[PATCH 1/2] msvcrt/tests: Add scanf tests for leading and trailing dash in scanset.

[Will Mainio]

Signed-off-by: Will Mainio 
---
 dlls/msvcrt/tests/scanf.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/dlls/msvcrt/tests/scanf.c b/dlls/msvcrt/tests/scanf.c
index 57d9cc0ba12..4b878e6d64e 100644
--- a/dlls/msvcrt/tests/scanf.c
+++ b/dlls/msvcrt/tests/scanf.c
@@ -186,6 +186,14 @@ static void test_sscanf( void )
     ok( ret == 1, "Error with format "%s"\n","%*[a-cd-dg-e]%c");
     ok( buffer[0] == 'h', "Error with "abcefgdh" "%c"\n", buffer[0]);

+    ret = p_sscanf("-123","%[-0-9]s",buffer);
+    ok( ret == 1, "Error with format "%s"\n","%[-0-9]s");
+    ok( strncmp("-123",buffer,strlen(buffer)) == 0, "Error with "-123" "%s"\n",buffer);
+
+    ret = p_sscanf("-321","%[0-9-]s",buffer);
+    ok( ret == 1, "Error with format "%s"\n","%[0-9-]s");
+    ok( strncmp("-321",buffer,strlen(buffer)) == 0, "Error with "-321" "%s"\n",buffer);
+
     buffer1[0] = 'b';
     ret = p_sscanf("a","%s%s", buffer, buffer1);
     ok( ret == 1, "expected 1, got %u\n", ret);

[PATCH 2/2] msvcrt: Fix scanf with leading dash in scanset.

[Will Mainio]

Signed-off-by: Will Mainio 
---
 dlls/msvcrt/scanf.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/dlls/msvcrt/scanf.h b/dlls/msvcrt/scanf.h
index 3b4686c1107..b0254ef9224 100644
--- a/dlls/msvcrt/scanf.h
+++ b/dlls/msvcrt/scanf.h
@@ -639,7 +639,7 @@ _FUNCTION_ {
                     while(*format && (*format != ']')) {
    		/* According to msdn:
    		 * "Note that %[a-z] and %[z-a] are interpreted as equivalent to %[abcde...z]." */
-			if((*format == '-') && (*(format + 1) != ']')) {
+			if((*format == '-') && (*(format - 1) != '[') && (*(format + 1) != ']')) {
    		    if ((*(format - 1)) < *(format + 1))
    			RtlSetBits(&bitMask, *(format - 1) +1 , *(format + 1) - *(format - 1));
    		    else

[Gijs Vermeulen]

Hi Will,

You will either need to reorder your patches or add todo_wine to the ok()'s
that are supposed to fail and remove them in your second patch.

Kind regards,
Gijs

On Fri, Oct 30, 2020, 12:39 Will Mainio  wrote:
> Signed-off-by: Will Mainio
>
> dlls/msvcrt/tests/scanf.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
>
> diff --git a/dlls/msvcrt/tests/scanf.c b/dlls/msvcrt/tests/scanf.c
> index 57d9cc0ba12..4b878e6d64e 100644
> --- a/dlls/msvcrt/tests/scanf.c
> +++ b/dlls/msvcrt/tests/scanf.c
> @@ -186,6 +186,14 @@ static void test_sscanf( void )
>      ok( ret == 1, "Error with format "%s"\n","%*[a-cd-dg-e]%c");
>      ok( buffer[0] == 'h', "Error with "abcefgdh" "%c"\n", buffer[0]);
> - ret = p_sscanf("-123","%[-0-9]s",buffer);
> - ok( ret == 1, "Error with format "%s"\n","%[-0-9]s");
> - ok( strncmp("-123",buffer,strlen(buffer)) == 0, "Error with "-123"
>
> "%s"\n",buffer);
> -
> - ret = p_sscanf("-321","%[0-9-]s",buffer);
> - ok( ret == 1, "Error with format "%s"\n","%[0-9-]s");
> - ok( strncmp("-321",buffer,strlen(buffer)) == 0, "Error with "-321"
>
> "%s"\n",buffer);
> - buffer1[0] = 'b';
>    ret = p_sscanf("a","%s%s", buffer, buffer1);
>    ok( ret == 1, "expected 1, got %u\n", ret);
>
> --
> 2.28.0

[Marvin]

Hi,

While running your changed tests, I think I found new failures.
Being a bot and all I'm not very good at pattern recognition, so I might be
wrong, but could you please double-check?

