Python csv 模块,QUOTE_NONNUMERIC 实例源码

我们从Python开源项目中,提取了以下50个代码示例,用于说明如何使用csv.QUOTE_NONNUMERIC

项目:IM_Climate    作者:IMDProjects    | 项目源码 | 文件源码
def _writeToCSV(self):
        '''
        INFO
        ----
        Writes a 2-dimensional list to a CSV text file
        Comma-delimits values.  If there is no data, then there is no attempt to
        creat a file.

        RETURNS
        -------
        None

        '''
        if self._dataAsList:
            with open(self._filePathAndName,'w') as csvFile:
                writer = csv.writer(csvFile, lineterminator='\n', quoting=csv.QUOTE_NONNUMERIC )
                writer.writerows(self._dataAsList)
            csvFile.close()
项目:bobby_boy    作者:ohbobbyboy    | 项目源码 | 文件源码
def clear(filename):
    rows = []

    with open(filename, 'rb') as csvfile:
        reader = csv.reader(csvfile, delimiter=';')
        rows = [row for row in reader]

    with open(filename, 'wb') as csvfile:
        writer = csv.writer(csvfile, delimiter=';', quotechar='"', skipinitialspace=True,
                            quoting=csv.QUOTE_NONNUMERIC)  # QUOTE_MINIMAL vs QUOTE_NONNUMERIC

        for row in rows:
            if not row[0].startswith("#"):
                # ??????? ??????????? ??????? ? ????????? ?????????
                row[2] = row[2].strip(' ')
                writer.writerow(row)
项目:catalyst    作者:enigmampc    | 项目源码 | 文件源码
def get_csv_stats(stats, recorded_cols=None):
    """
    Create a CSV buffer from the stats DataFrame.

    Parameters
    ----------
    path: str
    stats: list[Object]
    recorded_cols: list[str]

    Returns
    -------

    """
    df, columns = prepare_stats(stats, recorded_cols=recorded_cols)

    return df.to_csv(
        None,
        columns=columns,
        # encoding='utf-8',
        quoting=csv.QUOTE_NONNUMERIC
    ).encode()
项目:SecuML    作者:ANSSI-FR    | 项目源码 | 文件源码
def getFeatures(self):
        csv_files = self.experiment.getFeaturesFilesFullpaths()
        features_names = []
        features = None
        for csv_file in csv_files:
            with open(csv_file, 'r') as f:
                header = f.readline().strip('\n').split(',')[1:]
                features_names += header
                current_features = list(list(rec) for rec in csv.reader(f,
                    quoting = csv.QUOTE_NONNUMERIC))
                if features is None:
                    features = [l[1:] for l in current_features]
                else:
                    features = [f1 + f2[1:] for f1, f2 in zip(features, current_features)]
        features = np.array(features)
        return features, features_names
项目:automated-reporting    作者:Waikato    | 项目源码 | 文件源码
def query_to_csv(cursor, cols, outname):
    """
    Turns the query into a CSV file for the GPA calculation.

    :param cursor: the database cursor
    :type cursor: DictCursor
    :param cols: the header names
    :type cols: list
    :param outname: the CSV output filename
    :type outname: str
    """

    logger.info("Generating CSV: {0}".format(outname))
    with open(outname, 'w') as outfile:
        writer = csv.writer(outfile, quoting=csv.QUOTE_NONNUMERIC)
        writer.writerow(cols.split(","))
        for row in cursor.fetchall():
            writer.writerow(row)
        outfile.flush()
    logger.info("Generated CSV ({0}) exists: ".format(outname, os.path.isfile(outname)))
项目:PyDataLondon29-EmbarrassinglyParallelDAWithAWSLambda    作者:SignalMedia    | 项目源码 | 文件源码
def writerow(self, row):
            def _check_as_is(x):
                return (self.quoting == csv.QUOTE_NONNUMERIC and
                        is_number(x)) or isinstance(x, str)

            row = [x if _check_as_is(x)
                   else pprint_thing(x).encode("utf-8") for x in row]

            self.writer.writerow([s for s in row])
            # Fetch UTF-8 output from the queue ...
            data = self.queue.getvalue()
            data = data.decode("utf-8")
            # ... and reencode it into the target encoding
            data = self.encoder.encode(data)
            # write to the target stream
            self.stream.write(data)
            # empty queue
            self.queue.truncate(0)
项目:PyDataLondon29-EmbarrassinglyParallelDAWithAWSLambda    作者:SignalMedia    | 项目源码 | 文件源码
def writerows(self, rows):
            def _check_as_is(x):
                return (self.quoting == csv.QUOTE_NONNUMERIC and
                        is_number(x)) or isinstance(x, str)

            for i, row in enumerate(rows):
                rows[i] = [x if _check_as_is(x)
                           else pprint_thing(x).encode("utf-8") for x in row]

            self.writer.writerows([[s for s in row] for row in rows])
            # Fetch UTF-8 output from the queue ...
            data = self.queue.getvalue()
            data = data.decode("utf-8")
            # ... and reencode it into the target encoding
            data = self.encoder.encode(data)
            # write to the target stream
            self.stream.write(data)
            # empty queue
            self.queue.truncate(0)
项目:IRCLogParser    作者:prasadtalasila    | 项目源码 | 文件源码
def save_csv(matrix, output_directory, output_file_name):
    """ 
        Saves the input matrix as a CSV File

