o b:@sdZddlmZddlZddlmZddlmZddlZddl m Z ddl m Z ddl mZmZdd lmZGd d d Zd d ZGdddZddiZddZGdddeeZddZdddZdddZedkrpedSdS)zGa similarities / code duplication command line tool and pylint checker )print_functionN) defaultdict)groupby)decoding_stream) IRawChecker) BaseCheckertable_lines_from_stats)Tablec@sTeZdZdZ    dddZdddZd d Zd d Zd dZddZ ddZ dS)Similarz,finds copy-pasted lines of code in a projectFcCs"||_||_||_||_g|_dSN) min_linesignore_commentsignore_docstringsignore_importslinesets)selfr rrrrV/var/www/secure340b-portal/env/lib/python3.10/site-packages/pylint/checkers/similar.py__init__$s  zSimilar.__init__NcCsX|dur|j}nt||j}z|jt|||j|j|jWdSty+YdSw)z(append a file to search for similaritiesN) readlinesrrappendLineSetrrrUnicodeDecodeError)rZstreamidstreamencodingrrrr append_stream1s   zSimilar.append_streamcCs||dS)zisz)Similar._display_sims..z)TOTAL lines=%s duplicates=%s percent=%.2fY@N)printr1sortedname _real_linesrstripsumr) rr,Znb_lignes_dupliqueesr(r+r4idxlineZnb_total_lignesrrrr\s&   zSimilar._display_simsccs|j}|j}|j}d}|j}|t|krxd}d} |||D]L} d} tt|||| D]'\} \\} } \} }| |krO| |krH| |||| fVt|| }n| rU| d7} q.| d7} | |krf| |||| fVt|| }q||7}|t|ksdSdS)z+find similarities in the two given linesetsrr/N)enumerate_strippedfindr r1 enumeratezipmax)rr)r*Zlines1Zlines2r@Zindex1r skipr(index2Z non_blank_line1line2rrr _find_commonss:   zSimilar._find_commonccsRt|jddD]\}}|j|ddD]}|||D]}|Vqqq dS)zWiterate on similarities among all files, by making a cartesian product Nr/)rArrI)rr=r4r*simrrrr#szSimilar._iter_sims)r FFFr ) __name__ __module__ __qualname____doc__rrr rrrIr#rrrrr !s  r c Cs|r!td|}dd|jD}ddt|dddD}d }g}d } t|d d D]J\} } | } |rX| sM| d sA| drM| d d} | dd } | rX| | rVd } d} |rd| | |}|rdd} |rp| dd d} | | q+|S)z\return lines with leading/trailing whitespace and any ignored code features removed css(|]}|jt|tjtjffVqdSr )lineno isinstanceastroidImport ImportFrom)r3noderrr s  z!stripped_lines..cSs$i|]\}}|tdd|DqS)css|]\}}|VqdSr r)r3rFZ is_importrrrrWz,stripped_lines...)all)r3rQZnode_is_import_grouprrr sz"stripped_lines..cSs|dS)Nrr)xrrrsz stripped_lines..)keyFNr/)startz"""z'''#r) rSparsejoinbodyrrAstrip startswithendswithgetsplitr) linesrrrtreeZnode_is_import_by_linenoZline_begins_importZcurrent_line_is_importZ strippedlines docstringrQr>rrrstripped_linessB    rlc@sbeZdZdZ   dddZddZddZd d Zd d Zd dZ dddZ ddZ ddZ dS)rz7Holds and indexes all the lines of a single source fileFcCs*||_||_t|||||_||_dSr )r9r:rl_stripped_lines _mk_index_index)rr9rirrrrrrrs zLineSet.__init__cCs d|jS)Nzr9rrrr__str__ zLineSet.__str__cCs t|jSr )r1r:rrrr__len__rrzLineSet.__len__cCs |j|Sr rm)rindexrrr __getitem__rrzLineSet.__getitem__cCs |j|jkSr rp)rotherrrr__lt__s zLineSet.__lt__cCst|Sr )idrrrr__hash__szLineSet.__hash__rccs@|}|r |j|d}n|j}|D] }||fV|d7}qdS)zgreturn an iterator on stripped lines, starting from a given index if specified, else 0 Nr/rt)rZstart_atr=rir>rrrr?s  zLineSet.enumerate_strippedcCs|j|dS)z7return positions of the given stripped line in this setr)rorg)rZ stripped_linerrrr@sz LineSet.findcCs2tt}t|jD] \}}|r|||q |S)zcreate the index for this set)rr"rArmr)rruline_nor>rrrrns zLineSet._mk_indexN)FFFr) rLrMrNrOrrqrsrvrxrzr?r@rnrrrrrs   rR0801)zSimilar lines in %s files %szduplicate-codezIndicates that a set of similar lines has been detected among multiple file. This usually means that the code should be refactored to avoid this duplication.cCs2gd}|t||d7}|t|dddddS)z/make a layout with some stats about duplication)rPnowprevious differencenb_duplicated_linespercent_duplicated_linesr r/)childrencolsZrheadersZcheadersN)rrr )sectstatsZ old_statsrirrrreport_similaritiess rc @seZdZdZefZdZeZddddddfd d d d d dfdd d d ddfddd d ddffZ dde ffZ d ddZ d!ddZ ddZddZddZdS)"SimilarCheckerzchecks for similarities and duplicated code. This computation may be memory / CPU intensive, so you should disable it if you experiment some problems. Z similaritiesmin-similarity-linesr intzz%Minimum lines number of a similarity.)defaulttypemetavarhelpignore-commentsTZynzz,Ignore comments when computing similarities.ignore-docstringsz.Ignore docstrings when computing similarities.ignore-importsFz+Ignore imports when computing similarities.ZRP0801Z DuplicationNcCs(t||tj|ddddd|_dS)Nr T)r rr)rrr r)rlinterrrrrNs  zSimilarChecker.__init__cCsnt||||||dkr|jj|_dS|dkr|jj|_dS|dkr*|jj|_dS|dkr5|jj|_dSdS)zmethod called to set an option (registered in the options list) overridden to report options setting to Similar rrrrN)r set_optionconfigZmin_similarity_linesr rrr)roptnamevalueactionoptdictrrrrUszSimilarChecker.set_optioncCsg|_|jjddd|_dS)z.rr.r} )argsr/rr6rN) r<rrrrr9r&r:r;Z add_messager1rb) rtotalZ duplicatedrr(r+msgr4r=r>rrrcloseus zSimilarChecker.closer )NN)rLrMrNrOrZ__implements__r9MSGSZmsgsoptionsrZreportsrrrrrrrrrrsP    '   rcCs|t|dS)z.required method to auto register this checker N)Zregister_checkerr)rrrrregisterr!rcCs$tdttdt|dS)z&display command line usage informationz*finds copy pasted blocks in a set of filesz~Usage: symilar [-d|--duplicates min_duplicated_lines] [-i|--ignore-comments] [--ignore-docstrings] [--ignore-imports] file1...N)r7sysexit)statusrrrusages rc Cs|dur tjdd}ddlm}d}d}d}d}d}d}||||\}} |D])\} } | d vr4t| }q'| d vrs.      z,=  p  (