According to the Python documentation on idioms and anti-idioms in relation to exceptions: "You should try to use as few except clauses in your code as you can — the ones you do use will usually be inside calls which should always succeed, or a catch-all in a main function." Taking this sentence in sections...
"You should try to use as few except clauses in your code as you can"
A bit confusing for a newbie like myself, I thought it was good practise in Python to use the EAFP style -i.e. many try and except statements. Or am I missing the point?
"the ones you do use will usually be inside calls which should always succeed"
I don't understand the point that's being made here.
"or a catch-all in a main function."
So it's good style to any let code that throws an exception to simply pass it up the call stack until it reaches the top level where you have really generic exception handling?
"You should try to use as few except clauses in your code as you can"
It's easy to litter your code with exceptions:
def solve_linear(mat1, mat2):
det1 = determinant(mat1)
det2 = determinant(mat2)
try:
return det1 / det2
except ZeroDivisionError:
raise NoSolution
Here, it's probably fine to let the ZeroDivisionError propagate. You don't need to catch it.
"the ones you do use will usually be inside calls which should always succeed"
For example, look at this code which reads a file, or returns a cached value. It normally succeeds, in spite of the KeyError exception:
def read_file(path):
try:
return cache[path]
except KeyError:
fp = open(path, 'rb')
data = fp.read()
fp.close()
cache[path] = data
return data
"or a catch-all in a main function."
If your program is interactive, you'll probably want to catch almost everything at the top level. Here's the top loop of an interactive command-line program:
def main():
while True:
try:
text = raw_input('enter command: ')
cmd = parse_command(text)
except EOFError:
return
except KeyboardInterrupt:
return
except ValueError:
print 'Syntax error'
continue
try:
run_cmd(cmd)
except SystemExit:
raise
except Exception:
traceback.print_exc()
Concerning the first point, the whole point of using exceptions is that you don't have to wrap every line in one! E.g. in C, errors are usually determined by the return value of a function call. So you have to check those after every call if you want to catch all errors. With Python you can group a (possibly large) block of statements that go together in a try/except block and only deal with all errors once.
The second point is that (if possible) you want to solve failures close to the point where they occur. E.g. you are reading data from a network and get zero bytes. In that case it is usually perfectly allright to wait and try again.
The last point is that sometimes an error is so big that it cannot be handled at a low level. E.g. if you are trying to open a file that not exist, it wil fail. And your program cannot do whatever it was going to do with the contents of the file. It is best to handle that at the top level of the program and ask the user for another file name or maybe quit the program.
I think the point is that exceptions should be used only for 'exceptional' circumstances. That is the meaning behind the use "in calls which will usually succeed". An example of this might be some computation that under some really weird circumstances ends having to do a division by zero. Then you can enclose that in a try / except statement to deal with that possibility.
The catch-all in a main function would be applicable to the same scenario. Say your calculations end up with a division by zero somewhere deep in the call stack. From this point you can't continue, so there's no point in having a try/except statement right at the point of failure. It would make more sense to just have one at a higher level where you can reasonably recover from the error.
The example they give in the documentation is an example of this. When calling 'get_status' you would expect the file to be there. If it doesn't then you have the except statement to deal with that (although, as mentioned, in that particular case the 'with' statement is much better).
The Python philosophy is typically to "ask forgiveness, not permission." But the idea isn't to use a try-except clause to catch ALL possible errors. Ideally, each try-except will only catch the error(s) that are relevant.
BAD (doesn't specify a specific exception type):
a = [1, 2, 3]
i = 3
try:
print a[i]
except:
print "Not found!"
GOOD (only handles the exception we expect to get):
a = [1, 2, 3]
i = 3
try:
print a[i]
except IndexError:
print "Not found!"
The reason this is important is so we don't obscure other possible code errors. What if, in this example, i was 1.8? The first example would print Not Found!, obscuring the real issue, but the second example would return TypeError: list indices must be integers, not float letting us know that there's a logical flaw in our code.
Related
How do you best handle multiple levels of methods in a call hierarchy that raise exceptions, so that if it is a fatal error the program will exit (after displaying an error dialog)?
I'm basically coming from Java. There I would simply declare any methods as throws Exception, re-throw it and catch it somewhere at the top level.
