o
0i> @ s d Z ddlZddlmZmZmZmZmZmZm Z m
Z
m Z m Z m
Z
mZmZm Z ddlmZ ddlmZ g dZG dd dZG d d
d
eZed
e G d d d eZed
e G dd deZede G dd dZG dd dZze W n ey e ZY nw G dd dZ dS )z+
csv.py - read/write/investigate CSV files
N)Error __version__writerreaderregister_dialectunregister_dialect get_dialect
list_dialectsfield_size_limit
QUOTE_MINIMAL QUOTE_ALLQUOTE_NONNUMERIC
QUOTE_NONE__doc__)Dialect)StringIO)r r r
r r r r excel excel_tabr
r r r r r Snifferr r
DictReader
DictWriter unix_dialectc @ sD e Zd ZdZdZdZdZdZdZdZ dZ
dZ dZ dd Z
dd ZdS ) r zDescribe a CSV dialect.
This must be subclassed (see csv.excel). Valid attributes are:
delimiter, quotechar, escapechar, doublequote, skipinitialspace,
lineterminator, quoting.
FNc C s | j tkrd| _| d S )NT) __class__r _valid _validateself r */opt/alt/python310/lib64/python3.10/csv.py__init__* s
zDialect.__init__c
C s4 zt | W d S ty } ztt|d }~ww N)_Dialect TypeErrorr str)r er r r r / s zDialect._validate)__name__
__module__ __qualname__r _namer delimiter quotecharZ
escapechar doublequoteskipinitialspacelineterminatorquotingr r r r r r r s r c @ ( e Zd ZdZdZdZdZdZdZe Z
dS )r z;Describe the usual properties of Excel-generated CSV files.,"TF
N) r& r' r( r r* r+ r, r- r. r r/ r r r r r 6 r c @ s e Zd ZdZdZdS )r zEDescribe the usual properties of Excel-generated TAB-delimited files. N)r& r' r( r r* r r r r r @ s r z excel-tabc @ r0 )r z:Describe the usual properties of Unix-generated CSV files.r1 r2 TF
N) r& r' r( r r* r+ r, r- r. r r/ r r r r r E r4 r Zunixc @ sD e Zd Z d ddZdd Zedd Zejd d Zd
d ZdS )
r Nr c O s< || _ || _|| _t||g|R i || _|| _d| _d S Nr ) _fieldnamesrestkeyrestvalr dialectline_num)r f
fieldnamesr9 r: r; argskwdsr r r r Q s
zDictReader.__init__c C s | S r! r r r r r __iter__Z s zDictReader.__iter__c C s> | j d u rzt| j| _ W n ty Y nw | jj| _| j S r! )r8 nextr
StopIterationr< r r r r r> ] s
zDictReader.fieldnamesc C s
|| _ d S r! )r8 )r valuer r r r> g s
c C s | j dkr| j t| j}| jj | _ |g kr t| j}|g kstt| j|}t| j}t|}||k r?||d || j< |S ||krR| j|d D ]}| j||< qJ|S r7 ) r< r> rB r dictziplenr9 r: )r rowdZlfZlrkeyr r r __next__k s"
zDictReader.__next__)NNNr ) r& r' r( r rA propertyr> setterrK r r r r r P s
r c @ s: e Zd Z dddZdd Zdd Zd
d Zd d
ZdS )r r raiser c O sH || _ || _| dvrtd| || _t||g|R i || _d S )N)rN ignorez-extrasaction (%s) must be 'raise' or 'ignore')r> r: lower
ValueError extrasactionr )r r= r> r: rR r; r? r@ r r r r s zDictWriter.__init__c C s t t| j| j}| |S r! )rE rF r> writerow)r headerr r r writeheader s
zDictWriter.writeheaderc sN j dkr j }|r tdddd |D fddjD S )NrN z(dict contains fields not in fieldnames: z, c S g | ]}t |qS r )repr).0xr r r
z,DictWriter._dict_to_list..c 3 s | ]
} |jV qd S r! )getr: )rX rJ rowdictr r r s z+DictWriter._dict_to_list..)rR keysr> rQ join)r r^ Z wrong_fieldsr r] r
_dict_to_list s
zDictWriter._dict_to_listc C s | j | |S r! )r rS rb )r r^ r r r rS s zDictWriter.writerowc C s | j t| j|S r! )r writerowsmaprb )r Zrowdictsr r r rc s zDictWriter.writerowsN)r rN r )r& r' r( r rU rb rS rc r r r r r s
r c @ s: e Zd ZdZdd Zd
ddZdd Zd d
Zd d ZdS )r ze
"Sniffs" the format of a CSV file (i.e. delimiter, quotechar)
Returns a Dialect object.
