Hi @krmmalik,
What is the need for import qualified
for DA.List
here?
When you start a new Daml file, by default you have access to all of the functions defined in a “special” module called Prelude. It’s only special in that it is automatically imported in all Daml files.
Daml comes with many other modules, however, but these have to be manually imported. There are multiple ways to do that. Schematically, assuming you have a file Module.daml
with content like:
module Module where
one = 1
two = 2
Obviously not a very useful one, but it will serve its purpose. Here are a few examples of how import
statements work:
import Module
test = script do
assert $ one == 1
assert $ two == 2
assert $ Module.one == 1
assert $ Module.two == 2
import Module as M
test = script do
assert $ one == 1
assert $ two == 2
assert $ M.one == 1
assert $ M.two == 2
import qualified Module as M
test = script do
assert $ M.one == 1
assert $ M.two == 2
import qualified Module
test = script do
assert $ Module.one == 1
assert $ Module.two == 2
import qualified Module (one)
test = script do
assert $ Module.one == 1
import Module (one)
test = script do
assert $ one == 1
assert $ Module.one == 1
So you have a few options:
- A naked
import Module
lets you access all the names in Module
either directly as name
or “qualified” as Module.name
.
- Adding an
as
to the import statement changes the prefix used when qualifying an imported name.
- Adding the
qualified
keyword restricts the imported name to only their qualified form.
- Adding a list of names to the import statement restricts the import to just those names.
Now imagine Module.daml
has:
module Module where
map = "map"
and you want to write:
import Module
test = script do
assert $ map == "map" -- compilation error
This is the case described by @anthony: the compiler doesn’t know which map
you’re referring to (there is one in Module
, but there is also one in Prelude
).
You can still qualify the names if you want:
import Module
test = script do
assert $ Module.map == "map"
assert $ Prelude.map (*2) [1, 3, 7] == [2, 6, 14]
So, bottom line:
- There is no error if you have multiple imports that add the same name to the local list of available names.
- There is an error if you try to use an unqualified version of a name that has been imported multiple times.
- You can always qualify names at the point of use.
So using a qualified import is not necessary to resolve the ambiguity: you could choose to only qualify the names that do have a conflict.
I simply prefer, as a matter of personal taste, to always qualify all of my imports. I think that makes the code easier to read because, when I look at some of my code, I immediately know where a name comes from. Other people think that makes the code harder to read because all the names have long prefixes. There isn’t really a “right” answer here (except, you know, mine), it really is a matter of personal preference.