:Qg]ddlZddlZddlZddlZddlmZddlmZddlZ ddl Z dZ dZ GddeZGdd eZGd d eZGd d eZGddeZy)N)chain) itemgetter c&eZdZdZddZdZddZy)TextEdgeatDefines a text edge coordinates relative to a left-bottom origin. (PDF coordinate space) Parameters ---------- x : float x-coordinate of the text edge. y0 : float y-coordinate of bottommost point. y1 : float y-coordinate of topmost point. align : string, optional (default: 'left') {'left', 'right', 'middle'} Attributes ---------- intersections: int Number of intersections with horizontal text rows. is_valid: bool A text edge is valid if it intersections with at least TEXTEDGE_REQUIRED_ELEMENTS horizontal text rows. cX||_||_||_||_d|_d|_y)NrF)xy0y1align intersectionsis_valid)selfr r r r s I/var/www/html/answerous/venv/lib/python3.12/site-packages/camelot/core.py__init__zTextEdge.__init__.s-  c t|jd}t|jd}t|jd}d|d|d|d|jd|j d S)Nz )roundr r r r r)rr r r s r__repr__zTextEdge.__repr__6sd $&&!  477A  477A 1#T"T"WTZZL VW X rc,tj|j||rr|j|jz|zt |jdzz |_||_|xjdz c_|jt kDrd|_yyy)zdUpdates the text edge's x and bottom y coordinates and sets the is_valid attribute. atolTN)npiscloser rr floatTEXTEDGE_REQUIRED_ELEMENTSr)rr r edge_tols r update_coordszTextEdge.update_coords>s ::dggr 1((4661A5t?Q?QTU?U9VVDFDG   ! # !!$>> $ ? 2rN)left2)__name__ __module__ __qualname____doc__rrr#rrrrs0  %rrcLeZdZdZd dZedZdZdZdZ dZ dZ d Z y ) TextEdgeszDefines a dict of left, right and middle text edges found on the PDF page. The dict has three keys based on the alignments, and each key's value is a list of camelot.core.TextEdge objects. c(||_gggd|_y)Nr$rightmiddle)r" _textedges)rr"s rrzTextEdges.__init__Rs  #%bArc^|j}|j}|||z dz z}|||d}||S)zYReturns the x coordinate of a text row based on the specified alignment. g@)r$r1r0)x0x1)textliner x_leftx_rightx_middlex_coords r get_x_coordzTextEdges.get_x_coordVs@ ++Wv-44!XHu~rct|j|D]+\}}tj|j|ds)|cSy)zkReturns the index of an existing text edge using the specified x coordinate and alignment. g?rN) enumerater2rrr )rr:r ites rfindzTextEdges.findasCtu56 EArzz"$$c2 rc|j||}|j}|j}t||||}|j|j |y)z)Adds a new text edge to the current dict.)r N)r;r r rr2append)rr6r r r r r?s raddz TextEdges.addjsM   Xu - [[ [[ aRu - %%b)rcdD]s}|j||}|j||}||j||<|j||j ||j |j uy)z2Updates an existing text edge in the current dict.r/N)r")r;r@rCr2r#r r")rr6r r:idxs rupdatezTextEdges.updatersv0 E&&x7G))GU+C{5)&s+99X[[4==:  rc|D]>}t|jjdkDs.