter" � candidates for PaperUnits "inches" � ZoomFactor "100" � AutoZoom on � ReportName "simulink-default.rpt" � comparison Block { � BlockType DiscretePulseGenerator � Name "Discrete Pulse\nGenerator" � Position [45, 25, 75, 55] � Simulink: Amplitude "1" � Period "2" � PulseWidth "1" � model (too big) PhaseDelay "0" � SampleTime "1" � block (too small) } � Block { � BlockType Product � system (just right!) Name "Product" � Ports [2, 1, 0, 0, 0] � Position [145, 67, 175, 98] � Inputs "2" � SaturateOnIntegerOverflow on � } � ... � } � ... �
neutral neutral 1 validated_neutral 1 validated_neutral neutral neutral action action 2 2 up up up up checked_action validated_up 1 checked_action validated_up 1 [reset] reset neutral_up_down [reset] reset neutral_up_down From2 From2 check_up check_up action action 3 3 down down down down checked_action validated_down checked_action validated_down [reset] reset [reset] reset From1 From1 check_down check_down mutually_exclusive mutually_exclusive 4 [reset] 4 [reset] reset Goto1 reset Goto1 even with raw text, find some subsystem clones � � � � � � � � but: 90% irrelevant Simulink internal “formatting” systems some identical systems only 70% same entirely missed exact copies displayed differently
Recommend
More recommend
