[libcamera-devel] [PATCH v4 3/9] test: yaml-parser: Test out-of-range checks on integer parsing

[Laurent Pinchart]

Hi Paul,

On Thu, Aug 18, 2022 at 09:21:01PM +0900, paul.elder at ideasonboard.com wrote:
> On Tue, Aug 16, 2022 at 04:54:08AM +0300, Laurent Pinchart via libcamera-devel wrote:
> > Add 16-bit integer parsing tests, including a test to verify the
> > out-of-range checks when parsing 32-bit integers as 16-bit values. That
> 
> I can't find the test for the out-of-range checks...? Or is it not added
> because it fails?

See below...

> > test currently fails.
> > 
> > Signed-off-by: Laurent Pinchart <laurent.pinchart at ideasonboard.com>
> > ---
> >  test/yaml-parser.cpp | 77 ++++++++++++++++++++++++++++++++++++++------
> >  1 file changed, 68 insertions(+), 9 deletions(-)
> > 
> > diff --git a/test/yaml-parser.cpp b/test/yaml-parser.cpp
> > index 803e70beb782..28f8cc8822b3 100644
> > --- a/test/yaml-parser.cpp
> > +++ b/test/yaml-parser.cpp
> > @@ -24,8 +24,10 @@ using namespace std;
> >  static const string testYaml =
> >  	"string: libcamera\n"
> >  	"double: 3.14159\n"
> > -	"uint32_t: 100\n"
> > -	"int32_t: -100\n"
> > +	"int16_t: -1000\n"
> > +	"uint16_t: 1000\n"
> > +	"int32_t: -100000\n"
> > +	"uint32_t: 100000\n"
> >  	"size: [1920, 1080]\n"
> >  	"list:\n"
> >  	"  - James\n"
> > @@ -74,6 +76,8 @@ protected:
> >  
> >  	enum class Type {
> >  		String,
> > +		Int16,
> > +		UInt16,
> >  		Int32,
> >  		UInt32,
> >  		Double,
> > @@ -86,8 +90,10 @@ protected:
> >  	{
> >  		bool isList = type == Type::List || type == Type::Size;
> >  		bool isScalar = !isList && type != Type::Dictionary;
> > -		bool isInteger = type == Type::Int32 || type == Type::UInt32;
> > -		bool isSigned = type == Type::Int32;
> > +		bool isInteger16 = type == Type::Int16 || type == Type::UInt16;
> > +		bool isInteger32 = type == Type::Int32 || type == Type::UInt32;
> > +		bool isInteger = isInteger16 || isInteger32;
> > +		bool isSigned = type == Type::Int16 || type == Type::Int32;
> >  
> >  		if ((isScalar && !obj.isValue()) || (!isScalar && obj.isValue())) {
> >  			std::cerr
> > @@ -118,6 +124,20 @@ protected:
> >  			return TestFail;
> >  		}
> >  
> > +		if (!isInteger16 && obj.get<int16_t>()) {
> > +			std::cerr
> > +				<< "Object " << name << " didn't fail to parse as "
> > +				<< "int16_t" << std::endl;
> > +			return TestFail;
> > +		}

Here. This fails with the int32_t and uint32_t, obj.get<int16_t>()
should return nullopt in those cases, and it doesn't.

> > +
> > +		if ((!isInteger16 || isSigned) && obj.get<uint16_t>()) {
> > +			std::cerr
> > +				<< "Object " << name << " didn't fail to parse as "
> > +				<< "uint16_t" << std::endl;
> > +			return TestFail;
> > +		}
> > +
> >  		if (!isInteger && obj.get<int32_t>()) {
> >  			std::cerr
> >  				<< "Object " << name << " didn't fail to parse as "
> > @@ -154,7 +174,8 @@ protected:
> >  	{
> >  		uint64_t unsignedValue = static_cast<uint64_t>(value);
> >  		std::string strValue = std::to_string(value);
> > -		bool isSigned = type == Type::Int32;
> > +		bool isInteger16 = type == Type::Int16 || type == Type::UInt16;
> > +		bool isSigned = type == Type::Int16 || type == Type::Int32;
> >  
> >  		/* All integers can be parsed as strings or double. */
> >  
> > @@ -174,6 +195,26 @@ protected:
> >  			return TestFail;
> >  		}
> >  
> > +		if (isInteger16) {
> > +			if (obj.get<int16_t>().value_or(0) != value ||
> > +			    obj.get<int16_t>(0) != value) {
> > +				std::cerr
> > +					<< "Object " << name << " failed to parse as "
> > +					<< "int16_t" << std::endl;
> > +				return TestFail;
> > +			}
> > +		}
> > +
> > +		if (isInteger16 && !isSigned) {
> > +			if (obj.get<uint16_t>().value_or(0) != unsignedValue ||
> > +			    obj.get<uint16_t>(0) != unsignedValue) {
> > +				std::cerr
> > +					<< "Object " << name << " failed to parse as "
> > +					<< "uint16_t" << std::endl;
> > +				return TestFail;
> > +			}
> > +		}
> > +
> >  		if (obj.get<int32_t>().value_or(0) != value ||
> >  		    obj.get<int32_t>(0) != value) {
> >  			std::cerr
> > @@ -231,8 +272,8 @@ protected:
> >  		}
> >  
> >  		std::vector<const char *> rootElemNames = {
> > -			"string", "double", "int32_t", "uint32_t", "size",
> > -			"list", "dictionary", "level1",
> > +			"string", "double", "int16_t", "uint16_t", "int32_t",
> > +			"uint32_t", "size", "list", "dictionary", "level1",
> >  		};
> >  
> >  		for (const char *name : rootElemNames) {
> > @@ -255,13 +296,31 @@ protected:
> >  			return TestFail;
> >  		}
> >  
> > +		/* Test int16_t object */
> > +		auto &int16Obj = (*root)["int16_t"];
> > +
> > +		if (testObjectType(int16Obj, "int16_t", Type::Int16) != TestPass)
> > +			return TestFail;
> > +
> > +		if (testIntegerObject(int16Obj, "int16_t", Type::Int16, -1000) != TestPass)
> > +			return TestFail;
> > +
> > +		/* Test uint16_t object */
> > +		auto &uint16Obj = (*root)["uint16_t"];
> > +
> > +		if (testObjectType(uint16Obj, "uint16_t", Type::UInt16) != TestPass)
> > +			return TestFail;
> > +
> > +		if (testIntegerObject(uint16Obj, "uint16_t", Type::UInt16, 1000) != TestPass)
> > +			return TestFail;
> > +
> >  		/* Test int32_t object */
> >  		auto &int32Obj = (*root)["int32_t"];
> >  
> >  		if (testObjectType(int32Obj, "int32_t", Type::Int32) != TestPass)
> >  			return TestFail;
> >  
> > -		if (testIntegerObject(int32Obj, "int32_t", Type::Int32, -100) != TestPass)
> > +		if (testIntegerObject(int32Obj, "int32_t", Type::Int32, -100000) != TestPass)
> >  			return TestFail;
> >  
> >  		/* Test uint32_t object */
> > @@ -270,7 +329,7 @@ protected:
> >  		if (testObjectType(uint32Obj, "uint32_t", Type::UInt32) != TestPass)
> >  			return TestFail;
> >  
> > -		if (testIntegerObject(uint32Obj, "uint32_t", Type::UInt32, 100) != TestPass)
> > +		if (testIntegerObject(uint32Obj, "uint32_t", Type::UInt32, 100000) != TestPass)
> >  			return TestFail;
> >  
> >  		/* Test double value */

-- 
Regards,

Laurent Pinchart