Assignment 1 -Exception handling and coroutines in µC++

Starting from:

$27

Assignment 1
This assignment introduces exception handling and coroutines in µC++. Use it to become familiar with these new facilities, and ensure you use these concepts in your assignment solution, i.e., writing a C-style solution for questions is unacceptable, and will receive little or no marks. (You may freely use the code from these example programs.)
1. Exceptfor the code handlingthe command-linearguments,transform the C++ programin Figure 1 from dynamic multi-level exits to:
(a) NO break/throw, one return per routine, return codes, and ﬂag variables. The return type of routines may be changed and may be a structure.
(b) break, multiple return, return codes, and NO ﬂag variables. The return type of routines may be changed and may be a structure.
(c) global status-ﬂag, break, multiple return, and NO ﬂag variables. The return type of routines may NOT be changed.
Output from the transformed programs must be identical to the original program.
2. (a) Except for the code handling the command-line arguments, transform the program in Figure 2, p. 3 replacing throw/catch with longjmp/setjmp. No additional parameters may be added to routine Ackermann. No dynamic allocation is allowed, but creation of a global variable is allowed. Note, type jmp buf is an array allowing instances to be passed to setjmp/longjmp without having to take the address of the argument. Output from the transformed program must be identical to the original program, except for one aspect, which you will discover in the transformed program.
(b) i. Explain why the output is not the same between the original and transformed program. ii. Compare the original and transformed programs with respect to performance by doing the following: • Recompile both the programs with preprocessor option -DNOOUTPUT to suppress output. • Time the executions using the time command: $ /usr/bin/time -f "%Uu %Ss %E" ./a.out 3.21u 0.02s 0:03.32
(Output from time differs depending on the shell, so use the system time command.) Compare the user time (3.21u) only, which is the CPU time consumed solely by the execution of user code (versus system and real time). • Use the program command-line arguments (as necessary) to adjust program execution into the range 1 to 100 seconds. (Timing results below 1 second are inaccurate.) Use the same commandline values for all experiments, if possible; otherwise, increase/decrease the arguments as necessary and scale the difference in the answer. • Run both the experimentsagain after recompilingthe programswith compiler optimizationturned on (i.e., compiler ﬂag -O2). • Include 4 timing results to validate the experiments. iii. State the performance difference (larger/smaller/by how much) between the original and transformed programs, and what caused the difference. iv. State the performance difference (larger/smaller/by how much) between the original and transformed programs when compiler optimization is used.
1
CS 343 - Assignment 1
2
#include <iostream #include <cstdlib // access: rand, srand using namespace std; #include <unistd.h // access: getpid
int times = 1000; // default value
void rtn1( int i ) { for ( int j = 0; j < times; j += 1 ) { if ( rand() % 10000 == 42 ) throw j; } } void rtn2( int i ) { for ( int j = times; j = 0; j -= 1 ) { if ( rand() % 10000 == 42 ) throw j; } } void g( int i ) { for ( int j = 0; j < times; j += 1 ) { if ( rand() % 2 == 0 ) rtn1( i ); else rtn2( i ); } if ( i % 2 ) rtn2( i ); rtn1( i ); } void f( int i ) { for ( int j = 0; j < times; j += 1 ) { g( i ); } if ( i % 2 ) g( i ); g( i ); } int main( int argc, char *argv[] ) { int seed = getpid(); // default value try { // process command-line arguments switch ( argc ) { case 3: times = stoi( argv[2] ); if ( times <= 0 ) throw 1; case 2: seed = stoi( argv[1] ); if ( seed <= 0 ) throw 1; case 1: break; // use defaults default: throw 1; } // switch } catch( ... ) { cout << "Usage: " << argv[0] << " [ seed ( 0) [ times ( 0) ] ]" << endl; exit( 1 ); } // try srand( seed ); // seed random number try { // begin program f( 3 ); cout << "seed:" << seed << " times:" << times << " complete" << endl; } catch( int rc ) { cout << "seed:" << seed << " times:" << times << " rc:" << rc << endl; } // try }
