U
    f/e                     @  s\   d dl mZ d dlZd dlmZmZmZ d dlm	Z	 d dl
Zd dlmZ G dd deZdS )    )annotationsN)FilePathOrBufferScalarStorageOptions)import_optional_dependency)BaseExcelReaderc                      s   e Zd ZdZd+ddd fddZedd	 Zdd
ddZeddddZeddddZ	ddddZ
ddddZdddddZddd d!Zddd"d#Zddd$d%Zdd&dd'd(Zddd)d*Z  ZS ),	ODFReadera"  
    Read tables out of OpenDocument formatted files.

    Parameters
    ----------
    filepath_or_buffer : str, path to be parsed or
        an open readable stream.
    storage_options : dict, optional
        passed to fsspec for appropriate URLs (see ``_get_filepath_or_buffer``)
    Nr   r   )filepath_or_bufferstorage_optionsc                   s   t d t j||d d S )NZodf)r
   )r   super__init__)selfr	   r
   	__class__ >/tmp/pip-unpacked-wheel-tiezk1ph/pandas/io/excel/_odfreader.pyr      s    zODFReader.__init__c                 C  s   ddl m} |S )Nr   )OpenDocument)odf.opendocumentr   )r   r   r   r   r   _workbook_class%   s    zODFReader._workbook_class)r	   c                 C  s   ddl m} ||S )Nr   )load)r   r   )r   r	   r   r   r   r   load_workbook+   s    zODFReader.load_workbookstr)returnc                 C  s   dS )z'Property for compat with other readers. r   )r   r   r   r   empty_value0   s    zODFReader.empty_valuez	list[str]c                 C  s&   ddl m} | j|}dd |D S )z4Return a list of sheet names present in the documentr   Tablec                 S  s   g | ]}| d qS name)getAttribute).0tr   r   r   
<listcomp>;   s     z)ODFReader.sheet_names.<locals>.<listcomp>)	odf.tabler   bookgetElementsByType)r   r   tablesr   r   r   sheet_names5   s    zODFReader.sheet_namesint)indexc                 C  s*   ddl m} | | | j|}|| S )Nr   r   )r#   r   Zraise_if_bad_sheet_by_indexr$   r%   )r   r)   r   r&   r   r   r   get_sheet_by_index=   s    
zODFReader.get_sheet_by_indexr   c                 C  s^   ddl m} | | | j|}|D ]}|d|kr&|  S q&|   td| dd S )Nr   r   r   zsheet z
 not found)r#   r   Zraise_if_bad_sheet_by_namer$   r%   r   close
ValueError)r   r   r   r&   tabler   r   r   get_sheet_by_nameD   s    

zODFReader.get_sheet_by_nameboolzlist[list[Scalar]])convert_floatr   c                   sr  ddl m}m}m} | j}| j}||h ||}d}	d}
g }|D ]} fdd|jD }d}g }|D ]f}|j|kr| ||}n| j}| 	|}|| jkr||7 }qj|
| jg|  d}|
|g|  qj|
t|k rt|}
| |}| |r|	|7 }	qF|
| jgg|	  d}	t|D ]}|| q&qF|D ].}t||
k r>|
| jg|
t|   q>|S )z9
        Parse an ODF Table into a list of lists
        r   )CoveredTableCell	TableCellTableRowc                   s   g | ]}|j  kr|qS r   )qname)r    x
cell_namesr   r   r"   f   s     
 z,ODFReader.get_sheet_data.<locals>.<listcomp>)r#   r1   r2   r3   r4   r%   
childNodes_get_cell_valuer   _get_column_repeatextendlen_get_row_repeat_is_empty_rowrangeappend)r   Zsheetr0   r1   r2   r3   Zcovered_cell_nameZtable_cell_nameZ
sheet_rowsZ
empty_rowsZmax_row_lenr-   Z	sheet_rowZsheet_cellsZempty_cellsZ	table_rowZ
sheet_cellvalueZcolumn_repeatZ
row_repeat_rowr   r6   r   get_sheet_dataQ   sF    






zODFReader.get_sheet_datac                 C  s"   ddl m} t|j|dfdS )z
        Return number of times this row was repeated
        Repeating an empty row appeared to be a common way
        of representing sparse rows in the table.
        r   TABLENSznumber-rows-repeated   odf.namespacesrF   r(   
attributesget)r   rC   rF   r   r   r   r=      s    zODFReader._get_row_repeatc                 C  s"   ddl m} t|j|dfdS )Nr   rE   znumber-columns-repeatedrG   rH   )r   cellrF   r   r   r   r:      s    zODFReader._get_column_repeatc                 C  s$   |j D ]}t|j dkr dS qdS )z4
        Helper function to find empty rows
        r   FT)r8   r<   )r   rC   columnr   r   r   r>      s    
zODFReader._is_empty_rowr   c                 C  s@  ddl m} t|dkrtjS |j|df}|dkrJt|dkrFdS dS |d krX| jS |d	krt|j|d
f}|rt	|}||kr|S |S |dkr|j|d
f}t|S |dkr| 
|S |dkr|j|d
f}t|S |dkr|j|df}t|S |dkr&tt|}| S |   td| d S )Nr   )OFFICENSz#N/Az
value-typebooleanTRUETFfloatrA   
percentagestringcurrencydatez
date-valuetimezUnrecognized type )rI   rN   r   npnanrJ   rK   r   rQ   r(   _get_cell_string_valuepdZto_datetimerV   r+   r,   )r   rL   r0   rN   Z	cell_typeZ
cell_valuevalZstampr   r   r   r9      sB    



zODFReader._get_cell_valuec           	      C  s   ddl m} ddlm} ddlm} | j}g }|jD ]^}t||r|j|krtt	|j
|dfd}|d|  q|| | q6|t| q6d|S )	z
        Find and decode OpenDocument text:s tags that represent
        a run length encoded sequence of space characters.
        r   )Element)TEXTNS)ScrG    r   )Zodf.elementr\   rI   r]   Zodf.textr^   r4   r8   
isinstancer(   rJ   rK   r@   rY   r   join)	r   rL   r\   r]   r^   Ztext_srA   fragmentspacesr   r   r   rY      s    


z ODFReader._get_cell_string_value)N)__name__
__module____qualname____doc__r   propertyr   r   r   r'   r*   r.   rD   r=   r:   r>   r9   rY   __classcell__r   r   r   r   r      s$    
=

(r   )
__future__r   ZnumpyrW   Zpandas._typingr   r   r   Zpandas.compat._optionalr   ZpandasrZ   Zpandas.io.excel._baser   r   r   r   r   r   <module>   s   