o e, @sdZddlZddlZddlmZddlmZddlmZddZdd Z d d Z d d Z ddZ dddZ dddddddddd ZdddZGdddeZdS)zZ This code is refer from: https://github.com/weizwx/html2docx/blob/master/htmldocx/h2d.py N)Document) BeautifulSoup) HTMLParsercCsgd}|jd|ddS)Nz table > trztable > thead > trztable > tbody > trztable > tfoot > trz, F recursive)selectjoin) table_souptable_row_selectorsr \D:\Projects\ConvertPro\env\Lib\site-packages\paddleocr/ppstructure/recovery/table_process.pyget_table_rowssrcCs|r |jddgddSgS)NthtdFr)find_all)rowr r r get_table_columns"srcCsLt|}|r t|dng}d}|D]}|jdd}|t|7}q||fS)Nrcolspan)rrattrsgetint)r rowscolsZ col_countcolrr r r get_table_dimensions'srcCsddd|jDS)N cSsg|]}t|qSr )str).0ir r r ;sz!get_cell_html..)r contents)soupr r r get_cell_html7sr$cCs$|j}||d|_|_dSN)Z_elementZ getparentremove_p) paragraphpr r r delete_paragraph>sr*FcCs@|r tdd|}|rtdd|}tdd|}tdd|S)aPRemove white space from a string. Args: string(str): The string to remove white space from. leading(bool, optional): Remove leading new lines when True. trailing(bool, optional): Remove trailing new lines when False. Returns: str: The input string with new line characters removed and white space squashed. Examples: Single or multiple new line characters are replaced with space. >>> remove_whitespace("abc\ndef") 'abc def' >>> remove_whitespace("abc\n\n\ndef") 'abc def' New line characters surrounded by white space are replaced with a single space. >>> remove_whitespace("abc \n \n \n def") 'abc def' >>> remove_whitespace("abc \n \n \n def") 'abc def' Leading and trailing new lines are replaced with a single space. >>> remove_whitespace("\nabc") ' abc' >>> remove_whitespace(" \n abc") ' abc' >>> remove_whitespace("abc\n") 'abc ' >>> remove_whitespace("abc \n ") 'abc ' Use ``leading=True`` to remove leading new line characters, including any surrounding white space: >>> remove_whitespace("\nabc", leading=True) 'abc' >>> remove_whitespace(" \n abc", leading=True) 'abc' Use ``trailing=True`` to remove trailing new line characters, including any surrounding white space: >>> remove_whitespace("abc \n ", trailing=True) 'abc' z ^\s*\n+\s*z \s*\n+\s*$z\s*\n\s*rz\s+)resub)stringleadingZtrailingr r r remove_whitespaceEs (r0bolditalic underlinestrikeZ superscriptZ subscript) bstrongZemr ussupr-rZCourier)codeprecsheZdZfddZdddZddZdd Zd d Zd d ZddZ dddZ ddZ ddZ Z S) HtmlToDocxcs4tddddd|_gd|_d|_d|_dS)NT)fix-htmlimagestablesstylesr)super__init__optionsr table_styleparagraph_styleself __class__r r rBs   zHtmlToDocx.__init__NcCsnggd|_|r ||_nt|_|jd|_|j|_d|_|jd|_|jd|_d|_ d|_ d|_ d|_ dS)N)spanlistr=Tr>r@Fr) tagsdocrrCbsdocumentinclude_tablesZinclude_imagesZinclude_stylesr(skipZskip_tagZinstances_to_skip)rGrOr r r set_initial_attrss"   zHtmlToDocx.set_initial_attrscCs|j|_|j|_dS)z1Copy settings from another instance of HtmlToDocxN)rDrE)rGotherr r r copy_settings_froms zHtmlToDocx.copy_settings_fromcCs<g}d}|D]}|r|d8}q||t|d}q|S)a Returns array containing only the highest level tables Operates on the assumption that bs4 returns child elements immediately after the parent element in `find_all`. If this changes in the future, this method will need to be updated :return: rrtable)appendlenr)rGZ tables_soupZ new_tablesnestrUr r r ignore_nested_tabless zHtmlToDocx.ignore_nested_tablescCs2t|ds d|_dS||jd|_d|_dS)Nr#FrUr)hasattrrPrYr#rr?Ztable_norFr r r get_tabless  zHtmlToDocx.get_tablescCs<|jrtrt|d|_t|j}|jr|||dS)N html.parser)rNrr#rrPr[feed)rGhtmlr r r run_processs   zHtmlToDocx.run_processcCsjt|tjjstdtjj|jd}|jdkrt|||| ||j js3|j ddSdS)Nz Second argument needs to be a %srr+) isinstancedocxrU_Cell ValueErrorZ paragraphstextr*rRr_rM add_paragraph)rGr^cellZunwanted_paragraphr r r add_html_to_cells    zHtmlToDocx.add_html_to_cellc Cs\z|r ||j_WdS|jr|j|j_WdSWdSty-}z td|jd|d}~ww)NzUnable to apply style .)r(stylerEKeyErrorrc)rGrier r r apply_paragraph_styles z HtmlToDocx.apply_paragraph_stylecCs2t|d}t|\}}|t||}|jd|_d}t|D]w\}} t| } d} | D]f} t| j dd} t| j dd}t | }| j dkrLd|}| || }|jd krf| d7} | || }|jd ksW| ||d| | d}||kr}||t}||||pd || | 7} q+|d7}qd S) a* To handle nested tables, we will parse tables manually as follows: Get table soup Create docx table Iterate over soup and fill docx table with new instances of this parser Tell HTMLParser to ignore any tags until the corresponding closing table tag r\z Table Gridrrrrowspanrz %sr+rN)rrZ add_tablerWr@ri enumeraterrrrr$namerfrdmerger<rTrg)rGr^rMr rZcols_lenrUZcell_rowindexrrZcell_colrrrmZ cell_htmlZ docx_cellZ cell_to_mergeZ child_parserr r r handle_tables<             zHtmlToDocx.handle_tablec Cs|jrdSd|jvrt|dd}|js|j|_||jd}|r/||d|dS|j ||_ |jd}|D]}d|vrO| |d}| |q=|jD]}|t vret |}t|j j|d|tvrrt|}||j j_qSdS)Nr;TahrefrJri)rQrLr0r(rMrerlrZ handle_linkZadd_runrunZparse_dict_stringZadd_styles_to_run font_stylessetattrZfont font_namesro) rGdatalinkspansrJritagZ font_styleZ font_namer r r handle_datas4        zHtmlToDocx.handle_datar%)__name__ __module__ __qualname__rBrRrTrYr[r_rgrlrrr} __classcell__r r rHr r<s    +r<)FF)__doc__r,rarZbs4r html.parserrrrrr$r*r0rvrxr<r r r r s4      6