Create an ASCII table from CSV

Question

The following code convert csv to ascii table.

#!/usr/bin/env python3
import sys
import argparse
import logging
import prettytable
import csv
parser = argparse.ArgumentParser()
parser.add_argument("infile", nargs="?", type=argparse.FileType("r"), default=sys.stdin)
parser.add_argument(
 '-d',
 '--debug',
 help="Print lots of debugging statements",
 action="store_const",
 dest="loglevel",
 const=logging.DEBUG,
 default=logging.WARNING,
)
parser.add_argument(
 '-v',
 '--verbose',
 help="Be verbose",
 action="store_const",
 dest="loglevel",
 const=logging.INFO,
)
args = parser.parse_args()
# taking input from stdin which is empty, so there is neither a stdin nor a file as argument
if sys.stdin.isatty() and args.infile.name == "<stdin>":
 sys.exit("Please give some input")
logging.basicConfig(level=args.loglevel)
# # Business Logic Here
table = None
content = csv.reader(args.infile, delimiter=',', quotechar='"')
for row in content:
 if table is None:
 table = prettytable.PrettyTable(row)
 else:
 table.add_row(row)
print(table)

Please let me know how I can make this code better.

• Ξένη Γήινος: You don't need a csv to print an ASCII table, Python has a builtin function that returns the character located the Unicode codepoint for any inputted integer in the range [0, 1114111]: chr. ASCII is codepoints [0, 255], in Python: range(256). Use a list comprehension to get the characters: [chr(i) for i in range(256)]. If you want a square: [[chr(16*i+j) for j in range(16)] for i in range(16)]
• Ahmad Ismail: @XeнεiΞэnвες This sounds interesting. can you please elaborate in the answer section. I could not completely understand what you are saying.
• Toby Speight: @XeнεiΞэnвες, I think you have misunderstood what the code does (admittedly the description is very brief). It seems to be about formatting an existing table using ASCII formatting (-, | and +), rather than making a table of the ASCII characters.

Answer 1

I don't understand the point of this test:

if sys.stdin.isatty() and args.infile.name == "<stdin>":
 sys.exit("Please give some input")

We want to disallow a particular file name, but only if we're connected to a tty? If you want to prompt when input is coming from stdin, then we shouldn't be testing the file name, but instead its properties. And I don't see why we should exit in this case (e.g. we might want to type directly, or copy-paste input). I suggest:

if args.infile.isatty():
 print("Please enter your data:")

We might want to add a hint such as Finish with control-D - I'll leave it as an exercise to discover the terminal's control bindings and print the correct keystroke.

---

We don't need to read one row at a time. prettytable can do that for us:

table = prettytable.from_csv(args.infile)

However, it requires the stream to be seekable, so if it's not, we'll need to read it into memory and pass that as a stream:

if not args.infile.seekable():
 from io import StringIO
 args.infile = StringIO(args.infile.read())
table = prettytable.from_csv(args.infile)

---

It's probably a good idea to print only the message from any exceptions, rather than an entire backtrace (unless the user asks for the extra detail):

try:
 if not args.infile.seekable():
 from io import StringIO
 args.infile = StringIO(args.infile.read())
 print(prettytable.from_csv(args.infile))
except Exception as e:
 logging.error(e, exc_info=(args.loglevel<=logging.DEBUG))
 exit(1)

---

Modified code

#!/usr/bin/env python3
import sys
import argparse
import logging
import prettytable
import csv
parser = argparse.ArgumentParser()
parser.add_argument(
 "infile", nargs="?",
 type=argparse.FileType("r"),
 default=sys.stdin
)
parser.add_argument(
 '-d', '--debug',
 help="Print lots of debugging statements",
 action="store_const", dest="loglevel", const=logging.DEBUG,
 default=logging.WARNING,
)
parser.add_argument(
 '-v', '--verbose',
 help="Be verbose",
 action="store_const", dest="loglevel", const=logging.INFO,
)
args = parser.parse_args()
logging.basicConfig(level=args.loglevel)
if args.infile.isatty():
 print("Please enter your data:")
try:
 if not args.infile.seekable():
 from io import StringIO
 args.infile = StringIO(args.infile.read())
 table = prettytable.from_csv(args.infile)
 print(table)
except Exception as e:
 logging.error(e, exc_info=(args.loglevel<=logging.INFO))
 exit(1)

• Ahmad Ismail: There is a comment above the line which i know is sort of incomprehensible. We can take input from either file (0) or stdin(1). So there are four combinations. 00, 01, 10, 11. The lines you mentioned deal with 00.
• Toby Speight: We take input from a File object, i.e. a stream, which could be connected to a terminal, a seekable file, a named pipe, etc. Your version fails if we attempt to read a file called <stdin> - e.g. python3 273995.py '<stdin>' refuses. On the other hand python3 273995.py /dev/tty doesn't issue the prompt where my version does. Do you have an example where your version does better than mine?
• Ahmad Ismail: I replaced everything bellow # # Business Logic Here with table = prettytable.from_csv(args.infile) and printed the table. It just works.
• Toby Speight: I had meant to show the resultant modified program; I've now added it to the answer for convenience.
• Ahmad Ismail: Thank you very much for the code.

Answer 2

A few things that I would change:

1. Use functions - move some of the logic to functions; this change will enable you to reuse your code more easily in the future. It will also make unit testing much easier.

2. sys.exit() will only print the error to standard error stream. It's preferable to use the logging package instead.

3. Handle and communicating errors better - maybe the CSV is corrupted? Maybe it's empty and then you will print "None" to the user who will not understand what is the problem, so add more checks, log on error and better communicate the errors to the user.

• Ahmad Ismail: Can you please add code to your answer.
• Toby Speight: Note that demonstrative code isn't required in an answer. If a review includes suggested alternatives, that can sometimes make it more helpful, but it's not always appropriate.