{-jn(gdZddlZddlmZddlmZmZmZmZm Z ddl Z ddl m Z ddlmZddlmZd d lmZmZd e jd e jd efdZ d8dedededed ef dZdZ d9deedeeded efdZ d:de e jeefde e jeefd efdZdZ dZ!d Z"d!Z#d"Z$d#Z%d$Z& d;de ee'ee'e'e'e'ffde ee'ee'e'e'e'ffd%ed&ed ee ee'ee'e'e'e'fff d'Z( dd6ed efd7Z2dS)?)get_sub_regions_ocr_resget_show_colorsorted_layout_boxesN)deepcopy)DictListOptionalTupleUnion)Image)convert_points_to_boxes) OCRResult)BLOCK_LABEL_MAPREGION_SETTINGS src_boxes ref_boxesreturncZg}t|}|dkrt|dkrtt|D]}||}tj|d|dddf}tj|d|dddf}tj|d|dddf}tj|d|dddf} ||z } | |z } tj| dk| dkzd} || |S)a Get the indices of source boxes that overlap with reference boxes based on a specified threshold. Args: src_boxes (np.ndarray): A 2D numpy array of source bounding boxes. ref_boxes (np.ndarray): A 2D numpy array of reference bounding boxes. Returns: match_idx_list (list): A list of indices of source boxes that overlap with reference boxes. rNrr )lenrangenpmaximumminimumwhereextend) rrmatch_idx_list src_boxes_numrnoref_boxx1y1x2y2pub_wpub_h match_idxs p/var/www/html/banglarbhumi/venv/lib/python3.11/site-packages/paddlex/inference/pipelines/layout_parsing/utils.pyget_overlap_boxes_idxr+!s)N NNMqS^^a//Y(( - -CnGGAJ !!!Q$88BGAJ !!!Q$88BGAJ !!!Q$88BGAJ !!!Q$88BGEGE%!) !:;;A>I  ! !) , , , , TFoverall_ocr_res object_boxes flag_withinreturn_match_idxci}g|d<g|d<g|d<g|d<|d}t||}tt|}tt |D]}|r ||vrd}n d}n ||vrd}nd}|r|d|d||d|d||d|d||d|d|dD]} t j|| || < |r||fn|S)aX Filters OCR results to only include text boxes within specified object boxes based on a flag. Args: overall_ocr_res (OCRResult): The original OCR result containing all text boxes. object_boxes (list): A list of bounding boxes for the objects of interest. flag_within (bool): If True, only include text boxes within the object boxes. If False, exclude text boxes within the object boxes. return_match_idx (bool): If True, return the list of matching indices. Returns: OCRResult: A filtered OCR result containing only the relevant text boxes. rec_polys rec_texts rec_scores rec_boxesTF)r2r4r5)r+listsetrrappendrarray) r-r.r/r0sub_regions_ocr_resoverall_text_boxesrbox_no flag_matchkeys r*rr;s$') $') $(* %') $(5*+=|LLN#n--..N.//00  #''! " ^++! "    , 3 3 ,V4    , 3 3 ,V4    - 4 4 -f5    , 3 3 ,V4   8FF#%8,?,D#E#EC    ! n-- r,ct|}|dkr|St|d}t|}g}g}g}d} ||krn||dd|dz kr?||ddd |zd z kr!||||dz }nm||ddd|zd z kr!||||dz }n.||z }||z }|||g}g}|dz }t|d }t|d }|r||z }|r||z }|S) z Sort text boxes in order from top to bottom, left to right Args: res: List of dictionaries containing layout information. w: Width of image. Returns: List of dictionaries containing sorted layout information. rc:|dd|ddfS)N block_bboxrrxs r*z%sorted_layout_boxes..sa oa.