Figure 1: Dynamic Multi-Level Exit
CS 343 - Assignment 1
3
#include <iostream #include <cstdlib using namespace std; #include <unistd.h // access: getpid
#ifdef NOOUTPUT #deﬁne PRT( stmt ) #else #deﬁne PRT( stmt ) stmt #endif // NOOUTPUT
PRT( struct T { ~T() { cout << "~"; } }; )
struct E {}; // exception type long int freq = 5; // exception frequency
long int Ackermann( long int m, long int n ) { PRT( T t; ) if ( m == 0 ) { if ( rand() % freq == 0 ) throw E(); return n + 1; } else if ( n == 0 ) { try { return Ackermann( m - 1, 1 ); } catch( E ) { PRT( cout << "E1 " << m << " " << n << endl ); if ( rand() % freq == 0 ) throw E(); } // try } else { try { return Ackermann( m - 1, Ackermann( m, n - 1 ) ); } catch( E ) { PRT( cout << "E2 " << m << " " << n << endl ); } // try } // if return 0; // recover by returning 0 } int main( int argc, const char *argv[] ) { long int Ackermann( long int m, long int n ); long int m = 4, n = 6, seed = getpid(); // default values try { // process command-line arguments switch ( argc ) { case 5: freq = stoi( argv[4] ); if ( freq <= 0 ) throw 1; case 4: seed = stoi( argv[3] ); if ( seed <= 0 ) throw 1; case 3: n = stoi( argv[2] ); if ( n < 0 ) throw 1; case 2: m = stoi( argv[1] ); if ( m < 0 ) throw 1; case 1: break; // use defaults default: throw 1; } // switch } catch( ... ) { cout << "Usage: " << argv[0] << " [ m ( 0) [ n ( 0) [ seed ( 0)" " [ freq ( 0) ] ] ] ]" << endl; exit( 1 ); } // try srand( seed ); // seed random number try { // begin program PRT( cout << m << " " << n << " " << seed << " " << freq << endl ); long int val = Ackermann( m, n ); PRT( cout << val << endl ); } catch( E ) { PRT( cout << "E3" << endl ); } // try }
Figure 2: Throw/Catch
CS 343 - Assignment 1
4
3. This question requires the use of µC++, which means compiling the program with the u++ command.
Write a semi-coroutine with the following public interface (you may only add a public destructor and private members):
Coroutine StringLiteral { // YOU ADD MEMBERS HERE void main(); // coroutine main public: Event Match {}; // last character match Event Error {}; // last character invalid void next( char c ) { ch = c; // communication input resume(); // activate }
};
which veriﬁes a string of characters corresponds to a C++ string-literal described by the following grammar:
string-literal : encoding-preﬁx-opt " s-char-sequence " encoding-preﬁx-opt : u8 | u | U | L | ε (empty string) s-char-sequence : s-char-sequence s-char | ε s-char : any character except double-quote ("), backslash (\), or new-line | escape-sequence escape-sequence : simple-escape-sequence | octal-escape-sequence | hexadecimal-escape-sequence simple-escape-sequence : \’ | \" | \? | \\ | \a | \b | \f | \n | \r | \t | \v octal-escape-sequence : \ octal-digit | \ octal-digit octal-digit | \ octal-digit octal-digit octal-digit hexadecimal-escape-sequence : \x hexadecimal-sequence hexadecimal-sequence : hexadecimal-sequence hexadecimal-digit | hexadecimal-digit
Each octal or hexadecimal escape sequence must be the longest sequence of characters that can constitute the escape sequence; hexadecimal digits may have upper or lower case alphabetic characters. For example:
valid strings invalid strings "" " "a\n" "a "abc" ab" "xyz\012" "\\\" "xyz\xc" "\9" L"www" l"www"
After creation, the coroutine is resumed with a series of characters from a string (one character at a time). The coroutine raises one of the following exceptions at its resumer:
• Match means the characters form a valid C++ string-literal. • Error means the last character resulted in a string that is not a C++ string-literal.
After the coroutine raises an exception, it must NOT be resumed again; sending more characters to the coroutine after this point is undeﬁned and should generate an error.
