o
    di                     @  s  d Z ddlmZ ddlmZ ddlmZmZ ddlZddlm	Z	 ddl
ZddlmZmZmZ ddlmZmZ ddlZdd	lmZmZ dd
lmZ ddlmZmZmZmZmZ ddlm Z m!Z! ddl"m#  m$  m%Z& ddl'm(Z( dddZ)dddZ*G dd dZ+G dd de(ej,Z-dS )a  
Read SAS7BDAT files

Based on code written by Jared Hobbs:
  https://bitbucket.org/jaredhobbs/sas7bdat

See also:
  https://github.com/BioStatMatt/sas7bdat

Partial documentation of the file format:
  https://cran.r-project.org/package=sas7bdat/vignettes/sas7bdat.pdf

Reference for binary data compression:
  http://collaboration.cmc.ec.gc.ca/science/rpn/biblio/ddj/Website/articles/CUJ/1992/9210/ross/ross.htm
    )annotations)abc)datetime	timedeltaN)cast)CompressionOptionsFilePath
ReadBuffer)EmptyDataErrorOutOfBoundsDatetime)	DataFrameisna)
get_handle)read_double_with_byteswapread_float_with_byteswapread_uint16_with_byteswapread_uint32_with_byteswapread_uint64_with_byteswap)Parserget_subheader_index)
ReaderBasesas_datetimefloatunitstrc                 C  sR   t | rtjS |dkrtdddt| d S |dkr%tdddt| d S td)Ns     )secondsd)dayszunit must be 'd' or 's')r   pdZNaTr   r   
ValueError)r   r    r#   J/app/.heroku/python/lib/python3.10/site-packages/pandas/io/sas/sas7bdat.py_parse_datetime<   s   r%   sas_datetimes	pd.Seriesreturnc                 C  sD   z	t j| |ddW S  ty!   | jt|d}tt j|}| Y S w )a  
    Convert to Timestamp if possible, otherwise to datetime.datetime.
    SAS float64 lacks precision for more than ms resolution so the fit
    to datetime.datetime is ok.

    Parameters
    ----------
    sas_datetimes : {Series, Sequence[float]}
       Dates or datetimes in SAS
    unit : {str}
       "d" if the floats represent dates, "s" for datetimes

    Returns
    -------
    Series
       Series of datetime64 dtype or datetime.datetime.
    z
1960-01-01)r   originr   )r!   to_datetimer   applyr%   r   Series)r&   r   Zs_seriesr#   r#   r$   _convert_datetimesJ   s   r.   c                   @  sH   e Zd ZU ded< ded< ded< ded< ded< ded	< dddZdS )_Columnintcol_idstr | bytesnamelabelformatbytesctypelengthr(   Nonec                 C  s(   || _ || _|| _|| _|| _|| _d S N)r1   r3   r4   r5   r7   r8   )selfr1   r3   r4   r5   r7   r8   r#   r#   r$   __init__l   s   

z_Column.__init__N)r1   r0   r3   r2   r4   r2   r5   r2   r7   r6   r8   r0   r(   r9   )__name__
__module____qualname____annotations__r<   r#   r#   r#   r$   r/   d   s   
 r/   c                   @  sJ  e Zd ZU dZded< ded< 								dYdZddZd[ddZd[ddZd[dd Zd\d!d"Z	d\d#d$Z
d]d&d'Zd^d*d+Zd_d,d-Zd`d/d0Zdad2d3Zd\d4d5Zdbd6d7Zd\d8d9Zd\d:d;Zdcd<d=Zdcd>d?Zdcd@dAZdcdBdCZdcdDdEZdcdFdGZdcdHdIZdcdJdKZdddedMdNZdOdP Zd]dQdRZdSdT ZdfdWdXZ dS )gSAS7BDATReadera  
    Read SAS files in SAS7BDAT format.

