o eÀ=ã@sˆdZddlZddlmZmZddlmZddlmZddlm Z dd l m Z dd l m Z dd l mZmZdd lmZGd d„de ƒZdS)a—Text Span object based on PDF raw dict extracted with ``PyMuPDF``. Data structure for Span refer to this `link `_:: { # raw dict --------------------------- 'bbox': (x0,y0,x1,y1), 'color': sRGB 'font': fontname, 'size': fontsize, 'flags': fontflags, 'chars': [ chars ], # added dict ---------------------------- 'text': text, 'style': [ { 'type': int, 'color': int, 'uri': str # for hyperlink }, ... ] } éN)ÚPtÚRGBColor)Úqné©ÚCharé)ÚElement)ÚRectType)Ú constants)ÚshareÚdocx)ÚShapecsàeZdZdZd)def‡fdd„ Zedd„ƒZejdd„ƒZd d „Z ed d „ƒZ d e fdd„Z de fdd„Zdd„Zdd„Z‡fdd„Zdef‡fdd„ Zd*dedefdd „Zd*dedefd!d"„Zd#d$„Zd%d&„Zd'd(„Z‡ZS)+ÚTextSpanzObject representing text span.NÚrawcsè|pi}| dd¡|_| dd¡|_dd„| dg¡Dƒ}dd„|Dƒ|_| dd ¡|_| d d ¡|_| d d ¡|_| d d¡|_| dd¡|_| dd¡|_ | dg¡|_ | dd¡|_ t ƒ  |¡d|j ¡vrr| tj¡dSdS)NÚcolorrÚflagscSsg|]}t|ƒ‘qS©r)Ú.0ÚcrrúFD:\Projects\ConvertPro\env\Lib\site-packages\pdf2docx/text/TextSpan.pyÚ 3sz%TextSpan.__init__..ÚcharscSsg|] }|jdkr|‘qS)Ú©r©rÚcharrrrr4sÚtextrÚfontÚsizeg(@Úascenderçð?Ú descenderçÚ line_heightéÿÿÿÿÚstyleÚ char_spacingZUNNAMED)ÚgetrrrÚ_textrrr r"r$r&r'ÚsuperÚ__init__ÚupperÚ_change_font_and_update_bboxr ZDEFAULT_FONT_NAME)Úselfrr©Ú __class__rrr+-s" ÿzTextSpan.__init__cCs"|jrd dd„|jDƒ¡S|jS)z:Get span text. Note joining chars is in a higher priority.rcSsg|]}|j‘qSrrrrrrrTsz!TextSpan.text..)rÚjoinr)©r.rrrrQs"z TextSpan.textcCs ||_dS)zLSet span text directly in case no chars are stores, e.g. restored from json.N)r))r.ÚvaluerrrrVs cCs"t ¡}|jD]}||jO}q|S)z,Calculate bbox based on contained instances.)ÚfitzZRectrÚbbox)r.r5rrrrÚcal_bbox[szTextSpan.cal_bboxcCs |jdkS)Nr%)r$r2rrrÚis_valid_line_heightas zTextSpan.is_valid_line_heightÚ font_namec Csì||_t |¡}|j|j|jd}||jjkr"|j|jj|9_|j\}}}}t dd||f¡}|j |j dj |j||jd\} } | j |jd} | j | }| j| }| ||||f¡|j D]} | j\}} }} |  ||||f¡qadS)aSet new font, and update font size, span/char bbox accordingly. It's generally used for span with unnamed fonts. See this `issue `_. In corner case, where the PDF file containing unnamed and not embedded fonts, the span bbox extracted from ``PyMuPDF`` is not correct. ``PyMuPDF`` provides feature to replace these unnamed fonts with specified fonts, then extract correct bbox from the updated PDF. Since we care less about the original PDF itself but its layout, the idea here is to set a default font for text spans with unnamed fonts, and estimate the updated bbox with method from ``fitz.TextWriter``. Args: font_name (str): Font name. )Úfontsizer)rr9ç@N)rr4ZFontÚ text_lengthrrr5ÚwidthZ TextWriterÚappendrÚoriginÚheightÚy0Úy1Ú update_bbox) r.r8rÚ new_lengthÚx0r@Úx1rAZtwÚrectÚ_Zbuffrrrrr-es*    ü   þz%TextSpan._change_font_and_update_bboxrcCs|j |¡| |¡dS)z%Add char and update bbox accordingly.N)rr=Ú union_bbox)r.rrrrÚadd–s z TextSpan.addcCsP|j}| d¡s dSt|ƒt| ¡ƒ}|j|dd…|_|j| ¡ddS)z3Remove blanks at the left side, but keep one blank.ú FrN©rFT)rÚ startswithÚlenÚlstriprrBr6©r.Z original_textZ num_blanksrrrrNœó zTextSpan.lstripcCsP|j}| d¡s dSt|ƒt| ¡ƒ}|jdd|…|_|j| ¡ddS)z4Remove blanks at the right side, but keep one blank.rJFNrrKT)rÚendswithrMÚrstriprrBr6rOrrrrR¨rPzTextSpan.rstripc s:tƒ ¡}| |j|j|j|j|j|j|j |j dœ¡|S)N)rrrr$rrr&r') r*ÚstoreÚupdaterrrr$rrr&r')r.