@!L/RSBT-Ur,)r>rTrAr rc|ddSNrArrBrCs r*rEz%sorted_layout_boxes..sqq/Ar,c|ddSrIrBrCs r*rEz%sorted_layout_boxes..s<1Cr,)rsortedr6r8) resw num_boxes sorted_boxes_boxesnew_resres_left res_rightis r*rrwsCIA~~ ##U#UVVVL ,  FGHI A >>  1Il #A &Q . .q ,'*QUQY66 OOF1I & & & FAA AY| $Q '!a%!) 3 3   VAY ' ' ' FAA x G y G NN6!9 % % %HI FA'*h$B$BCCCHy&D&DEEEI89 Nr, horizontalunionbbox1bbox2 directionc@d\}}|dkrd\}}t||||}t||||}||z }|dkrdS|dkr:t||||t||||z } n}|dkr/t||||z ||||z } nH|dkr/t||||z ||||z } ntd|d | dkr|| z nd S) a Calculate the IoU of lines between two bounding boxes. Args: bbox1 (List[float]): First bounding box [x_min, y_min, x_max, y_max]. bbox2 (List[float]): Second bounding box [x_min, y_min, x_max, y_max]. direction (str): direction of the projection, "horizontal" or "vertical". Returns: float: Line overlap ratio. Returns 0 if there is no overlap. )rrrU)rr rrVsmalllarge Invalid mode -, must be one of ['union', 'small', 'large'].)maxmin ValueError) rWrXrYmode start_index end_indexintersection_startintersection_endoverlap ref_widths r*"calculate_projection_overlap_ratiorjsn""KL  !% YU;/{1CDD5+U9-=>>!33G!||q wi(% *:;;c + k 2? ?    ) u[1 153CeKFX3X    ) u[1 153CeKFX3X   OD O O O   #,a--7Y  S8r,c6tj|tj}tj|tj}tj|d|d}tj|d|d}tj|d|d}tj|d|d}tjd||z }tjd||z }tj||tj} t |} t |} |dkr | | z| z } nK|dkrtj| | } n/|dkrtj| | } ntd |d | dkrd S| | z S) a Calculate the overlap ratio between two bounding boxes using NumPy. Args: bbox1 (np.ndarray, list or tuple): The first bounding box, format [x_min, y_min, x_max, y_max] bbox2 (np.ndarray, list or tuple): The second bounding box, format [x_min, y_min, x_max, y_max] mode (str): The mode of calculation, either 'union', 'small', or 'large'. Returns: float: The overlap ratio value between the two bounding boxes dtyperrr rrVr[r\r]r^r_)rr9float64rrmultiplycalculate_bbox_arearb) rWrXrc x_min_inter y_min_inter x_max_inter y_max_inter inter_width inter_height inter_area bbox1_area bbox2_arearef_areas r*calculate_overlap_ratior{s HU"* - - -E HU"* - - -E*U1XuQx00K*U1XuQx00K*U1XuQx00K*U1XuQx00K*Q k 9::K:a{!:;;L[,bjIIIJ$U++J$U++J w *Z7 :j*55 :j*55 OD O O O   1}}s   r,cl|stdtj|}tj|dddf}tj|dddf}tj|dddf}tj|dddf}tj||||gS)a? Calculate the minimum enclosing bounding box for a list of bounding boxes. Args: bboxes (list): A list of bounding boxes represented as lists of four integers [x1, y1, x2, y2]. Returns: list: The minimum enclosing bounding box represented as a list of four integers [x1, y1, x2, y2]. z$The list of bounding boxes is empty.Nrrr r)rbrr9rar`)bboxes bboxes_arraymin_xmin_ymax_xmax_ys r* calculate_minimum_enclosing_bboxrs A?@@@8F##L F<1% & &E F<1% & &E F<1% & &E F<1% & &E 8UE5%0 1 11r,cFttjd|S)z#check if the char is english letterz ^[A-Za-z]$boolrematchchars r*is_english_letterr(s -- . ..r,cFttjd|S)zcheck if the char is numericz^[\d]+$rrs r* is_numericr-s T** + ++r,chd}||vS)z check if the char is non-breaking punctuation Args: char (str): character to check Returns: bool: True if the char is non-breaking punctuation > “、，：；,-"':;rB)rnon_breaking_punctuationss r*is_non_breaking_punctuationr2s" ! ! ! , ,,r,c tttt|\}}}}d|d|d|d|d|d S)Nz imgs/img_in__box__z.jpg)r6mapint)labelboxx_miny_minx_maxy_maxs r*construct_img_pathrMsQ!