We Keep Coding

Save CSV file using python

I am able to change the data to lowercase and remove all the punctuation but I have trouble saving the corrected data in CSV file.
import csv
import re
import os
input_file=raw_input("Name of the CSV file:")
output_file=raw_input("Output Name:")
reg_test=input_file
result = ''
with open(input_file,'r') as csvfile:
with open(output_file,'w') as csv_out_file:
filereader = csv.reader(csvfile)
filewriter =csv.writer(csv_out_file)
for row in filereader:
row = re.sub('[^A-Za-z0-9]+', '', str(row))
result += row + ','
lower = (result).lower()
csvfile.close()
csv_out_file.close()

You do not have to close the files, this is done automatically after the context of the with statement is over and you have to actually write something after you create the csv.writer, e.g. with writerow:
import csv
import re
input_file = 'in.csv'
output_file = 'out.csv'
with open(input_file, 'r') as csvfile, open(output_file, 'w') as csv_out_file:
filereader = csv.reader(csvfile)
filewriter = csv.writer(csv_out_file)
for row in filereader:
new_row = re.sub('[^A-Za-z0-9]+', '', str(row)) # manipulate the row
filewriter.writerow([new_row.lower()]) # write the new row to the out file
# the files are closed automatically after the context of the with statement is over
This saves the manipulated content of the first csv file to the second.

Python CSV writer - writing columns in new csv file up to maximum number of fields in csv files

I have 200 CSV files in my folder.
What I am trying to do is read first row of each files and write in new csv.
And on top, I want to write [file,field1,field2,...fieldn]
n is maximum number of fields.
import csv
import glob
list=[]
hel=[]
files=glob.glob('C:/dataset/*.csv')
with open('test.csv', 'w',newline='') as testfile:
csv_writer = csv.writer(testfile)
for file in files:
with open(file, 'r') as infile:
file=file[file.rfind('\\')+1:]
file=file.strip('.csv')
reader = csv.reader(infile)
headers = next(reader)
hel.append((len(headers)))
max(hel)
lst = [file] + headers
csv_writer.writerow(lst)
It came out that maximum number of fields of 200 files are 255.
So on top of new csv file, I want to write file, field1, field2 ... field 255.
How can I do this?
import csv
import glob
list=[]
hel=[]
files=glob.glob('C:/dataset/*.csv')
with open('test.csv', 'w',newline='') as testfile:
csv_writer = csv.writer(testfile)
for file in files:
with open(file, 'r') as infile:
file=file[file.rfind('\\')+1:]
file=file.strip('.csv')
reader = csv.reader(infile)
headers = next(reader)
hel.append((len(headers)))
b=['field{}'.format(i) for i in range(1,max(hel)+1)]
lst = [file] + headers
csv_writer.writerow(lst)
Now b is list that looks like this ['field1','field2'...'field255']
I need to insert 'file' before 'field1' and write that row on the top of new csv file. Writing code after csv_writer.writerow(lst) gives me csv file with 'field1','field2'.. every other line. How can I fix this problem

You first need to read all your input files to determine the maximum number of fields is 255. Then you need to construct a list of field names to write into the output file (just once, not in a loop):
['field{}'.format(i) for i in range(1, 256)]
You can pass that list to the csv module to write it.

Read the field count and first line from each file before writing the file.
import glob
from itertools import chain
import os
from os.path import splitext, basename
def first_line(filepath):
with open(filepath) as f:
return next(f)
def write_test_file(dest_file_path, source_path_name):
source_paths = glob.glob(source_path_name)
first_lines = list(map(first_line, source_paths))
max_count = max(l.count(",") for l in first_lines)
field_names = map("field{}".format, range(1, max_count + 2))
header = ",".join(chain(["file"], field_names)) + os.linesep
file_names = (splitext(basename(p))[0] for p in source_paths)
content = chain([header], map(",".join, zip(file_names, first_lines)))
with open(dest_file_path, 'w') as testfile:
testfile.write("".join(content))
write_test_file('test.csv', 'C:/dataset/*.csv')

