Re: [patch] Support counted strings and embedded nulls in value.string

[Tom Tromey <tromey at redhat dot com>]

Phil> This patch adds two related elements to the pretty-printer python
Phil> support.  This patch already exists in the archer-tromey-python branch
Phil> of the archer repository, and here is my submission for the FSF
Phil> repository.

Thanks.

Phil> Index: c-lang.c

[...]

Phil>  /* Obtain a C string from the inferior storing it in a newly allocated

[...]

Phil> +   buffer in BUFFER, which should be freed by the caller.  If LENGTH
Phil> +   is specified at -1, the string is read until a null character of

This should mention *LENGTH, not LENGTH.
I think this comment (and the similar comment in language.h) would be
clearer if it mentioned that LENGTH is both an in- and out-parameter.

Phil> Index: python/python-prettyprint.c
Phil> + /* Apply a pretty-printer for the varobj code.  PRINTER_OBJ is the
Phil> +    print object.  It must have a 'to_string' method (but this is
Phil> +    checked by varobj, not here) which takes no arguments and
Phil> +    returns a string.  The printer will return a value and in the case
Phil> +    of a Python string being returned, this function will return a
Phil> +    PyObject containing the string.  For any other type, *REPLACEMENT is
Phil> +    set to the replacement value and this function returns NULL.  On
Phil> +    error, *REPLACEMENT is set to NULL and this function also returns
Phil> +    NULL.  */

I noticed recently that this entire comment is shifted over by a space.
Could you move it back?

The code bits are approved with those two changes.
It still needs a doc review.

Tom