Full results can be found at:
https://testbot.winehq.org/JobDetails.pl?Key=81265

Your paranoid android.

=== build (build log) ===

error: corrupt patch at line 54
Task: Patch failed to apply

=== debiant (build log) ===

Task: Patch failed to apply

=== debiant (build log) ===

Task: Patch failed to apply

[Will Mainio]

Hi Gijs,

I will reorder the patches and submit a v2. Thanks for the heads up.

Cheers

On Fri, Oct 30, 2020, at 1:00 PM, Gijs Vermeulen wrote:
> Hi Will,
>
> You will either need to reorder your patches or add todo_wine to the ok()'s that are supposed to fail and remove them in your second patch.
>
> Kind regards,
> Gijs
>
> On Fri, Oct 30, 2020, 12:39 Will Mainio  wrote:
> > Signed-off-by: Will Mainio
> >
> > dlls/msvcrt/tests/scanf.c | 8 ++++++++
> >  1 file changed, 8 insertions(+)
> >
> > diff --git a/dlls/msvcrt/tests/scanf.c b/dlls/msvcrt/tests/scanf.c
> > index 57d9cc0ba12..4b878e6d64e 100644
> > --- a/dlls/msvcrt/tests/scanf.c
> > +++ b/dlls/msvcrt/tests/scanf.c
> > @@ -186,6 +186,14 @@ static void test_sscanf( void )
> >      ok( ret == 1, "Error with format "%s"\n","%*[a-cd-dg-e]%c");
> >      ok( buffer[0] == 'h', "Error with "abcefgdh" "%c"\n", buffer[0]);
> > - ret = p_sscanf("-123","%[-0-9]s",buffer);
> > - ok( ret == 1, "Error with format "%s"\n","%[-0-9]s");
> > - ok( strncmp("-123",buffer,strlen(buffer)) == 0, "Error with "-123" "%s"\n",buffer);
> > -
> > - ret = p_sscanf("-321","%[0-9-]s",buffer);
> > - ok( ret == 1, "Error with format "%s"\n","%[0-9-]s");
> > - ok( strncmp("-321",buffer,strlen(buffer)) == 0, "Error with "-321" "%s"\n",buffer);
> > - buffer1[0] = 'b';
> >    ret = p_sscanf("a","%s%s", buffer, buffer1);
> >    ok( ret == 1, "expected 1, got %u\n", ret);
> >
> > -- 
> > 2.28.0

[Piotr Caban]

Hi Will,

There are more problems with current implementation. The '-' character 
can be also added to the set if it's not defining a range even if it's 
not the first or last character. You can see it running following test:
sscanf("123", "%[1-2-4]s", buf);
This format will look for 1-2 range, '-' character and '4' (not for 1-2 
range and 2-4 range).