    Args:
        matrix(list):  an array containing data to be saved
        output_drectory(str): location to save graph
        output_file_name(str): name of the csv file to be saved

    Returns:
       null

    """
    if config.DEBUGGER:
        print "Generating", (output_file_name + ".csv")
    check_if_dir_exists(output_directory) #create output directory if doesn't exist
    output_file = output_directory + "/" + output_file_name +".csv"
    with open(output_file, 'wb') as myfile:
        wr = csv.writer(myfile, quoting=csv.QUOTE_NONNUMERIC)
        if matrix is not None:
            for col in matrix:
                wr.writerow(col)
项目:PyCS    作者:COSMOGRAIL    | 项目源码 | 文件源码
def writecsv(estimates, filepath, append=False):
    """
    Write a CSV file of estimates.
    If append = True, I do NOT overwrite an existing file, but append to it !
    """
    if append:
        f = open(filepath, "a")
    else:
        f = open(filepath, "w")
    writer = csv.writer(f, delimiter=',', quotechar='"', quoting=csv.QUOTE_NONNUMERIC)
    for est in estimates:
        #est.td = "%.5f" % (est.td)
        #est.tderr = "%.5f" % (est.tderr)
        #est.ms = "%.5f" % (est.ms)
        #est.timetaken = "%.5f" % (est.timetaken)
        writer.writerow(est.aslist()) 
    f.close()
    print "Wrote %i estimates into %s" % (len(estimates), filepath)
项目:graph_tools_networkx    作者:AirtrafficAnalysisGroup    | 项目源码 | 文件源码
def build_graph_for_file(file_path, dir_name, name):
    data = open(file_path, 'r')
    G=nx.DiGraph()
    rows = csv.reader(data, quoting=csv.QUOTE_NONNUMERIC)
    next(rows) #skip the header
    for row in rows:
        row_fil = list(filter(lambda x: type(x) is float, row))
        if G.has_node(row_fil[0]) is not True:
            G.add_node(row_fil[0], market_id=row_fil[1])
        if G.has_node(row_fil[2]) is not True:
            G.add_node(row_fil[2], market_id=row_fil[3])
        if G.has_edge(row_fil[0], row_fil[2]):
            old = G.get_edge_data(row_fil[0], row_fil[2])
            G.add_edge(row_fil[0], row_fil[2], num_of_people=old['num_of_people'] + row_fil[4], total_price=old['total_price'] + row_fil[5])
        else:
            G.add_edge(row_fil[0], row_fil[2], num_of_people=row_fil[4], total_price=row_fil[5])

    output_file_path = ('graphs/' + name + '.gexf') 
    nx.write_gexf(G, output_file_path)
项目:pretix-sepadebit    作者:pretix    | 项目源码 | 文件源码
def render(self, form_data: dict):
        output = io.StringIO()
        writer = csv.writer(output, quoting=csv.QUOTE_NONNUMERIC, delimiter=",")
        tz = pytz.timezone(self.event.settings.timezone)

        headers = [
            _('Order code'), _('Order date'), _('Invoices'), _('SEPA export date'), _('Payment amount')
        ]

        writer.writerow(headers)

        qs = SepaExportOrder.objects.filter(export__event=self.event).order_by('export__datetime').select_related(
            'export', 'order'
        ).prefetch_related('order__invoices')
        for seo in qs:
            row = [
                seo.order.code,
                seo.order.datetime.astimezone(tz).strftime('%Y-%m-%d'),
                ', '.join([i.number for i in seo.order.invoices.all()]),
                seo.export.datetime.astimezone(tz).strftime('%Y-%m-%d %H:%M:%S'),
                seo.amount,
            ]
            writer.writerow(row)

        return 'sepaexports.csv', 'text/csv', output.getvalue().encode("utf-8")
项目:AestheticDril    作者:rhodochrosiite    | 项目源码 | 文件源码
def _build_file(self):
        self.total_rows = 0 

        #Get recent tweets from dril and add to new file
        for status in tweepy.Cursor(api.user_timeline, 'dril', since_id=self.since).items():
            self.total_rows += self._process_status(status)

