xjr& *dZddlmZddlmZmZmZmZddlm Z e dZ e dZ dZ dZdZd Zd ed ed eeeee ffd eeeee fffdZeeddZeeddZeeddZdZdZdZdZeeeeedZieeeegZeeeefededediZded ee fdZdeded e fdZdeded e fd Z!deded e fd!Z"deded e fd"Z#d#S)$z2Module contains the score calculation algorithems.)partial)DictListUnioncast) SCORE_INDICESz-infinfg{Gztg{Gzg? char_start char_stop hash_tablereturnc |}|fdtt|t|dzD|S)aGenerate index mapping for `bonus` calculation. Args: char_start: Starting char of the range. char_stop: Ending char of the range. value: Value to give to the range of char. hash_table: Base dictionary to add the mapping. Returns: A dictionary containing the given range with provided index. Examples: >>> _char_range_with("a", "d", 1, {}) {'a': 1, 'b': 1, 'c': 1, 'd': 1} c3:K|]}t|fVdS)N)chr).0uni_charvalues ;C:\Projects\Historical\venv\Lib\site-packages\pfzy/score.py z#_char_range_with.."sB  X)copyupdaterangeord)r r rr s ` r_char_range_withrsu$""Jc*oos9~~/ABB razAZ09g?g?gffffff?g333333?)/-_ .rhaystackcd}g}|D]Q}|tt|d|d|}R|S)aCalculate bonus score for the given haystack. The bonus are applied to each char based on the previous char. When previous char is within the `BONUS_MAP` then additional bonus are applied to the current char due to it might be the start of a new word. When encountered a mix case character, if the current char is capitalised then if the previous char is normal case or within `BONUS_MAP`, additional bounus are applied. Args: haystack: String to calculate bonus. Returns: A list of float matching the length of the given haystack with each index representing the bonus score to apply. Examples: >>> _bonus("asdf") [0.9, 0, 0, 0] >>> _bonus("asdf asdf") [0.9, 0, 0, 0, 0, 0.8, 0, 0, 0] >>> _bonus("asdf aSdf") [0.9, 0, 0, 0, 0, 0.8, 0.7, 0, 0] >>> _bonus("asdf/aSdf") [0.9, 0, 0, 0, 0, 0.9, 0.7, 0, 0] r#r)append BONUS_STATES BONUS_INDEXget)r) prev_charbonuschars r_bonusr2<s^:I E \+//$":":;?? 1MMNNN Lrneedlec< t|t|c} t|}|r|}|dks| kr#tt t |fS fdt |D} fdt |D}t |D]}t}||dz krtnt}t D]} |||| krt} |dkr| tz|| z} nK| dkrEt||dz | dz || z||dz | dz tz} | ||| <t| ||zx||| <}t||| <||zx||| <}ˌ|dz dz } }d} dt |D} |dkr| dkr| s||| ||| krZ||| tkrC|dko1| dko+||| ||dz | dz tzk} | | |<| dz} n | dz} | dk|dz}|dk||dz  dz | fS)afUse fzy logic to calculate score for `needle` within the given `haystack`. 2 2D array to track the score. 1. The running score (`running_score`) which represents the best score for the current position. 2. The result score (`result_score`) which tracks to overall best score that could be for the current positon. With every consequtive match, additional bonuse score are given and for every non matching char, a negative gap score is applied. After the score is calculated, the final matching score will be stored at the last position of the `result_score`. Backtrack the result by comparing the 2 2D array to find the corresponding indices. Args: needle: Substring to find in haystack. haystack: String to be searched and scored. Returns: A tuple of matching score with a list of matching indices. rcBg|]}dtDS)cg|]}dSrrr%s r z%_score...(((q(((rrrr% haystack_lens rr:z_score..s=(((-.((E,''((((((rcBg|]}dtDS)cg|]}dSr7r8r9s rr:z%_score...r;rr<r=s rr:z_score..s='''-.((E,''((('''rrFcg|]}dSr7r8r9s rr:z_score..s,,,Qq,,,r) lenr2islowerlower SCORE_MAXlistr SCORE_MINSCORE_GAP_TRAILINGSCORE_GAP_INNERSCORE_GAP_LEADINGmaxSCORE_MATCH_CONSECUTIVE)r3r) needle_len bonus_score running_score result_scorei prev_score gap_scorejscorematch_requiredindicesr>s @r_scorerXasM* #6{{CMMJ ""K~~$>>##Q* 44$uZ001111((((27 2C2C(((M ''''27 2C2C'''L:  II +,zA~*=*=&&? |$$ I IAayHQK''!66 11KNBEE!VV$QU+AE2[^C%a!e,QU36MME ', a #25eZ)=S2T2TT Q"ZZ&/ a #2>> _subsequence("as", "bbwi") False >>> _subsequence("as", "bbaiws") True >>> _subsequence("sa", "bbaiws") False TrrF)rDfind)r3r)offsetr1s r _subsequencer\sr$||~~x~~'7'7HF t FtV,,q0 Q;;55  4rcTt||rt||StdfS)aUse fzy matching algorithem to match needle against haystack. Note: The `fzf` unordered search is not supported for performance concern. When the provided `needle` is not a subsequence of `haystack` at all, then `(-inf, None)` is returned. See Also: https://github.com/jhawthorn/fzy/blob/master/src/match.c Args: needle: Substring to find in haystack. haystack: String to be searched and scored against. Returns: A tuple of matching score with a list of matching indices. Examples: >>> fzy_scorer("ab", "acb") (0.89, [0, 2]) >>> fzy_scorer("ab", "acbabc") (0.98, [3, 4]) >>> fzy_scorer("ab", "wc") (-inf, None) N)r\rXrG)r3r)s r fzy_scorerr^s04FH%%fh'''$rcg}d}||}}|dD]f}|s|||}|dkr tdfcSt |}|t |||z||z }g|sd|fS|ddz|dz d|ddzz zd|ddzz z|fS)aUMatch needle against haystack using :meth:`str.find`. Note: Scores may be negative but the higher the score, the higher the match rank. `-inf` score means no match found. See Also: https://github.com/aslpavel/sweep.py/blob/3f4a179b708059c12b9e5d76d1eb3c70bf2caadc/sweep.py#L837 Args: needle: Substring to find in haystack. haystack: String to be searched and scored against. Returns: A tuple of matching score with a list of matching indices. Example: >>> substr_scorer("ab", "awsab") (-1.3, [3, 4]) >>> substr_scorer("ab", "abc") (0.5, [0, 1]) >>> substr_scorer("ab", "iop") (-inf, None) >>> substr_scorer("ab", "asdafswabc") (-1.6388888888888888, [7, 8]) >>> substr_scorer(" ", "asdf") (0, []) rr&Nrr()rDsplitrZrGrBextendr)r3r)rWr[rMs r substr_scorerrcs:G F||~~x~~'7'7HF,,s##  vv.. A::d? " " "[[ uVVj%899:::* 'z "+/GAJ &'!wqzA~*>>gbkTUoAVV rN)$__doc__ functoolsrtypingrrrr pfzy.typesrfloatrGrErJrHrIrLstrintr lower_with upper_with digit_withSCORE_MATCH_SLASHSCORE_MATCH_WORDSCORE_MATCH_CAPITALSCORE_MATCH_DOT BONUS_MAPr,r-r2rXboolr\r^rcr8rrrts88************$$$$$$ E&MM E%LL  #8::aAzz!R?P?P1Q1Q#R#RSS "S"T%[""""JaC3aC#aC-aCaCaCaCH:scm@1#11111111r