openstack/deb-python-taskflow
Code Issues Proposed changes

Use an appropriate `extract_traceback` limit

Browse Source 
When looking for where this modules ``run/load`` ends and
the user code begins we only need to search back at most
three levels to find this information; so to avoid getting
many more frames (especially when they won't be used) then
we need to just apply a sensible limit on the levels we get.

Change-Id: I9186e8844c92697b7e3f850328b6b39afba0150d
This commit is contained in:
Joshua Harlow
2014-12-07 10:50:41 -08:00
committed by Joshua Harlow
parent cd664bdd3b
commit 50b866cceb
1 changed files with 3 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
taskflow/engines/helpers.py
View File

@@ -84,8 +84,9 @@ def _extract_engine(**kwargs):
# Figure out where our code ends and the calling code begins (this is
# needed since this code is called from two functions in this module,
# which means the stack level will vary by one depending on that).
finder = itertools.takewhile(lambda frame: frame[0] in _FILE_NAMES,
reversed(traceback.extract_stack()))
finder = itertools.takewhile(
lambda frame: frame[0] in _FILE_NAMES,
reversed(traceback.extract_stack(limit=3)))
stacklevel = sum(1 for _frame in finder)
decorator = deprecation.renamed_kwarg('engine_conf', 'engine',
version="0.6",