The so-called CSV (Comma Separated Values) format is the most common import andexport format for spreadsheets and databases. CSV format was used for manyyears prior to attempts to describe the format in a standardized way inRFC 4180. The lack of a well-defined standard means that subtle differencesoften exist in the data produced and consumed by different applications. Thesedifferences can make it annoying to process CSV files from multiple sources.Still, while the delimiters and quoting characters vary, the overall format issimilar enough that it is possible to write a single module which canefficiently manipulate such data, hiding the details of reading and writing thedata from the programmer.
Thecsv module implements classes to read and write tabular data in CSVformat. It allows programmers to say, “write this data in the format preferredby Excel,” or “read data from this file which was generated by Excel,” withoutknowing the precise details of the CSV format used by Excel. Programmers canalso describe the CSV formats understood by other applications or define theirown special-purpose CSV formats.
Thecsv module’sreader andwriter objects read andwrite sequences. Programmers can also read and write data in dictionary formusing theDictReader andDictWriter classes.
See also
Thecsv module defines the following functions:
Return a reader object which will iterate over lines in the givencsvfile.csvfile can be any object which supports theiterator protocol and returns astring each time its__next__() method is called —file objects and list objects are both suitable. Ifcsvfile is a file object,it should be opened withnewline=''.[1] An optionaldialect parameter can be given which is used to define a set of parametersspecific to a particular CSV dialect. It may be an instance of a subclass oftheDialect class or one of the strings returned by thelist_dialects() function. The other optionalfmtparams keyword argumentscan be given to override individual formatting parameters in the currentdialect. For full details about the dialect and formatting parameters, seesectionDialects and Formatting Parameters.
Each row read from the csv file is returned as a list of strings. Noautomatic data type conversion is performed unless theQUOTE_NONNUMERIC formatoption is specified (in which case unquoted fields are transformed into floats).
A short usage example:
>>>importcsv>>>withopen('eggs.csv',newline='')ascsvfile:...spamreader=csv.reader(csvfile,delimiter=' ',quotechar='|')...forrowinspamreader:...print(', '.join(row))Spam, Spam, Spam, Spam, Spam, Baked BeansSpam, Lovely Spam, Wonderful Spam
Return a writer object responsible for converting the user’s data into delimitedstrings on the given file-like object.csvfile can be any object with awrite() method. Ifcsvfile is a file object, it should be opened withnewline=''[1]. An optionaldialectparameter can be given which is used to define a set of parameters specific to aparticular CSV dialect. It may be an instance of a subclass of theDialect class or one of the strings returned by thelist_dialects() function. The other optionalfmtparams keyword argumentscan be given to override individual formatting parameters in the currentdialect. For full details about the dialect and formatting parameters, seesectionDialects and Formatting Parameters. To make itas easy as possible to interface with modules which implement the DB API, thevalueNone is written as the empty string. While this isn’t areversible transformation, it makes it easier to dump SQL NULL data values toCSV files without preprocessing the data returned from acursor.fetch* call.All other non-string data are stringified withstr() before being written.
A short usage example:
importcsvwithopen('eggs.csv','w',newline='')ascsvfile:spamwriter=csv.writer(csvfile,delimiter=' ',quotechar='|',quoting=csv.QUOTE_MINIMAL)spamwriter.writerow(['Spam']*5+['Baked Beans'])spamwriter.writerow(['Spam','Lovely Spam','Wonderful Spam'])
Associatedialect withname.name must be a string. Thedialect can be specified either by passing a sub-class ofDialect, orbyfmtparams keyword arguments, or both, with keyword arguments overridingparameters of the dialect. For full details about the dialect and formattingparameters, see sectionDialects and Formatting Parameters.
Delete the dialect associated withname from the dialect registry. AnError is raised ifname is not a registered dialect name.
Return the dialect associated withname. AnError is raised ifname is not a registered dialect name. This function returns an immutableDialect.
Return the names of all registered dialects.
