Escaping slashes (double backslash plague)

[Harry George]

Peter Hansen <peter at engcorp.com> writes:

> Aloysio Figueiredo wrote:
> > 
> > I need to replace every ocurrence of '/' in s by '\/'
> > in order to create a file named s. My first attempt
> > was:
> > 
> > s = '\/'.join(s.split('/'))
> > 
> > but it doesn't work:
> > 
> > >>> s = 'a/b'
> > >>> s = '\/'.join(s.split('/'))
> > >>> s
> > 'a\\/b'
> > >>> repr(s)
> > "'a\\\\/b'"
> > >>>
> > 
> > '\/'.join() escapes the backslashes and I don't know why.
> 
> It does not, although *you* are not escaping the backslash
> yourself, and that is dangerous.  Get in the habit of always
> escaping your own backslashes, so that if you ever happen
> to use a backslash followed by one of the characters which _is_
> a valid escape sequence, you won't get confused.
> 
>  '\/' == '\\/'
> 
> but 
> 
>  '\t' != '\\t'
> 
> The first example shows two ways of writing a string with the blackslash
> character followed by a forward slash.  The second example shows a TAB
> character on the left, but a backslash plus the letter 't', on the right.
> 
> As for your apparent automatic escaping of backslashes: when you show
> results in an interactive session by just typing the expression, such as
> when you do ">>> s"   you will see the repr() of the value, not the actual
> content.  Use print instead and you'll see the difference:
> 
> >>> print s
> 
> This is all covered pretty well, I think, by the Python tutorials and
> such.  Have you gone through those?
> 
> -Peter

Did someone already mention os.path?  Since this is about filenames,
that is the best cross-platform colution.


-- 
harry.g.george at boeing.com
6-6M31 Knowledge Management
Phone: (425) 342-5601