Is there some way to get lint error on missed commas in literal list of strings?
Example:
exceptions = ["banana", "pineapple", "apple"
"pen"]
You may think this list contains 4 items, but truth be told! "apple" and "pen" are joined into "applepen".
I'm terrified of these omitted commas. Is there some lint tool to help me find them?
Example 2:
exceptions = ["Carmichael",
"Vanessa" # <--- Spot the missing comma
"Ford"]
I'm not sure what kind of source analyze tool you're using, so I can only propose a suggestion. However, it would be too long for a comment, so I wrote a proof-of-concept script.
The idea is to look at the source code with Python's tokenize module, which generates tokens from Python expressions. If well-formed Python code contains implicitly continued string literals, it will show up as a STRING token followed by NL.
For example, let's use the following source file source.py as a test case.
x = ("a"
"b" # some trailing spaces
# Coment line
"c"
""
# The following is an explicit continuation
"d" \
"e")
Running the command python check.py < source.py on the file generates:
1:8: implicit continuation:
x = ("a"
~~~^
2:35: implicit continuation:
"b" # some trailing spaces
~~~^
4:3: implicit continuation:
"c"
~~~^
5:2: implicit continuation:
""
^
The program, check.py, is just a proof-of-concept and it does not check syntax errors or other edge cases:
import sys
import tokenize
LOOKNEXT = False
tok_gen = tokenize.generate_tokens(sys.stdin.readline)
for tok, tok_str, start, end, line_text in tok_gen:
if tok == tokenize.STRING:
LOOKNEXT = True
continue
if LOOKNEXT and (tok == tokenize.NL):
warn_header = "%d:%d: implicit continuation: " % start
print >> sys.stderr, warn_header
print >> sys.stderr, line_text
indents = start[1] - 3
if indents >= 0:
print >> sys.stderr, "%s~~~^" % (" " * indents)
else:
print >> sys.stderr, "%s^" % (" " * start[1])
LOOKNEXT = False
Hopefully the idea might help you extend your lint tool or IDE for your purpose.
SublimeText3 with plugin flake8, that is a wrapper around other python lint plugins, can fix it.
Else, you can make a script that count ((number of ")/2)-1 and commas in a line, and if result dont match, add a coma.
EDIT:
Explanaition of what i'm saying:
def countQuotes(string):
return string.count('"')
def countCommas(string):
return string.count(',')
files = os.listdir('your/directory')
for filename in files:
if filename.endswith(".py"):
with fileinput.FileInput("your/directory"+"/"+filename, inplace=True, backup='.bak') as fileContent:
for line in fileContent:
if '"' in line:
numQuotes = countQuotes(line)
numCommas = countCommas(line)
if(numQuotes == 2 and ']' in line):
if(numCommas != 0):
#error, must delete a comma in right place and print line
else:
print(line)
if(numQuotes == 2 and ']' not in line):
if(numCommas != 1):
#error, must add a comma in right place and print line
else:
print(line)
if(numQuotes > 2):
if(numCommas > (numQuotes//2)-1)
#error, must delete a comma in right place and print line
elif(numCommas < (numQuotes//2)-1)
#error, must add a comma in right place and print line
else:
print(line)
else:
print(line)
This method must work, just think where you must insert or delete the comma to finally have the format you want.
