[Libguestfs] [PATCH libguestfs 01/23] generator.ml: define new String-like type: "Device"

[Jim Meyering]

Richard W.M. Jones wrote:

> On Wed, Aug 12, 2009 at 06:52:37PM +0200, Jim Meyering wrote:
>> @@ -4474,7 +4475,8 @@ check_state (guestfs_h *g, const char *caller)
>>         | args ->
>>             List.iter (
>>               function
>> -             | String n ->
>> +	     (* FIXME *)
>> +             | Device n | String n ->
>>                   pr "  args.%s = (char *) %s;\n" n n
>>               | OptString n ->
>>                   pr "  args.%s = %s ? (char **) &%s : NULL;\n" n n n
>
> What was the FIXME here?  Looks OK to me.

Sorry about that.  Those were notes to self to revisit.
They're gone in the final result.
I had already removed a few others from intermediate
patches but didn't look back far enough.
I can remove those as well, if you wish.

>> @@ -4683,7 +4685,8 @@ and generate_daemon_actions () =
>>                  * allow device name translation.  This is safe because
>>                  * we can modify the string (passed from RPC).
>>                  *)
>> -             | String n
>> +	       (* FIXME *)
>> +             | Device n | String n
>>               | OptString n -> pr "  char *%s;\n" n
>>               | StringList n -> pr "  char **%s;\n" n
>>               | Bool n -> pr "  int %s;\n" n
>
> And here?  I guess doing device name translation?
>
>> @@ -6222,9 +6228,11 @@ and generate_prototype ?(extern = true) ?(static = false) ?(semicolon = true)
>>      in
>>      List.iter (
>>        function
>> +      | Device n
>>        | String n
>>        | OptString n ->
>>            next ();
>> +(* FIXME *)
>>            if not in_daemon then pr "const char *%s" n
>>            else pr "char *%s" n
>>        | StringList n ->
>
> Also unclear what the FIXME is here?
>
> - - -
>
> Anyway, the patch looks fine to me.  ACK.
>
> Rich.