Py Lec S 21 5 - ExceptionBasic PDF

Preview

Summary

ExceptionBasic...

Description

Error Handling

Error types 





Syntax error:  When: caught by Python interpreter  How: Python sends SyntaxError message to STDERR Runtime error:  When: an “unexecutable” condition occurs  Note: Often caused by invalid input from file, user, etc.  How: Python “raises an exception” Logic error:  When: discovered during program testing  Note: Can be caused by incorrect variable name, indentation, operator precedence, loop (off-by-one)  How: mitigated by using PDL/test plan



Purpose of error handling Purpose of error handling: prevent program from stopping due to unexpected condition at runtime.  Instead:  either terminate gracefully  or handle error appropriately & continue  Two complimentary approaches:  Defensive programming:  Code for anticipated errors  Example: zero division error  Exception handling:  Handle exceptional conditions raised by Python during program execution  Example: Missing file, DB connectivity  Note: What falls under “anticipated” and “exceptional” can be debatable. 



Error handling example 

Defensive coding a = int(input(“Num1: ”)) b = int(input(“Num2: “)) if b != 0: print(a/b) else: print(“Error: Division by zero!”)



Unhandled exception 

Unexecutable condition not handled a = int(input(“Num1: ”)) b = int(input(“Num2: “)) # user input: 0 print(a/b)



Named exception is raised; program aborts Traceback (most recent call last): File "./calc.py", line 5, in print(a/b) ZeroDivisionError: division by zero



Handling exceptional errors  

“Try” to execute code, and if an “exception” during runtime occurs, “handle” it. Benefits:  Program code & error handling code are distinct: can improve code readability & maintenance.  Error handling code can be written anywhere in the process hierarchy: separation of error handling responsibilities.  “Exceptions” are named Python objects that include relevant information regarding the unexecutable condition.



Exception handling process Raising an exception: Python “raises an exception” when encountering a runtime error. 2. Handling an exception: Python expects an exception to be “handled” by program.  If handled, Python continues program execution.  If not handled, Python terminates program with error message to STDERR.  Python provides clauses for exception handling:  try/except: required clauses for error handling  else, finally: optional clauses for error handling  raise: optional clause to explicitly raise exceptions 1.

 Example:

Program flow with try & except



Code try: # attempt execution of statements x=2/0 # exception raised except: # if exception raised execute statements print(“Exception raised!”)



Output Exception raised!



Program flow 





The try clause is used to raise exceptions  All statements in the try clause are executed.  If no exception occurs, the statements in the except clause are not executed. The except clause is used to handle raised exceptions  When an exception occurs in the try clause the rest of the statements in the try clause are not executed.  Instead, all statements in the except clause are executed. Program execution  Program execution continues after the try & except block.

 Example:

Program flow with try & except



Code try: # attempt execution of statements print(“Hello …”) # executed x=2/0 # exception raised print(“… world”) # not executed except: # if exception raised execute statements print(“Exception raised!”) print(“~ Done with experiment ~”)



Output Hello … Exception raised! ~ Done with experiment ~



Example: Exception handling 



Using unnamed exception try: return 2/0 except: # unnamed exception print(“Error”) Using named exception try: return 2/0 except ZeroDivisionError: # named exception print(“Error: zero division”)



Error handling options   

Continue program if error recoverable. Terminate program gracefully if error unrecoverable. Pass on error handling to higher-level function.  Error state usually communicated to higher-level function with return value.  Return value must be tested in calling function.  Return value can be ignored, potentially further compounding error condition.



Exception handling actions 





Handle quietly at the point of the raised exception (and keep running).  Note: The pass statement does “nothing” but is considered valid for exception handling. Make the exception visible by:  displaying an error message  logging  re-raising an exception for calling process Manage the exception in a program specific way.  Example: Use default values if appropriate values have not been provided.



Exception handling actions – example 

Inform & handle try: x=2/0 except ZeroDivisionError: print(“Cannot divide by zero!”) x=0 print(“Result:”, x)

Unhandled exception & traceback 





When an exception is not handled in the current process, the process stops and passes the exception to the calling process, which passes it to the calling process until:  Either the exception is handled  Or program execution stops Python displays a “Traceback”, which follows the exception from the point of occurrence up the invocation chain. Example:  Main program calls function F  Function F raises exception  If exception is not handled in F, it is passed to main  If exception is not handled in main, program stops: the error displays the path the exception traveled.



Code def divide() 2/0 def calc() divide() calc() # main -> calc -> divide: exception



Traceback error message Traceback (most recent call last): File “./main.py", line 9, in calc() File “./main.py", line 7, in calc divide() File “./main.py", line 5, in divide 2/0 ZeroDivisionError: division by zero...