    Parameters
    ----------
    path_or_buf : path name or buffer
        Name of SAS file or file-like object pointing to SAS file
        contents.
    index : column identifier, defaults to None
        Column to use as index.
    convert_dates : bool, defaults to True
        Attempt to convert dates to Pandas datetime values.  Note that
        some rarely used SAS date formats may be unsupported.
    blank_missing : bool, defaults to True
        Convert empty strings to missing values (SAS uses blanks to
        indicate missing character variables).
    chunksize : int, defaults to None
        Return SAS7BDATReader object for iterations, returns chunks
        with given number of lines.
    encoding : str, 'infer', defaults to None
        String encoding acc. to Python standard encodings,
        encoding='infer' tries to detect the encoding from the file header,
        encoding=None will leave the data in binary format.
    convert_text : bool, defaults to True
        If False, text variables are left as raw bytes.
    convert_header_text : bool, defaults to True
        If False, header text, including column names, are left as raw
        bytes.
    r0   _int_lengthzbytes | None_cached_pageNTinferpath_or_bufFilePath | ReadBuffer[bytes]convert_datesboolblank_missing	chunksize
int | Noneencoding
str | Noneconvert_textconvert_header_textcompressionr   r(   r9   c
           
   	   C  s   || _ || _|| _|| _|| _|| _|| _d| _d| _g | _	g | _
g | _g | _g | _d | _g | _g | _g | _d| _d| _d| _t|dd|	d| _| jj| _| j| j| j| j| j| j| j| jd g	| _ z| !  | "  W d S  t#yx   | $   w )Nzlatin-1    r   rbF)Zis_textrP   )%indexrG   rI   rJ   rL   rN   rO   default_encodingrP   column_names_rawcolumn_namescolumn_formatscolumns%_current_page_data_subheader_pointersrC   _column_data_lengths_column_data_offsets_column_types_current_row_in_file_indexZ_current_row_on_page_indexr   handleshandle_path_or_buf_process_rowsize_subheader_process_columnsize_subheader_process_subheader_counts_process_columntext_subheader_process_columnname_subheader#_process_columnattributes_subheader_process_format_subheader_process_columnlist_subheader_subheader_processors_get_properties_parse_metadata	Exceptionclose)
r;   rE   rS   rG   rI   rJ   rL   rN   rO   rP   r#   r#   r$   r<      sT   
zSAS7BDATReader.__init__
np.ndarrayc                 C     t j| jt jdS )z5Return a numpy int64 array of the column data lengthsdtype)npasarrayrZ   int64r;   r#   r#   r$   column_data_lengths      z"SAS7BDATReader.column_data_lengthsc                 C  ro   )z0Return a numpy int64 array of the column offsetsrp   )rr   rs   r[   rt   ru   r#   r#   r$   column_data_offsets   rw   z"SAS7BDATReader.column_data_offsetsc                 C  s   t j| jt ddS )zj
        Returns a numpy character array of the column types:
           s (string) or d (double)
        ZS1rp   )rr   rs   r\   rq   ru   r#   r#   r$   column_types   s   zSAS7BDATReader.column_typesc                 C  s   | j   d S r:   )r^   rm   ru   r#   r#   r$   rm      s   zSAS7BDATReader.closec                 C  s  | j d | j d| _| jdttj tjkrtd| tj	tj
}|tjkr:d| _d| _tj| _tj| _nd| _tj| _tj| _d| _| tjtj}|tjkrYtj}nd}| tjtj}|dkrqd	| _tjd
k| _n	d| _tjdk| _| tjtjd }|tj v rtj | | _!| j"dkr| j!| _"nd| d| _!t#ddd}| $tj%| tj&}|t'j(|dd | _)| $tj*| tj+}|t'j(|dd | _,| -tj.| tj/| _0| j | j0d }|  j|7  _t| j| j0krtd| -tj1| tj2| _3d S )Nr   i   z'magic number mismatch (not a SAS file?)T   F      <big>littlerD   zunknown (code=)r   r   r   r*   z*The SAS7BDAT file appears to be truncated.)4r`   seekreadrC   lenconstmagicr"   _read_bytesZalign_1_offsetZalign_1_lengthZu64_byte_checker_valueU64rB   Zpage_bit_offset_x64_page_bit_offsetZsubheader_pointer_length_x64_subheader_pointer_lengthZpage_bit_offset_x86Zsubheader_pointer_length_x86Zalign_2_offsetZalign_2_lengthZalign_1_checker_valueZalign_2_valueZendianness_offsetZendianness_length
byte_ordersys	byteorderneed_byteswapZencoding_offsetZencoding_lengthZencoding_namesZinferred_encodingrL   r   _read_floatZdate_created_offsetZdate_created_lengthr!   Zto_timedeltaZdate_createdZdate_modified_offsetZdate_modified_lengthZdate_modified
_read_uintZheader_size_offsetZheader_size_lengthheader_lengthZpage_size_offsetZpage_size_length_page_length)r;   bufZalign1epochxr#   r#   r$   rj      sd   





