Qde+ddlZddlmZGddZGddeZdZdZd Zed d d gZ Gd dZ dS)N) namedtupleceZdZdZd dZedZedZdZdZ eee Z dZ d Z d Z d Zd ZdS) AlignedSenta# Return an aligned sentence object, which encapsulates two sentences along with an ``Alignment`` between them. Typically used in machine translation to represent a sentence and its translation. >>> from nltk.translate import AlignedSent, Alignment >>> algnsent = AlignedSent(['klein', 'ist', 'das', 'Haus'], ... ['the', 'house', 'is', 'small'], Alignment.fromstring('0-3 1-2 2-0 3-1')) >>> algnsent.words ['klein', 'ist', 'das', 'Haus'] >>> algnsent.mots ['the', 'house', 'is', 'small'] >>> algnsent.alignment Alignment([(0, 3), (1, 2), (2, 0), (3, 1)]) >>> from nltk.corpus import comtrans >>> print(comtrans.aligned_sents()[54]) 'So why should EU arm...'> >>> print(comtrans.aligned_sents()[54].alignment) 0-0 0-1 1-0 2-2 3-4 3-5 4-7 5-8 6-3 7-9 8-9 9-10 9-11 10-12 11-6 12-6 13-13 :param words: Words in the target language sentence :type words: list(str) :param mots: Words in the source language sentence :type mots: list(str) :param alignment: Word-level alignments between ``words`` and ``mots``. Each alignment is represented as a 2-tuple (words_index, mots_index). :type alignment: Alignment Nc||_||_|tg|_dSt |tusJ||_dSN)_words_mots Alignment alignmenttype)selfwordsmotsr s 2lib/python3.11/site-packages/nltk/translate/api.py__init__zAlignedSent.__init__/sI   &r]]DNNN ??i////&DNNNc|jSr)rr s rrzAlignedSent.words8s {rc|jSr)r rs rrzAlignedSent.mots<s zrc|jSr) _alignmentrs r_get_alignmentzAlignedSent._get_alignment@s rc~tt|jt|j|||_dSr)_check_alignmentlenrrr)r r s r_set_alignmentzAlignedSent._set_alignmentCs/TZ#di..)DDD#rcddd|jDz}ddd|jDz}d|d|d|jdS)z_ Return a string representation for this ``AlignedSent``. :rtype: str z[%s]z, c3 K|] }d|zV dSz'%s'N.0ws r z'AlignedSent.__repr__..Os&#D#D1FQJ#D#D#D#D#D#Drc3 K|] }d|zV dSrr r!s rr$z'AlignedSent.__repr__..Ps&"B"B!6A:"B"B"B"B"B"Brz AlignedSent())joinrr r)r rrs r__repr__zAlignedSent.__repr__Isu $))#D#D #D#D#DDDE"B"Btz"B"B"BBBCCeCCtCCtCCCCrcd}|dz }|jD]}|d|d|dz }|jD]}|d|d|dz }|jD]'\}}|d|j|d|j|dz }(tt |jd z D]4}|d |j||j|d zz }5tt |jd z D]4}|d |j||j|d zz }5|d d d|jDzz }|d d d|jDzz }|dz }|S)z< Dot representation of the aligned sentence zgraph align { znode[shape=plaintext] "z_source" [label="z"] z_target" [label="z _source" -- "z _target" z)"{}_source" -- "{}_source" [style=invis] z)"{}_target" -- "{}_target" [style=invis] z{rank = same; %s}  c3 K|] }d|zV dS)z "%s_source"Nr r!s rr$z&AlignedSent._to_dot..us'.V.VQ}q/@.V.V.V.V.V.Vrc3 K|] }d|zV dS)z "%s_target"Nr r!s rr$z&AlignedSent._to_dot..vs'.U.UQ}q/@.U.U.U.U.U.Ur})rr rrangerformatr')r sr#uvis r_to_dotzAlignedSent._to_dotTs  && 2 2A 1Q11111 1AA 2 2A 1Q11111 1AAO M MDAq LT[^LL$*Q-LLL LAAs4;''!