mò 0™ÜDc@s dZdkZdefd„ƒYZedjojdkZdkZeeiƒdjo^dZ yAe e dƒZ e i ƒZ e iƒeƒZeie ƒZWnde GHnXnáeeiƒdjoÊgZx“eieidƒD]{ZeiieideƒZ e GHyAe e dƒZ e i ƒZ e iƒeƒZeie ƒZWnd e GHnXqëWd Zx!eD]Zed7ZeGeGHqwWnndS( s¸ Modul AcronmyDecomposer Programmier-Gesellenstueck Author: Branimira Nikolova SS 2006 This module surches text for Akronym-Expansion Pairs and trys to match the them to each other. NtAcrocBsGtZd„Zd„Zd„Zd„Zd„Zd„Zd„ZRS(NcCsdd|_d|_d|_d|_d|_hddddg<dd<d d d g<|_dS( Ntit2ttottwotsecondt3tthirdt4tfortfourth(tselftacronymt expansiont candidatet expCandidatetpositiont digitDict(R ((tC:\AD\src\acronymDecomposer.pyt__init__s      cCsd|i|ifS(Ns%s, %s(R R R (R ((Rt__str__sc Cs«tddddƒ}|i|ƒ}g}d}d}t i |ƒ}x_t t|ƒƒD]K}||iiƒp+||iiƒp|i||iƒodt||iƒjo djnp×||i|idjo¸||di|idjoš||di|id jow||i|_||_|i|i|i||iƒ|_|i|i|i|iƒ}|o|i|ƒq˜q—q›||di|idjo+||d i|id jog|_||i|id joqXn||i|_xf||di|id joF||di|id joPq|d7}|ii||ƒqWt|iƒdjoqXqâ|iiƒ|i|i|i|iƒ}|o|i|ƒqâq—q›||did jo“||didjo{||i|_|d|_|i|i|i||iƒ|_|i|i|i|iƒ}|o|i|ƒq—q›||didjoî||did joÖg|_||i|_||_|d 7}xC||i|id joPqð|ii||ƒ|d7}qðWt|iƒdjoqXn|iiƒ|i|i|i|iƒ}|o|i|ƒq—q›qŸq£qXqXW|S(sé Surches for Acronym-candidates: uppercase or capitalized tokens - in parentheses - infront of parentheses - inftont of ", or" - after ", or" Returns Candidates list. ttokenttags([A-Z][a-zA-Z]+)ii tnounitopentcloseitdigitt delimiterstort,N(tTokentNonetnewTokenR t tagTextWithMTttextt tokenListt resultListtresulttmixtretcompiletregextrangetlentitRtistitletisuppertmatchRt tagSetDictRRtcomputeSearchSpaceRt checkTrueExptappendtreverse( R R"R)R&R$R R#R%R,((RtfindAcroCandidates~ ?+ $   0 $0    %c Cs®g}t|iƒdjot|iƒd}nt|iƒd}|id}||}xJ||jo<||i |djoPn|i ||ƒ|d8}q`W|S(s[ Computes the search space for Acronym-Candidates: - if the Acronym-Candidate is longer than 5 characters the searchspace is definedto be the lenght of the Acronym-Candidate+5; - if it is shorter than 5 characters the searchspace is the lenght*2 Returns the Acronym-Candiate as list of AD_Token instances. iiRiN( RR+R Rt searchSpaceRtstarttendR#RR0R3( R RRR#R0R8RR6R7((RR1Šs   cCsTd}x$|iD]}d||if}qWd}d}g}d} d}g} d} x|i D]}|i|ƒqaW|iƒx;|t|ƒdjo¬|i|i|djo|d7}q…q£|i|i}||iƒ|diƒjo"| i|i|iƒ|} Pq£|d7}|t|iƒdjoPq£n\||iƒ|i|iiƒjo“| |||i|fjo|d7}q£| i|i|iƒ|||i|f} |d7}|d7}|djo|d7}d}q£n¢||iƒok|d7}|i|i|djoN|||i|ijo/| i|i|iƒ|d7}|d7}qzq£|i|||i|iƒo/| i|i|iƒ|d7}|d7}q£|i|||i|ƒo‡|d7}t||ƒ}x,|i|||!D]} | i| iƒq!W|d7}||d7}|t|ƒdjo |} Pqzq£Pn&|d7}|t|iƒjoPn|t|ƒjoPq…q…Wt| ƒt|i ƒjodSnm| }d} x6|djo(d| |i|if} |d8}qóWtƒ}|i |_| |_ |GH|SdS( s¹ Compares the Acronym-Candidate and the Expansion-Candidate backwards. Each character from the Acronym-Candidate must appear in one of the tokens in the Expansion-Candidate in the same order as in the Acronym-Candidate; the first character of the Acronym-Candidate must match a character in the initial position of the first word in the Expansion-Candidate. Returns an Acro instance. Rs%s %siitstopTagsiRN(R"R RtiRtitctitstclistR/tfoundR t expansionEndRt characterR3R4R+RR0tfirsttlowertisdigittcheckDigitDicttcountFirstChartinttstepR,tfullRtnewAR (R RRR0R"R;R<R@R=R?R/RHR R,RGR:RIR>RA((RR2«s¢    !  '      !          cCsD||iiƒjo*|iƒ|i|jodSq@dSndS(s^Check if the digit in the acronym stands for a word like "for"(4) or "to"(2). iiN(tacR RtkeystecRB(R RJRL((RRD2s cCsd}t|ƒd}|||djoP||GHxª|D]8}||d|idjo|iGH|d7}q;q;Wng||doW||GHxK|||!D]8}||d|idjo|iGH|d7}q¡q¡Wn|t ||ƒjodSndSdS(s If the acronym starts with a digit, or some digit is inside of the acronym, count the words in the expansion, which start with the character preceding or following the digit and if the number of the characters is equal to the digit return 1. iiiÿÿÿÿN( tcountR+RLtecEndRJtacPosR,RtecPosRF(R RJRORLRPRMRNR,((RRE?s(  ( t__name__t __module__RRR5R1R2RDRE(((RRs  k ! ‡ t__main__is\t\testFile.txttrs %s not found!sError while reading %s!i(t__doc__R'RRRQtsystosR+targvtfilenameRtfhtreadR"RtnewAcroR5R%tlistdirtftpathtjointnumtitem( R\R^RYRVR'RaR%RbR"RZRWR((Rt? sD ÿR