:Qg ddlmZddlZddlZddlmZmZmZmZddl m Z ddl m Z m Z ddlmZddlmZmZmZmZmZmZmZmZmZmZddlmZdd lmZdd lm Z m!Z!dd l"m#Z#dd l$m%Z%m&Z&dd l'm(Z(m)Z)m*Z*m+Z+m,Z,m-Z-e%ej\e ddddddd ddZ/ d ddZ0 ddZ1ddZ2 d ddZ3 d ddZ4d dZ5y)!) annotationsN)IOAnyCallableLiteral)add_chunking_strategy)auto_paragraph_grouper clean_bullets)CoordinateSystem) AddressElementElementMetadata EmailAddressFooterHeaderListItem NarrativeTextTextTitle) read_txt_file)FileType)PARAGRAPH_PATTERNUNICODE_BULLETS_RE) exactly_one)apply_metadataget_last_modified_date)is_bulleted_textis_email_addressis_possible_narrative_textis_possible_numbered_listis_possible_titleis_us_city_state_ziptext)fileencodingr#paragraph_grouperdetection_originc  ||jdk(r|s|sgSt|||d}|t||\}}n |t||\}}n | t|}|durn| ||}n t |}t |}g} t |r t|nd} || _|D]W} | j} | st| r"t| } tj| | _ | j| Y| S)aPartition a .txt documents into its constituent paragraph elements. If paragraphs are below "min_partition" or above "max_partition" boundaries, they are combined or split. Parameters ---------- filename A string defining the target filename path. file A file-like object using "rb" mode --> open(filename, "rb"). encoding The encoding method used to decode the input bytes when drawn from `filename` or `file`. Defaults to "utf-8". text The string representation of the .txt document. paragrapher_grouper A str -> str function for fixing paragraphs that are interrupted by line breaks for formatting purposes. N)filenamer$r#)r*r%)r$r%F) last_modified)striprrstrr _split_by_paragraphrrr'_is_empty_bulletelement_from_textcopydeepcopymetadataappend) r*r$r%r#r&r'kwargs file_text file_contentelementsr3ctextelements X/var/www/html/answerous/venv/lib/python3.12/site-packages/unstructured/partition/text.pypartition_textr<(s> DJJLB.tH 48I+XQ) +I) I E!  &%i0 *95 &y1L H:B,X6H!1H%  )%0'.G#}}X6G  OOG $ % Oct||rt|||St||rt|||St |rt |}t |||St|r t|St|rt|||St|rt |||St|rt|||St|rt|||St!|||S)N)r# coordinatescoordinate_systemr#)_is_in_header_positionr_is_in_footer_positionrrr rrrr"r r rrr!rr)r#r?r@ clean_texts r;r0r0os k+<=#/  -> ?#/  $ "4( #/  $ && d ##/  #4 (#/  $D )#/  4 #/  #/  r=c\td|Dt|z }||jz S)Nc3&K|] }|d yw)N).0 coordinates r; z)_get_height_percentage..s<* 1  BBr=c.||yt||}||kDS)zZChecks to see if the position of the text indicates that the text belongs to a footer.FrPr?r@ thresholdheight_percentages r;rCrCs,/7.{ros" --7@   ;2K<Y B"FJ#)BB B B  B D B!BBBBN;?157 7 77 /7  7 ~,0,', ,C )7 ). ) ) )" )7 ). ) ) ):r=