o Õj“ã@sPdZddlZddlmZddlmZddlmZddlmZGd d „d eƒZ dS) ak Objects representing PDF path (stroke and filling) extracted by ``page.get_drawings()``. This method is new since ``PyMuPDF`` 1.18.0, with both pdf raw path and annotations like Line, Square and Highlight considered. * https://pymupdf.readthedocs.io/en/latest/page.html#Page.get_drawings * https://pymupdf.readthedocs.io/en/latest/faq.html#extracting-drawings éNé)ÚImagesExtractor)Ú lazyproperty)Ú Collectioné)ÚPathc @speZdZdZdefdd„Zedd„ƒZedd„ƒZ d d „Z d d „Z   dde de de de de f dd„Z dS)ÚPathszA collection of paths.ÚrawscCsBdd|jj|jjf}|D]}t|ƒ}|j |¡sq | |¡q |S)z>Initialize paths from raw data get by ``page.get_drawings()``.r)ÚparentÚwidthÚheightrÚbboxÚ intersectsÚappend)Úselfr ÚrectÚrawÚpath©rúU/var/www/html/Deteccion_Ine/venv/lib/python3.10/site-packages/pdf2docx/shape/Paths.pyÚrestores  z Paths.restorecCs"t ¡}|jD]}||jO}q|S)N)ÚfitzÚRectÚ _instancesr )rr Úinstancerrrr sz Paths.bboxcCs|jD]}|js dSqdS)z@It is iso-oriented when all contained segments are iso-oriented.FT)rÚis_iso_oriented)rrrrrr's zPaths.is_iso_orientedcCs4|jsdS| ¡}|jD]}| |¡q | ¡dS)zePlot paths for debug purpose. Args: page (fitz.Page): ``PyMuPDF`` page. N)rZ new_shapeÚplotÚcommit)rÚpageZcanvasrrrrr/s  z Paths.plotcCs*g}|jD] }|js q| | ¡¡q|S)zˆConvert contained paths to ISO strokes or rectangular fills. Returns: list: A list of ``Shape`` raw dicts. )rrÚextendÚ to_shapes)rZshapesrrrrr <s  zPaths.to_shapesérç@Úmin_svg_gap_dxÚmin_svg_gap_dyÚmin_wÚmin_hÚclip_image_res_ratioc Cs g}|jr| | ¡¡|gfSg}t|jjƒ}| ||||¡} dtdtfdd„} dd„| Dƒ} |j D] } t | | ƒD]\\} }}| j   | ¡rR| | |ƒsP|  | ¡nq:q3t | | ƒD]2\\} }}|jr}| | ¡¡|D]}|  |jt |¡d|d¡qkqY|  |jt | ¡d|d¡qY||fS) a¶Convert paths to iso-oriented shapes or images. The semantic type of path is either table/text style or vector graphic. This method is to: * detect svg regions -> exist at least one non-iso-oriented path * convert svg to bitmap by clipping page * convert the rest paths to iso-oriented shapes for further table/text style parsing Args: min_svg_gap_dx (float): Merge svg if the horizontal gap is less than this value. min_svg_gap_dy (float): Merge svg if the vertical gap is less than this value. min_w (float): Ignore contours if the bbox width is less than this value. min_h (float): Ignore contours if the bbox height is less than this value. clip_image_res_ratio (float, optional): Resolution ratio of clipped bitmap. Defaults to 3.0. Returns: tuple: (list of shape raw dict, list of image raw dict). rÚcontourscSs&|D]}t |¡ |j¡rdSqdS)NTF)rrÚcontainsr )rr(r rrrÚcontained_in_inner_contoursqsz?Paths.to_shapes_and_images..contained_in_inner_contourscSsg|]}tƒ‘qSr)r)Ú.0Ú_rrrÚ wsz.Paths.to_shapes_and_images..T)r Zrm_imager')rrr rr Z page_engineZdetect_svg_contoursrÚlistrÚzipr rrZclip_page_to_dictrr)rr#r$r%r&r'Z iso_shapesZimagesZieÚgroupsr*Z group_pathsrr Z inner_bboxesÚpathsZsvg_bboxrrrÚto_shapes_and_imagesJs<   þ€ þÿ þzPaths.to_shapes_and_imagesN)r!r!rrr")Ú__name__Ú __module__Ú __qualname__Ú__doc__r.rrr Úpropertyrrr Úfloatr2rrrrrs2   ûÿþýüûr) r6rZimage.ImagesExtractorrZ common.sharerZcommon.CollectionrrrrrrrÚs