        #Put content of old file in new file
        #This is kind of messy uhhh
        try:
            #Open things for reading and writing
            readFile = open('data/dril.csv', 'rt', encoding='utf-8')
            writeFile = open('data/new.csv', 'at', encoding='utf-8')

            read = reader(readFile)
            write = writer(writeFile, delimiter=',', quoting=QUOTE_NONNUMERIC) #Uhhhhmmmmmhmh mmmm

            for row in read:
                write.writerow([int(row[0]), row[1]])
                self.total_rows += 1
        except IOError:
            print('Failed to open file (1) [okay if this is the first time running]')

        #Rename the new file to be the old file
        os.rename('data/new.csv', 'data/dril.csv')
项目:edata    作者:ap-Codkelden    | 项目源码 | 文件源码
def make_csv(edata, verbose=None):
    fieldnames = ["id", 'doc_number', 'doc_date', 'doc_v_date', "trans_date",
                  "amount", "payer_edrpou", "payer_name", 'payer_account',
                  "payer_mfo", "payer_bank", "recipt_edrpou", "recipt_name",
                  'recipt_account', "recipt_mfo", "recipt_bank", "region_id",
                  'doc_add_attr', "payment_details",]
    try:
        with open('edata.csv', 'w') as csvfile:
            writer = csv.DictWriter(
                csvfile, fieldnames=fieldnames,
                delimiter=';', quoting=csv.QUOTE_NONNUMERIC
                )
            writer.writeheader()
            for row in edata:
                writer.writerow(row)
        if verbose:
            sys.stdout.write("{} ?????? ????????\n".format(len(edata)))
    except:
        raise
项目:django-admin-reports    作者:simplyopen-it    | 项目源码 | 文件源码
def to_csv(self, fileobj, header=False, totals=False, delimiter=';',
               quotechar='"', quoting=csv.QUOTE_NONNUMERIC,
               escapechar='', extra_rows=None, **kwargs):
        writer = csv.writer(fileobj, delimiter=str(delimiter),
                            quotechar=str(quotechar), quoting=quoting,
                            escapechar=str(escapechar), **kwargs)
        if extra_rows is not None:
            writer.writerows(extra_rows)
        if header:
            writer.writerow([name.encode(settings.DEFAULT_CHARSET) for name, _ in self.get_fields()])
        for record in self.iter_results():
            writer.writerow([elem.encode(settings.DEFAULT_CHARSET) if isinstance(elem, unicode) else elem
                             for elem in record])
        if totals and self.get_has_totals():
            writer.writerow(self.totals)
项目:fb-page-chat-download    作者:eisenjulian    | 项目源码 | 文件源码
def run(self):
        output = open(self.output, 'w', newline="\n", encoding="utf-8")
        threads = requests.get(self.uri).json()
        if 'error' in threads:
            print(threads)
            return

        fieldnames = ['from_id', 'from', 'time', 'message', 'attachments', 'shares', 'url']
        self.writer = csv.DictWriter(output, dialect='excel', fieldnames=fieldnames, extrasaction='ignore', quoting=csv.QUOTE_NONNUMERIC)
        self.writer.writerow(dict((n, n) for n in fieldnames))
        self.scrape_thread_list(threads, 5)
        output.close()
项目:pymotw3    作者:reingart    | 项目源码 | 文件源码
def __init__(self, outputFile):
        self.writer = csv.writer(
            outputFile,
            quoting=csv.QUOTE_NONNUMERIC,
        )
        self.group_name = ''
项目:BitcoinExchangeFH    作者:Aurora-Team    | 项目源码 | 文件源码
def insert(self, table, columns, types, values, primary_key_index=[], is_orreplace=False, is_commit=True):
        """
        Insert into the table
        :param table: Table name
        :param columns: Column array
        :param types: Type array
        :param values: Value array
        :param primary_key_index: An array of indices of primary keys in columns,
                          e.g. [0] means the first column is the primary key
        :param is_orreplace: Indicate if the query is "INSERT OR REPLACE"
        """
        ret = True
        file_path = os.path.join(self.file_directory, table + ".csv")
        if len(columns) != len(values):
            return False

        self.lock.acquire()
        if not os.path.isfile(file_path):
            ret = False
        else:
            with open(file_path, "a+") as csvfile:
                writer = csv.writer(csvfile, lineterminator='\n', quotechar='\"', quoting=csv.QUOTE_NONNUMERIC)
                writer.writerow(values)
        self.lock.release()

        if not ret:
            raise Exception("File (%s) has not been created.")

