Programming language exercises (I)

Can be in here Download Books, notes and answers .
welcome star/fork To participate in My compiler project

1.1

With C++ Language as an example ：

a. Lexical errors

int a=$; //$ Report errors 

b. Grammar mistakes

int c = .+ a; //. Wrong number 

c. Static semantic error

int a=1;
int c=a+b;//b No definition 

d. Dynamic semantic error

Access rights error , You may not report an error , But actually C++：“ Generate some code , Perform appropriate checks at runtime … The implementation of some rules may bring high 、 An unacceptable price , The implementation of language may not check them at all ”, in other words , This procedure " It's simply the wrong program "
The compiler is just as far as possible , Some dynamic semantic errors are detected , Static semantic errors can be detected . How much can be detected , It depends on the implementation of the language .

class Object
{
public:
	int a;
};

int main() {
	Object *b = (Object*)new Object;
	delete(b);
	cout << (*b).a; 
}

e. Errors that violate language definitions

such , Instead of calling destructors, just free memory , The result will be an error that violates the language definition without the compiler running and reporting an error .

class Object
{
public:
	int a;
};

int main() {
	Object *b = (Object*)new Object;
	free(b);
	cout << (*b).a;
}

1.2

Compiled and interpreted languages have their own advantages . Interpretive languages have better performance , Compiled languages can bring more flexibility , And it can make a better diagnosis of the program .

1.3

1.4

Pascal division algorithm

The time complexity is more reduced and higher .

1.5

make Dependency management will lead to some unnecessary compilation
such as ：
Modification of local variables ; Local symbol table changes but does not affect other documents ; Character changes but no word changes, etc

1.6

lint Check C Potential errors in the program , Include （ But not limited to, ） Suspicious type combinations 、 Unused variables 、 Unreachable code and non portable code .lint It will produce a series of diagnostic information that programmers need to read carefully from beginning to end . Use lint Are the benefits of ：1. It can check for errors that have been missed by the compiler ; 2. You can associate many files for error checking and code analysis , It has strong flexibility

1.8

I hope c++ Of static Assignment can be more convenient
But because c++ Of static Shared by all objects , Can only assign values outside the class
If you assign a value in the class , It is equivalent to changing this value every time you create a new object , It will definitely increase the complexity

1.9

Report errors
No,
If the semantic check of array out of bounds is added , Will increase O(n) Time complexity of .

1.11

Because we have to judge grammar , semantics , Whether it is right , Also check the boundary and access rights .
What can be found is abnormal errors , Cross the border, for example , Null pointer , overflow ; And logical errors cannot be found .