A little bit of effort enables us to propagate valid typing annotations, making
subsequent typing re-inference easier (and avoiding a traversal just to remove
type annotations)
since it doesn't correctly propagate and update types. (Ideally it would, but
otherwise it would be better to remove the type annotations on the fly instead
of introducing wrong ones then cleaning them up).
The primary use-case for this was to be able to run computations on a list of
structures, then return an updated list with some fields in the structures
modified : that is what we need for distribution of tax amounts among household
members, for example.
This patch has a few components:
- Addition of a test as an example for tax distributions
- Added a transformation, performed during desugaring, that -- where lists are
syntactically expected, i.e. after the `among` keyword -- turns a (syntactic)
tuple of lists into a list of tuples ("zipping" the lists)
- Arg-extremum transformation was also fixed to use an intermediate list instead
of computing the predicate twice
- For convenience, allow to bind multiple variables in most* list
operations (previously only `let in` and functions allowed it)
- Fixed the printer for tuples to differentiate them from lists
*Note: tuples are not yet allowed on the left-hand side of filters and
arg-extremums for annoying syntax conflict reasons.
Always generate the version through git when possible, and encode that within
the binaries so that `catala --version` does'nt give misleading information.
Previously we used dune's builtin functionality, but that resorts to a hack at
install time which is unpleasant and doesn't work with our use of `opam
install`.
The cost is a re-linking of catala_utils and the binaries upon git commit, which
is hardly noticeable.
With the new EAppOp node, the typing of operator application has changed and this causes some problems
with the example in test_scope/good/nothing.catala_en when retyping the LCalc to go to Scalc. This
commit changes the weird thing that for generating Scalc, the "typed" parameter was set to "untyped".
As part of making tuples first-class citizens, expliciting the arity upon
function application was needed (so that a function of two args can
transparently -- in the surface language -- be applied to either two arguments
or a pair).
It was decided to actually explicit the whole type of arguments because the cost
is the same, and this is consistent with lambda definitions.
A related change done here is the replacement of the `EOp` node for operators by
an "operator application" `EAppOp` node, enforcing a pervasive invariant that
operators are always directly applied. This makes matches terser, and highlights
the fact that the treatment of operator application is almost always different
from function application in practice.
Added a new type safety invariant to ensure that the type `TDefault` can only appear in certain positions,
* On the left-hand side of an arrow with arity 1, as the type of a scope (for scope calls).
* At the root of the type tree (outside a default).
* On the right-hand side of the arrow at the root of the type (occurs for rentrant variables).
This is crucial to maintain the safety of the type system, as demonstrated in the formal development.
The invariant was checked on all tests cases and on family and housing benefits.
Adjusted inversion invariant about app to handle external objects as well.
This changes the `decl_ctx` to be toplevel only, with flattened references to
uids for most elements. The module hierarchy, which is still useful in a few
places, is kept separately.
Module names are also changed to UIDs early on, and support for module aliases
has been added (needs testing).
This resolves some issues with lookup, and should be much more robust, as well
as more convenient for most lookups.
The `decl_ctx` was also extended for string ident lookups, which avoids having
to keep the desugared resolution structure available throughout the compilation
chain.
- Add a `-I` option that allows defined modules to be available from other
directories
- Add reporting of the number of successful / failed tests
- Locate the project root, and always run the commands from there
The way nested priorities are encoded use `< < excs | true :- nested > :- x >`,
which imply that `nested` can actually be ∅ ; to cope with this, the typing of
default terms is made more generic (the return type is now the same as the
`cons` type `'a`, rather than `<'a>`). For the general case, we add an explicit
`EPureDefault` node which just encapsulates its argument (a `return`, in monad
terminology).
it is useful e.g. to be able to print intermediate ASTs when they don't type, to
debug the typing errors. This is better than commenting the typing line each
time.
Note that the option is not available on all targets (esp. not for ocaml and
python outputs ; it's allowed on the interpreters for debugging purposes but I'm
not sure if that's a good idea)
this is more consistent, avoids empty stamp files and should make things simpler
overall.
The slightly tricky `-C` option is added to Catala so that it can be run from
_build while paths to source and destination files are still specified relative
to CWD (Ninja doesn't provide string manipulation, so otherwise we would have to
explicit both the `_build/dir/` and `dir/` versions of each path).
* Obsolete code for included tests has been removed
* The engine uses a proper lexer and is much simplified
* An inline test in the middle of the file now only "sees" the file up to that
point. This fixes an issue where we had spurious errors when a type error was
added at the end of a file, and it would pop up in tests before it. This makes
files including many tests much more practical.
* diffing and resetting the tests has been reintroduced (done at the moment in
Ninja, but for more control (count number of failed tests, etc.) we could put it
back into Clerk at some point
* The Catala CLI can now take an input from stdin (with the possibility to link
a (possibly fake) on-disk file for error reporting and file locations ; this
is useful for running tests)
We need a concrete intermediate target for e.g. transitive uses of `> Include`
for Ninja to correctly handle them.
Of course we could also unroll all transitive dependencies, but meh.
Note also that now tests now just generate the outputs but facilities for
diffing and resetting are temporarily absent.
Rather than require all files to be listed on the command-line (and having to
check consistency with `> Using` directives), the main catala CLI is now a bit
more clever.
⇒ There is a new assumption that a module name definition must match the file
name (up to case and extension) — with appropriate error handling to enforce it.
In exchange, `> Using` directives are now used to more transparently lookup the
appropriate `.catala_*` interfaces and the compiled artifacts for the used modules (handling transitive dependencies), with just standard `-I` flags for when they need to be looked up in different places.
A recent patch introduced custom terms in expressions manipulated by the
interpreter. For typing reasons, a traversal is done to extend the supplied
expression with these custom terms — it's functionally the identity, but, due to
the fact that the type-checker can't infer covariance of our AST terms on their
phantom parameter, playing by the rules imposes a full traversal + rebuild.
Without resorting to a (constrained) `Obj.magic`, this patch avoids extra
intermediate conversions, which is enough to cut out the huge extra cost we were
incurring.
Closes#516