        return True
项目:Python-Programs    作者:OmkarPathak    | 项目源码 | 文件源码
def csvWrite(filePath, data):
    with open(filePath, 'a') as fd:
        writer = csv.writer(fd, delimiter=',', quoting=csv.QUOTE_NONNUMERIC)
        writer.writerow(data)
项目:dati-ckan-docker    作者:italia    | 项目源码 | 文件源码
def get_csv(self):
        csvout = StringIO()
        csvwriter = csv.writer(
            csvout,
            dialect='excel',
            quoting=csv.QUOTE_NONNUMERIC
        )
        csvwriter.writerow(self.column_names)
        for row in self.rows:
            row_formatted = []
            for cell in row:
                if isinstance(cell, datetime.datetime):
                    cell = cell.strftime('%Y-%m-%d %H:%M')
                elif isinstance(cell, (int, long)):
                    cell = str(cell)
                elif isinstance(cell, (list, tuple)):
                    cell = str(cell)
                elif cell is None:
                    cell = ''
                else:
                    cell = cell.encode('utf8')
                row_formatted.append(cell)
            try:
                csvwriter.writerow(row_formatted)
            except Exception, e:
                raise Exception("%s: %s, %s"%(e, row, row_formatted))
        csvout.seek(0)
        return csvout.read()
项目:zippy    作者:securesystemslab    | 项目源码 | 文件源码
def test_write_quoting(self):
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"')
        self.assertRaises(csv.Error,
                          self._write_test,
                          ['a',1,'p,q'], 'a,1,p,q',
                          quoting = csv.QUOTE_NONE)
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"',
                         quoting = csv.QUOTE_MINIMAL)
        self._write_test(['a',1,'p,q'], '"a",1,"p,q"',
                         quoting = csv.QUOTE_NONNUMERIC)
        self._write_test(['a',1,'p,q'], '"a","1","p,q"',
                         quoting = csv.QUOTE_ALL)
        self._write_test(['a\nb',1], '"a\nb","1"',
                         quoting = csv.QUOTE_ALL)
项目:zippy    作者:securesystemslab    | 项目源码 | 文件源码
def test_read_quoting(self):
        self._read_test(['1,",3,",5'], [['1', ',3,', '5']])
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quotechar=None, escapechar='\\')
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quoting=csv.QUOTE_NONE, escapechar='\\')
        # will this fail where locale uses comma for decimals?
        self._read_test([',3,"5",7.3, 9'], [['', 3, '5', 7.3, 9]],
                        quoting=csv.QUOTE_NONNUMERIC)
        self._read_test(['"a\nb", 7'], [['a\nb', ' 7']])
        self.assertRaises(ValueError, self._read_test,
                          ['abc,3'], [[]],
                          quoting=csv.QUOTE_NONNUMERIC)
项目:oil    作者:oilshell    | 项目源码 | 文件源码
def test_write_quoting(self):
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"')
        self._write_error_test(csv.Error, ['a',1,'p,q'],
                               quoting = csv.QUOTE_NONE)
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"',
                         quoting = csv.QUOTE_MINIMAL)
        self._write_test(['a',1,'p,q'], '"a",1,"p,q"',
                         quoting = csv.QUOTE_NONNUMERIC)
        self._write_test(['a',1,'p,q'], '"a","1","p,q"',
                         quoting = csv.QUOTE_ALL)
        self._write_test(['a\nb',1], '"a\nb","1"',
                         quoting = csv.QUOTE_ALL)
项目:oil    作者:oilshell    | 项目源码 | 文件源码
def test_write_float(self):
        # Issue 13573: loss of precision because csv.writer
        # uses str() for floats instead of repr()
        orig_row = [1.234567890123, 1.0/7.0, 'abc']
        f = StringIO()
        c = csv.writer(f, quoting=csv.QUOTE_NONNUMERIC)
        c.writerow(orig_row)
        f.seek(0)
        c = csv.reader(f, quoting=csv.QUOTE_NONNUMERIC)
        new_row = next(c)
        self.assertEqual(orig_row, new_row)
项目:oil    作者:oilshell    | 项目源码 | 文件源码
def test_read_quoting(self):
        self._read_test(['1,",3,",5'], [['1', ',3,', '5']])
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quotechar=None, escapechar='\\')
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quoting=csv.QUOTE_NONE, escapechar='\\')
        # will this fail where locale uses comma for decimals?
        self._read_test([',3,"5",7.3, 9'], [['', 3, '5', 7.3, 9]],
                        quoting=csv.QUOTE_NONNUMERIC)
        self._read_test(['"a\nb", 7'], [['a\nb', ' 7']])
        self.assertRaises(ValueError, self._read_test,
                          ['abc,3'], [[]],
                          quoting=csv.QUOTE_NONNUMERIC)
项目:python2-tracer    作者:extremecoders-re    | 项目源码 | 文件源码
def test_write_quoting(self):
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"')
        self._write_error_test(csv.Error, ['a',1,'p,q'],
                               quoting = csv.QUOTE_NONE)
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"',
                         quoting = csv.QUOTE_MINIMAL)
        self._write_test(['a',1,'p,q'], '"a",1,"p,q"',
                         quoting = csv.QUOTE_NONNUMERIC)
        self._write_test(['a',1,'p,q'], '"a","1","p,q"',
                         quoting = csv.QUOTE_ALL)
        self._write_test(['a\nb',1], '"a\nb","1"',
                         quoting = csv.QUOTE_ALL)
项目:python2-tracer    作者:extremecoders-re    | 项目源码 | 文件源码
def test_write_float(self):
        # Issue 13573: loss of precision because csv.writer
        # uses str() for floats instead of repr()
        orig_row = [1.234567890123, 1.0/7.0, 'abc']
        f = StringIO()
        c = csv.writer(f, quoting=csv.QUOTE_NONNUMERIC)
        c.writerow(orig_row)
        f.seek(0)
        c = csv.reader(f, quoting=csv.QUOTE_NONNUMERIC)
        new_row = next(c)
        self.assertEqual(orig_row, new_row)
项目:python2-tracer    作者:extremecoders-re    | 项目源码 | 文件源码
def test_read_quoting(self):
        self._read_test(['1,",3,",5'], [['1', ',3,', '5']])
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quotechar=None, escapechar='\\')
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quoting=csv.QUOTE_NONE, escapechar='\\')
        # will this fail where locale uses comma for decimals?
        self._read_test([',3,"5",7.3, 9'], [['', 3, '5', 7.3, 9]],
                        quoting=csv.QUOTE_NONNUMERIC)
        self._read_test(['"a\nb", 7'], [['a\nb', ' 7']])
        self.assertRaises(ValueError, self._read_test,
                          ['abc,3'], [[]],
                          quoting=csv.QUOTE_NONNUMERIC)
项目:Effective-Amazon-Machine-Learning    作者:PacktPublishing    | 项目源码 | 文件源码
def run_experiment(experiment):
    '''
    This function is the main function
    It takes a string 'experiment' that drives the naming of the files

