Kurt Rose

hmm this turned out to be embarrassingly complicated ```python >>> glom( (obj_list, c_list), (Invoke(zip).star(T), [(Assign(T[0]['classC'], Spec(T[1])), T[0])]) ) [{'classA': 'A1', 'classB': 'B1', 'classC': 'C1'}, {'classA': 'A2', 'classB': 'B2', 'classC': 'C2'},...

```python (Invoke(zip).star(T), [(Assign(T[0]['classC'], T[1]), T[0])]) ``` what can we do to clean this up? * Zip in reduction.py * Assign val eval Spec or T, not just Spec ```python (Zip(),...

another approach... ```python (Invoke(zip).star(T), [(Let(class_c=T[1], T[0], Assign('classC', S['class_c'])]) ```

maybe another approach -- don't mutate the old but just move forward also doesn't mutate the old ```python >>> glom( (obj_list, c_list), (Invoke(zip).star(T), [Invoke(dict).specs(T[0], classC=T[1])]) ) [{'classA': 'A1', 'classB': 'B1',...

totally agree -- maybe building into the scope?

not to necro the old thread, but we just had a release that added `Fill` which solves this more neatly: ```python target={ 'issues' : [ { 'id': '999999', 'summary': 'this...

here's one that would be useful: ``` >>> from glom import S, T, glom >>> data = { "a": 1, "b": [ { "c": 1, "d": 5 }, { "c":...

oh hey that's a really neat idea -- does it only go one level or is this a recursive walk thing? sorry for the slow response!

hmm.... what we probably need to make this work coherently is some generic notion of path akin to the hierarchy of tags in XML it is a little bit tricky,...

I'm really trying to push towards gloms-that-transform-gloms there are a few bits and pieces to get set up, but the idea would be that "inverse of (certain types of) gloms"...