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

Re: [PATCH] rebase -x: sanity check command

From
Johannes Schindelin <johannes.schindelin@gmx.de>
Date
Jan 29, 2019, 15:32 UTC
Message-ID
<nycvar.QRO.7.76.6.1901291632070.41@tvgsbejvaqbjf.bet>
In-Reply-To
<13f14cc6-e437-3f18-7f24-1d6edda1cb23@talktalk.net>
Hi Phillip,

[sorry for the double-send, I dropped the Cc: list by mistake the first time I replied]

On Tue, 29 Jan 2019, Phillip Wood wrote:
Show 20 quoted lines
> On 28/01/2019 22:03, Johannes Schindelin wrote:
> > Hi Phillip,
> > 
> > On Mon, 28 Jan 2019, Phillip Wood wrote:
> > 
> >> From: Phillip Wood <phillip.wood@dunelm.org.uk>
> >>
> >> If the user gives an empty argument to --exec then the rebase starts to
> >> run before erroring out with
> >>
> >>   error: missing arguments for exec
> >>   error: invalid line 2: exec
> >>   You can fix this with 'git rebase --edit-todo' and then run 'git rebase --continue'.
> >>   Or you can abort the rebase with 'git rebase --abort'.
> > 
> > And that's the same if you specify an incorrect command.
> 
> Not quite, the issue is that the todo list is invalid, not that the
> command fails - it never gets that far. Those errors are coming from
> parse_insn_line() and parse_insn_buffer().
Ah! I misunderstood "then the rebase starts to run" part.
So please let me withdraw my objections to catch this error.

However, I still would highly recommend to use `strspn()` to simplify the code.