c C s g d| _ d S )N)r1 r5 ; :) preferredr r r r r s zSniffer.__init__Nc C sd | ||\}}}}|s| ||\}}|stdG dd dt}||_||_|p+d|_||_|S )zI
Returns a dialect (or None) corresponding to the sample
z Could not determine delimiterc @ s e Zd ZdZdZeZdS )z Sniffer.sniff..dialectZsniffedr3 N)r& r' r( r) r. r r/ r r r r r; s r; r2 )_guess_quote_and_delimiter_guess_delimiterr r r, r* r+ r- )r sample
delimitersr+ r, r* r- r; r r r sniff s
z
Sniffer.sniffc C s g }dD ]}t |t jt jB }||}|r nq|s dS i }i }d}|j} |D ]\}
| d d } |
| } | rA|| dd || < z | d d } |
| } W n tyV Y q)w | rk|du sa| |v rk|| dd || < z| d d } W n ty| Y q)w |
| r|d7 }q)t||jd }
|rt||jd }|| |k}|d
krd }nd }d}t d t ||
d
t j}|
|rd}nd}|
|||fS )a
Looks for text enclosed between two identical quotes
(the probable quotechar) which are preceded and followed
by the same character (the probable delimiter).
For example:
,'some text',
The quote with the most wins, same with the delimiter.
If there is no quotechar the delimiter can't be determined
this way.
)zI(?P[^\w\n"\'])(?P ?)(?P["\']).*?(?P=quote)(?P=delim)zG(?:^|\n)(?P["\']).*?(?P=quote)(?P[^\w\n"\'])(?P ?)zG(?P[^\w\n"\'])(?P ?)(?P["\']).*?(?P=quote)(?:$|\n)z-(?:^|\n)(?P["\']).*?(?P=quote)(?:$|\n))r FNr r quote delimNZspacerJ r6 r z]((%(delim)s)|^)\W*%(quote)s[^%(delim)s\n]*%(quote)s[^%(delim)s\n]*%(quote)s\W*((%(delim)s)|$))rp rn TF) recompileDOTALL MULTILINEfindall
groupindexr\ KeyErrormaxescapesearch)r datarl ZmatchesZrestrZregexpZquotesdelimsspacesrw mnrJ r+ rp r- Z dq_regexpr, r r r ri sj
z"Sniffer._guess_quote_and_delimiterc C s t td|d}dd tdD }tdt|}d}i }i }i }d|} }
| t|k r-|d7 }|| |
D ] } |D ]} || i }
| | }|
|dd |
|< |
|| < q.
r ro c S s | d S )Nro r )rY r r r G s z*Sniffer._guess_delimiter..rq c s s | ]}|d V qdS )ro Nr )rX itemr r r r_ L s z+Sniffer._guess_delimiter..g ?g?g{Gz?z%c )r r c S s g | ]\}}||fqS r r )rX kvr r r rZ v s )listfiltersplitrangeminrG r\ countr` itemsry removesumfloatrh sort)r r| rl asciiZ chunkLengthZ iterationZ
charFrequencyZmodesr} startendlinecharZ
metaFrequencyZfreqr ZmodeListtotalZ consistencyZ thresholdr r rp r- rI r r r rj s
3
zSniffer._guess_delimiterc
C sj t t|| |}t|}t|}i }t|D ]}d ||< qd}|D ]L}|dkr+ nE|d7 }t||kr6q#t| D ]2} t}
z|
|| W n t t
fyY t|| }
Y nw |
|| krn|| d u rk|
|| < q<|| = q s* @
2