CS 343 - Assignment 1
5
Write a program stringliteral that checks if a string is a C++ string-literal. The shell interface to the stringliteral program is as follows:
stringliteral [ inﬁle ]
(Square brackets indicate optional command line parameters, and do not appear on the actual command line.) If no input ﬁle name is speciﬁed, input comes from standard input. Output is sent to standard output. For any speciﬁed command-line ﬁle, check it exists and can be opened. You may assume I/O reading and writing do not result in I/O errors.
The program should:
• read a line from the ﬁle, • create a StringLiteral coroutine, • pass characters from the input line to the coroutine one at time. • print an appropriate message when the coroutine returns exception Match or Error, or if there are no more characters to send. • check for extra characters, • terminate the coroutine, and • repeat these steps for each line in the ﬁle.
For every non-empty input line, print the line, how much of the line is parsed, and the string yes if the string is a valid C++ string-literal and the string no otherwise. If there are extra characters (including whitespace) on a line after parsing, print these characters with an appropriate warning. Print an appropriate warning for an empty input line, i.e., a line containing only ’\n’. The following is some example output:
’""’ : ’""’ yes ’"a\n"’ : ’"a\n"’ yes ’"abc"’ : ’"abc"’ yes ’"xyz\012"’ : ’"xyz\012"’ yes ’"xyz\xc"’ : ’"xyz\xc"’ yes ’L"www"’ : ’L"www"’ yes ’"abc"xyz’ : ’"abc"’ yes -- extraneous characters ’xyz’ ’’ : Warning! Blank line. ’"’ : ’"’ no ’"a’ : ’"a’ no ’ab"’ : ’a’ no -- extraneous characters ’b"’ ’"\\\"’ : ’"\\\"’ no ’"\9"’ : ’"\9’ no -- extraneous characters ’"’ ’l"www"’ : ’l’ no -- extraneous characters ’"www"’
Assume a valid C++ string-literal starts at the beginning of the input line, i.e., there is no leading whitespace. See the C library routine isxdigit(c), which returns true if character c is a hexadecimal digit.
WARNING: When writing coroutines, try to reduce or eliminate execution “state” variables and control-ﬂow statements using them. A state variable contains information that is not part of the computation and exclusively used for control-ﬂow purposes (like ﬂag variables). Use of execution state variables in a coroutine usually indicates you are not using the ability of the coroutine to remember prior execution information. Little or no marks will be given for solutions explicitly managing “state” variables. See Section 3.1.3 in the Course Notes for details on this issue.
Submission Guidelines
Please follow these guidelines very carefully. Review the Assignment Guidelines and C++ Coding Guidelines before starting each assignment. Each text ﬁle, i.e., *.*txt ﬁle, must be ASCII text and not exceed 500 lines in length, where a line is a maximum of 120 characters. Programs should be divided into separate compilation units, i.e., *.{h,cc,C,cpp} ﬁles, where applicable. Use the submit command to electronically copy the following ﬁles to the course account.
1. q1ﬂags.{cc,C,cpp}, q1globalstatusﬂag.{cc,C,cpp}, q1returncodes.{cc,C,cpp} – code for question 1, p. 1. No program documentation needs to be present in your submitted code. No test, user or system documentation
CS 343 - Assignment 1
6
is to be submitted for this question. Output for this question is checked via a marking program, so it must match exactly with the given program.
2. q2longjmp.{cc,C,cpp} – code for question 2a, p. 1. No program documentation needs to be present in your submitted code. No test, user or system documentation is to be submitted for this question. Output for this question is checked via a marking program, so it must match exactly with the given program.
3. q2longjmp.txt – contains the information required by question 2b, p. 1.
4. q3*.{h,cc,C,cpp} – code for question 3, p. 4. Split your code across *.h and *.{cc,C,cpp} ﬁles as needed. Program documentation must be present in your submitted code. No user or system documentation is to be submitted for this question. Output for this question is checked via a marking program, so it must match exactly with the given program.
5. q3*.testdoc – test documentation for question 3, which includes the input and output of your tests. Poor documentation of how and/or what is tested can results in a loss of all marks allocated to testing.