|j|@y)zMGenerates the text edges dict based on horizontal text rows. rN)lenget_textstriprF)r textlinestls rgeneratezTextEdges.generate~s: B2;;=&&()A- B rc(td|jdDtd|jdDtd|jdDd}t|jt d d }|j|S) zReturns the list of relevant text edges (all share the same alignment) based on which list intersects horizontal text rows the most. c3NK|]}|js|jywNrr.0r?s r z)TextEdges.get_relevant..s!%'bkk  %%r$c3NK|]}|js|jywrPrQrRs rrTz)TextEdges.get_relevant..s!%'r{{  rUr0c3NK|]}|js|jywrPrQrRs rrTz)TextEdges.get_relevant..s!%'   rUr1r/rkeyr)sumr2maxitemsr)rintersections_sumrelevant_aligns r get_relevantzTextEdges.get_relevants +/??6+B+/??7+C+/??8+D  .446JqMJ1M~..rcZd}|jdi}|D]}|js|s2d||j|j|j|jf<Ed}|D]+}|j|dk\s|j|dks)|}n|2d||j|j|j|jf<|j ||dt |j|dt|d|jt|d|jf}d||<d} |D]} | | j| jz z } d}|D]+}| j|dk\s| j|dks)|}n|T|j |t | j|dt | j|dt|d| jt|d| jf}d||<| tt|z } i} |D]}d| ||| <| S) zpReturns a dict of interesting table areas on the PDF page calculated using relevant text edges. cl|dtz }|dtz }|dtz}|d|dzz}||||fS)Nrrr)TABLE_AREA_PADDING)areaaverage_row_heightr4r r5r s rpadz&TextEdges.get_table_areas..padsRa--Ba--Ba--Ba-11BB# #rc4|j |jfSrP)r r )r?s rz+TextEdges.get_table_areas..s~rrXNrrbrr) sortrr r r popminr[r4r5r rH) rrKrelevant_textedgesrg table_areasr?foundre updated_areasum_textline_heightrLaverage_textline_heighttable_areas_paddeds rget_table_areaszTextEdges.get_table_areass?  $ $=> $ 9B{{">BKruubddBEE :; E +"55DG+a0@$(E! " }BF RTT255"$$$>?#.!!HuQx0a"$$/a"%%0 ( 59 L1+ 9:  1B 255255= 0 E# 55DG#a(8 E    &uQx(uQx(a"%%(a"%%(   -1 L)! 1"#6c)n8M"M  JDEI s4)@A B J"!rNr%) r'r(r)r*r staticmethodr;r@rCrFrMr_rtr+rrr-r-Ls@ B*  /.G"rr-c`eZdZdZdZdZedZejdZedZ y)Cella\Defines a cell in a table with coordinates relative to a left-bottom origin. (PDF coordinate space) Parameters ---------- x1 : float x-coordinate of left-bottom point. y1 : float y-coordinate of left-bottom point. x2 : float x-coordinate of right-top point. y2 : float y-coordinate of right-top point. Attributes ---------- lb : tuple Tuple representing left-bottom coordinates. lt : tuple Tuple representing left-top coordinates. rb : tuple Tuple representing right-bottom coordinates. rt : tuple Tuple representing right-top coordinates. left : bool Whether or not cell is bounded on the left. right : bool Whether or not cell is bounded on the right. top : bool Whether or not cell is bounded on the top. bottom : bool Whether or not cell is bounded on the bottom. hspan : bool Whether or not cell spans horizontally. vspan : bool Whether or not cell spans vertically. text : string Text assigned to cell. c||_||_||_||_||f|_||f|_||f|_||f|_d|_d|_ d|_ d|_ d|_ d|_ d|_y)NF)r5r x2y2lbltrbrtr$r0topbottomhspanvspan_textrr5r rzr{s rrz Cell.