Thanks,
Piotr

On 10/30/20 1:23 PM, Will Mainio wrote:
> Hi Gijs,
>
> I will reorder the patches and submit a v2. Thanks for the heads up.
>
> Cheers
>
> On Fri, Oct 30, 2020, at 1:00 PM, Gijs Vermeulen wrote:
> > Hi Will,
> >
> > You will either need to reorder your patches or add todo_wine to the ok()'s that are supposed to fail and remove them in your second patch.
> >
> > Kind regards,
> > Gijs
> >
> > On Fri, Oct 30, 2020, 12:39 Will Mainio  wrote:
> > > Signed-off-by: Will Mainio
> > >
> > > dlls/msvcrt/tests/scanf.c | 8 ++++++++
> > >   1 file changed, 8 insertions(+)
> > >
> > > diff --git a/dlls/msvcrt/tests/scanf.c b/dlls/msvcrt/tests/scanf.c
> > > index 57d9cc0ba12..4b878e6d64e 100644
> > > --- a/dlls/msvcrt/tests/scanf.c
> > > +++ b/dlls/msvcrt/tests/scanf.c
> > > @@ -186,6 +186,14 @@ static void test_sscanf( void )
> > >       ok( ret == 1, "Error with format "%s"\n","%*[a-cd-dg-e]%c");
> > >       ok( buffer[0] == 'h', "Error with "abcefgdh" "%c"\n", buffer[0]);
> > > - ret = p_sscanf("-123","%[-0-9]s",buffer);
> > > - ok( ret == 1, "Error with format "%s"\n","%[-0-9]s");
> > > - ok( strncmp("-123",buffer,strlen(buffer)) == 0, "Error with "-123" "%s"\n",buffer);
> > > -
> > > - ret = p_sscanf("-321","%[0-9-]s",buffer);
> > > - ok( ret == 1, "Error with format "%s"\n","%[0-9-]s");
> > > - ok( strncmp("-321",buffer,strlen(buffer)) == 0, "Error with "-321" "%s"\n",buffer);
> > > -
> > > buffer1[0] = 'b';
> > > ret = p_sscanf("a","%s%s", buffer, buffer1);
> > > ok( ret == 1, "expected 1, got %u\n", ret);
> > >
> > > -- 
> > > 2.28.0

[Will Mainio]

Hi Piotr,

unfortunately I am not very familiar with C so I can't implement that but I can add tests with dashes in various locations.
Should that go into a v3 or into a separate patch?

Cheers

On Fri, Oct 30, 2020, at 1:45 PM, Piotr Caban wrote:
> Hi Will,
>
> There are more problems with current implementation. The '-' character 
> can be also added to the set if it's not defining a range even if it's 
> not the first or last character. You can see it running following test:
> sscanf("123", "%[1-2-4]s", buf);
> This format will look for 1-2 range, '-' character and '4' (not for 1-2 
> range and 2-4 range).
>
> Thanks,
> Piotr
>
> On 10/30/20 1:23 PM, Will Mainio wrote:
> > Hi Gijs,
> >
> > I will reorder the patches and submit a v2. Thanks for the heads up.
> >
> > Cheers
> >
> > On Fri, Oct 30, 2020, at 1:00 PM, Gijs Vermeulen wrote:
> > > Hi Will,
> > >
> > > You will either need to reorder your patches or add todo_wine to the ok()'s that are supposed to fail and remove them in your second patch.
> > >
> > > Kind regards,
> > > Gijs
> > >
> > > On Fri, Oct 30, 2020, 12:39 Will Mainio  wrote:
> > > > Signed-off-by: Will Mainio
> > > >
> > > > dlls/msvcrt/tests/scanf.c | 8 ++++++++
> > > >   1 file changed, 8 insertions(+)
> > > >
> > > > diff --git a/dlls/msvcrt/tests/scanf.c b/dlls/msvcrt/tests/scanf.c
> > > > index 57d9cc0ba12..4b878e6d64e 100644
> > > > --- a/dlls/msvcrt/tests/scanf.c
> > > > +++ b/dlls/msvcrt/tests/scanf.c
> > > > @@ -186,6 +186,14 @@ static void test_sscanf( void )
> > > >       ok( ret == 1, "Error with format "%s"\n","%*[a-cd-dg-e]%c");
> > > >       ok( buffer[0] == 'h', "Error with "abcefgdh" "%c"\n", buffer[0]);
> > > > - ret = p_sscanf("-123","%[-0-9]s",buffer);
> > > > - ok( ret == 1, "Error with format "%s"\n","%[-0-9]s");
> > > > - ok( strncmp("-123",buffer,strlen(buffer)) == 0, "Error with "-123" "%s"\n",buffer);
> > > > -
> > > > - ret = p_sscanf("-321","%[0-9-]s",buffer);
> > > > - ok( ret == 1, "Error with format "%s"\n","%[0-9-]s");
> > > > - ok( strncmp("-321",buffer,strlen(buffer)) == 0, "Error with "-321" "%s"\n",buffer);
> > > > -
> > > > buffer1[0] = 'b';
> > > > ret = p_sscanf("a","%s%s", buffer, buffer1);
> > > > ok( ret == 1, "expected 1, got %u\n", ret);
> > > >
> > > > -- 
> > > > 2.28.0

[Will Mainio]

Never mind that, it was simple enough for me to implement.
I submitted a v3 that can handle dashes in any location.