+,,  A =DD A AE" AA s4:*++  A =DD 1  1q5! AA "chh.V.V$+.V.V.V&V&V WW "chh.U.U$*.U.U.U&U&U VV Srcl|d}d} tjdd|zgtjtjtj}n"#t $r}t d|d}~wwxYw||\}}|dS)zR Ipython magic : show SVG representation of this ``AlignedSent``. utf8svgdotz-T%s)stdinstdoutstderrz0Cannot find the dot binary from Graphviz packageN) r6encode subprocessPopenPIPEOSError Exception communicatedecode)r dot_string output_formatprocesseouterrs r _repr_svg_zAlignedSent._repr_svg_|s\\^^**622   W &./ o!! GG  W W WNOOUV V W&&z22Szz&!!!s;A'' B1BBcd|jdddz}d|jdddz}d|d|dS)zn Return a human-readable string representation for this ``AlignedSent``. :rtype: str r,Nz...z 'z'>)r'rr )r sourcetargets r__str__zAlignedSent.__str__s_ $+&&ss+e3$*%%crc*U2999v9999rcft|j|j|jS)zm Return the aligned sentence pair, reversing the directionality :rtype: AlignedSent )rr rrinvertrs rrSzAlignedSent.inverts( 4:t{DO4J4J4L4LMMMrr)__name__ __module__ __qualname____doc__rpropertyrrrrr r(r6rLrQrSr rrrrs>''''XX$$$88I D D D&&&P"""&:::NNNNNrrcTeZdZdZdZedZdZdZd dZ dZ d Z d Z dS) r ac A storage class for representing alignment between two sequences, s1, s2. In general, an alignment is a set of tuples of the form (i, j, ...) representing an alignment between the i-th element of s1 and the j-th element of s2. Tuples are extensible (they might contain additional data, such as a boolean to indicate sure vs possible alignments). >>> from nltk.translate import Alignment >>> a = Alignment([(0, 0), (0, 1), (1, 2), (2, 2)]) >>> a.invert() Alignment([(0, 0), (1, 0), (2, 1), (2, 2)]) >>> print(a.invert()) 0-0 1-0 2-1 2-2 >>> a[0] [(0, 1), (0, 0)] >>> a.invert()[2] [(2, 1), (2, 2)] >>> b = Alignment([(0, 0), (0, 1)]) >>> b.issubset(a) True >>> c = Alignment.fromstring('0-0 0-1') >>> b == c True ct||}|tgkrtd|Dnd|_d|_|S)Nc3&K|] }|dV dSrNr r"ps rr$z$Alignment.__new__..s&++!++++++rr) frozenset__new__max_len_index)clspairsr s rr`zAlignment.__new__sV  e,,/3y}}/D/DC++d++++++!   rcXtd|DS)a Read a giza-formatted string and return an Alignment object. >>> Alignment.fromstring('0-0 2-1 9-2 21-3 10-4 7-5') Alignment([(0, 0), (2, 1), (7, 5), (9, 2), (10, 4), (21, 3)]) :type s: str :param s: the positional alignments in giza format :rtype: Alignment :return: An Alignment object corresponding to the string representation ``s``. c,g|]}t|Sr ) _giza2pair)r"as r z(Alignment.fromstring..s;;;A*Q--;;;r)r split)rdr2s r fromstringzAlignment.fromstrings);;;;;<<.s;>>A1Q41,122.>>>>>>r)r rs rrSzAlignment.inverts!>>>>>>>>rNc0t}|js||s.tt t |j}|D],}|d|j|D-t|S)z Work out the range of the mapping from the given positions. If no positions are specified, compute the range of the entire mapping. c3 K|] \}}|V dSrr )r"_fs rr$z"Alignment.range..s&66tq!666666r)setrcrnlistr0rupdatesorted)r positionsimager^s rr0zAlignment.ranges {       6U3t{#3#34455I 7 7A LL66t{1~666 6 6 6 6e}}rc&dt|zS)M Produce a Giza-formatted string representing the alignment. z Alignment(%r))r{rs rr(zAlignment.