__init__s{r(r(r(r(      rc t|jd}t|jd}t|jd}t|jd}d|d|d|d|d S)Nrz >rN) r'r(r)r*rrpropertyrsetterrr+rrrwrwsR'R"9 [[..??rrwczeZdZdZdZdZdZedZedZ dZ ddZ d Z d Z d Zd Zd ZdZdZy)TableaDefines a table with coordinates relative to a left-bottom origin. (PDF coordinate space) Parameters ---------- cols : list List of tuples representing column x-coordinates in increasing order. rows : list List of tuples representing row y-coordinates in decreasing order. Attributes ---------- df : :class:`pandas.DataFrame` shape : tuple Shape of the table. accuracy : float Accuracy with which text was assigned to the cell. whitespace : float Percentage of whitespace in the table. order : int Table number on PDF page. page : int PDF page number. c||_||_|Dcgc]*}|Dcgc]}t|d|d|d|dc},c}}|_d|_d|_d|_d|_d|_d|_ ycc}wcc}}w)Nrr)rr) colsrowsrwcellsdfshapeaccuracy whitespaceorderpage)rrrrcs rrzTable.__init__Ts{  LPQqTBtAaD!A$!ad3BQ     CQs A<!A7 A<7A<cPd|jjd|jdS)N>**+74::,a@@rc|j|jk(r|j|jkry|j|jkryy)NT)rr)rothers r__lt__z Table.__lt__bs= 99 "zzEKK' 99uzz ! "rcg}|jD]9}|j|Dcgc]}|jjc};|Scc}w)z1Returns two-dimensional list of strings in table.)rrBrrJ)rdrowcells rdataz Table.dataisH :: :C HHC8Ddiioo'8 9 :9s!A ct|jdt|jd|j|jd}|S)zlReturns a parsing report with %accuracy, %whitespace, table number on page and page number. r)rrrr)rrrrr)rreports rparsing_reportzTable.parsing_reportqs<dmmQ/3ZZII    rcn|jD]%}|D]}dx|_x|_x|_|_ '|S)zSets all table edges to True.T)rr$r0rr)rrrs r set_all_edgeszTable.set_all_edgessK:: GC GBFF FDJFDK G G rc  |D]}t|jDcgc]%\}}tj|d|d|r|'c}}}t|jDcgc]%\}}tj|d|d|r|'c}}}t|jDcgc]%\}}tj|d|d|r|'c}}}|s|d} |dgk(rv|d} |r/|d} | | ksd|j | | _| dz } | | kr"t|j} | | ks6d|j | | _| dz } | | kr"Z|gk(rt|jdz } |r0|d} | | ksd|j | | _| dz } | | kr"t|j} | | ksd|j | | _| dz } | | kr"|d} |rJ|d} | | ksd|j | | _d|j | | dz _| dz } | | kr<:t|j} | | ksVd|j | | _d|j | | dz _| dz } | | kr<|D]} t|jDcgc]%\}}tj| d|d|r|'c}}}t|jDcgc]%\}}tj| d|d|r|'c}}}t|jDcgc]%\}}tj| d|d|r|'c}}}|s|d} |dgk(rv|d} |r/|d} | | ksd|j | | _ | dz } | | kr"t|j} | | ks6d|j | | _ | dz } | | kr"Z|gk(rt|jdz } |r0|d} | | ksd|j | | _ | dz } | | kr"t|j} | | ksd|j | | _ | dz } | | kr"|d} |rJ|d} | | ksd|j | | _ d|j | dz | _ | dz } | | kr<:t|j} | | ksVd|j | | _ d|j | dz | _ | dz } | | kr<|Scc}}wcc}}wcc}}wcc}}wcc}}wcc}}w)aNSets a cell's edges to True depending on whether the cell's coordinates overlap with the line's coordinates within a tolerance. Parameters ---------- vertical : list List of detected vertical lines. horizontal : list List of detected horizontal lines. rrrbrTr) r=rrrrrr$rHr0rr) rvertical horizontal