Following system colour scheme Selected dark colour scheme Selected light colour scheme

Python Enhancement Proposals

PEP 553 – Built-in breakpoint()

Author:
Barry Warsaw <barry at Python.org>
Status:
Final
Type:
Standards Track
Created:
05-Sep-2017
Python-Version:
3.7
Post-History:
05-Sep-2017, 07-Sep-2017, 13-Sep-2017
Resolution:
Python-Dev message

Table of Contents

Abstract

This PEP proposes adding a new built-in function calledbreakpoint()which enters a Python debugger at the point of the call. Additionally, two new names are added to thesysmodule to make the choice of which debugger is entered configurable.

Rationale

Python has long had a great debugger in its standard library calledpdb. Setting a break point is commonly written like this:

foo()
importpdb;pdb.set_trace()
bar()

Thus after executingfoo()and before executingbar(),Python will enter the debugger. However this idiom has several disadvantages.

  • It’s a lot to type (27 characters).
  • It’s easy to typo. The PEP author often mistypes this line, e.g. omitting the semicolon, or typing a dot instead of an underscore.
  • It ties debugging directly to the choice of pdb. There might be other debugging options, say if you’re using an IDE or some other development environment.
  • Python linters (e.g. flake8[linters]) complain about this line because it contains two statements. Breaking the idiom up into two lines complicates its use because there are more opportunities for mistakes at clean up time. I.e. you might forget to delete one of those lines when you no longer need to debug the code.

Python developers also have many other debuggers to choose from, but remembering how to invoke them can be problematic. For example, even when IDEs have user interface for setting breakpoints, it may still be more convenient to just edit the code. The APIs for entering the debugger programmatically are inconsistent, so it can be difficult to remember exactly what to type.

We can solve all these problems by providing a universal API for entering the debugger, as proposed in this PEP.

Proposal

The JavaScript language provides adebuggerstatement[js-debugger]which enters the debugger at the point where the statement appears.

This PEP proposes a new built-in function calledbreakpoint() which enters a Python debugger at the call site. Thus the example above would be written like so:

foo()
breakpoint()
bar()

Further, this PEP proposes two new name bindings for thesys module, calledsys.breakpointhook()and sys.__breakpointhook__.By default,sys.breakpointhook() implements the actual importing and entry intopdb.set_trace(), and it can be set to a different function to change the debugger that breakpoint()enters.

sys.__breakpointhook__is initialized to the same function as sys.breakpointhook()so that you can always easily reset sys.breakpointhook()to the default value (e.g. by doing sys.breakpointhook=sys.__breakpointhook__). This is exactly the same as how the existingsys.displayhook()/sys.__displayhook__and sys.excepthook()/sys.__excepthook__work[hooks].

The signature of the built-in isbreakpoint(*args,**kws).The positional and keyword arguments are passed straight through tosys.breakpointhook() and the signatures must match or aTypeErrorwill be raised. The return fromsys.breakpointhook()is passed back up to, and returned from breakpoint().

The rationale for this is based on the observation that the underlying debuggers may accept additional optional arguments. For example, IPython allows you to specify a string that gets printed when the break point is entered[i Python -embed].As of Python 3.7, the pdb module also supports an optionalheaderargument[pdb-header].

Environment variable

The default implementation ofsys.breakpointhook()consults a new environment variable calledPYTHONBREAKPOINT.This environment variable can have various values:

  • PYTHONBREAKPOINT=0disables debugging. Specifically, with this value sys.breakpointhook()returnsNoneimmediately.
  • PYTHONBREAKPOINT=(i.e. the empty string). This is the same as not setting the environment variable at all, in which casepdb.set_trace() is run as usual.
  • PYTHONBREAKPOINT=some.importable.callable.In this case, sys.breakpointhook()imports thesome.importablemodule and gets the callableobject from the resulting module, which it then calls. The value may be a string with no dots, in which case it names a built-in callable, e.g.PYTHONBREAKPOINT=int.(Guido has expressed the preference for normal Python dotted-paths, not setuptools-style entry point syntax[syntax].)