Returns the current maximum field size allowed by the parser. Ifnew_limit isgiven, this becomes the new limit.
Thecsv module defines the following classes:
Create an object which operates like a regular reader but maps theinformation read into a dict whose keys are given by the optionalfieldnames parameter. Thefieldnames parameter is asequence whose elements are associated with the fields of theinput data in order. These elements become the keys of the resultingdictionary. If thefieldnames parameter is omitted, the values in thefirst row of thecsvfile will be used as the fieldnames. If the row readhas more fields than the fieldnames sequence, the remaining data is added asa sequence keyed by the value ofrestkey. If the row read has fewerfields than the fieldnames sequence, the remaining keys take the value ofthe optionalrestval parameter. Any other optional or keyword argumentsare passed to the underlyingreader instance.
Create an object which operates like a regular writer but maps dictionariesonto output rows. Thefieldnames parameter is asequence of keys that identify the order in which values in thedictionary passed to thewriterow() method are written to thecsvfile. The optionalrestval parameter specifies the value to bewritten if the dictionary is missing a key infieldnames. If thedictionary passed to thewriterow() method contains a key not found infieldnames, the optionalextrasaction parameter indicates what action totake. If it is set to'raise' aValueError is raised. If it isset to'ignore', extra values in the dictionary are ignored. Any otheroptional or keyword arguments are passed to the underlyingwriterinstance.
Note that unlike theDictReader class, thefieldnames parameterof theDictWriter is not optional. Since Python’sdictobjects are not ordered, there is not enough information available to deducethe order in which the row should be written to thecsvfile.
TheDialect class is a container class relied on primarily for itsattributes, which are used to define the parameters for a specificreader orwriter instance.
Theexcel class defines the usual properties of an Excel-generated CSVfile. It is registered with the dialect name'excel'.
Theexcel_tab class defines the usual properties of an Excel-generatedTAB-delimited file. It is registered with the dialect name'excel-tab'.
Theunix_dialect class defines the usual properties of a CSV filegenerated on UNIX systems, i.e. using'\n' as line terminator and quotingall fields. It is registered with the dialect name'unix'.
New in version 3.2.
TheSniffer class is used to deduce the format of a CSV file.
TheSniffer class provides two methods:
An example forSniffer use:
withopen('example.csv')ascsvfile:dialect=csv.Sniffer().sniff(csvfile.read(1024))csvfile.seek(0)reader=csv.reader(csvfile,dialect)# ... process CSV file contents here ...
Thecsv module defines the following constants:
Instructswriter objects to only quote those fields which containspecial characters such asdelimiter,quotechar or any of the characters inlineterminator.
Instructswriter objects to quote all non-numeric fields.
Instructs the reader to convert all non-quoted fields to typefloat.
Instructswriter objects to never quote fields. When the currentdelimiter occurs in output data it is preceded by the currentescapecharcharacter. Ifescapechar is not set, the writer will raiseError ifany characters that require escaping are encountered.
Instructsreader to perform no special processing of quote characters.
Thecsv module defines the following exception:
Raised by any of the functions when an error is detected.
To make it easier to specify the format of input and output records, specificformatting parameters are grouped together into dialects. A dialect is asubclass of theDialect class having a set of specific methods and asinglevalidate() method. When creatingreader orwriter objects, the programmer can specify a string or a subclass oftheDialect class as the dialect parameter. In addition to, or insteadof, thedialect parameter, the programmer can also specify individualformatting parameters, which have the same names as the attributes defined belowfor theDialect class.
Dialects support the following attributes:
A one-character string used to separate fields. It defaults to','.
Controls how instances ofquotechar appearing inside a field should bethemselves be quoted. WhenTrue, the character is doubled. WhenFalse, theescapechar is used as a prefix to thequotechar. Itdefaults toTrue.
On output, ifdoublequote isFalse and noescapechar is set,Error is raised if aquotechar is found in a field.
