Re: [PATCH 2/2] hex: drop sha1_to_hex()

From: Junio C Hamano <gitster@pobox.com>
Date: Nov 13, 2019, 01:09 UTC
Message-ID: <xmqqbltgeg58.fsf@gitster-ct.c.googlers.com>
In-Reply-To: <20191112121537.GB12074@sigill.intra.peff.net>

Jeff King <peff@peff.net> writes:

Show 34 quoted lines

> On Tue, Nov 12, 2019 at 08:49:44PM +0900, Junio C Hamano wrote:
>
>> Jeff King <peff@peff.net> writes:
>> 
>> >>     @@ cache.h: int hex_to_bytes(unsigned char *binary, const char *hex, size_t len);
>> >>        * buffers, making it safe to make multiple calls for a single statement, like:
>> >>        *
>> >>      - *   printf("%s -> %s", sha1_to_hex(one), sha1_to_hex(two));
>> >>     -+ *   printf("%s -> %s", oid_to_hex(one), oid_to_hex(two));
>> >>     ++ *   printf("%s -> %s", hash_to_hex(one), hash_to_hex(two));
>> >>        */
>> >>       char *hash_to_hex_algop_r(char *buffer, const unsigned char *hash, const struct git_hash_algo *);
>> >>       char *oid_to_hex_r(char *out, const struct object_id *oid);
>> >
>> > This one-liner leaves the types of "one" and "two" unspecified. :) So
>> > it's not wrong to use hash_to_hex(), but maybe it's better to be pushing
>> > people towards oid_to_hex() as their first choice? It probably doesn't
>> > matter too much either way.
>> 
>> The pre-context of that comment reads:
>> 
>>  * Convert a binary hash to its hex equivalent. The `_r` variant is reentrant,
>>  * and writes the NUL-terminated output to the buffer `out`, which must be at
>>  * least `GIT_MAX_HEXSZ + 1` bytes, and returns a pointer to out for
>>  * convenience.
>> 
>> so I think the intent of the example that used to use sha1_to_hex()
>> has been the raw bytestring that is the object name, not its form
>> that is encapsulated in the "struct object_id()".
>
> I guess you're keying on the phrase "binary hash" there (the
> "GIT_MAX_HEXZ" bits only apply to the "_r" variants anyway). I'd read it
> as encompassing all of the functions below, including oid_to_hex(). But
> I'm OK with it either way.

Yes. "binary hash" is about "unsigned char[]". I think that's historical accident---we added "struct object_id *" variants without updating the comment.

Here is another try. The "everybody uses oid_to_hex" in the log message has also been updated.

1:  8a030f1796 ! 1:  b19f3fe9dd hex: drop sha1_to_hex()
    @@ Metadata
      ## Commit message ##
         hex: drop sha1_to_hex()
     
    -    There's only a single caller left of sha1_to_hex(), since everybody now
    -    uses oid_to_hex() instead. This case is in the sha1dc wrapper, where we
    -    print a hex sha1 when we find a collision. This one will always be sha1,
    -    regardless of the current hash algorithm, so we can't use oid_to_hex()
    -    here. In practice we'd probably not be running sha1 at all if it isn't
    -    the current algorithm, but it's possible we might still occasionally
    +    There's only a single caller left of sha1_to_hex(), since everybody
    +    that has an object name in "unsigned char[]" now uses hash_to_hex()
    +    instead.
    +
    +    This case is in the sha1dc wrapper, where we print a hex sha1 when
    +    we find a collision. This one will always be sha1, regardless of the
    +    current hash algorithm, so we can't use hash_to_hex() here. In
    +    practice we'd probably not be running sha1 at all if it isn't the
    +    current algorithm, but it's possible we might still occasionally
         need to compute a sha1 in a post-sha256 world.
     
         Since sha1_to_hex() is just a wrapper for hash_to_hex_algop(), let's
    @@ Commit message
         it (and as with sha1_to_hex_r() in the previous patch, we'll drop the
         coccinelle transformations, too).
     
    -    The sha1_to_hex() function is mentioned in a comment; we can easily swap
    -    that out for oid_to_hex() to give a better example. It's also mentioned
    -    in some test vectors in t4100, but that's not runnable code, so there's
    -    no point in trying to clean it up.
    +    The sha1_to_hex() function is mentioned in a comment; we can easily
    +    swap that out for oid_to_hex() to give a better example.  Also
    +    update the comment that was left stale when we added "struct
    +    object_id *" as a way to name an object and added functions to
    +    convert it to hex.
    +
    +    The function is also mentioned in some test vectors in t4100, but
    +    that's not runnable code, so there's no point in trying to clean it
    +    up.
     
         Signed-off-by: Jeff King <peff@peff.net>
         Signed-off-by: Junio C Hamano <gitster@pobox.com>
     
      ## cache.h ##
    +@@ cache.h: int get_oid_hex(const char *hex, struct object_id *sha1);
    + int hex_to_bytes(unsigned char *binary, const char *hex, size_t len);
    + 
    + /*
    +- * Convert a binary hash to its hex equivalent. The `_r` variant is reentrant,
    ++ * Convert a binary hash in "unsigned char []" or an object name in
    ++ * "struct object_id *" to its hex equivalent. The `_r` variant is reentrant,
    +  * and writes the NUL-terminated output to the buffer `out`, which must be at
    +  * least `GIT_MAX_HEXSZ + 1` bytes, and returns a pointer to out for
    +  * convenience.
     @@ cache.h: int hex_to_bytes(unsigned char *binary, const char *hex, size_t len);
       * The non-`_r` variant returns a static buffer, but uses a ring of 4
       * buffers, making it safe to make multiple calls for a single statement, like:
       *
     - *   printf("%s -> %s", sha1_to_hex(one), sha1_to_hex(two));
    ++ *   printf("%s -> %s", hash_to_hex(one), hash_to_hex(two));
     + *   printf("%s -> %s", oid_to_hex(one), oid_to_hex(two));
       */
      char *hash_to_hex_algop_r(char *buffer, const unsigned char *hash, const struct git_hash_algo *);