Cheers

On Fri, Oct 30, 2020, at 6:54 PM, Will Mainio wrote:
> Hi Piotr,
>
> unfortunately I am not very familiar with C so I can't implement that 
> but I can add tests with dashes in various locations.
> Should that go into a v3 or into a separate patch?
>
> Cheers
>
> On Fri, Oct 30, 2020, at 1:45 PM, Piotr Caban wrote:
> > Hi Will,
> >
> > There are more problems with current implementation. The '-' character 
> > can be also added to the set if it's not defining a range even if it's 
> > not the first or last character. You can see it running following test:
> > sscanf("123", "%[1-2-4]s", buf);
> > This format will look for 1-2 range, '-' character and '4' (not for 1-2 
> > range and 2-4 range).
> >
> > Thanks,
> > Piotr
> >
> > On 10/30/20 1:23 PM, Will Mainio wrote:
> > > Hi Gijs,
> > >
> > > I will reorder the patches and submit a v2. Thanks for the heads up.
> > >
> > > Cheers
> > >
> > > On Fri, Oct 30, 2020, at 1:00 PM, Gijs Vermeulen wrote:
> > > > Hi Will,
> > > >
> > > > You will either need to reorder your patches or add todo_wine to the ok()'s that are supposed to fail and remove them in your second patch.
> > > >
> > > > Kind regards,
> > > > Gijs
> > > >
> > > > On Fri, Oct 30, 2020, 12:39 Will Mainio  wrote:
> > > > > Signed-off-by: Will Mainio
> > > > >
> > > > > dlls/msvcrt/tests/scanf.c | 8 ++++++++
> > > > >   1 file changed, 8 insertions(+)
> > > > >
> > > > > diff --git a/dlls/msvcrt/tests/scanf.c b/dlls/msvcrt/tests/scanf.c
> > > > > index 57d9cc0ba12..4b878e6d64e 100644
> > > > > --- a/dlls/msvcrt/tests/scanf.c
> > > > > +++ b/dlls/msvcrt/tests/scanf.c
> > > > > @@ -186,6 +186,14 @@ static void test_sscanf( void )
> > > > >       ok( ret == 1, "Error with format "%s"\n","%*[a-cd-dg-e]%c");
> > > > >       ok( buffer[0] == 'h', "Error with "abcefgdh" "%c"\n", buffer[0]);
> > > > > - ret = p_sscanf("-123","%[-0-9]s",buffer);
> > > > > - ok( ret == 1, "Error with format "%s"\n","%[-0-9]s");
> > > > > - ok( strncmp("-123",buffer,strlen(buffer)) == 0, "Error with "-123" "%s"\n",buffer);
> > > > > -
> > > > > - ret = p_sscanf("-321","%[0-9-]s",buffer);
> > > > > - ok( ret == 1, "Error with format "%s"\n","%[0-9-]s");
> > > > > - ok( strncmp("-321",buffer,strlen(buffer)) == 0, "Error with "-321" "%s"\n",buffer);
> > > > > -
> > > > > buffer1[0] = 'b';
> > > > > ret = p_sscanf("a","%s%s", buffer, buffer1);
> > > > > ok( ret == 1, "expected 1, got %u\n", ret);
> > > > >
> > > > > -- 
> > > > > 2.28.0

[Piotr Caban]

Hi Will,

I have sent simplified version of your patch as v4.