However, Python is different. My Python code basically looks like the below.
EDIT: added much simpler code...
Main entry function (plugin.py):
def main(catalog):
print "Executing main(catalog)... "
# instantiate generator
gen = JpaAnnotatedClassGenerator(options)
# run generator
try:
gen.generate_bar() # doesn't bubble up
except ValueError as error:
Utilities.show_error("Error", error.message, "OK", "", "")
return
... usually do the real work here if no error
JpaAnnotatedClassGenerator class (engine.py):
class JpaAnnotatedClassGenerator:
def generate_bar(self):
self.generate_value_error()
def generate_value_error(self):
raise ValueError("generate_value_error() raised an error!")
I'd like to return to the caller with an exception that is to be thrown back to that ones call until it reaches the outermost try-except to display an error dialog with the exception's message.
QUESTION:
How is this best done in Python? Do I really have to repeat try-except for every method being called?
BTW: I am using Python 2.6.x and I cannot upgrade due to being bound to MySQL Workbench that provides the interpreter (Python 3 is on their upgrade list).
If you don't catch an exception, it bubbles up the call stack until someone does. If no one catches it, the runtime will get it and die with the exception error message and a full traceback. IOW, you don't have to explicitely catch and reraise your exception everywhere - which would actually defeat the whole point of having exceptions. Actually, despite being primarily used for errors / unexpected conditions, exceptions are first and foremost a control flow tool allowing to break out of the normal execution flow and pass control (and some informations) to any arbitrary place up in the call stack.
From this POV your code seems mostlt correct (caveat: I didn't bother reading the whole thing, just had a quick look), except (no pun indented) for a couple points:
First, you should define your own specific exception class(es) instead of using the builtin ValueError (you can inherit from it if it makes sense to you) so you're sure you only catch the exact exceptions you expect (quite a few layers "under" your own code could raise a ValueError that you didn't expect).
Then, you may (or not, depending on how your code is used) also want to add a catch-all top-level handler in your main() function so you can properly log (using the logger module) all errors and eventually free resources, do some cleanup etc before your process dies.
As a side note, you may also want to learn and use proper string formatting, and - if perfs are an issue at least -, avoid duplicate constant calls like this:
elif AnnotationUtil.is_embeddable_table(table) and AnnotationUtil.is_secondary_table(table):
# ...
elif AnnotationUtil.is_embeddable_table(table):
# ...
elif AnnotationUtil.is_secondary_table(table):
# ...
Given Python's very dynamic nature, neither the compiler nor runtime can safely optimize those repeated calls (the method could have been dynamically redefined between calls), so you have to do it yourself.
EDIT:
When trying to catch the error in the main() function, exceptions DON'T bubble up, but when I use this pattern one level deeper, bubbling-up seems to work.
You can easily check that it works correctly with a simple MCVE:
def deeply_nested():
raise ValueError("foo")
def nested():
return deeply_nested()
def firstline():
return nested()
def main():
try:
firstline()
except ValueError as e:
print("got {}".format(e))
else:
print("you will not see me")
if __name__ == "__main__":
main()
It appears the software that supplies the Python env is somehow treating the main plugin file in a wrong way. Looks I will have to check the MySQL Workbench guys
Uhu... Even embedded, the mechanism expection should still work as expected - at least for the part of the call stack that depends on your main function (can't tell what happens upper in the call stack). But given how MySQL treats errors (what about having your data silently truncated ?), I wouldn't be specially suprised if they hacked the runtime to silently pass any error in plugins code xD
It is fine for errors to bubble up
Python's exceptions are unchecked, meaning you have no obligation to declare or handle them. Even if you know that something may raise, only catch the error if you intend to do something with it. It is fine to have exception-transparent layers, which gracefully abort as an exception bubbles through them:
def logged_get(map: dict, key: str):
result = map[key] # this may raise, but there is no state to corrupt
# the following is not meaningful if an exception occurred
# it is fine for it to be skipped by the exception bubbling up
print(map, '[%s]' % key, '=>', result)
return result
In this case, logged_get will simply forward any KeyError (and others) that are raised by the lookup.
If an outer caller knows how to handle the error, it can do so.
