listing exceptions programmatically - Python

Question

Is there any way to programmatically determine which exceptions an object or method might raise?

Like dir(obj) lists available methods, I'm looking for the equivalent dir_exceptions(obj).

As far as I know, the only way to achieve this would be to parse the source.

Consider that the method might call library functions which themselves might raise arbitrary exceptions. — Russell Borogove
– Russell Borogove, Commented Nov 22, 2011 at 19:13
possible duplicate of Python and Exceptions (see comments on Noufal's answer) — user395760
– user395760, Commented Nov 22, 2011 at 19:27

Noufal Ibrahim · Accepted Answer · 2011-11-22 18:58:40Z

2

I don't think this is possible. An exception is a runtime phenomenon and you'll know what it possible (or what happens) only while running. Why would you want to do this though?

answered Nov 22, 2011 at 18:58

Noufal Ibrahim

73.2k13 gold badges140 silver badges174 bronze badges

Sign up to request clarification or add additional context in comments.

10 Comments

Civilian Over a year ago

I bet wants to do something like: try { blah() } catch dir_exceptions(blah) { }

Achim Over a year ago

It's definitively not possible. In Python it would even be possible to generate the exception class on the fly, based on some strange random data. That makes no sense of course, but it would be possible. ;-)

nmichaels Over a year ago

@Civilian: Python is not Java. The Python equivalent of that is try: blah <newline/tab> except: blah

Josh Bleecher Snyder Over a year ago

@nmichaels or (usually) better: except Exception:.

user395760 Over a year ago

@nmichaels: I'm giving the benefit of doubt and assume he's using braces for blocks simply because they are far easier to write and read in SO comments due to the lack of newlines. At least I do it whenever I need to embed multi-line code snippets in comments. And I don't even know Java ;)

|

Matt Fenwick · Accepted Answer · 2011-11-22 19:05:02Z

2

It looks like you'll have to trust the code's developers on this one: if they did a good job, the method/class documentation should list all the exceptions that could be raised.

answered Nov 22, 2011 at 19:05

Matt Fenwick

49.3k24 gold badges130 silver badges198 bronze badges

Comments

jaime · Accepted Answer · 2011-11-22 19:13:43Z

1

No, there is not a practical way to do this.

Most python developers derive from Exception, so if you're not sure, just catch Exception.

try:
    some_secret_code()
except Exception:
    print 'oops, something happened'

If you're thinking that you can import a module and poke around looking for things derived from Exception, that won't quite work either. What about that python nut that does this ->

exec "raise SystemExit()"

I'm not sure that there is a non-practical way to accomplish this.

answered Nov 22, 2011 at 19:13

jaime

2,3641 gold badge20 silver badges24 bronze badges

Comments

inspectorG4dget · Accepted Answer · 2011-11-22 19:09:26Z

-1

I don't think this is possible either, but if you trust that the programmer has named their exceptions with "Exception" or "Error" in the name, then you could do a dir on the class and search for elements that end with "Exception" or "Error". Aside from that (which is pretty hacky in itself), I don't see a straightforward/native/idiomatic way to do this

answered Nov 22, 2011 at 19:09

inspectorG4dget

115k30 gold badges159 silver badges253 bronze badges

1 Comment

Russell Borogove Over a year ago

Exceptions are instances of their own classes, and it would be extremely unusual to make an exception class or instance a member of another class. Usually you create an exception locally and anonymously and raise it immediately: raise ValueError( "illegal value" )

Collectives™ on Stack Overflow

listing exceptions programmatically - Python

4 Answers 4

10 Comments

Comments

Comments

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

10 Comments

Comments

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related