Python csv - replace any columns with specified value

I have the following input file with a header row:
test_in.csv
LON,LAT,NUMBER,STREET,UNIT,CITY,DISTRICT,REGION,POSTCODE,ID,HASH
-72.5708234,41.4155142,39,HICKORY LA,,,,,,,8a0df668e0d49b02
-72.5647745,41.4160301,1213,KILLINGWORTH RD,,,,,,,b3ecaab86e476f46
I need to replace any of the columns with a specified string
for example CITY column's data should be replaced from "" to "MyCity"
My code only outputs the header and first row
python test_forcefld.py test_in.csv MyCity CITY out_test.csv
import csv
import sys
in_file_name = sys.argv[1]
force_data = sys.argv[2]
force_fld = sys.argv[3]
out_file_name = sys.argv[4]
# First read top row/header from input file
fieldnames = []
for filename in [in_file_name]:
with open(filename, "rb") as f_in:
reader = csv.reader(f_in)
headers = next(reader)
for h in headers:
fieldnames.append(h)
#print headers to output file
with open(out_file_name, 'w') as fou:
dw = csv.DictWriter(fou, delimiter=',', fieldnames=fieldnames)
dw.writeheader()
f_in2 = open(in_file_name, "rb")
reader2 = csv.DictReader(f_in2) # Uses the field names in this file
datarow = next(reader2)
datarow[force_fld] = force_data
with open(out_file_name, 'wa') as fou:
dw2 = csv.DictWriter(fou, delimiter=',', fieldnames=fieldnames)
dw2.writeheader()
dw2.writerow(data row)
Output shows
LON,LAT,NUMBER,STREET,UNIT,CITY,DISTRICT,REGION,POSTCODE,ID,HASH
-72.5708234,41.4155142,39,HICKORY LA,,MyCity,,,,,8a0df668e0d49b02

Your code is a little difficult to read, but assuming datarow is a dictionary containing your records:
In your last row, change
dw2.writerow(datarow)
Into
dw2.writerows(datarow)
While you're at it, you should also consider using datarow.keys() for your fieldnames, for conciseness.

This should do it, you just need pandas:
import pandas as pd
df = pd.read_csv(in_file_name, sep=',')
df['CITY'].fillna('MyCity', inplace=True)
And to save it:
df.to_csv(out_file_name)

You can try somthing like this in order to have your desired file:
I'm assuming your input file is called f_input.txt and your output file is called f_output.txt:
data = list(k.rstrip().split(',') for k in open("f_input.txt", 'r'))
with open("f_output.txt", 'a+') as f:
f.write(",".join(data[0]) + '\n')
for k in data[1:]:
# Modify the positions of k[:n] + your data + k[n+1]
# if you need to handle another position
f.write(",".join(k[:6]) + "MyCity" + ",".join(k[7:]) + "\n")

This worked in the end:
import csv
import sys
in_file_name = sys.argv[1]
force_data = sys.argv[2]
force_fld = sys.argv[3]
out_file_name = sys.argv[4]
# First read top row/header from input file
fieldnames = []
for filename in [in_file_name]:
with open(filename, "rb") as f_in:
reader = csv.reader(f_in)
headers = next(reader)
for h in headers:
fieldnames.append(h)
f_in2 = open(in_file_name, "r")
#print headers to output file
fou = open(out_file_name, 'wa')
dw = csv.DictWriter(fou, delimiter=',', fieldnames=fieldnames)
dw.writeheader()
reader2 = csv.DictReader(f_in2) # Uses the field names in this file
for row in reader2:
row[force_fld] = force_data
dw2 = csv.DictWriter(fou, delimiter=',', fieldnames=fieldnames)
dw2.writerow(row)

How to not just add a new first column to csv but alter the header names

I would like to do the following
read a csv file, Add a new first column, then rename some of the columns
then load the records from csv file.
Ultimately, I would like the first column to be populated with the file
name.
I'm fairly new to Python and I've kind of worked out how to change the fieldnames however, loading the data is a problem as it's looking for the original fieldnames which no longer match.
Code snippet
import csv
import os
inputFileName = "manifest1.csv"
outputFileName = os.path.splitext(inputFileName)[0] + "_modified.csv"
with open(inputFileName, 'rb') as inFile, open(outputFileName, 'wb') as outfile:
r = csv.DictReader(inFile)
fieldnames = ['MapSvcName','ClientHostName', 'Databasetype', 'ID_A', 'KeepExistingData', 'KeepExistingMapCache', 'Name', 'OnPremisePath', 'Resourcestype']
w = csv.DictWriter(outfile,fieldnames)
w.writeheader()
*** Here is where I start to go wrong
# copy the rest
for node, row in enumerate(r,1):
w.writerow(dict(row))
Error
File "D:\Apps\Python27\ArcGIS10.3\lib\csv.py", line 148, in _dict_to_list
+ ", ".join([repr(x) for x in wrong_fields]))
ValueError: dict contains fields not in fieldnames: 'Databases [xsi:type]', 'Resources [xsi:type]', 'ID'
Would like to some assistance to not just learn but truly understand what I need to do.
Cheers and thanks
Peter
Update..
I think I've worked it out
import csv
import os
inputFileName = "manifest1.csv"
outputFileName = os.path.splitext(inputFileName)[0] + "_modified.csv"
with open(inputFileName, 'rb') as inFile, open(outputFileName, 'wb') as outfile:
r = csv.reader(inFile)
w = csv.writer(outfile)
header = next(r)
header.insert(0, 'MapSvcName')
#w.writerow(header)
next(r, None) # skip the first row from the reader, the old header
# write new header
w.writerow(['MapSvcName','ClientHostName', 'Databasetype', 'ID_A', 'KeepExistingData', 'KeepExistingMapCache', 'Name', 'OnPremisePath', 'Resourcestype'])
prevRow = next(r)
prevRow.insert(0, '0')
w.writerow(prevRow)
for row in r:
if prevRow[-1] == row[-1]:
val = '0'
else:
val = prevRow[-1]
row.insert(0,val)
prevRow = row
w.writerow(row)