A one-character string used by the writer to escape thedelimiter ifquotingis set toQUOTE_NONE and thequotechar ifdoublequote isFalse. On reading, theescapechar removes any special meaning fromthe following character. It defaults toNone, which disables escaping.
The string used to terminate lines produced by thewriter. It defaultsto'\r\n'.
Note
Thereader is hard-coded to recognise either'\r' or'\n' asend-of-line, and ignoreslineterminator. This behavior may change in thefuture.
A one-character string used to quote fields containing special characters, suchas thedelimiter orquotechar, or which contain new-line characters. Itdefaults to'"'.
Controls when quotes should be generated by the writer and recognised by thereader. It can take on any of theQUOTE_* constants (see sectionModule Contents) and defaults toQUOTE_MINIMAL.
Reader objects (DictReader instances and objects returned by thereader() function) have the following public methods:
Return the next row of the reader’s iterable object as a list, parsed accordingto the current dialect. Usually you should call this asnext(reader).
Reader objects have the following public attributes:
A read-only description of the dialect in use by the parser.
The number of lines read from the source iterator. This is not the same as thenumber of records returned, as records can span multiple lines.
DictReader objects have the following public attribute:
If not passed as a parameter when creating the object, this attribute isinitialized upon first access or when the first record is read from thefile.
Writer objects (DictWriter instances and objects returned bythewriter() function) have the following public methods. Arow must bea sequence of strings or numbers forWriter objects and a dictionarymapping fieldnames to strings or numbers (by passing them throughstr()first) forDictWriter objects. Note that complex numbers are writtenout surrounded by parens. This may cause some problems for other programs whichread CSV files (assuming they support complex numbers at all).
Write therow parameter to the writer’s file object, formatted according tothe current dialect.
Write all therows parameters (a list ofrow objects as described above) tothe writer’s file object, formatted according to the current dialect.
Writer objects have the following public attribute:
A read-only description of the dialect in use by the writer.
DictWriter objects have the following public method:
Write a row with the field names (as specified in the constructor).
New in version 3.2.
The simplest example of reading a CSV file:
importcsvwithopen('some.csv',newline='')asf:reader=csv.reader(f)forrowinreader:print(row)
Reading a file with an alternate format:
importcsvwithopen('passwd',newline='')asf:reader=csv.reader(f,delimiter=':',quoting=csv.QUOTE_NONE)forrowinreader:print(row)
The corresponding simplest possible writing example is:
importcsvwithopen('some.csv','w',newline='')asf:writer=csv.writer(f)writer.writerows(someiterable)
Sinceopen() is used to open a CSV file for reading, the filewill by default be decoded into unicode using the system defaultencoding (seelocale.getpreferredencoding()). To decode a fileusing a different encoding, use theencoding argument of open:
importcsvwithopen('some.csv',newline='',encoding='utf-8')asf:reader=csv.reader(f)forrowinreader:print(row)
The same applies to writing in something other than the system defaultencoding: specify the encoding argument when opening the output file.
Registering a new dialect:
importcsvcsv.register_dialect('unixpwd',delimiter=':',quoting=csv.QUOTE_NONE)withopen('passwd',newline='')asf:reader=csv.reader(f,'unixpwd')
A slightly more advanced use of the reader — catching and reporting errors:
importcsv,sysfilename='some.csv'withopen(filename,newline='')asf:reader=csv.reader(f)try:forrowinreader:print(row)exceptcsv.Errorase:sys.exit('file {}, line {}: {}'.format(filename,reader.line_num,e))
And while the module doesn’t directly support parsing strings, it can easily bedone:
importcsvforrowincsv.reader(['one,two,three']):print(row)
Footnotes
| [1] | (1,2) Ifnewline='' is not specified, newlines embedded inside quoted fieldswill not be interpreted correctly, and on platforms that use\r\n linendingson write an extra\r will be added. It should always be safe to specifynewline='', since the csv module does its own(universal) newline handling. |
14.2.configparser — Configuration file parser
Enter search terms or a module, class or function name.
