j d dl Z d dlmZ d dlmZ d dlmZ d dlm Z m
Z
ddl m Z m
Z
dj d Zd
e
d efd Zdd
e
d
ed e
fdZd
e
fdZded efdZded efdZy) N)BytesIO)BinaryIO) ElementTree)
BeautifulSoupTag )OMML_NS oMath2Latex )z z{0}tagreturnc t j t j t | }|j t dz }t | j }|S )z
Converts an OMML (Office Math Markup Language) tag to LaTeX format.
Args:
tag (Tag): A BeautifulSoup Tag object representing the OMML element.
Returns:
str: The LaTeX representation of the OMML element.
oMath) ET
fromstringMATH_ROOT_TEMPLATEformatstrfindr r
latex)r math_root math_elementr s l/root/aria/tools/markitdown-venv/lib/python3.12/site-packages/markitdown/converter_utils/docx/pre_process.py_convert_omath_to_latexr ! sJ
0 7 7C ABI >>'G"34L % + +E L blockc t d }|rdt | dndt | d|_ t d }|j | |S )a
Creates a replacement tag for an OMML (Office Math Markup Language) element.
Args:
tag (Tag): A BeautifulSoup Tag object representing the "oMath" element.
block (bool, optional): If True, the LaTeX will be wrapped in double dollar signs for block mode. Defaults to False.
Returns:
Tag: A BeautifulSoup Tag object representing the replacement element.
zw:tnamez$$$zw:r)r r stringappend)r r t_tagr_tags r _get_omath_tag_replacementr% 4 s]
UOE
$S
) *"-(-.a
0
L
UOE LL Lr c H | j dk( rPt d }| j d D ] }|j t |d | j | y | j dk( r | j t | d y t
d| j )
a
Replaces OMML (Office Math Markup Language) elements with their LaTeX equivalents.
Args:
tag (Tag): A BeautifulSoup Tag object representing the OMML element. Could be either "oMathPara" or "oMath".
Raises:
ValueError: If the tag is not supported.
oMathParazw:pr r T)r FzNot supported tag: N)r r find_allr" r% replace_with
ValueError)r p_tag child_tags r _replace_equationsr- J s xx; g. LI LL3ITJ K L W 3CuEF.sxxj9::r contentc t | j d }|j d D ]
}t | |j d D ]
}t | t | j S )a
Pre-processes the math content in a DOCX -> XML file by converting OMML (Office Math Markup Language) elements to LaTeX.
This preprocessed content can be directly replaced in the DOCX file -> XMLs.
Args:
content (bytes): The XML content of the DOCX file as bytes.
Returns:
bytes: The processed content with OMML elements replaced by their LaTeX equivalents, encoded as bytes.
xml)featuresr' r )r decoder( r- r encode)r. soupr s r _pre_process_mathr5 c sj )E :D}}[) 3 }}W% 3 t9 r
input_docxc X t }g d}t j | d 5 }|j D ci c] }||j | }}t j |d 5 }|j
|_ |j
D ]: \ }}||v r t | }|j || )|j || < ddd ddd |j d |S c c}w # t $ r |j || Y w xY w# 1 sw Y JxY w# 1 sw Y NxY w)a
Pre-processes a DOCX file with provided steps.
The process works by unzipping the DOCX file in memory, transforming specific XML files
(such as converting OMML elements to LaTeX), and then zipping everything back into a
DOCX file without writing to disk.
Args:
input_docx (BinaryIO): A binary input stream representing the DOCX file.
Returns:
BinaryIO: A binary output stream representing the processed DOCX file.
)zword/document.xmlzword/footnotes.xmlzword/endnotes.xmlr)modewNr ) r zipfileZipFilenamelistreadcommentitemsr5 writestr Exceptionseek) r6 output_docxpre_process_enable_files zip_inputr files
zip_outputr. updated_contents r pre_process_docxrJ v s' )K
# . 7)8A8J8J8LMy~~d++MM
__[s
3
7z!*!2!2J !&
7
g33;*;G*D"++D/B
''g6
7
77 Q ! N % ;"++D':;
7
77 7sX D C.D +,D C35D D .D 3D DD DD D D))F)r; ior typingr xml.etreer r bs4r r math.ommlr r
joinr r r boolr% r- bytesr5 rJ r r rT s ' " +WW . &C ,;C ;2 u && &h &r