zSAS7BDATReader._get_propertiesr   c                 C  s(   | j | jpdd}|jr|   t|S )Nr   )nrows)r   rJ   emptyrm   StopIteration)r;   dar#   r#   r$   __next__8  s
   zSAS7BDATReader.__next__offsetwidthc                 C  sN   | j d usJ |dkrt| j || jS |dkrt| j || jS |   td)Nr{   rz   zinvalid float width)rC   r   r   r   rm   r"   r;   r   r   r#   r#   r$   r   @  s   

zSAS7BDATReader._read_floatc                 C  s~   | j d usJ |dkr| |dd S |dkrt| j || jS |dkr+t| j || jS |dkr7t| j || jS |   td)Nr   r      r{   rz   zinvalid int width)rC   r   r   r   r   r   rm   r"   r   r#   r#   r$   r   O  s"   


zSAS7BDATReader._read_uintr8   c                 C  sB   | j d usJ || t| j kr|   td| j |||  S )NzThe cached page is too small.)rC   r   rm   r"   r;   r   r8   r#   r#   r$   r   c  s
   zSAS7BDATReader._read_bytesr2   c                 C  s   |  | ||dS )N     )_convert_header_textr   rstripr   r#   r#   r$   _read_and_convert_header_textj  s   z,SAS7BDATReader._read_and_convert_header_textc                 C  sV   d}|s)| j | j| _t| jdkrd S t| j| jkr!td|  }|rd S d S )NFr   z2Failed to read a meta data page from the SAS file.)r`   r   r   rC   r   r"   _process_page_meta)r;   doner#   r#   r$   rk   o  s   zSAS7BDATReader._parse_metadatac                 C  sZ   |    tjtjtjg }| j|v r|   | jtjk}| jtjk}t|p+|p+| j	g kS r:   )
_read_page_headerr   page_meta_typesZpage_amd_typepage_mix_type_current_page_type_process_page_metadatapage_data_typerH   rY   )r;   ptZis_data_pageZis_mix_pager#   r#   r$   r   y  s   
z!SAS7BDATReader._process_page_metac                 C  s^   | j }tj| }| |tjtj@ | _tj| }| |tj| _	tj
| }| |tj| _d S r:   )r   r   Zpage_type_offsetr   Zpage_type_lengthZpage_type_mask2r   Zblock_count_offsetZblock_count_lengthZ_current_page_block_countZsubheader_count_offsetZsubheader_count_length_current_page_subheaders_count)r;   
bit_offsetZtxr#   r#   r$   r     s   



z SAS7BDATReader._read_page_headerc                 C  s  | j }t| jD ]}tj| }|| j|  }| || j}|| j7 }| || j}|| j7 }| |d}|d7 }| |d}|dksG|tjkrHq| 	|| j}	t
|	}
| j|
 }|d u r|tjdfv }|tjk}| jrx|rx|rx| j||f q|   td|	 ||| qd S )Nr   r   zUnknown subheader signature )r   ranger   r   Zsubheader_pointers_offsetr   r   rB   Ztruncated_subheader_idr   r   ri   Zcompressed_subheader_idZcompressed_subheader_typerP   rY   appendrm   r"   )r;   r   ir   Ztotal_offsetZsubheader_offsetZsubheader_lengthZsubheader_compressionZsubheader_typeZsubheader_signatureZsubheader_indexZsubheader_processorf1f2r#   r#   r$   r     s@   




z%SAS7BDATReader._process_page_metadatac                 C  s   | j }|}|}| jr|d7 }|d7 }n|d7 }|d7 }| |tj|  || _| |tj|  || _| |tj|  || _	| |tj
|  || _tj| }| || || _| |d| _| |d| _d S )Ni  i  ib  iz  r   )rB   r   r   r   Zrow_length_offset_multiplierZ
row_lengthZrow_count_offset_multiplier	row_countZcol_count_p1_multipliercol_count_p1Zcol_count_p2_multipliercol_count_p2Z'row_count_on_mix_page_offset_multiplierZ_mix_page_row_count_lcs_lcp)r;   r   r8   int_lenZ
lcs_offsetZ
lcp_offsetZmxr#   r#   r$   ra     s4   