    - it generates the filenames for the datasources, schemas, recipes
    - it sends a randomized version of the data to an s3 bucket
    - creates the train, valid and test datasources
    - creates the model
    - creates the evaluation
    '''
    # start by generating all teh filenames for the current experiment
    filenames = generate_filenames(experiment)

    # load the data form local, shuffle and save back to original file
    df = pd.read_csv(filepath + filenames['main'])
    df = df.reindex(np.random.permutation(df.index))
    df.to_csv(filepath + filenames['main'], quoting= csv.QUOTE_NONNUMERIC, index=False)
    # sends the original file to s3
    os.system("aws s3 cp %s%s %s "% (filepath, filenames['main'], s3_path) )

    # write cli JSON
    create_dsrc("train", 0, 60)
    create_dsrc("valid", 60, 80)
    create_dsrc("test", 80, 100)

    create_model()
    create_eval("valid")
    create_eval("test")


# ----------------------------------------------------------------------------
#  datasource, model, evaluation and batch predictions functions
#  These functions all do the same thing
#  1. write the JSON parameters to a JSON formatted file in json_path folder
#  2. execute the AWS CLI command that will create the object: datasource, model, ...
# ----------------------------------------------------------------------------
项目:web_ctp    作者:molebot    | 项目源码 | 文件源码
def test_write_quoting(self):
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"')
        self.assertRaises(csv.Error,
                          self._write_test,
                          ['a',1,'p,q'], 'a,1,p,q',
                          quoting = csv.QUOTE_NONE)
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"',
                         quoting = csv.QUOTE_MINIMAL)
        self._write_test(['a',1,'p,q'], '"a",1,"p,q"',
                         quoting = csv.QUOTE_NONNUMERIC)
        self._write_test(['a',1,'p,q'], '"a","1","p,q"',
                         quoting = csv.QUOTE_ALL)
        self._write_test(['a\nb',1], '"a\nb","1"',
                         quoting = csv.QUOTE_ALL)
项目:web_ctp    作者:molebot    | 项目源码 | 文件源码
def test_read_quoting(self):
        self._read_test(['1,",3,",5'], [['1', ',3,', '5']])
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quotechar=None, escapechar='\\')
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quoting=csv.QUOTE_NONE, escapechar='\\')
        # will this fail where locale uses comma for decimals?
        self._read_test([',3,"5",7.3, 9'], [['', 3, '5', 7.3, 9]],
                        quoting=csv.QUOTE_NONNUMERIC)
        self._read_test(['"a\nb", 7'], [['a\nb', ' 7']])
        self.assertRaises(ValueError, self._read_test,
                          ['abc,3'], [[]],
                          quoting=csv.QUOTE_NONNUMERIC)
项目:qal    作者:OptimalBPM    | 项目源码 | 文件源码
def _quotestr_to_constants(_str):
        if _str is None:
            return csv.QUOTE_NONE
        elif _str.upper() == "MINIMAL":
            return csv.QUOTE_MINIMAL
        elif _str.upper() == "ALL":
            return csv.QUOTE_ALL
        elif _str.upper() == "NONNUMERIC":
            return csv.QUOTE_NONNUMERIC
        else:
            raise Exception("Error in _quotestr_to_constants: " + str(_str) + " is an invalid quotestr.")
项目:pandahouse    作者:kszucs    | 项目源码 | 文件源码
def to_csv(df):
    data = df.to_csv(header=False, index=False, encoding='utf-8',
                     quoting=csv.QUOTE_NONNUMERIC, escapechar='\\')
    if PY3:
        return data.encode('utf-8')
    else:
        return data
项目:PyDataLondon29-EmbarrassinglyParallelDAWithAWSLambda    作者:SignalMedia    | 项目源码 | 文件源码
def test_to_csv_quoting(self):
        df = DataFrame({'A': [1, 2, 3], 'B': ['foo', 'bar', 'baz']})