%c#smm!4!4E5% I% I Ie I Ie I Ie I Ie I I IIr,c|||||f}|jdkrdS|jdkrtj|S|jddkr)tj|dddfddddfStj|S)z Crop `original_img` to [y_min:y_max, x_min:x_max], clamped to image bounds. Returns a C-contiguous array suitable for `PIL.Image.fromarray`. rNr r.)sizendimrascontiguousarrayshape) original_imgrrrrcrops r*_crop_image_region_for_pilrRs e U5[0 1D yA~~t yA~~#D))) z"~#DbqbM#ttt)$<===  % %%r,c g}t|jdt|jd}}|D]8}|dtdvr|d}tt t|d\}}} } t dt ||}t dt | |} t dt ||}t dt | |} | |ks| |krt||d} t|||| | } | tj | } | | | |||| | f|dd:|S)Nrrr image_labels coordinatescore)pathimgrrr) rrrr6rr`rarrr fromarrayr8)rlayout_det_objs imgs_in_dochrMdet_objrrrrrimg_path arr_for_pilrs r* gather_imgsrasK |!!$ % %s<+=a+@'A'AqA" 7 ~> > >G$E)-c#w|7L.M.M)N)N &E5%3ua==))E3ua==))E3ua==))E3ua==))E~~%)%1FGGH4eUE5K"/+..C   $"#(%">$W-     r,ratiosmallerct|}t|}t||d}||kr||kr|s||kr|sdSdSdS)a Determine if the overlap area between two bounding boxes exceeds a given ratio and return the smaller (or larger) bounding box based on the `smaller` flag. Args: bbox1 (Union[List[int], Tuple[int, int, int, int]]): Coordinates of the first bounding box [x_min, y_min, x_max, y_max]. bbox2 (Union[List[int], Tuple[int, int, int, int]]): Coordinates of the second bounding box [x_min, y_min, x_max, y_max]. ratio (float): The overlap ratio threshold. smaller (bool): If True, return the smaller bounding box; otherwise, return the larger one. Returns: Optional[Union[List[int], Tuple[int, int, int, int]]]: The selected bounding box or None if the overlap ratio is not exceeded. r[rcrr N)rpr{)rWrXrrarea1area2 overlap_ratios r*_get_minbox_if_overlap_by_ratiorse*  & &E  & &E+E5wGGGMu UNNwNEUNN7N11 4r,?blocks thresholdc@t}t|}g}t|dD]\}}t|dzt |dD]}|d|}||vs||vrt |d|d||} | e|ddk} |ddk} | | kr(| r|n|} ||d| n | dkr|n|} || t|d D] } |d| = |S) a  Remove overlapping blocks based on a specified overlap ratio threshold. Args: blocks (List[Dict[str, List[int]]]): List of block dictionaries, each containing a 'block_bbox' key. threshold (float): Ratio threshold to determine significant overlap. smaller (bool): If True, the smaller block in overlap is removed. Returns: Tuple[List[Dict[str, List[int]]], List[Dict[str, List[int]]]]: A tuple containing the updated list of blocks and a list of dropped blocks. boxesrr)rNrimageT)reverse) r7r enumeraterrrr8addrK)rrrdropped_indexesoverlap_image_blocksrTblock1jblock2overlap_box_indexis_block1_imageis_block2_image drop_indexindexs r*remove_overlap_blocksrsueeO f  Fvg//00 6q1uc&/2233 0 0AG_Q'FO##qO';'; ?|$|$ !!!  !,"(/W"<"(/W"<"o55&5!