Allow comment parsing to not die from "in family __" comment.

It doesn't actually _parse_ the new comment to figure out what
family the thing comes from, but in most contexts it is not needed.
At least now it doesn't crash. 
Also added the new phrase to at least one item in the testing examples
so the unit tests could catch this.

Co-authored-by: Richard West <[email protected]>
Co-authored-by: Sevy Harris <[email protected]>

Author: rwest

rmgpy/data/kinetics/family.py

@@ -4487,6 +4487,7 @@ def extract_source_from_comments(self, reaction):
         template_matches = re.search(template_pattern, full_comment_string)
         if autogen_node_matches is not None:  # autogenerated trees
             template_str = autogen_node_matches.group(1).split('Multiplied by reaction path degeneracy')[0].strip()
+            template_str = template_str.split('in family')[0].strip()
             tokens = template_str.split()
             if len(tokens) == 2:  # The node was probably split because wordwrap was turned off
                 assert len(template_str) > 115, 'The node name is too short to have been broken up by the chemkin writer'

test/rmgpy/test_data/parsing_data/chem_annotated.inp

@@ -259,7 +259,7 @@ HCCO(23)(+M)=O(2)+C2H(21)(+M)                       1.000e+00 0.000     0.000
 ! Reaction index: Chemkin #25; RMG #97
 ! Template reaction: Disproportionation
 ! Flux pairs: CH3CHCH3, C3H6(28); C2H5(27), C2H6; 
-! Estimated from node Root_Ext-2R!H-R_2R!H->C_4R->C
+! Estimated from node Root_Ext-2R!H-R_2R!H->C_4R->C in family Disproportionation.
 ! Multiplied by reaction path degeneracy 6.0
 C2H5(27)+CH3CHCH3<=>C2H6+C3H6(28)                   3.000e+11 0.000     0.000