        buf = StringIO()
        df.to_csv(buf, index=False, quoting=csv.QUOTE_NONNUMERIC)

        result = buf.getvalue()
        expected = ('"A","B"\n'
                    '1,"foo"\n'
                    '2,"bar"\n'
                    '3,"baz"\n')

        self.assertEqual(result, expected)

        # quoting windows line terminators, presents with encoding?
        # #3503
        text = 'a,b,c\n1,"test \r\n",3\n'
        df = pd.read_csv(StringIO(text))
        buf = StringIO()
        df.to_csv(buf, encoding='utf-8', index=False)
        self.assertEqual(buf.getvalue(), text)

        # testing if quoting parameter is passed through with multi-indexes
        # related to issue #7791
        df = pd.DataFrame({'a': [1, 2], 'b': [3, 4], 'c': [5, 6]})
        df = df.set_index(['a', 'b'])
        expected = '"a","b","c"\n"1","3","5"\n"2","4","6"\n'
        self.assertEqual(df.to_csv(quoting=csv.QUOTE_ALL), expected)
项目:PyDataLondon29-EmbarrassinglyParallelDAWithAWSLambda    作者:SignalMedia    | 项目源码 | 文件源码
def test_to_csv_unicodewriter_quoting(self):
        df = DataFrame({'A': [1, 2, 3], 'B': ['foo', 'bar', 'baz']})

        buf = StringIO()
        df.to_csv(buf, index=False, quoting=csv.QUOTE_NONNUMERIC,
                  encoding='utf-8')

        result = buf.getvalue()
        expected = ('"A","B"\n'
                    '1,"foo"\n'
                    '2,"bar"\n'
                    '3,"baz"\n')