Thanks,
Piotr

On 10/31/20 7:13 PM, Will Mainio wrote:
> Never mind that, it was simple enough for me to implement.
> I submitted a v3 that can handle dashes in any location.
>
> Cheers
>
> On Fri, Oct 30, 2020, at 6:54 PM, Will Mainio wrote:
> > Hi Piotr,
> >
> > unfortunately I am not very familiar with C so I can't implement that
> > but I can add tests with dashes in various locations.
> > Should that go into a v3 or into a separate patch?
> >
> > Cheers
> >
> > On Fri, Oct 30, 2020, at 1:45 PM, Piotr Caban wrote:
> > > Hi Will,
> > >
> > > There are more problems with current implementation. The '-' character
> > > can be also added to the set if it's not defining a range even if it's
> > > not the first or last character. You can see it running following test:
> > > sscanf("123", "%[1-2-4]s", buf);
> > > This format will look for 1-2 range, '-' character and '4' (not for 1-2
> > > range and 2-4 range).
> > >
> > > Thanks,
> > > Piotr
> > >
> > > On 10/30/20 1:23 PM, Will Mainio wrote:
> > > > Hi Gijs,
> > > >
> > > > I will reorder the patches and submit a v2. Thanks for the heads up.
> > > >
> > > > Cheers
> > > >
> > > > On Fri, Oct 30, 2020, at 1:00 PM, Gijs Vermeulen wrote:
> > > > > Hi Will,
> > > > >
> > > > > You will either need to reorder your patches or add todo_wine to the ok()'s that are supposed to fail and remove them in your second patch.
> > > > >
> > > > > Kind regards,
> > > > > Gijs
> > > > >
> > > > > On Fri, Oct 30, 2020, 12:39 Will Mainio  wrote:
> > > > > > Signed-off-by: Will Mainio
> > > > > >
> > > > > > dlls/msvcrt/tests/scanf.c | 8 ++++++++
> > > > > >    1 file changed, 8 insertions(+)
> > > > > >
> > > > > > diff --git a/dlls/msvcrt/tests/scanf.c b/dlls/msvcrt/tests/scanf.c
> > > > > > index 57d9cc0ba12..4b878e6d64e 100644
> > > > > > --- a/dlls/msvcrt/tests/scanf.c
> > > > > > +++ b/dlls/msvcrt/tests/scanf.c
> > > > > > @@ -186,6 +186,14 @@ static void test_sscanf( void )
> > > > > >        ok( ret == 1, "Error with format "%s"\n","%*[a-cd-dg-e]%c");
> > > > > >        ok( buffer[0] == 'h', "Error with "abcefgdh" "%c"\n", buffer[0]);
> > > > > > - ret = p_sscanf("-123","%[-0-9]s",buffer);
> > > > > > - ok( ret == 1, "Error with format "%s"\n","%[-0-9]s");
> > > > > > - ok( strncmp("-123",buffer,strlen(buffer)) == 0, "Error with "-123" "%s"\n",buffer);
> > > > > > -
> > > > > > - ret = p_sscanf("-321","%[0-9-]s",buffer);
> > > > > > - ok( ret == 1, "Error with format "%s"\n","%[0-9-]s");
> > > > > > - ok( strncmp("-321",buffer,strlen(buffer)) == 0, "Error with "-321" "%s"\n",buffer);
> > > > > > -
> > > > > >  buffer1[0] = 'b';
> > > > > >  ret = p_sscanf("a","%s%s", buffer, buffer1);
> > > > > >  ok( ret == 1, "expected 1, got %u\n", ret);
> > > > > >
> > > > > > -- 
> > > > > > 2.28.0

[Will Mainio]

Hi Piotr,

that's great, thank you!

