v5`ddlmZddlmZddlmZ ddlZn#e $r ddl m Z dZYnwxYwgdZ Gd d eZ Gd d eZ Gd deZe Ze ZeZdS))SequenceMatcher) find_ngrams)BaseSimilarityN)array)lcsseqlcsstrratcliff_obershelpLCSSeqLCSStrRatcliffObershelpc2eZdZdZd dZdZdZdZd ZdS) r zplongest common subsequence similarity https://en.wikipedia.org/wiki/Longest_common_subsequence_problem rNTc>||_|p|j|_||_dSN)qval_ident test_funcexternal)selfrrrs Flib/python3.11/site-packages/textdistance/algorithms/sequence_based.py__init__zLCSSeq.__init__s" "1dk  c~trCtjt|dztdzftj}n+fdt t|dzD}t |D]~\}}t D]i\}}||kr!|||dz||dz|dz<,t ||dz||||dz||dz|dz<jd}t|t}}|dkr|dkr|||||dz |kr|dz}nY||||||dz kr|dz}n2||dz |dz ksJ||dz |z}|dz}|dz}|dkr|dk|S)z https://github.com/chrislit/abydos/blob/master/abydos/distance/_lcsseq.py http://www.dis.uniroma1.it/~bonifaci/algo/LCSSEQ.py http://rosettacode.org/wiki/Longest_common_subsequence#Dynamic_Programming_8 r)dtypec Xg|]&}tddgtdzz'S)Lrr)rlen).0_seq2s r z#LCSSeq._dynamic..)s4WWWQuS1#TQ"788WWWrr)numpyzerosrintrange enumeratemax) rseq1r!lengthsichar1jchar2results ` r_dynamiczLCSSeq._dynamic s  Xk3t99q=#d))a-"@ RRRGGWWWW%D TU BVBVWWWG"$ V VHAu%dOO V V5E>>,3AJqMA,=GAEN1q5)),/Aq0A71:aRSeCT,U,UGAEN1q5))  V4yy#d))11ffaqz!}Aq 111QA'!*QU"333QAE{d1q5k1111a!ev-QQ1ffa rct|st|dS|jd|Dr"|dd}d|D}|||zSt|d}t|D]D\}}|d||ddfz||dzdz}t |||gt }E|S)Nrcg|] }|d S)rss rr"z%LCSSeq._recursive..Es555aAbE555rr4c"g|] }|dd S)Nr4r5r6s rr"z%LCSSeq._recursive..Gs 333A3B3333rrkey)alltyperr(r)r)r sequencescmr,r7sss r _recursivezLCSSeq._recursiveBs9~~ (%4 ! %%'' ' 4>559555 6 (! R A33333I4#a' ' D1   i(( - -DAq2A2!CRC&+iA.??BTT2YN,,,AArcl|sdS|j|}t|dkr |j|S|j|S)Nr#r)_get_sequencesrr1rArr=s r__call__zLCSSeq.__call__OsM 2'D'3 y>>Q   4=), ,"4?I. .rc&t||SrrrDs r similarityzLCSSeq.similarityX44#$$$r)rNT) __name__ __module__ __qualname____doc__rr1rArErHr5rrr r so!!!!   D   ///%%%%%rr c*eZdZdZdZdZdZdZdS)r z(longest common substring similarity ct||}|dt|dt|}||j|j|jzS)N)abr)_SequenceMatcherfind_longest_matchrrPsize)rs1s2matchermatchs r _standartzLCSStr._standart_sT"R2...**1c"ggq#b''BB%'57UZ//00rct|t}t|}t|ddD];}t||D](}d|}|D]}||vrn |ccS)<t |S)Nr9rr4r#)minrr'rjoinr<)rr=shortlengthnsubseqseqs r_customzLCSStr._customdsI3'''Uvq"%% " "A%eQ// " "$""CS(()"MMMMM  "tE{{}}rct|sdSt|}|dkrdS|dkr|dS|j|}|dkr0tt t|dkr |j|S|j|S)Nr#rrr)r;rrCr)maprYrb)rr=r^s rrEzLCSStr.__call__qs9~~ 2Y Q;;2 Q;;Q< 'D'3 Q;;3s3 2233c99!4>9- -t|Y''rc&t||SrrGrDs rrHzLCSStr.similarityrIrN)rJrKrLrMrYrbrErHr5rrr r \sZ111    ( ( (%%%%%rr c$eZdZdZdZdZdZdS)ra+Ratcliff-Obershelp similarity This follows the Ratcliff-Obershelp algorithm to derive a similarity measure: 1. Find the length of the longest common substring in sequences. 2. Recurse on the strings to the left & right of each this substring in sequences. The base case is a 0 length common substring, in which case, return 0. Otherwise, return the sum of the current longest common substring and the left & right recursed sums. 3. Multiply this length by 2 and divide by the sum of the lengths of sequences. https://en.wikipedia.org/wiki/Gestalt_Pattern_Matching https://github.com/Yomguithereal/talisman/blob/master/src/metrics/distance/ratcliff-obershelp.js https://xlinux.nist.gov/dads/HTML/ratcliffObershelp.html cdS)Nrr5rDs rmaximumzRatcliffObershelp.maximumsqrct|tdkrdSfd|D}fd|D}|j|z|j|zS)NrcJg|]}|d| Srfind)rr7r`s rr"z+RatcliffObershelp._find..s,888!OQVVF^^O$888rcPg|]"}||zd#Srrl)rr7r^r`s rr"z+RatcliffObershelp._find..s3@@@166&>>F*++,@@@r)r r_find)rr=beforeafterr^r`s @@rrozRatcliffObershelp._finds9%V Q;;18888i888@@@@@i@@@tz6"V+jdj%.@@@rc|j|}||St|}ttt|}|j|}||j|z|z Sr) quick_answerrsumrerCro)rr=r0scountecounts rrEzRatcliffObershelp.__call__sf""I.  MYSi(())'D'3   I..77rN)rJrKrLrMrirorEr5rrrrsN AAA88888rr)difflibrrRutilsrbaser_BaseSimilarityr$ ImportErrorr__all__r r rr r r r5rrr}sJ777777 333333LLLL EEE     C%C%C%C%C%_C%C%C%L$%$%$%$%$%_$%$%$%N$8$8$8$8$8$8$8$8N  &&((s  ))