[PATCH] pathspec: give better message for submodule related pathspec error

[PATCH] pathspec: give better message for submodule related pathspec error

[Stefan Beller]

Every once in a while someone complains to the mailing list to have
run into this weird assertion[1].

The usual response from the mailing list is link to old discussions[2],
and acknowledging the problem stating it is known.

For now just improve the user visible error message.

[1] https://www.google.com/search?q=item-%3Enowildcard_len
[2] http://git.661346.n2.nabble.com/assert-failed-in-submodule-edge-case-td7628687.html
    https://www.spinics.net/lists/git/msg249473.html

Signed-off-by: Stefan Beller <sbeller@google.com>
---

If you were following the mailing list closely today, you may sense
that I am cleaning up stalled branches. :)

I think such a hot fix is warranted given how often we had reports
on the mailing list.

Thanks,
Stefan

 pathspec.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/pathspec.c b/pathspec.c
index 22ca74a126..d522f43331 100644
--- a/pathspec.c
+++ b/pathspec.c
@@ -313,8 +313,11 @@ static unsigned prefix_pathspec(struct pathspec_item *item,
 	}
 
 	/* sanity checks, pathspec matchers assume these are sane */
-	assert(item->nowildcard_len <= item->len &&
-	       item->prefix         <= item->len);
+	if (item->nowildcard_len <= item->len &&
+	    item->prefix         <= item->len)
+		die (_("Path leads inside submodule '%s', but the submodule "
+		       "was not recognized, i.e. not initialized or deleted"),
+		       ce->name);
 	return magic;
 }
 
-- 
2.11.0.196.gee862f456e.dirty

Re: [PATCH] pathspec: give better message for submodule related pathspec error

[Jeff King]

On Tue, Dec 27, 2016 at 04:05:59PM -0800, Stefan Beller wrote:

> Every once in a while someone complains to the mailing list to have
> run into this weird assertion[1].

If people are running into it, it definitely should not be an assert,
nor a die("BUG"). It should be a regular die(), which your patch does.
So this is definitely a good step, even if the ultimate goal may be to
handle the case more gracefully (I say that without having even read the
background, or knowing what the right handling would be).

But...

> diff --git a/pathspec.c b/pathspec.c
> index 22ca74a126..d522f43331 100644
> --- a/pathspec.c
> +++ b/pathspec.c
> @@ -313,8 +313,11 @@ static unsigned prefix_pathspec(struct pathspec_item *item,
>  	}
>  
>  	/* sanity checks, pathspec matchers assume these are sane */
> -	assert(item->nowildcard_len <= item->len &&
> -	       item->prefix         <= item->len);
> +	if (item->nowildcard_len <= item->len &&
> +	    item->prefix         <= item->len)
> +		die (_("Path leads inside submodule '%s', but the submodule "
> +		       "was not recognized, i.e. not initialized or deleted"),
> +		       ce->name);

Don't you need to flip the logic here? An assert() triggers when the
condition is not true, but an "if" does the opposite. So "assert(X)"
should always become "if (!X) die(...)".

-Peff

[PATCHv2] pathspec: give better message for submodule related pathspec error

[Stefan Beller]

Every once in a while someone complains to the mailing list to have
run into this weird assertion[1].

The usual response from the mailing list is link to old discussions[2],
and acknowledging the problem stating it is known.

For now just improve the user visible error message.

[1] https://www.google.com/search?q=item-%3Enowildcard_len
[2] http://git.661346.n2.nabble.com/assert-failed-in-submodule-edge-case-td7628687.html
    https://www.spinics.net/lists/git/msg249473.html

Signed-off-by: Stefan Beller <sbeller@google.com>
---

Peff wrote:
> Don't you need to flip the logic here? An assert() triggers when the
> condition is not true, but an "if" does the opposite. So "assert(X)"
> should always become "if (!X) die(...)".

Duh! and it should compile as well. 

