Convenience facilities related to Python functions.

Latest release 20220619: @trace: rename some parameters, add with_caller.

Function callif(doit, func, *a, **kw)

Call func(*a,**kw) if doit is true otherwise just print it out.

The parameter func may be preceeded optionally by a dict containing modes. The current modes are:

• 'print': the print function, default the builtin print

Function callmethod_if(o, method, default=None, a=None, kw=None)

Call the named method on the object o if it exists.

If it does not exist, return default (which defaults to None). Otherwise call getattr(o, method)(*a, **kw). a defaults to (). kw defaults to {}.

Function derived_from(property_name)

A property which must be recomputed if the revision of another property exceeds the snapshot revision.

Function derived_property(func, original_revision_name='_revision', lock_name='_lock', property_name=None, unset_object=None)

A property which must be recomputed if the reference revision (attached to self) exceeds the snapshot revision.

Function func_a_kw_fmt(func, *a, **kw)

Prepare a percent-format string and associated argument list describing a call to func(*a,**kw). Return format,args.

The func argument can also be a string, presumably a prepared description of func such as funccite(func).

Function funccite(func)

Return a citation for a function (name and code location).

Function funcname(func)

Return a name for the supplied function func. Several objects do not have a name attribute, such as partials.

Function prop(func)

A substitute for the builtin @property.

The builtin @property decorator lets internal AttributeErrors escape. While that can support properties that appear to exist conditionally, in practice this is almost never what I want, and it masks deeper errors. Hence this wrapper for @property that transmutes internal AttributeErrors into RuntimeErrors.

Function returns_bool(func)

Decorator for functions which should return Booleans.

Function returns_str(func)

Decorator for functions which should return strings.

Function returns_type(func, basetype)

Decrator which checks that a function returns values of type basetype.

Function trace(*da, **dkw)

Decorator to report the call and return of a function.

Function yields_str(func)

Decorator for generators which should yield strings.

Function yields_type(func, basetype)

Decorator which checks that a generator yields values of type basetype.

Release Log

Release 20220619: @trace: rename some parameters, add with_caller.

Release 20220311.1: New callif(doit[,modes],func,*a,**kw) to call or print a function, used in -n/--dry-run stuff sometimes.

Release 20220311:

• @trace: new pprint=False option to use pprint.pformat instead of repr for the return value.
• @trace: bugfix use of retval parameter.

Release 20210913: New func_a_kw_fmt(func,*a,**kw) imported from cs.pfx, hooked it into @trace.

Release 20210906: funcname: special case functools.partial, return concise name.

Release 20210717:

• Move @trace from debug to py.func, defer log call imports to avoid loops.
• Drop cs.pfx requirement, import opportunisticly with fallback.

Release 20200518: funcname: prefer func.__qualname__ over func.__name__

Release 20190729: funccite: handle callables with no code attribute.

Release 20190108: Break import loop. Use cs.py3.raise_from for portability.

Release 20181231:

• Some type specific convenience wrappers for yields_type and returns_type.
• Bugfix for @prop.

Release 20170906.1: Minor tweaks.

Release 20170906: Bugfix for @prop.

Release 20170608: New decorator @prop which works just like the builtin @property, but transmutes internal AttributeError into RuntimeError, unmasking many bugs.

Release 20160828: Use "install_requires" instead of "requires" in DISTINFO.

Release 20150115: First PyPI release.