So, just call self.create_collection_embeddable_class_stub the way you do.
It is fine for errors to kill the application
Even if nothing handles an error, the interpreter does. You get a stack trace, showing what went wrong and where. Fatal errors of the kind "only happens if there is a bug" can "safely" bubble up to show what went wrong.
In fact, exiting the interpreter and assertions use this mechanism as well.
>>> assert 2 < 1, "This should never happen"
Traceback (most recent call last):
File "<string>", line 1, in <module>
AssertionError: This should never happen
For many services, you can use this even in deployment - for example, systemd would log that for a Linux system service. Only try to suppress errors for the outside if security is a concern, or if users cannot handle the error.
It is fine to use precise errors
Since exceptions are unchecked, you can use arbitrary many without overstraining your API. This allows to use custom errors that signal different levels of problems:
class DBProblem(Exception):
"""Something is wrong about our DB..."""
class DBEntryInconsistent(DBProblem):
"""A single entry is broken"""
class DBInconsistent(DBProblem):
"""The entire DB is foobar!"""
It is generally a good idea not to re-use builtin errors, unless your use-case actually matches their meaning. This allows to handle errors precisely if needed:
try:
gen.generate_classes(catalog)
except DBEntryInconsistent:
logger.error("aborting due to corrupted entry")
sys.exit(1)
except DBInconsistent as err:
logger.error("aborting due to corrupted DB")
Utility.inform_db_support(err)
sys.exit(1)
# do not handle ValueError, KeyError, MemoryError, ...
# they will show up as a stack trace
It is conventional to use pass statement in python like the following piece of code.
try:
os.makedirs(dir)
except OSError:
pass
So, 'pass' bascially does not do anything here. In this case, why would we still put a few codes like this in the program? I am confused. Many thanks for your time and attention.
It's for the parser. If you wrote this:
try:
# Code
except Error:
And then put nothing in the except spot, the parser would signal an error because it would incorrectly identify the next indentation level. Imagine this code:
def f(x):
try:
# Something
except Error:
def g(x):
# More code
The parser was expecting a statement with a greater indentation than the except statement but got a new top-level definition. pass is simply filler to satisfy the parser.
This is in case you want the code to continue right after the lines in the try block. If you won't catch it - it either skips execution until it is caught elsewhere - or fails the program altogether.
Suppose you're creating a program that attempts to print to a printer, but also prints to the standard output - you may not want it to file if the printer is not available:
try:
print_to_printer('hello world')
except NoPrinterError:
pass # no printer - that's fine
print("hello world")
If you would not use a try-catch an error would stop execution until the exception is caught (or would fail the program) and nothing would be printed to standard output.
The pass is used to tell the program what to do when it catches an error. In this particular case you're pretty much ignoring it. So you're running your script and if you experience an error keep going without worrying as to why and how.
That particular case is when you are definite on what is expected. There are other cases where you can break and end the program, or even assign the error to a variable so you can debug your program by using except Error as e.
try:
os.makedirs(dir)
except OSError:
break
or:
try:
os.makedirs(dir)
except OSError as e:
print(str(e))
try:
# Do something
except:
# again some code
# few more code
There are two uses of pass. First, and most important use :- if exception arises for the code under try, the execution will jump to except block. And if you have nothing inside the except block, it will throw IndentationError at the first place. So, to avoid this error, even if you have nothing to do when exception arises, you need to put pass inside except block.
The second use, if you have some more code pieces after the try-except block (e.g. again some code and few more code), and you don't put pass inside except, then that code piece will not be executed (actually the whole code will not be executed since compiler will throw IndentationError). So, in order to gracefully handle the scenario and tell the interpreter to execute the lines after except block, we need to put pass inside the except block, even though we don't want to do anything in case of exception.
So, here pass as indicated from name, handles the except block and then transfers the execution to the next lines below the except block.
I'm new on Flask, when writing view, i wander if all errors should be catched. If i do so, most of view code should be wrappered with try... except. I think it's not graceful.
for example.
#app.route('/')
def index():
try:
API.do()
except:
abort(503)
Should i code like this? If not, will the service crash(uwsgi+lnmp)?