        self.assertEqual(result, expected)
项目:PyDataLondon29-EmbarrassinglyParallelDAWithAWSLambda    作者:SignalMedia    | 项目源码 | 文件源码
def _stringify(s, encoding, errors):
    if s is None:
        return ''
    if isinstance(s, unicode):
        return s.encode(encoding, errors)
    elif isinstance(s, (int , float)):
        pass #let csv.QUOTE_NONNUMERIC do its thing.
    elif not isinstance(s, str):
        s=str(s)
    return s
项目:python-exasol    作者:EXASOL    | 项目源码 | 文件源码
def pandasWriteCallback(data, outputFile, **kw):
    """Write callback for Pandas data frames"""
    # import only when required
    global pandas; import pandas # pylint: disable=F0401
    if not isinstance(data, pandas.DataFrame):
        raise TypeError("pandas.DataFrame expected as first argument")
    data.to_csv(outputFile, header = False, index = False, quoting = csv.QUOTE_NONNUMERIC, **kw)
项目:neighborhood_mood_aws    作者:jarrellmark    | 项目源码 | 文件源码
def _stringify(s, encoding, errors):
    if s is None:
        return ''
    if isinstance(s, unicode):
        return s.encode(encoding, errors)
    elif isinstance(s, (int , float)):
        pass #let csv.QUOTE_NONNUMERIC do its thing.
    elif not isinstance(s, str):
        s=str(s)
    return s
项目:neuromorphic_odour_classifier    作者:iulialexandra    | 项目源码 | 文件源码
def __init__(self, logPath, logFileName, serialPort, fieldnames):
        self.logPath = logPath
        self.logFileName = logFileName
        # open the output file (for appending)
        self.out = open(os.path.join(logPath, logFileName), 'w' )
        self.fieldnames = fieldnames
        self.writer = csv.DictWriter(self.out, fieldnames = fieldnames, 
                      dialect = 'excel', quoting = csv.QUOTE_NONNUMERIC)
        self.writer.writeheader()
        self.serialPort = serialPort
        self._connect_arduino()
项目:pefile.pypy    作者:cloudtracer    | 项目源码 | 文件源码
def test_write_quoting(self):
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"')
        self.assertRaises(csv.Error,
                          self._write_test,
                          ['a',1,'p,q'], 'a,1,p,q',
                          quoting = csv.QUOTE_NONE)
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"',
                         quoting = csv.QUOTE_MINIMAL)
        self._write_test(['a',1,'p,q'], '"a",1,"p,q"',
                         quoting = csv.QUOTE_NONNUMERIC)
        self._write_test(['a',1,'p,q'], '"a","1","p,q"',
                         quoting = csv.QUOTE_ALL)
        self._write_test(['a\nb',1], '"a\nb","1"',
                         quoting = csv.QUOTE_ALL)
项目:pefile.pypy    作者:cloudtracer    | 项目源码 | 文件源码
def test_write_float(self):
        # Issue 13573: loss of precision because csv.writer
        # uses str() for floats instead of repr()
        orig_row = [1.234567890123, 1.0/7.0, 'abc']
        f = StringIO()
        c = csv.writer(f, quoting=csv.QUOTE_NONNUMERIC)
        c.writerow(orig_row)
        f.seek(0)
        c = csv.reader(f, quoting=csv.QUOTE_NONNUMERIC)
        new_row = next(c)
        self.assertEqual(orig_row, new_row)
项目:pefile.pypy    作者:cloudtracer    | 项目源码 | 文件源码
def test_read_quoting(self):
        self._read_test(['1,",3,",5'], [['1', ',3,', '5']])
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quotechar=None, escapechar='\\')
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quoting=csv.QUOTE_NONE, escapechar='\\')
        # will this fail where locale uses comma for decimals?
        self._read_test([',3,"5",7.3, 9'], [['', 3, '5', 7.3, 9]],
                        quoting=csv.QUOTE_NONNUMERIC)
        self._read_test(['"a\nb", 7'], [['a\nb', ' 7']])
        self.assertRaises(ValueError, self._read_test,
                          ['abc,3'], [[]],
                          quoting=csv.QUOTE_NONNUMERIC)
项目:ouroboros    作者:pybee    | 项目源码 | 文件源码
def test_write_quoting(self):
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"')
        self._write_error_test(csv.Error, ['a',1,'p,q'],
                               quoting = csv.QUOTE_NONE)
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"',
                         quoting = csv.QUOTE_MINIMAL)
        self._write_test(['a',1,'p,q'], '"a",1,"p,q"',
                         quoting = csv.QUOTE_NONNUMERIC)
        self._write_test(['a',1,'p,q'], '"a","1","p,q"',
                         quoting = csv.QUOTE_ALL)
        self._write_test(['a\nb',1], '"a\nb","1"',
                         quoting = csv.QUOTE_ALL)
项目:ouroboros    作者:pybee    | 项目源码 | 文件源码
def test_read_quoting(self):
        self._read_test(['1,",3,",5'], [['1', ',3,', '5']])
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quotechar=None, escapechar='\\')
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quoting=csv.QUOTE_NONE, escapechar='\\')
        # will this fail where locale uses comma for decimals?
        self._read_test([',3,"5",7.3, 9'], [['', 3, '5', 7.3, 9]],
                        quoting=csv.QUOTE_NONNUMERIC)
        self._read_test(['"a\nb", 7'], [['a\nb', ' 7']])
        self.assertRaises(ValueError, self._read_test,
                          ['abc,3'], [[]],
                          quoting=csv.QUOTE_NONNUMERIC)
项目:ndk-python    作者:gittor    | 项目源码 | 文件源码
def test_write_quoting(self):
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"')
        self.assertRaises(csv.Error,
                          self._write_test,
                          ['a',1,'p,q'], 'a,1,p,q',
                          quoting = csv.QUOTE_NONE)
        self._write_test(['a',1,'p,q'], 'a,1,"p,q"',
                         quoting = csv.QUOTE_MINIMAL)
        self._write_test(['a',1,'p,q'], '"a",1,"p,q"',
                         quoting = csv.QUOTE_NONNUMERIC)
        self._write_test(['a',1,'p,q'], '"a","1","p,q"',
                         quoting = csv.QUOTE_ALL)
        self._write_test(['a\nb',1], '"a\nb","1"',
                         quoting = csv.QUOTE_ALL)
项目:ndk-python    作者:gittor    | 项目源码 | 文件源码
def test_write_float(self):
        # Issue 13573: loss of precision because csv.writer
        # uses str() for floats instead of repr()
        orig_row = [1.234567890123, 1.0/7.0, 'abc']
        f = StringIO()
        c = csv.writer(f, quoting=csv.QUOTE_NONNUMERIC)
        c.writerow(orig_row)
        f.seek(0)
        c = csv.reader(f, quoting=csv.QUOTE_NONNUMERIC)
        new_row = next(c)
        self.assertEqual(orig_row, new_row)
项目:ndk-python    作者:gittor    | 项目源码 | 文件源码
def test_read_quoting(self):
        self._read_test(['1,",3,",5'], [['1', ',3,', '5']])
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quotechar=None, escapechar='\\')
        self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']],
                        quoting=csv.QUOTE_NONE, escapechar='\\')
        # will this fail where locale uses comma for decimals?
        self._read_test([',3,"5",7.3, 9'], [['', 3, '5', 7.3, 9]],
                        quoting=csv.QUOTE_NONNUMERIC)
        self._read_test(['"a\nb", 7'], [['a\nb', ' 7']])
        self.assertRaises(ValueError, self._read_test,
                          ['abc,3'], [[]],
                          quoting=csv.QUOTE_NONNUMERIC)
项目:Sentiment-Analysis-Twitter    作者:ayushoriginal    | 项目源码 | 文件源码
def printFreqDistCSV( dist, filename='' ):
    n_samples = len(dist.keys())
    n_repeating_samples = sum([ 1 for (k,v) in dist.items
        () if v>1 ])
    n_outcomes = dist._N
    print '%-12s %-12s %-12s'%( 'Samples', 'RepSamples', 'Outcomes' )
    print '%12d %12d %12d'%( n_samples, n_repeating_samples, n_outcomes )

