From: Dmitry Ivankov <divanorama@gmail.com>
Date: Mon, 02 Apr 2012 06:56:40 GMT
Subject: Re: [PATCH] fast-import: catch garbage after marks in from/merge
Message-ID: <CA+gfSn_8J-HzNjLMi2fXn1XQNA9wx3EVuiseq3pjy0nP-odb5A@mail.gmail.com>
In-Reply-To: <20120402001354.GA12651@padd.com>

On Mon, Apr 2, 2012 at 6:13 AM, Pete Wyckoff <pw@padd.com> wrote:
> jrnieder@gmail.com wrote on Sun, 01 Apr 2012 18:12 -0500:
>> Hi Pete,
>>
>> Pete Wyckoff wrote:
>>
>> >     from :1M 100644 :103 hello.c
>> >
>> > It is missing a newline and should be:
>> >
>> >     from :1
>> >     M 100644 :103 hello.c
>>
>> Good idea; thanks.
>>
>> I agree that this at least deserves a warning and probably should
>> error out.
>>
>> [...]
>> > --- a/fast-import.c
>> > +++ b/fast-import.c
>> > @@ -2537,8 +2537,16 @@ static int parse_from(struct branch *b)
>> >             hashcpy(b->branch_tree.versions[0].sha1, t);
>> >             hashcpy(b->branch_tree.versions[1].sha1, t);
>> >     } else if (*from == ':') {
>> > -           uintmax_t idnum = strtoumax(from + 1, NULL, 10);
>> > -           struct object_entry *oe = find_mark(idnum);
>> > +           char *eptr;
>> > +           uintmax_t idnum = strtoumax(from + 1, &eptr, 10);
>> > +           struct object_entry *oe;
>> > +           if (eptr) {
>> > +                   for (; *eptr && isspace(*eptr); eptr++) ;
>> > +                   if (*eptr)
>> > +                           die("Garbage after mark: %s",
>>
>> The implementation seems more complicated than it needs to be.  Why
>> allow whitespace after the mark number?
>
> Fear of breaking existing fast-import users that might happen
> to have stray whitespace, or \r\n terminators.
>
> Other similar fast-import are less forgiving, such as
> parse_cat_blob.  Maybe we should generalize and enforce its
> approach to parsing marks.

Docs say that "fast-import is very strict about its input", so
probably it is ok to both deny trailing spaces and fix all other
strtoumax()-es.

>
>                -- Pete