__repr__s--rcZddt|DS)rr,c30K|]}d|ddzVdS)z%d-%dNrsr r]s rr$z$Alignment.__str__..s.>>A!BQB%>>>>>>r)r'r{rs rrQzAlignment.__str__s+xx>>>>>>>>rcdt|jdzD|_|D](}|j|d|)dS)z Build a list self._index such that self._index[i] is a list of the alignments originating from word i. cg|]}gSr r )r"rvs rrjz*Alignment._build_index..s888ar888rr+rN)r0rbrcappend)r r^s rrnzAlignment._build_indexsb 985Q#7#7888  ( (A K!  $ $Q ' ' ' ' ( (rr) rTrUrVrWr` classmethodrlrorSr0r(rQrnr rrr r s2  = =[ =,,,???    ... ??? (((((rr cn|d\}}t|t|fSN-rkint) pair_stringr5js rrhrhs/   S ! !DAq q663q66>rcp|d\}}}t|t|fSrr)rr5rr^s r _naacl2pairr s1$$GAq! q663q66>rct|tusJtfd|Dstdtfd|DstddS)ab Check whether the alignments are legal. :param num_words: the number of source language words :type num_words: int :param num_mots: the number of target language words :type num_mots: int :param alignment: alignment to be checked :type alignment: Alignment :raise IndexError: if alignment falls outside the sentence c3DK|]}d|dcxkokncVdSr\r )r"pair num_wordss rr$z#_check_alignment..sC>>DqDG''''i''''>>>>>>rz&Alignment is outside boundary of wordsc3XK|]$}|ddupd|dcxkokncV%dS)r+Nrr )r"rnum_motss rr$z#_check_alignment..sPPPdtAw$9!tAw"9"9"9"9"9"9"9"9PPPPPPrz%Alignment is outside boundary of motsN)r r all IndexError)rrr s`` rrrs  ??i ' ' ' ' >>>>I>>> > >CABBB PPPPiPPP P PB@AAABBrPhraseTableEntry trg_phraselog_probc*eZdZdZdZdZdZdZdS) PhraseTablezs In-memory store of translations for a given phrase, and the log probability of the those translations c,t|_dSr)dict src_phrasesrs rrzPhraseTable.__init__,s66rc|j|S)a Get the translations for a source language phrase :param src_phrase: Source language phrase of interest :type src_phrase: tuple(str) :return: A list of target language phrases that are translations of ``src_phrase``, ordered in decreasing order of likelihood. Each list element is a tuple of the target phrase and its log probability. :rtype: list(PhraseTableEntry) rr src_phrases rtranslations_forzPhraseTable.translations_for/s ++rct||}||jvr g|j|<|j|||j|dddS)z :type src_phrase: tuple(str) :type trg_phrase: tuple(str) :param log_prob: Log probability that given ``src_phrase``, ``trg_phrase`` is its translation :type log_prob: float )rrc|jSr)r)rIs rz!PhraseTable.add..Ks rT)rpreverseN)rrrsort)r rrrentrys raddzPhraseTable.add>sx!JJJJ T- - -+-D Z ( $++E222 $)).B.BD)QQQQQrc||jvSrrrs r __contains__zPhraseTable.__contains__MsT---rN)rTrUrVrWrrrrr rrrr&s] """ , , , R R R.....rr) r? collectionsrrr_r rhrrrrr rrrs """"""PNPNPNPNPNPNPNPNf_(_(_(_(_( _(_(_(D  BBB*:0<2LMM(.(.(.(.(.(.(.(.(.(.r