<1J(//w 0KLLLL&71&<&<!J##J///- 02666## 7OE " " Mr,bboxctj|}tj|}t|jdkr|nt |gd}t|jdkr|nt |gd}t |d|d}t |d|d}t |d|d}t |d|d}||ks||krdS|dkrtj||||gS|dkr,tj||g||g||g||ggtjStd ) a_ Compute the intersection of two bounding boxes, supporting both 4-coordinate and 8-coordinate formats. Args: bbox1 (tuple): The first bounding box, either in 4-coordinate format (x_min, y_min, x_max, y_max) or 8-coordinate format (x1, y1, x2, y2, x3, y3, x4, y4). bbox2 (tuple): The second bounding box in the same format as bbox1. return_format (str): The format of the output intersection, either 'bbox' or 'poly'. Returns: tuple or None: The intersection bounding box in the specified format, or None if there is no intersection. rrr rNrpolyrlz.return_format must be either 'bbox' or 'poly'.) rr9rrrr`raint16rb) rWrX return_formatrect1rect2rqrrrsrts r*get_bbox_intersectionrsy HUOOE HUOOE%%**EE0G0P0PQR0SE%%**EE0G0P0PQR0SEeAha))KeAha))KeAha))KeAha))Kk!![K%?%?txk; LMMM & xk*k*k*k*   (    IJJJr,c |\}}}} |\} } } } ddddd}| |z |z | |z |z || z |z | | z |z g}t|}t|}|||}t|dkr|gfSt dD]}||}|dd}||||<gg}}|D]w}t ||d}|t jd d kr||I|t jd d kr||xt|dkrt|dkr|D]}|}|\}}}} |\} } } } | |z |z | |z |z || z |z | | z |z g}t|}||}||}||||<t|||||\}}t|dkrfd |D}t|}nA| |t|}|||}||fS)aG Shrink the supplement region bbox according to the reference region bbox and match the block bboxes. Args: supplement_region_bbox (list): The supplement region bbox. ref_region_bbox (list): The reference region bbox. image_width (int): The width of the image. image_height (int): The height of the image. block_idxes_set (set): The indexes of the blocks that intersect with the region bbox. block_bboxes (dict): The dictionary of block bboxes. Returns: list: The new region bbox and the matched block idxes. r rrr)rrr rNr[r#match_block_overlap_ratio_thresholdg?#split_block_overlap_ratio_thresholdg?c g|] }| SrBrB).0idx block_bboxess r* z1shrink_supplement_region_bbox..TsLLLCl3/LLLr,) rrarrrr{rgetr8r`shrink_supplement_region_bboxrremove)supplement_region_bboxref_region_bbox image_width image_heightblock_idxes_setrr#r$r%r&x1_primey1_primex2_primey2_primeindex_conversion_mapedge_distance_listedge_distance_list_tmp min_distance src_indexr dst_indextmp_region_bboxiner_block_idxessplit_block_idxes block_idxrsplit_block_idxsplit_block_bbox max_distance iner_idxesmatched_bboxess ` r*rrs4,,NBB-<*Hh( Q133 B+% B,& h+% h,&  &&899)**L$%7%=%=l%K%KLI ?q  %r)) 1XX2U2U(3 03%4Y%? ".0"+( 4 4I3i!8wM25s  !'' 2222!45s"""((333  1 $ $$%%))'8!!O'3O'D$%4NBB=M:Hh(!B+5!B,6h+5h,6 *& $''9#:#:L 2 8 8 F FI 4Y ?I1A)1LOI.2O''#$($ 33/OZ:!++ ,LLLL;KLLLN%En%U%U " E " ) ), 7 7 7566L,-?-E-El-S-STII !