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

Re: [PATCH v2 02/11] bisect--helper: introduce new `write_in_file()` function

From
Junio C Hamano <gitster@pobox.com>
Date
Apr 3, 2020, 05:13 UTC
Message-ID
<xmqqeet51794.fsf@gitster.c.googlers.com>
In-Reply-To
<20200321161020.22817-3-mirucam@gmail.com>
Miriam Rubio <mirucam@gmail.com> writes:
Show 24 quoted lines
> Let's refactor code adding a new `write_in_file()` function
> that opens a file for writing a message and closes it.
>
> This helper will be used in later steps and makes the code
> simpler and easier to understand.
>
> Mentored-by: Christian Couder <chriscool@tuxfamily.org>
> Mentored-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
> Signed-off-by: Miriam Rubio <mirucam@gmail.com>
> ---
>  builtin/bisect--helper.c | 27 +++++++++++++++++++++------
>  1 file changed, 21 insertions(+), 6 deletions(-)
>
> diff --git a/builtin/bisect--helper.c b/builtin/bisect--helper.c
> index 1f81cff1d8..e949ea74e2 100644
> --- a/builtin/bisect--helper.c
> +++ b/builtin/bisect--helper.c
> @@ -74,6 +74,26 @@ static int one_of(const char *term, ...)
>  	return res;
>  }
>  
> +static int write_in_file(const char *filepath, const char *mode, const char *format,...)
> +{
> +	FILE *fp = NULL;

It is crystal clear in this concise helper function that fp will never be used without getting assigned the returned value from fopen(), so I do not think there is any need to initialize it to NULL.

I'd use "path", not "filepath" (which I do not think we use anywhere in our codebase), if I were writing this function, by the way.

Show 5 quoted lines
> +	va_list args;
> +	int res = 0;
> +
> +	if (!strcmp(mode, "a") && !strcmp(mode, "w"))
> +		return error_errno(_("wrong writing mode '%s'"), mode);

I do not see where you saw a failure from a call to system library function, which would make 'errno' variable valid at this point, so I am puzzled. By using error_errno(), whose error status are you trying to show?

Puzzled.  Shouldn't it be just error(_("..."), mode)?
> +	fp = fopen(filepath, mode);
> +	if (!fp)
> +		return error_errno(_("could not open file '%s'"), filepath);

This one would show why fopen() failed, so error_errno() would be good. Does it help us help the users if they can tell us which mode we failed to write to the file? Something like

	cannot open file '%s' in mode '%s'
perhaps?
Show 5 quoted lines
> +	va_start(args, format);
> +	res = vfprintf(fp, format, args);
> +	va_end(args);
> +	if (!res)
> +		return error_errno(_("could not write to file '%s'"), filepath);

This would show errors from vfprintf(), which is good. However, you fail to fclose the FILE when this return is hit, which is not good.

Show 27 quoted lines
> +	return fclose(fp);
> +}
> +
>  static int check_term_format(const char *term, const char *orig_term)
>  {
>  	int res;
> @@ -104,7 +124,6 @@ static int check_term_format(const char *term, const char *orig_term)
>  
>  static int write_terms(const char *bad, const char *good)
>  {
> -	FILE *fp = NULL;
>  	int res;
>  
>  	if (!strcmp(bad, good))
> @@ -113,12 +132,8 @@ static int write_terms(const char *bad, const char *good)
>  	if (check_term_format(bad, "bad") || check_term_format(good, "good"))
>  		return -1;
>  
> -	fp = fopen(git_path_bisect_terms(), "w");
> -	if (!fp)
> -		return error_errno(_("could not open the file BISECT_TERMS"));
> +	res = write_in_file(git_path_bisect_terms(), "w", "%s\n%s\n", bad, good);
>  
> -	res = fprintf(fp, "%s\n%s\n", bad, good);
> -	res |= fclose(fp);
>  	return (res < 0) ? -1 : 0;
>  }
Previous: Miriam RubioNext: Miriam Rubio
Message 7 of 18 in “Finish converting git bisect to C part 2”
  1. 00/11 Finish converting git bisect to C part 2Miriam Rubio, Mar 21, 2020
  2. 01/11 bisect--helper: fix `cmd_*()` function switch default returnMiriam Rubio, Mar 21, 2020
  3. Junio C HamanoApr 3, 2020
  4. Christian CouderApr 3, 2020
  5. Junio C HamanoApr 3, 2020
  6. 02/11 bisect--helper: introduce new `write_in_file()` functionMiriam Rubio, Mar 21, 2020
  7. Junio C HamanoApr 3, 2020
  8. 03/11 bisect--helper: reimplement `bisect_next` and `bisect_auto_next` shell functions in CMiriam Rubio, Mar 21, 2020
  9. Junio C HamanoApr 3, 2020
  10. Miriam R.Apr 23, 2020
  11. 05/11 bisect--helper: retire `--bisect-clean-state` subcommandMiriam Rubio, Mar 21, 2020
  12. 04/11 bisect--helper: finish porting `bisect_start()` to CMiriam Rubio, Mar 21, 2020
  13. 06/11 bisect--helper: retire `--next-all` subcommandMiriam Rubio, Mar 21, 2020
  14. 10/11 bisect--helper: retire `--write-terms` subcommandMiriam Rubio, Mar 21, 2020
  15. 09/11 bisect--helper: retire `--check-expected-revs` subcommandMiriam Rubio, Mar 21, 2020
  16. 07/11 bisect--helper: reimplement `bisect_autostart` shell function in CMiriam Rubio, Mar 21, 2020
  17. 08/11 bisect--helper: reimplement `bisect_state` & `bisect_head` shell functions in CMiriam Rubio, Mar 21, 2020
  18. 11/11 bisect--helper: retire `--bisect-autostart` subcommandMiriam Rubio, Mar 21, 2020

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.