z)SAS7BDATReader._process_rowsize_subheaderc                 C  sX   | j }||7 }| ||| _| j| j | jkr*td| j d| j d| j d d S d S )Nz Warning: column count mismatch (z + z != z)
)rB   r   column_countr   r   print)r;   r   r8   r   r#   r#   r$   rb     s   
z,SAS7BDATReader._process_columnsize_subheaderc                 C     d S r:   r#   r   r#   r#   r$   rc     s   z(SAS7BDATReader._process_subheader_countsc           	      C  s  || j 7 }| |tj}| ||}|d| d}| j| t| jdkrd}tj	D ]}||v r5|}q-|| _
|| j 8 }|d }| jrI|d7 }| || j}|d}|dkrwd| _|d }| jrg|d7 }| || j}|d| j | _nB|tjkr|d	 }| jr|d7 }| || j}|d| j | _n"| jdkrd| _|d }| jr|d7 }| || j}|d| j | _t| d
r| | j| _d S d S d S )Nr   r   r   rQ      r{           (   creator_proc)rB   r   r   Ztext_block_size_lengthr   r   rU   r   r   Zcompression_literalsrP   r   r   r   r   Zrle_compressionhasattrr   )	r;   r   r8   Ztext_block_sizer   Z	cname_rawZcompression_literalclZoffset1r#   r#   r$   rd     sT   






"z,SAS7BDATReader._process_columntext_subheaderc                 C  s   | j }||7 }|d|  d d }t|D ]Q}|tj|d   tj }|tj|d   tj }|tj|d   tj }| |tj}	| |tj	}
| |tj
}| j|	 }||
|
|  }| j| | qd S )Nr      rz   r   )rB   r   r   Zcolumn_name_pointer_lengthZ!column_name_text_subheader_offsetZcolumn_name_offset_offsetZcolumn_name_length_offsetr   Z!column_name_text_subheader_lengthZcolumn_name_offset_lengthZcolumn_name_length_lengthrU   rV   r   r   )r;   r   r8   r   Zcolumn_name_pointers_countr   Ztext_subheaderZcol_name_offsetZcol_name_lengthidx
col_offsetZcol_lenZname_rawcnamer#   r#   r$   re     sB   
z,SAS7BDATReader._process_columnname_subheaderc           
      C  s   | j }|d|  d |d  }t|D ]Y}|| tj ||d   }|d|  tj ||d   }|d|  tj ||d   }| ||}	| j|	 | |tj	}	| j
|	 | |tj}	| j|	dkridnd qd S )Nr   r   rz   r      d   s)rB   r   r   Zcolumn_data_offset_offsetZcolumn_data_length_offsetZcolumn_type_offsetr   r[   r   Zcolumn_data_length_lengthrZ   Zcolumn_type_lengthr\   )
r;   r   r8   r   Zcolumn_attributes_vectors_countr   Zcol_data_offsetZcol_data_lenZ	col_typesr   r#   r#   r$   rf   8  s,   
z2SAS7BDATReader._process_columnattributes_subheaderc                 C  r   r:   r#   r   r#   r#   r$   rh   R  s   z,SAS7BDATReader._process_columnlist_subheaderc                 C  sx  | j }|tj d|  }|tj d|  }|tj d|  }|tj d|  }|tj d|  }|tj d|  }	| |tj	}
t
|
t| jd }| |tj}| |tj}| |tj}t
|t| jd }| |tj}| |	tj}| j| }| ||||  }| j| }| ||||  }t| j}t|| j| ||| j| | j| }| j| | j| d S )N   r   )rB   r   Z)column_format_text_subheader_index_offsetZcolumn_format_offset_offsetZcolumn_format_length_offsetZ(column_label_text_subheader_index_offsetZcolumn_label_offset_offsetZcolumn_label_length_offsetr   Z)column_format_text_subheader_index_lengthminr   rU   Zcolumn_format_offset_lengthZcolumn_format_length_lengthZ(column_label_text_subheader_index_lengthZcolumn_label_offset_lengthZcolumn_label_length_lengthr   rX   r/   rV   r\   rZ   rW   r   )r;   r   r8   r   Ztext_subheader_formatZcol_format_offsetZcol_format_lenZtext_subheader_labelZcol_label_offsetZcol_label_lenr   Z
format_idxZformat_startZ
format_lenZ	label_idxZlabel_startZ	label_lenZlabel_namesZcolumn_labelZformat_namesZcolumn_formatZcurrent_column_numbercolr#   r#   r$   rg   V  sX   


