Re: Single vs. Double Quotes

[Ben Abbott]

On Oct 16, 2011, at 3:52 PM, Rik wrote:

> On 10/16/2011 08:53 AM, Ben Abbott wrote:
>> On Oct 15, 2011, at 5:07 PM, John W. Eaton wrote:
>> 
>>> On 15-Oct-2011, Michael Goffioul wrote:
>>> 
>>> | On Sat, Oct 15, 2011 at 7:08 PM, John W. Eaton <address@hidden> wrote:
>>> | > On 15-Oct-2011, Michael Goffioul wrote:
>>> | >
>>> | > | way.This is could be achieved as easily as:
>>> | > |
>>> | > | strrep (delimiter_string, '\n', "\n")
>>> | >
>>> | > I'm not sure, but I don't think it is quite this simple.  WDMD if the
>>> | > delimiter string is set to '\\n'?  Is this recognized as a
>>> | > two-character sequence backslash and n, or is it backslash and NL?
>>> | >
>>> | > In *printf, I think it is backslash and n, so it would be unfortunate
>>> | > if the delimiter string in strread did something different.
>>> | 
>>> | You're right, this was just a quick hack I considered, but far from a
>>> | good solution. Your other solution is better. I also thought about
>>> | that, but was afraid of other side-effects due to the use of sprintf.
>>> | Doubling % characters should prevent them.
>>> 
>>> Another possibility is do_string_escapes.  It should still be applied
>>> only to single-quoted strings in strread, but it saves you the step of
>>> doubling the % characters and it only does backslash-style escape
>>> processing, so it might be faster than sprintf.
>>> 
>>> jwe
>> 
>> Is there a robust way to determine whether a string is single-quoted or 
>> double-quoted?
> 
> Ben,
> 
> I'm not sure if this is the absolutely perfect way to do it but ...
> 
> typeinfo ("abc")
> ans = string
> typeinfo ('abc')
> ans = sq_string
> 
> The 'sq' stands for Single Quote.
> 
> Cheers,
> Rik

Are delimiter, whitespace, and endofline the only inputs that need this 
treatment.

If so, is the attached satisfactory?

Ben

changeset.patch
Description: Binary data