Attempting to merge three columns in CSV, updating original CSV

Some example data:
title1|title2|title3|title4|merge
test|data|here|and
test|data|343|AND
",3|data|343|and
My attempt at coding this:
import csv
import StringIO
storedoutput = StringIO.StringIO()
fields = ('title1', 'title2', 'title3', 'title4', 'merge')
with open('file.csv', 'rb') as input_csv:
reader = csv.DictReader(input_csv, fields, delimiter='|')
for counter, row in enumerate(reader):
counter += 1
#print row
if counter != 1:
for field in fields:
if field == "merge":
row['merge'] = ("%s%s%s" % (row["title1"], row["title3"], row["title4"]))
print row
storedoutput.writelines(','.join(map(str, row)) + '\n')
contents = storedoutput.getvalue()
storedoutput.close()
print "".join(contents)
with open('file.csv', 'rb') as input_csv:
input_csv = input_csv.read().strip()
output_csv = []
output_csv.append(contents.strip())
if "".join(output_csv) != input_csv:
with open('file.csv', 'wb') as new_csv:
new_csv.write("".join(output_csv))
Output should be
title1|title2|title3|title4|merge
test|data|here|and|testhereand
test|data|343|AND|test343AND
",3|data|343|and|",3343and
For your reference upon running this code the first print it prints the rows as I would hope then to appear in the output csv. However the second print prints the title row x times where x is the number of rows.
Any input or corrections or working code would be appreciated.

I think we can make this a lot simpler. Dealing with the rogue " was a bit of a nuisance, I admit, because you have to work hard to tell Python you don't want to worry about it.
import csv
with open('file.csv', 'rb') as input_csv, open("new_file.csv", "wb") as output_csv:
reader = csv.DictReader(input_csv, delimiter='|', quoting=csv.QUOTE_NONE)
writer = csv.DictWriter(output_csv, reader.fieldnames, delimiter="|",quoting=csv.QUOTE_NONE, quotechar=None)
merge_cols = "title1", "title3", "title4"
writer.writeheader()
for row in reader:
row["merge"] = ''.join(row[col] for col in merge_cols)
writer.writerow(row)
produces
$ cat new_file.csv
title1|title2|title3|title4|merge
test|data|here|and|testhereand
test|data|343|AND|test343AND
",3|data|343|and|",3343and
Note that even though you wanted the original file updated, I refused. Why? It's a bad idea, because then you can destroy your data while working on it.
How can I be so sure? Because that's exactly what I did when I first ran your code, and I know better. ;^)

That double quote in the last line is definitely messing up the csv.DictReader().
This works:
new_lines = []
with open('file.csv', 'rb') as f:
# skip the first line
new_lines.append(f.next().strip())
for line in f:
# the newline and split the fields
line = line.strip().split('|')
# exctract the field data you want
title1, title3, title4 = line[0], line[2], line[3]
# turn the field data into a string and append in to the rest
line.append(''.join([title1, title3, title4]))
# save the new line for later
new_lines.append('|'.join(line))
with open('file.csv', 'w') as f:
# make one long string and write it to the new file
f.write('\n'.join(new_lines))

import csv
import StringIO
stored_output = StringIO.StringIO()
with open('file.csv', 'rb') as input_csv:
reader = csv.DictReader(input_csv, delimiter='|', quoting=csv.QUOTE_NONE)
writer = csv.DictWriter(stored_output, reader.fieldnames, delimiter="|",quoting=csv.QUOTE_NONE, quotechar=None)
merge_cols = "title1", "title3", "title4"
writer.writeheader()
for row in reader:
row["merge"] = ''.join(row[col] for col in merge_cols)
writer.writerow(row)
contents = stored_output.getvalue()
stored_output.close()
print contents
with open('file.csv', 'rb') as input_csv:
input_csv = input_csv.read().strip()
if input_csv != contents.strip():
with open('file.csv', 'wb') as new_csv:
new_csv.write("".join(contents))