git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH 4/7] add object_as_type helper for casting objects

From
Jeff King <peff@peff.net>
Date
Jul 11, 2014, 16:59 UTC
Message-ID
<20140711165956.GA7856@sigill.intra.peff.net>
In-Reply-To
<53BFC066.2090102@ramsay1.demon.co.uk>
On Fri, Jul 11, 2014 at 11:45:58AM +0100, Ramsay Jones wrote:
Show 15 quoted lines
> > @@ -1729,9 +1729,8 @@ static enum peel_status peel_object(const unsigned char *name, unsigned char *sh
> >  
> >  	if (o->type == OBJ_NONE) {
> >  		int type = sha1_object_info(name, NULL);
> > -		if (type < 0)
> > +		if (type < 0 || !object_as_type(o, type, 0))
> --------------------------------------------------------^^^
> 
> It is not possible here for object_as_type() to issue an error()
> report, but I had to go look at the code to check. (It would have
> been a change in behaviour, otherwise). So, it doesn't really matter
> what you pass to the quiet argument, but setting it to 1 _may_ help
> the next reader. (No, I'm not convinced either; the only reason I
> knew it had anything to do with error messages was because I had
> just read the code ...) Hmm, dunno.

Right, as I mentioned in the commit message, the type-check part of object_type is a noop here. In that sense you could write this as just:

  object_as_type(o, type. 1);

and ignore the return value. However, I'd rather err on the side of checking for and reporting the error, even if we expect it to do nothing right now. That decouples the two functions, and if object_as_type ever learns to report other errors, then we automatically do the right thing here.

-Peff
Previous: Ramsay JonesNext: Jeff King
Message 31 of 35 in “alloc.c: remove the redundant commit_count variable”
  1. 2/2 alloc.c: remove the redundant commit_count variableRamsay Jones, Jul 10, 2014
  2. Jeff KingJul 11, 2014
  3. Ramsay JonesJul 11, 2014
  4. Jeff KingJul 11, 2014
  5. Ramsay JonesJul 11, 2014
  6. 0/7 ensure index is set for all OBJ_COMMIT objects variableJeff King, Jul 11, 2014
  7. 1/7 alloc.c: remove the alloc_raw_commit_node() functionJeff King, Jul 11, 2014
  8. 2/7 move setting of object->type to alloc_* functionsJeff King, Jul 11, 2014
  9. Ramsay JonesJul 12, 2014
  10. Jeff KingJul 12, 2014
  11. Jeff KingJul 13, 2014
  12. 1/8 alloc.c: remove the alloc_raw_commit_node() functionJeff King, Jul 13, 2014
  13. Junio C HamanoJul 15, 2014
  14. 2/8 alloc: write out allocator definitionsJeff King, Jul 13, 2014
  15. Junio C HamanoJul 15, 2014
  16. 3/8 move setting of object->type to alloc_* functionsJeff King, Jul 13, 2014
  17. Junio C HamanoJul 15, 2014
  18. 4/8 parse_object_buffer: do not set object typeJeff King, Jul 13, 2014
  19. 5/8 add object_as_type helper for casting objectsJeff King, Jul 13, 2014
  20. 6/8 alloc: factor out commit indexJeff King, Jul 13, 2014
  21. 7/8 object_as_type: set commit indexJeff King, Jul 13, 2014
  22. 8/8 diff-tree: avoid lookup_unknown_objectJeff King, Jul 13, 2014
  23. Ramsay JonesJul 13, 2014
  24. Jeff KingJul 14, 2014
  25. Ramsay JonesJul 14, 2014
  26. Ramsay JonesJul 12, 2014
  27. Jeff KingJul 12, 2014
  28. 3/7 parse_object_buffer: do not set object typeJeff King, Jul 11, 2014
  29. 4/7 add object_as_type helper for casting objectsJeff King, Jul 11, 2014
  30. Ramsay JonesJul 11, 2014
  31. Jeff KingJul 11, 2014
  32. 5/7 alloc: factor out commit indexJeff King, Jul 11, 2014
  33. 6/7 object_as_type: set commit indexJeff King, Jul 11, 2014
  34. 7/7 diff-tree: avoid lookup_unknown_objectJeff King, Jul 11, 2014
  35. Ramsay JonesJul 11, 2014

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.