I think this is worth a try.
How far do we want to go with this? Update the docs? Update the grammar and/or code? I think the former is probably good enough for now to see if it takes, and if it does then we can talk about updating code to not confuse people.
+1 on advocating the term “generator comprehension” as a synonym for
“generator expression” in future documentation and instructional
The long version:
If we were starting from scratch, maybe it would makes sense to use “
builder” or something else, rather than comprehension. But we’re not
starting from scratch.
And “list comprehension” has become a pretty universal (or at least
common) term in the python world. So we really can’t expect to re-name
And despite the term having its roots in an esoteric corner of
mathematics, I’ve never seen anyone confused by it — it’s just a new
word for a new thing to most folks.
Which is actually why “generator expression” is a bit confusing — it
seems to have more in common with comprehensions than with generators
(As in: the things functions with a yield in them produce)
In fact, the term “generator” is a bit confusing itself — due to it
being a common English word, it seems to refer to the generic idea of
“a thing that generates values on the fly”, which, of course it can be
in Python. But it actually is a specific type. And you can make a
“thing that generates values in the fly” with a regular class that
follows the iterator protocol.
So re-using the word in “generator expression” just adds a bit to the
confusion. (Even though it’s technically correct, it does “make” an
actual generator object, doesn’t it? (On a phone, so I can’t check)
but that’s kind of an implementation detail.
Also: if you google: “python list comprehension” you get a LOT of hits
to tutorials, etc.
So it is WAY too late to change that name.
When you google “python generator expression” the top hit is the
python docs, and the next few are about comparing generator
expressions and list comprehensions.
So I think we could start a new naming scheme for those.
‘cause then we’d get lost of questions about the difference between
generator expressions and generator comprehensions ;-)