Thanks,
Stefan

 pathspec.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/pathspec.c b/pathspec.c
index 22ca74a126..4724d522f2 100644
--- a/pathspec.c
+++ b/pathspec.c
@@ -313,8 +313,11 @@ static unsigned prefix_pathspec(struct pathspec_item *item,
 	}
 
 	/* sanity checks, pathspec matchers assume these are sane */
-	assert(item->nowildcard_len <= item->len &&
-	       item->prefix         <= item->len);
+	if (item->nowildcard_len > item->len ||
+	    item->prefix         > item->len)
+		die (_("Path leads inside submodule '%s', but the submodule "
+		       "was not recognized, i.e. not initialized or deleted"),
+		       item->original);
 	return magic;
 }
 
-- 
2.11.0.196.gee862f456e.dirty

Re: [PATCHv2] pathspec: give better message for submodule related pathspec error

[Brandon Williams]

On 12/28, Stefan Beller wrote:
> Every once in a while someone complains to the mailing list to have
> run into this weird assertion[1].
> 
> The usual response from the mailing list is link to old discussions[2],
> and acknowledging the problem stating it is known.
> 
> For now just improve the user visible error message.
> 
> [1] https://www.google.com/search?q=item-%3Enowildcard_len
> [2] http://git.661346.n2.nabble.com/assert-failed-in-submodule-edge-case-td7628687.html
>     https://www.spinics.net/lists/git/msg249473.html
> 
> Signed-off-by: Stefan Beller <sbeller@google.com>
> ---
> 
> Peff wrote:
> > Don't you need to flip the logic here? An assert() triggers when the
> > condition is not true, but an "if" does the opposite. So "assert(X)"
> > should always become "if (!X) die(...)".
> 
> Duh! and it should compile as well. 
> 
> Thanks,
> Stefan
> 
>  pathspec.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/pathspec.c b/pathspec.c
> index 22ca74a126..4724d522f2 100644
> --- a/pathspec.c
> +++ b/pathspec.c
> @@ -313,8 +313,11 @@ static unsigned prefix_pathspec(struct pathspec_item *item,
>  	}
>  
>  	/* sanity checks, pathspec matchers assume these are sane */
> -	assert(item->nowildcard_len <= item->len &&
> -	       item->prefix         <= item->len);
> +	if (item->nowildcard_len > item->len ||
> +	    item->prefix         > item->len)
> +		die (_("Path leads inside submodule '%s', but the submodule "
> +		       "was not recognized, i.e. not initialized or deleted"),
> +		       item->original);
>  	return magic;
>  }

Turns out I should comment on the most recent version of the patch :P
This looks better to me. (It resolves the issue with using a variable
not in scope).

-- 
Brandon Williams

Re: [PATCH] pathspec: give better message for submodule related pathspec error

[Brandon Williams]

On 12/27, Stefan Beller wrote:
> Every once in a while someone complains to the mailing list to have
> run into this weird assertion[1].
> 
> The usual response from the mailing list is link to old discussions[2],
> and acknowledging the problem stating it is known.
> 
> For now just improve the user visible error message.
> 
> [1] https://www.google.com/search?q=item-%3Enowildcard_len
> [2] http://git.661346.n2.nabble.com/assert-failed-in-submodule-edge-case-td7628687.html
>     https://www.spinics.net/lists/git/msg249473.html
> 
> Signed-off-by: Stefan Beller <sbeller@google.com>
> ---
> 
> If you were following the mailing list closely today, you may sense
> that I am cleaning up stalled branches. :)
> 
> I think such a hot fix is warranted given how often we had reports
> on the mailing list.
> 
> Thanks,
> Stefan
> 
>  pathspec.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/pathspec.c b/pathspec.c
> index 22ca74a126..d522f43331 100644
> --- a/pathspec.c
> +++ b/pathspec.c
> @@ -313,8 +313,11 @@ static unsigned prefix_pathspec(struct pathspec_item *item,
>  	}
>  
>  	/* sanity checks, pathspec matchers assume these are sane */
> -	assert(item->nowildcard_len <= item->len &&
> -	       item->prefix         <= item->len);
> +	if (item->nowildcard_len <= item->len &&
> +	    item->prefix         <= item->len)
> +		die (_("Path leads inside submodule '%s', but the submodule "
> +		       "was not recognized, i.e. not initialized or deleted"),
> +		       ce->name);
>  	return magic;

I haven't been following everything on the list these past couple days,
but are we sure this is caused by submodules?  Also variable 'ce'
shouldn't be in scope here.

-- 
Brandon Williams