Thanks, Dscho

Show 120 quoted lines
> > In both cases, I would probably heed the second line of the advice: git
> > rebase --abort.
> > 
> >> Instead check for empty commands before starting the rebase.
> >>
> >> Also check that the command does not contain any newlines as the
> >> todo-list format is unable to cope with multiline commands. Note that
> >> this changes the behavior, before this change one could do
> >>
> >> git rebase --exec='echo one
> >> exec echo two'
> >>
> >> and it would insert two exec lines in the todo list, now it will error
> >> out.
> > 
> > This, however, makes a ton of sense to me.
> > 
> >> diff --git a/builtin/rebase.c b/builtin/rebase.c
> >> index 00de70365e..b6c54b03c1 100644
> >> --- a/builtin/rebase.c
> >> +++ b/builtin/rebase.c
> >> @@ -793,6 +793,24 @@ static void set_reflog_action(struct rebase_options *options)
> >>  	strbuf_release(&buf);
> >>  }
> >>  
> >> +static int check_exec_cmd(const char *cmd)
> >> +{
> >> +	int non_blank = 0;
> >> +
> >> +	while (*cmd) {
> >> +		if (*cmd == '\n')
> >> +			return error(_("exec commands cannot contain newlines"));
> >> +		if (!isspace(*cmd))
> >> +			non_blank = 1;
> >> +		cmd++;
> >> +	}
> >> +
> >> +	if (non_blank)
> >> +		return 0;
> > 
> > We are not in a performance critical path here, so I would prefer the
> > readability of this code:
> > 
> > 	if (strchr(cmd, '\n'))
> > 		return error(...);
> > 
> > And if you *really* must,
> > 
> > 	/* Does the command consist purely of whitespace? */
> > 	if (!cmd[strspn(cmd, " \t\r\n")])
> > 		return error(...);
> > 
> > But as I suggested also in a reply to Junio's answer: where would we stop
> > to validate the commands?
> 
> I'm not trying to validate the command (and I don't think we can/should)
> - just generate a todo list that can be parsed.
> 
> Best Wishes
> 
> Phillip
> 
> > 
> > Ciao,
> > Dscho
> > 
> >> +
> >> +	return error(_("empty exec command"));
> >> +}
> >> +
> >>  int cmd_rebase(int argc, const char **argv, const char *prefix)
> >>  {
> >>  	struct rebase_options options = {
> >> @@ -1130,6 +1148,10 @@ int cmd_rebase(int argc, const char **argv, const char *prefix)
> >>  		}
> >>  	}
> >>  
> >> +	for (i = 0; i < exec.nr; i++)
> >> +		if (check_exec_cmd(exec.items[i].string))
> >> +			exit(1);
> >> +
> >>  	if (!(options.flags & REBASE_NO_QUIET))
> >>  		argv_array_push(&options.git_am_opts, "-q");
> >>  
> >> diff --git a/t/t3404-rebase-interactive.sh b/t/t3404-rebase-interactive.sh
> >> index 7a440e08d8..c98f64eb2d 100755
> >> --- a/t/t3404-rebase-interactive.sh
> >> +++ b/t/t3404-rebase-interactive.sh
> >> @@ -147,6 +147,25 @@ test_expect_success 'rebase -i with the exec command checks tree cleanness' '
> >>  	git rebase --continue
> >>  '
> >>  
> >> +test_expect_success 'rebase -x with empty command fails' '
> >> +	test_when_finished "git rebase --abort ||:" &&
> >> +	test_must_fail git rebase -x "" @ 2>actual &&
> >> +	test_write_lines "error: empty exec command" >expected &&
> >> +	test_i18ncmp expected actual &&
> >> +	test_must_fail git rebase -x " " @ 2>actual &&
> >> +	test_i18ncmp expected actual
> >> +'
> >> +
> >> +LF='
> >> +'
> >> +test_expect_success 'rebase -x with newline in command fails' '
> >> +	test_when_finished "git rebase --abort ||:" &&
> >> +	test_must_fail git rebase -x "a${LF}b" @ 2>actual &&
> >> +	test_write_lines "error: exec commands cannot contain newlines" \
> >> +			 >expected &&
> >> +	test_i18ncmp expected actual
> >> +'
> >> +
> >>  test_expect_success 'rebase -i with exec of inexistent command' '
> >>  	git checkout master &&
> >>  	test_when_finished "git rebase --abort" &&
> >> -- 
> >> 2.20.1
> >>
> >>
> 
> 
Previous: Phillip WoodNext: Phillip Wood
Message 8 of 31 in “rebase -x: sanity check command”
  1. rebase -x: sanity check commandPhillip Wood, Jan 28, 2019
  2. Junio C HamanoJan 28, 2019
  3. Johannes SchindelinJan 28, 2019
  4. Phillip WoodJan 29, 2019
  5. Johannes SchindelinJan 29, 2019
  6. Johannes SchindelinJan 28, 2019
  7. Phillip WoodJan 29, 2019
  8. Johannes SchindelinJan 29, 2019
  9. rebase -x: sanity check commandPhillip Wood, Jan 29, 2019
  10. Junio C HamanoJan 29, 2019
  11. Johannes SchindelinJan 30, 2019
  12. Ævar Arnfjörð BjarmasonFeb 13, 2019
  13. rebase: remove the rebase.useBuiltin settingÆvar Arnfjörð Bjarmason, Feb 13, 2019
  14. Johannes SchindelinFeb 13, 2019
  15. Junio C HamanoFeb 13, 2019
  16. rebase: fix regression in rebase.useBuiltin=false test modeÆvar Arnfjörð Bjarmason, Feb 13, 2019
  17. Junio C HamanoFeb 13, 2019
  18. Phillip WoodFeb 14, 2019
  19. rebase: remove the rebase.useBuiltin settingÆvar Arnfjörð Bjarmason, Mar 14, 2019
  20. Johannes SchindelinMar 14, 2019
  21. Ævar Arnfjörð BjarmasonMar 14, 2019
  22. rebase: remove the rebase.useBuiltin settingÆvar Arnfjörð Bjarmason, Mar 15, 2019
  23. Johannes SchindelinMar 15, 2019
  24. Ævar Arnfjörð BjarmasonMar 15, 2019
  25. Junio C HamanoMar 18, 2019
  26. Phillip WoodMar 18, 2019
  27. rebase: remove the rebase.useBuiltin settingÆvar Arnfjörð Bjarmason, Mar 18, 2019
  28. Phillip WoodMar 19, 2019
  29. rebase: remove transitory rebase.useBuiltin setting & envÆvar Arnfjörð Bjarmason, Mar 23, 2021
  30. Junio C HamanoMar 23, 2021
  31. Johannes SchindelinMar 23, 2021

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.