#3 33r,c||S|\}}}}|\}}}} tt||}tt||}tt||}tt|| } |||| g}|S)zUpdate region box with bbox)rrar`) r region_boxr#r$r%r& x1_region y1_region x2_region y2_regions r*update_region_boxr ^s NBB1;.Iy)YCI&&''ICI&&''ICI&&''ICI&&''IY 9=J r,formula_res_listocr_rescT|D]#}ttt|d\}}}}||f||f||f||fg}|d||d}|d||djdkrt j|dg|d<n&t j|d|dgf|d<|dd|d||d d %d S) zConvert formula result to OCR result format Args: formula_res_list (List): Formula results ocr_res (dict): OCR result Returns: ocr_res (dict): Updated OCR result dt_polys rec_formular3r5r rec_labelsformular2r4rN)r6rrr8rrr9vstack) r r formula_resrrrr poly_pointsformula_res_texts r*!convert_formula_res_to_ocr_formatrpsR((( %)#c;z3J*K*K%L%L"ueU EN EN EN EN   "";/// +M : ##$4555 ;  $ ) )#%8[-D,E#F#FGK #%9% J(?'@A$$GK   $$Y/// ##K000 $$Q'''''((r,cjtt|\}}}}t||z ||z z}|S)zCalculate bounding box area)rfloatabs)rr#r$r%r&areas r*rprps8%%NBB RBG$ % %D Kr,c>|\}}|\}}||z dz||z dzzdzS)z/Calculate euclidean distance between two pointsr g?rB)point1point2r#r$r%r&s r*caculate_euclidean_distr s4 FB FB "WNb2g!^ + 33r,c&d}d}|j}|j}|j}|j}||j}|j} t |j| z dk} |dk} ||jko ||jk} | rCt |j|}t|j|}t || z dk} d} nt |j|jz } ||z dk}| r$|r"| r | t|j|jkrd}n ||z dkrd}||z dkrd}||fS)a Get segment start flag and end flag based on previous block Args: block (Block): Current block prev_block (Block): Previous block Returns: seg_start_flag (bool): Segment start flag seg_end_flag (bool): Segment end flag TN rrF) start_coordinateend_coordinateseg_start_coordinateseg_end_coordinate num_of_linesrrar`width)block prev_blockseg_start_flag seg_end_flagcontext_left_coordinatecontext_right_coordinater%r&num_of_prev_linespre_block_seg_end_coordinateprev_end_space_smallprev_lines_more_than_oneoverlap_blocks edge_distancecurrent_start_space_smalls r* get_seg_flagr6sNL#4$3 51&3'1'D$  ),HH I IB N $5q#8  $j&? ? G(:+FF   T&)+-D'' #(+)+C(( $,/KKLLrQ !MM 69R RSSM$8;R$RUW$W! ! #) #) #J$4ek B BBB"N "9 9B > >"N"44r99 < ''r,rc |rdddddddddd d }nr8r;)r>r8r;)r83r;)rBr;)r@rLr;)5r;)rHr;) doc_titledoc_title_textparagraph_titlesub_paragraph_titlevision vision_titlevision_footnote normal_text cross_layoutcross_referencerLrJ table_title)r8r8r>r; figure_title chart_titlerPtext vertical_textinline_formular)rr8rr;display_formulaabstractcontent)(\r;seal)rarar;table)rcrr;rfigure)rer;)r8rcr;)r8r;)chart referencereference_content algorithm)r)r order_label label_colors default_colors r*rrsT/ -23#7)03,/3   ! 3!  -! / !  0!  /!  3!  %!  .!  /!  '! /!! " ,#! & )'! * (+! . '/! 2 (3! 6 )7! 8*-!5-A! ! ! D)M   E= 1 11r,)TF)rUrV)rV)T)rT)r)F)3__all__rcopyrtypingrrr r r numpyrPILr componentsr ocr.resultrsettingrrndarrayr+rrrrstrrjr6tupler{rrrrrrrrrrrrr dictrrpr r6rrBr,r*r}sK     55555555555555000000""""""55555555RZBJ4:" 99999 9  9999x444t"  ,9,9 ;,9 ;,9,9  ,9,9,9,9d .!.! T5( ).! T5( ).! .!.!.!.!b2224/// ,,, ---6JJJ & & &H  cE#sC"455 6 cE#sC"455 6   eDIuS#sC%7889: FRV00 c49n% &0380JN0 4S$s)^$ %tDd3i,@'A AB0000f+K+K+K+K\W4 W4W4W4W4t$((t((((>444@(@(@(F2222#22U222222222222r,