My file a contains the text,
bcd\\\\.
With bash, i read the file and print characters from 4th to 8th position as,
tmp=$(cat a)
echo "${tmp:3:4}"
It prints,
\\\\
All happy. Now i use python's array slicing to print characters from 4th to 8th position as,
>>> f = open('a')
>>> v=f.read()
>>> v[3:7]
It prints,
'\\\\\\\\'
Why does bash and python behave differently when there are backslashes?
It is a matter of how python displays strings. Observe:
>>> f = open('a')
>>> v=f.read()
>>> v[3:7]
'\\\\\\\\'
>>> print v[3:7]
\\\\
When displaying v[3:7], the backslashes are escaped. When printing, print v[3:7], they are not escaped.
The line in your file should end with a newline character. In that case, observe:
>>> v[-1]
'\n'
>>> print v[-1]
>>>
The newline character is displayed as a backslash-n. It prints as a newline.
The results for tab are similar:
>>> s='a\tb'
>>> s
'a\tb'
>>> print s
a b
__repr__ and __str__, when you print something out, you get the return from the __str__ function, but when you just have it echoed on the console, you get the __repr__ return value