You only catch what you can handle. The word "handle" means "do something useful with" not merely "print a message and die". The print-and-die is already handled by the exception mechanism and probably does it better than you will.
For example, this is not handling an exception usefully:
denominator = 0
try:
y = x / denominator
except ZeroDivisionError:
abort(503)
There is nothing useful you can do, and the abort is redundant as that's what uncaught exceptions will cause to happen anyway. Here is an example of a useful handling:
try:
config_file = open('private_config')
except IOError:
config_file = open('default_config_that_should_always_be_there')
but note that if the second open fails, there is nothing useful to do so it will travel up the call stack and possibly halt the program. What you should never do is have a bare except: because it hides information about what faulted where. This will result in much head scratching when you get a defect report of "all it said was 503" and you have no idea what went wrong in API.do().
Try / except blocks that can't do any useful handling clutter up the code and visually bury the main flow of execution. Languages without exceptions force you to check every call for an error return if only to generate an error return yourself. Exceptions exist in part to get rid of that code noise.
The Python Tutorial states:
The try ... except statement has an optional else clause, which,
when present, must follow all except clauses. It is useful for code
that must be executed if the try clause does not raise an exception.
For example:
for arg in sys.argv[1:]:
try:
f = open(arg, 'r')
except IOError:
print 'cannot open', arg
else:
print arg, 'has', len(f.readlines()), 'lines'
f.close()
The use of the else clause is better than adding additional code
to the try clause because it avoids accidentally catching an exception
that wasn’t raised by the code being protected by the try ... except
statement.
Question 1> Ａｆｔｅｒ reading the above document, I still don't get the idea why we cannot simply move the code from else clause into try clause.
Question 2> How does try clause can accidentally catch an exception since all those catches are done in the except clause, right?
You could put the else code in the try suite, but then you'd catch any exceptions that might be raised there. If you didn't intend that to happen, it would be "accidental," hence the wording of the document you linked to.
Best practice is to put as little code as possible in a try block, so that when an error occurs, you know what operation caused it and can handle it appropriately. If you have five lines of code in a try block and only expect one of them to ever raise an exception, your exception-handling code will be ill-prepared when an exception occurs in a line you didn't expect it to. Better in that case to let the exception be raised than handle it the wrong way.
If you move the code from the else into the try then that becomes part of the "critical path" which can raise an exception. If f.readlines() raises some sort of exception (perhaps an I/O error while reading the file because of a bad sector on the disk) then that error will be conflated with the one error that you currently catch. (Technically the "cannot open" error message would be wrong at that point ... because opening a file can succeed when reading it later fails; in fact opening it must succeed before you can even get an I/O error while processing it).
Normally you'd use a pattern more like:
foo = None
try:
# some code to access/load/initialize foo from an external source
except ...:
# handle various types of file open/read, database access, etc errors
else:
foo = something
... so that your subsequently run code and simply check if foo is None and use it or
work around it's unavailability in whatever way you see fit.
Answer for both questions is similar,
If you move code to the try clause, then you can not be sure from where the exception is coming from. Thus if you have another line of code that produces an unexpected IOError you could end searching for a problem where there is not.
So to better disect your code you want to simplify as much as possible the lines in the try making the catch as especific as possible.
1) Of course you could just move code from the else clause into the try clause. You could move it outside of the try block entirely, but this allows extra flexibility, and further modulation of the code. Also, perhaps specificity with the errors being caught. You could list a whole load of different exceptions that might be likely to happen, each with different statements. The stuff in the else clause would still only happen if no exception was raised, after the execution of the final line in the try block. e.g. printing a successful return message.
Also, try clauses add some extra CPU overhead with regards to garbage collection management and error tracing, so anything outside of the try clause is not protected in the same manner and may run more efficiently.
2) Error catching is quite specific. e.g. The except clause in your above example will only be run if an IOError is raised when running the f = open(arg,'r') line. If you want to catch any form of exception, use except Exception:.