joint_tolvr>rjkJLKhs r set_edgeszTable.set_edgess: A &dii0Aq::adAaDy9A&dii0Aq::adAaDy9A&dii0Aq::adAaDy9A !AQCxaD!Aa%04 1 a(-Qa%DIIAa%04 1 a(-Qa%b NQ&!Aa%15 1 a(.Qa%DIIAa%15 1 a(.Qa%aD!Aa%04 1 a(-59 1 a!e,2Qa% DIIAa%04 1 a(-59 1 a!e,2Qa%o: x: A &dii0Aq::adAaDy9A&dii0Aq::adAaDy9A&dii0Aq::adAaDy9A !AQCxaD!Aa%/3 1 a(,Qa%DIIAa%/3 1 a(,Qa%b NQ&!Aa%26 1 a(/Qa%DIIAa%26 1 a(/Qa%aD!Aa%/3 1 a(,6: 1q5)!,3Qa% DIIAa%/3 1 a(,6: 1q5)!,3Qa%o: x k  d  s#*T1%*T7+*T=6*U<*U *Uctt|jD]F}d|j|d_d|j|t|j dz _Htt|j D]F}d|jd|_d|jt|jdz |_H|S)z Sets table border edges to True.Trr) rangerHrrr$rr0rr)rrrs r set_borderzTable.set_border ss499~& ;A$(DJJqM!  !6:DJJqM#dii.1, - 3 ;s499~& #!"( !rrc^eZdZdZdZdZdZdZedZ e dZ d d Z d Z d d Zy) TableListzDefines a list of camelot.core.Table objects. Each table can be accessed using its index. Attributes ---------- n : int Number of tables in the list. c||_yrP_tables)rtabless rrzTableList.__init__s  rcPd|jjd|jdS)Nrz n=r)rr'nrs rrzTableList.__repr__s&4>>**+3tvvha88rc,t|jSrP)rHrrs r__len__zTableList.__len__s4<<  rc |j|SrPr)rrEs r __getitem__zTableList.__getitem__s||C  rc t|d|S)Nto_)getattr)tablers r _format_funczTableList._format_funcsuA3i((rct|SrP)rHrs rrz TableList.ns 4yrNc @|jd}|jd}|jd}|jD]\}|d|jd|j|}tj j ||}|j||} | |^y)Ndirnamerootext-page-r)getrrrosrrr) rrrrrrrfilenamefilepath to_formats r _write_filezTableList._write_files**Y'zz&!jj\\ Evejj\ cUKHww||GX6H))%3I h   rc h|jd}|jd}|jd}|jd}tjjtjj ||dz}t j |dd5}|jD]q}|d |jd |j|} tjj|| } |j| tjj| s dddy#1swYyxYw) Nrrrr.ziprT allowZip64rr) rrrrrzipfileZipFilerrrrbasename) rrrrrrzipnamezrrrs r _compress_dirzTableList._compress_dirszz&!**Y'zz&!jj'',,rwwt4d;fD __Wcd ; >q >"V6%**WU[[M#O77<<:"''"2"28"<= > > > >s BD((D1ctjj|}tjj|}tjj |\}}|rt j }||||d}|dvr*|jdd|i||r|jdi|y y |dk(r$tjj||} tj| } |jD]<} d| jd| j} | jj!| | d>| j#|rtjjtjj||d z} t%j&| d d 5}|j)| tjj| d d d y y |dk(rtjj||} |jD]} | j+| |rtjjtjj||d z} t%j&| d d 5}|j)| tjj| d d d y y y #1swYy xYw#1swYy xYw)a:Exports the list of tables to specified file format. Parameters ---------- path : str Output filepath. f : str File format. Can be csv, json, excel, html and sqlite. compress : bool Whether or not to add files to a ZIP archive. )rrrr)csvjsonhtmlrexcelrrrrr rTr Nsqliter+)rrrrsplitexttempfilemkdtemprrrrrrrrrrrr r rr)rrrcompressrrrrrrrrrrrs rexportzTableList.exportsR''//$'77##D)GG$$X. c &&(G7DM ' ' D   +q +F +""",V, '\ww||GX6H^^H-F S$UZZL }E !!&Z'!R S KKM'',,rwwt'2J;>KrP)rF)r'r(r)r*rrrrrurrrrrrr+rrrrsS9!!))  >+Brr)rrr r itertoolsroperatorrnumpyrpandasrr!rdobjectrr-rwrrr+rrr"sw  4%v4%nX"X"vM?6M?`SFSl `B`Br