This environment variable allows external processes to control how breakpoints are handled. Some uses cases include:

  • Completely disabling all accidentalbreakpoint()calls pushed to production. This could be accomplished by settingPYTHONBREAKPOINT=0in the execution environment. Another suggestion by reviewers of the PEP was to setPYTHONBREAKPOINT=sys.exitin this case.
  • IDE integration with specialized debuggers for embedded execution. The IDE would run the program in its debugging environment withPYTHONBREAKPOINT set to their internal debugging hook.

PYTHONBREAKPOINTis re-interpreted every timesys.breakpointhook()is reached. This allows processes to change its value during the execution of a program and havebreakpoint()respond to those changes. It is not considered a performance critical section since entering a debugger by definition stops execution. Thus, programs can do the following:

os.environ['PYTHONBREAKPOINT']='foo.bar.baz'
breakpoint()# Imports foo.bar and calls foo.bar.baz()

Overridingsys.breakpointhookdefeats the default consultation of PYTHONBREAKPOINT.It is up to the overriding code to consult PYTHONBREAKPOINTif they want.

If access to thePYTHONBREAKPOINTcallable fails in any way (e.g. the import fails, or the resulting module does not contain the callable), a RuntimeWarningis issued, and no breakpoint function is called.

Note that as with all otherPYTHON*environment variables, PYTHONBREAKPOINTis ignored when the interpreter is started with -E.This means the default behavior will occur (i.e.pdb.set_trace()will run). There was some discussion about alternatively treatingPYTHONBREAKPOINT=0when-Eas in effect, but the opinions were inconclusive, so it was decided that this wasn’t special enough for a special case.

Implementation

A pull request exists with the proposed implementation[impl].

While the actual implementation is in C, the Python pseudo-code for this feature looks roughly like the following:

# In builtins.
defbreakpoint(*args,**kws):
importsys
missing=object()
hook=getattr(sys,'breakpointhook',missing)
ifhookismissing:
raiseRuntimeError('lost sys.breakpointhook')
returnhook(*args,**kws)

# In sys.
defbreakpointhook(*args,**kws):
importimportlib,os,warnings
hookname=os.getenv('PYTHONBREAKPOINT')
ifhooknameisNoneorlen(hookname)==0:
hookname='pdb.set_trace'
elifhookname=='0':
returnNone
modname,dot,funcname=hookname.rpartition('.')
ifdot=='':
modname='builtins'
try:
module=importlib.import_module(modname)
hook=getattr(module,funcname)
except:
warnings.warn(
'Ignoring unimportable $PYTHONBREAKPOINT:{}'.format(
hookname),
RuntimeWarning)
returnNone
returnhook(*args,**kws)

__breakpointhook__=breakpointhook

Rejected alternatives

A new keyword

Originally, the author considered a new keyword, or an extension to an existing keyword such asbreakhere.This is rejected on several fronts.

  • A brand new keyword would require a__future__to enable it since almost any new keyword could conflict with existing code. This negates the ease with which you can enter the debugger.
  • An extended keyword such asbreakhere,while more readable and not requiring a__future__would tie the keyword extension to this new feature, preventing more useful extensions such as those proposed in PEP 548.
  • A new keyword would require a modified grammar and likely a new bytecode. Each of these makes the implementation more complex. A new built-in breaks no existing code (since any existing module global would just shadow the built-in) and is quite easy to implement.

sys.breakpoint()

Why notsys.breakpoint()?Requiring an import to invoke the debugger is explicitly rejected becausesysis not imported in every module. That just requires more typing and would lead to:

importsys;sys.breakpoint()

which inherits several of the problems this PEP aims to solve.

Version History

  • 2019-10-13
    • Add missingreturnNoneinexceptclause to pseudo-code.
  • 2017-09-13
    • ThePYTHONBREAKPOINTenvironment variable is made a first class feature.
  • 2017-09-07
    • debug()renamed tobreakpoint()
    • Signature changed tobreakpoint(*args,**kws)which is passed straight through tosys.breakpointhook().

References


Source:https://github / Python /peps/blob/main/peps/pep-0553.rst

Last modified:2023-09-09 17:39:29 GMT