I know using below code to ignore a certain exception, but how to let the code go back to where it got exception and keep executing? Say if the exception 'Exception' raises in do_something1, how to make the code ignore it and keep finishing do_something1 and process do_something2? My code just go to finally block after process pass in except block. Please advise, thanks.
try:
do_something1
do_something2
do_something3
do_something4
except Exception:
pass
finally:
clean_up
EDIT:
Thanks for the reply. Now I know what's the correct way to do it. But here's another question, can I just ignore a specific exception (say if I know the error number). Is below code possible?
try:
do_something1
except Exception.strerror == 10001:
pass
try:
do_something2
except Exception.strerror == 10002:
pass
finally:
clean_up
do_something3
do_something4
There's no direct way for the code to go back inside the try-except block. If, however, you're looking at trying to execute these different independant actions and keep executing when one fails (without copy/pasting the try/except block), you're going to have to write something like this:
actions = (
do_something1, do_something2, #...
)
for action in actions:
try:
action()
except Exception, error:
pass
update. The way to ignore specific exceptions is to catch the type of exception that you want, test it to see if you want to ignore it and re-raise it if you dont.
try:
do_something1
except TheExceptionTypeThatICanHandleError, e:
if e.strerror != 10001:
raise
finally:
clean_up
Note also, that each try statement needs its own finally clause if you want it to have one. It wont 'attach itself' to the previous try statement. A raise statement with nothing else is the correct way to re-raise the last exception. Don't let anybody tell you otherwise.
What you want are continuations which python doesn't natively provide. Beyond that, the answer to your question depends on exactly what you want to do. If you want do_something1 to continue regardless of exceptions, then it would have to catch the exceptions and ignore them itself.
if you just want do_something2 to happen regardless of if do_something1 completes, you need a separate try statement for each one.
try:
do_something1()
except:
pass
try:
do_something2()
except:
pass
etc. If you can provide a more detailed example of what it is that you want to do, then there is a good chance that myself or someone smarter than myself can either help you or (more likely) talk you out of it and suggest a more reasonable alternative.
This is pretty much missing the point of exceptions.
If the first statement has thrown an exception, the system is in an indeterminate state and you have to treat the following statement as unsafe to run.
If you know which statements might fail, and how they might fail, then you can use exception handling to specifically clean up the problems which might occur with a particular block of statements before moving on to the next section.
So, the only real answer is to handle exceptions around each set of statements that you want to treat as atomic
you could have all of the do_something's in a list, and iterate through them like this, so it's no so wordy. You can use lambda functions instead if you require arguments for the working functions
work = [lambda: dosomething1(args), dosomething2, lambda: dosomething3(*kw, **kwargs)]
for each in work:
try:
each()
except:
pass
cleanup()
Exceptions are usually raised when a performing task can not be completed in a manner intended by the code due to certain reasons. This is usually raised as exceptions. Exceptions should be handled and not ignored. The whole idea of exception is that the program can not continue in the normal execution flow without abnormal results.
What if you write a code to open a file and read it? What if this file does not exist?
It is much better to raise exception. You can not read a file where none exists. What you can do is handle the exception, let the user know that no such file exists. What advantage would be obtained for continuing to read the file when a file could not be opened at all.
In fact the above answers provided by Aaron works on the principle of handling your exceptions.
I posted this recently as an answer to another question. Here you have a function that returns a function that ignores ("traps") specified exceptions when calling any function. Then you invoke the desired function indirectly through the "trap."
def maketrap(*exceptions):
def trap(func, *args, **kwargs):
try:
return func(*args, **kwargs)
except exceptions:
return None
return trap
# create a trap that ignores all exceptions
trapall = maketrap(Exception)
# create a trap that ignores two exceptions
trapkeyattrerr = maketrap(KeyError, AttributeError)
# Now call some functions, ignoring specific exceptions
trapall(dosomething1, arg1, arg2)
trapkeyattrerr(dosomething2, arg1, arg2, arg3)
In general I'm with those who say that ignoring exceptions is a bad idea, but if you do it, you should be as specific as possible as to which exceptions you think your code can tolerate.
Python 3.4 added contextlib.suppress(), a context manager that takes a list of exceptions and suppresses them within the context:
with contextlib.suppress(IOError):
print('inside')
print(pathlib.Path('myfile').read_text()) # Boom
print('inside end')
print('outside')
Note that, just as with regular try/except, an exception within the context causes the rest of the context to be skipped. So, if an exception happens in the line commented with Boom, the output will be:
inside
outside