	z(SAS7BDATReader._process_format_subheaderr   c                 C  s   |d u r| j d ur| j }n|d u r| j}t| jdkr#|   td|dkr0| j| jkr0t S t|| j| j }| j	d}| j	d}t
j||ftd| _t
j|d| ft
jd| _d| _t| }|| |  }| jd urw|| j}|S )Nr   zNo columns to parse from filer   r   rp   rz   )rJ   r   r   r\   rm   r
   r]   r   r   countrr   r   object_string_chunkZzerosZuint8_byte_chunk_current_row_in_chunk_indexr   r   _chunk_to_dataframerS   Z	set_index)r;   r   ndnsprsltr#   r#   r$   r     s*   

zSAS7BDATReader.readc                 C  s   g | _ | j| j| _t| jdkrdS t| j| jkr3|   dt| jdd| jdd}t||   | j	t
jv rA|   | j	t
jt
jt
jg vrQ|  S dS )Nr   Tz-failed to read complete page from file (read r   z of z bytes)F)rY   r`   r   r   rC   r   rm   r"   r   r   r   r   r   r   r   _read_next_page)r;   msgr#   r#   r$   r     s,   
zSAS7BDATReader._read_next_pagec                 C  sj  | j }| j}t|| |}i }d\}}t| jD ]}| j| }| j| dkrl| j|d d f j| jd d}	t	j
|	tj|d||< | jrg| j| tjv rVt|| d||< n| j| tjv rgt|| d||< |d7 }q| j| dkrt	j
| j|d d f |d	||< | jr| jd ur| || j||< |d7 }q|   td
t| j|  t|| j|dd}
|
S )N)r   r   r   r   rp   )rq   rS   r   r   r   )rS   zunknown column type F)rX   rS   copy)r   r]   r   r   rV   r\   r   viewr   r!   r-   rr   Zfloat64rG   rW   r   Zsas_date_formatsr.   Zsas_datetime_formatsr   rN   rL   _decode_stringr   rm   r"   reprr   )r;   nmZixr   jsZjbjr3   Zcol_arrZdfr#   r#   r$   r     s2   
 
 
z"SAS7BDATReader._chunk_to_dataframec                 C  s   | | jp| jS r:   )decoderL   rT   r;   br#   r#   r$   r     s   zSAS7BDATReader._decode_stringr   r6   c                 C  s   | j r| |S |S r:   )rO   r   r   r#   r#   r$   r     s   
z#SAS7BDATReader._convert_header_text)NTTNNTTrD   )rE   rF   rG   rH   rI   rH   rJ   rK   rL   rM   rN   rH   rO   rH   rP   r   r(   r9   )r(   rn   )r(   r9   )r(   r   )r   r0   r   r0   )r   r0   r   r0   r(   r0   )r   r0   r8   r0   )r   r0   r8   r0   r(   r2   )r(   rH   )r   r0   r8   r0   r(   r9   r:   )r   rK   r(   r   )r   r6   r(   r2   )!r=   r>   r?   __doc__r@   r<   rv   rx   ry   rm   rj   r   r   r   r   r   rk   r   r   r   ra   rb   rc   rd   re   rf   rh   rg   r   r   r   r   r   r#   r#   r#   r$   rA      sP   
 
?




F









+



-
!

7
rA   )r   r   r   r   )r&   r'   r   r   r(   r'   ).r   
__future__r   collectionsr   r   r   r   typingr   numpyrr   Zpandas._typingr   r   r	   Zpandas.errorsr
   r   Zpandasr!   r   r   Zpandas.io.commonr   Zpandas.io.sas._byteswapr   r   r   r   r   Zpandas.io.sas._sasr   r   Zpandas.io.sas.sas_constantsiosasZsas_constantsr   Zpandas.io.sas.sasreaderr   r%   r.   r/   IteratorrA   r#   r#   r#   r$   <module>   s(    