Cheers

On Mon, Nov 2, 2020, at 4:02 PM, Piotr Caban wrote:
> Hi Will,
>
> I have sent simplified version of your patch as v4.
>
> Thanks,
> Piotr
>
> On 10/31/20 7:13 PM, Will Mainio wrote:
> > Never mind that, it was simple enough for me to implement.
> > I submitted a v3 that can handle dashes in any location.
> >
> > Cheers
> >
> > On Fri, Oct 30, 2020, at 6:54 PM, Will Mainio wrote:
> > > Hi Piotr,
> > >
> > > unfortunately I am not very familiar with C so I can't implement that
> > > but I can add tests with dashes in various locations.
> > > Should that go into a v3 or into a separate patch?
> > >
> > > Cheers
> > >
> > > On Fri, Oct 30, 2020, at 1:45 PM, Piotr Caban wrote:
> > > > Hi Will,
> > > >
> > > > There are more problems with current implementation. The '-' character
> > > > can be also added to the set if it's not defining a range even if it's
> > > > not the first or last character. You can see it running following test:
> > > > sscanf("123", "%[1-2-4]s", buf);
> > > > This format will look for 1-2 range, '-' character and '4' (not for 1-2
> > > > range and 2-4 range).
> > > >
> > > > Thanks,
> > > > Piotr
> > > >
> > > > On 10/30/20 1:23 PM, Will Mainio wrote:
> > > > > Hi Gijs,
> > > > >
> > > > > I will reorder the patches and submit a v2. Thanks for the heads up.
> > > > >
> > > > > Cheers
> > > > >
> > > > > On Fri, Oct 30, 2020, at 1:00 PM, Gijs Vermeulen wrote:
> > > > > > Hi Will,
> > > > > >
> > > > > > You will either need to reorder your patches or add todo_wine to the ok()'s that are supposed to fail and remove them in your second patch.
> > > > > >
> > > > > > Kind regards,
> > > > > > Gijs
> > > > > >
> > > > > > On Fri, Oct 30, 2020, 12:39 Will Mainio  wrote:
> > > > > > > Signed-off-by: Will Mainio 
> > > > > > > ---
> > > > > > >    dlls/msvcrt/tests/scanf.c | 8 ++++++++
> > > > > > >    1 file changed, 8 insertions(+)
> > > > > > >
> > > > > > > diff --git a/dlls/msvcrt/tests/scanf.c b/dlls/msvcrt/tests/scanf.c
> > > > > > > index 57d9cc0ba12..4b878e6d64e 100644
> > > > > > > --- a/dlls/msvcrt/tests/scanf.c
> > > > > > > +++ b/dlls/msvcrt/tests/scanf.c
> > > > > > > @@ -186,6 +186,14 @@ static void test_sscanf( void )
> > > > > > >        ok( ret == 1, "Error with format "%s"\n","%*[a-cd-dg-e]%c");
> > > > > > >        ok( buffer[0] == 'h', "Error with "abcefgdh" "%c"\n", buffer[0]);
> > > > > > >
> > > > > > > +    ret = p_sscanf("-123","%[-0-9]s",buffer);
> > > > > > > +    ok( ret == 1, "Error with format "%s"\n","%[-0-9]s");
> > > > > > > +    ok( strncmp("-123",buffer,strlen(buffer)) == 0, "Error with "-123" "%s"\n",buffer);
> > > > > > > +
> > > > > > > +    ret = p_sscanf("-321","%[0-9-]s",buffer);
> > > > > > > +    ok( ret == 1, "Error with format "%s"\n","%[0-9-]s");
> > > > > > > +    ok( strncmp("-321",buffer,strlen(buffer)) == 0, "Error with "-321" "%s"\n",buffer);
> > > > > > > +
> > > > > > >        buffer1[0] = 'b';
> > > > > > >        ret = p_sscanf("a","%s%s", buffer, buffer1);
> > > > > > >        ok( ret == 1, "expected 1, got %u\n", ret);
> > > > > > > -- 
> > > > > > > 2.28.0
> > > > > > >
> > > > > > >

[Marvin]

Hi,

While running your changed tests, I think I found new failures.
Being a bot and all I'm not very good at pattern recognition, so I might be
wrong, but could you please double-check?

Full results can be found at:
https://testbot.winehq.org/JobDetails.pl?Key=81262

Your paranoid android.

=== debiant (32 bit report) ===

msvcrt:
scanf.c:190: Test failed: Error with format "%[-0-9]s"

=== debiant (32 bit French report) ===

msvcrt:
scanf.c:190: Test failed: Error with format "%[-0-9]s"

=== debiant (32 bit Japanese:Japan report) ===

msvcrt:
scanf.c:190: Test failed: Error with format "%[-0-9]s"

=== debiant (32 bit Chinese:China report) ===

msvcrt:
scanf.c:190: Test failed: Error with format "%[-0-9]s"

=== debiant (32 bit WoW report) ===

msvcrt:
scanf.c:190: Test failed: Error with format "%[-0-9]s"

=== debiant (64 bit WoW report) ===

msvcrt:
scanf.c:190: Test failed: Error with format "%[-0-9]s"