6. Use the following Makeﬁle to compile the programs for questions 1, 2 and 3:
CXX = u++ # compiler CXXFLAGS = -g -Wall -MMD -std=c++11 # compiler ﬂags MAKEFILE NAME = ${ﬁrstword ${MAKEFILE LIST}}# makeﬁle name
OBJECTS01 = q1exception.o # optional build of given program EXEC01 = exception # given executable name
OBJECTS1 = # object ﬁles forming 1st executable with preﬁx “q1” EXEC1 = ﬂags # 1st executable name
OBJECTS2 = # object ﬁles forming 2nd executable with preﬁx “q1” EXEC2 = globalstatusﬂag # 2nd executable name
OBJECTS3 = # object ﬁles forming 3rd executable with preﬁx “q1” EXEC3 = returncodes # 3rd executable name
OBJECTS02 = q2throwcatch.o # optional build of given program EXEC02 = throwcatch # given executable name
OBJECTS4 = # object ﬁles forming 4th executable with preﬁx “q2” EXEC4 = longjmp # 4th executable name
OBJECTS5 = # object ﬁles forming 5th executable with preﬁx “q3” EXEC5 = stringliteral # 5th executable name
OBJECTS = ${OBJECTS1} ${OBJECTS2} ${OBJECTS3} ${OBJECTS4} ${OBJECTS5} DEPENDS = ${OBJECTS:.o=.d} EXECS = ${EXEC1} ${EXEC2} ${EXEC3} ${EXEC4} ${EXEC5}
#############################################################
.PHONY : all clean
all : ${EXECS} # build all executables
${EXEC01} : ${OBJECTS01} # optional build of given program g++-6 ${CXXFLAGS} $^ -o $@
q1%.o : q1%.cc # change compiler 1st executable, ADJUST SUFFIX (for .C/.cpp) g++-6 ${CXXFLAGS} -c $< -o $@
${EXEC1} : ${OBJECTS1} # compile and link 1st executable g++-6 ${CXXFLAGS} $^ -o $@
CS 343 - Assignment 1
7
${EXEC02} : ${OBJECTS02} # optional build of given program g++-6 ${CXXFLAGS} $^ -o $@
${EXEC2} : ${OBJECTS2} # compile and link 2nd executable g++-6 ${CXXFLAGS} $^ -o $@
${EXEC3} : ${OBJECTS3} # compile and link 3rd executable g++-6 ${CXXFLAGS} $^ -o $@
q2%.o : q2%.cc # change compiler 2nd executable, ADJUST SUFFIX (for .C/.cpp) g++-6 ${CXXFLAGS} -c $< -o $@
${EXEC4} : ${OBJECTS4} # compile and link 4th executable g++-6 ${CXXFLAGS} $^ -o $@
${EXEC5} : ${OBJECTS5} # compile and link 5th executable ${CXX} ${CXXFLAGS} $^ -o $@
#############################################################
${OBJECTS} : ${MAKEFILE NAME} # OPTIONAL : changes to this ﬁle = recompile
-include ${DEPENDS} # include *.d ﬁles containing program dependences
clean : # remove ﬁles that can be regenerated rm -f *.d *.o ${EXEC01} ${EXEC02} ${EXECS}
This makeﬁle is used as follows:
$ make exception $ exception ... $ make ﬂags $ ﬂags ... $ make globalstatusﬂag $ globalstatusﬂag ... $ make returncodes $ returncodes ... $ make longjmp $ longjmp ... $ make stringliteral $ stringliteral ...
Put this Makeﬁle in the directory with the programs, name the source ﬁles as speciﬁed above, and then type make exception, make ﬂags, make globalstatusﬂag, make returncodes, make longjmp, or make stringliteral in the directory to compile the programs. This Makeﬁle must be submitted with the assignment to build the program, so it must be correct. Use the web tool Request Test Compilation to ensure you have submitted the appropriate ﬁles, your makeﬁle is correct, and your code compiles in the testing environment.
Follow these guidelines. Your grade depends on it!

More products