Úresr/rrrS´s ø zTextSpan.storercstƒj|||dddS)Nr)ÚstrokeÚfillr<)r*Úplot)r.Úpagerr/rrrXÃsz TextSpan.plotTrFÚ horizontalc sĈj|j@}|jr |gSg}ˆr|jj|_|jj|_n |jj|_|jj|_‡‡fdd„}tt|t|j ƒƒƒ}|r>|ddnd}t |ƒ}t ||dƒ} |dkr€ˆr_|jj|jj|j|jjf} n |jj|j|jj|jjf} |  ¡  | ¡} |j d|…| _ | | ¡|dkr¨|j|j|j|jf} |  ¡  | ¡} |j || …| _ |  ˆˆ¡| | ¡| t |j ƒkràˆr¿|j|jj|jj|jjf} n |jj|jj|jj|jf} |  ¡  | ¡} |j | d…| _ | | ¡|S)aSplit span with the intersection: span-intersection-span. Args: rect (Shape): Target shape to split this text span. horizontal (bool, optional): Text direction. Defaults to True. Returns: list: Split text spans. cs|d ˆˆ¡S)Nr)Zcontained_in_rect)Úitems©rZrFrrÚêsz TextSpan.split..rr%N)r5Zis_emptyr@rArDrEÚlistÚfilterÚ enumeraterrMÚmaxÚcopyrBr=Ú_parse_text_format) r.rFrZZintsecZ split_spansÚfZ index_charsÚposÚlengthZpos_endr5Z split_spanrr\rÚsplitÆsD         zTextSpan.splitcCs8| tj¡s | tj¡rdS| tj¡r#|j |j|j|j dœ¡dS|r'dnd}|j |d|j |}|j |d|j |}t |j |d|j |ƒ}|d|kri|jt   d¡krh| |tj¡rhtj|_n|d |krttj|_nd ||kr‚d |krˆnntj|_|jsdS|j|jd œ}|j |¡dS) a=Parse text style based on the position to a rect shape. Args: rect (Shape): Target rect shape reprenting potential text style. horizontal (bool, optional): Horizontal text direction. Defaults to True. Returns: bool: Parsed text style successfully or not. F)ÚtyperÚuriTrrrgà?)rrrgÐ?gffffffÖ?gè?)rhr)Z equal_to_typer ZBORDERZSHADINGÚ HYPERLINKr&r=rhrrir5Úabsr Z rgb_valueÚ get_main_bboxr Z FACTOR_MAJORÚ HIGHLIGHTÚ UNDERLINEÚSTRIKEZ is_determined)r.rFrZÚidxZh_rectZh_spanÚdr&rrrrcs6  ý  €    þ zTextSpan._parse_text_formatcCsv| |j¡r | ¡S| |j¡stƒS| ¡}|j ¡| d¡|jD]}| |t j ¡r8|j  |¡|  |¡q$|S)z‰Create new TextSpan object with chars contained in given bbox. Args: rect (fitz.Rect): Target bbox. )r#r#r#r#) Úcontainsr5rbÚ intersectsrrÚclearrBrlr Z FACTOR_A_HALFr=rH)r.rFÚspanrrrrrsPs       €zTextSpan.intersectscCsp|jD]}|dtjjkr|j ¡rt ||d|j¡}nq| |j¡}|  |¡|j r6t  ||j ¡dSdS)a€Add text span to a docx paragraph, and set text style, e.g. font, color, underline, hyperlink, etc. .. note:: Hyperlink and its style is parsed separately from pdf. For instance, regarding a general hyperlink with an underline, the text and uri is parsed as hyperlink itself, while the underline is treated as a normal text style. rhriN) r&r rjr3rÚstripr Z add_hyperlinkZadd_runÚ_set_text_formatr'Zset_char_spacing)r.Z paragraphr&Údocx_runrrrÚ make_docxks €  ÿzTextSpan.make_docxcCs.t|jd@ƒ|_t|jd@ƒ|_t|jd@ƒ|_|j}||j_|jjj   t dƒ|¡t t  |j¡Ž|jj_t|jdƒd}t|ƒ|j_|j|}t|dƒdkrWt ||¡|jD]:}|d}|tjjkrot ||d ¡qZ|tjjkrŠ|j|d krd |j_qZt ||d ¡qZ|tjjkr”d |j_ qZd S) z/Set text format for ``python-docx.run`` object.rréz w:eastAsiar:r!g{®Gáz„?rhrTN)!ÚboolrZ superscriptÚitalicÚboldrÚnameZ_elementZrPrZrFontsÚsetrrr Z rgb_componentrÚrgbÚroundrrrkr Zset_char_scalingr&r rmr3Zset_char_shadingrnÚ underlineZset_char_underlineroÚstrike)r.rxr8Z font_sizeÚscaler&Útrrrrwƒs0         €îzTextSpan._set_text_format)N)T)Ú__name__Ú __module__Ú __qualname__Ú__doc__Údictr+ÚpropertyrÚsetterr6r7Ústrr-rrIrNrRrSÚtuplerXrr{rgrcrsryrwÚ __classcell__rrr/rr+s*$   1 M=r)r‰r4Z docx.sharedrrZ docx.oxml.nsrrZcommon.Elementr Z common.sharer Úcommonr r r Z shape.ShaperrrrrrÚs