    if( len(filename)>0 and '_'!=filename[0] ):
        with open( filename, 'w' ) as fcsv:
            distwriter = csv.writer( fcsv, delimiter=',', quotechar='"', quoting=csv.QUOTE_NONNUMERIC )

            for (key,value) in dist.items():
                distwriter.writerow( [key, value] ) #print key, '\t,\t', dist[key]
项目:Sentiment-Analysis-Twitter    作者:ayushoriginal    | 项目源码 | 文件源码
def oldStats2CSV( in_file, fileprefix=''):
    if fileprefix == '':
        fileprefix = in_file.rstrip('_stats.txt')
    fp = open( in_file, 'r' )
    fq = open( fileprefix+'_statsnew.txt', 'w' )

    line = ''
    line_start = 0
    line_count = 20
    line_end   = line_start+line_count
    for line_num in range(line_start, line_end):   # write Statistics
        line = fp.readline()
        fq.write( line )

    for section in [1,2,3]:
        line_start = line_end
        line_count = 2
        line_end   = line_start+line_count
        for line_num in range( line_start, line_end ):
            line = fp.readline()
            fq.write( line )

        line_start = line_end
        line_count = [int(l) for l in line.split() if l.isdigit()][0]
        line_end = line_start+line_count
        fr = open( fileprefix+'_%dgrams.csv'%section, 'w')
        fwrt = csv.writer( fr, delimiter=',', quotechar='"', quoting=csv.QUOTE_NONNUMERIC )
        for line_num in range( line_start, line_end ):  # write unigrams
            line = fp.readline()
            row = line.split('\t,\t')
            row[0] = row[0].strip()
            row[1] = int(row[1])
            fwrt.writerow( row )
        fr.close()

    fp.close()
    fq.close()
项目:piwik-to-bigquery    作者:afterlastangel    | 项目源码 | 文件源码
def write_sql_to_file(
            self, sql, bind, last_id_name=None,
            with_header=True, delimiter=',',
            quotechar='"', quoting=csv.QUOTE_NONNUMERIC):
        conn = self.connection
        cur = conn.cursor(pymysql.cursors.SSCursor)
        cur.execute(sql, bind)
        fields_name = []
        schemas = []
        count = 0
        last_id_column = None
        last_id = None
        sql_type = []
        description = cur.description
        for column in description:
            fields_name.append(column[0])
            sql_type.append(column[1])
            schemas.append(
                SchemaField(column[0], self.get_type(column[1])))
            if column[0] == last_id_name:
                last_id_column = count
            count += 1

        ofile = open(self.full_path_name(), 'wb')
        csv_writer = csv.writer(
            ofile, delimiter=delimiter, quotechar=quotechar, quoting=quoting)
        if with_header:
            csv_writer.writerow(fields_name)
        while True:
            results = cur.fetchmany(size=1000)
            if len(results) != 0:
                for result in results:
                    data = self.process_binary(result, sql_type, description)
                    csv_writer.writerow(data)
                    if last_id_column is not None:
                        last_id = result[last_id_column]
            else:
                break
        cur.close()
        ofile.close()
        return schemas, last_id
项目:adaptor    作者:e-stat-api    | 项目源码 | 文件源码
def _merge_data(appId, ids, group_by, ext):
    eStatAPI._['appId'] = appId
    aggregate = request.args.get('aggregate') if request.args.get(
        'aggregate') is not None else ''
    data = eStatAPI.merge_data(ids, group_by, aggregate)
    eStatAPI.path['tmp_merge'] = eStatAPI.path['tmp'] + '.'.join(
        [eStatAPI._['appId'], ''.join([l for l in random.choice(eStatAPI.random_str)]), 'csv'])
    data.to_csv(eStatAPI.path['tmp_merge'],
                quoting=csv.QUOTE_NONNUMERIC, index=None)
    tmp_csv = eStatAPI.cmd_line(eStatAPI.build_cmd(
        ['cat', eStatAPI.path['tmp_merge']]))
    eStatAPI.cmd_line(eStatAPI.build_cmd(['rm', eStatAPI.path['tmp_merge']]))
    return eStatAPI.response(eStatAPI.get_output(tmp_csv, ext), ext)