Agreed. My feeling is that the mention of Jinja in the PEP probably shouldn’t be taken as anything too specific, but it did generate some interesting discussion about how tags might be defined in real-life applications. I’m still not sure if there’s a case for lazy evaluation in tagged strings (“replicating Jinja” clearly isn’t one 
), but I am still convinced that very few[1] people would use a tag that required you to wrap all your expressions in lambda.
I still wish the PEP had some better examples of actual use cases, though.
I was going to say “no-one”, but I don’t want to annoy @charliermarsh again
↩︎