Re: [PATCH] Allow export to export svn:eol-style=native files with a user suplied eol marker.

Greg Hudson <ghudson@MIT.EDU> writes:

> On Sun, 2004-04-04 at 15:53, Philip Martin wrote:
>> +static void
>> +get_eol_style (svn_subst_eol_style_t *style,
>> + const char **eol,
>> + const char *value,
>> + const char *native_eol)
>> +{
>> + svn_subst_eol_style_from_value (style, eol, value);
>> + if (native_eol && *style == svn_subst_eol_style_native)
>> + svn_subst_eol_style_from_value (NULL, eol, native_eol)
>> +}
>
> Hm. That erroneously accepts "native", but I guess that's no big deal
> if the command-line client is filtering for that.

I already suggested that we allow "native" for consistency ;) If we
did that we drop the literal check in main.c and simply use the result
of svn_subst_eol_style_from_value to determine whether to raise an
error.

> Also, we don't document that NULL is an acceptable first argument to
> svn_subst_eol_style_from_value. We should fix the documentation before
> adding any (more) call sites which rely on that bit of defensiveness.

Or just pass &dummy instead of NULL, none of the other callers pass
NULL. I didn't read the docs, I just came across the function when
grepping for literals like "CR" etc.

-- 
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org