| Age | Commit message (Collapse) | Author |
|---|
|
We remove the special error printing pre-processing in favor of just
calling the standard printers.
Error printing has been a bit complex for a while due to an incomplete
migration to a new printing scheme based on registering exception
printers; this PR should alleviate that by completing the registration
approach.
After this cleanup, it should not be ever necessary for normal
functions to worry a lot about catching errors and re-raising them,
unless they have some very special needs.
This change also allows to consolidate the `explainErr` and `himsg`
modules into one, removing the need to export the error printing
functions. Ideally we would make the contents of `himsg` more
localized, but this can be done in a gradual way.
|
|
We cleanup a few imports on Declare, and indeed we find a suspicious
exception `AlreadyDeclared` present in `CErrors` where it should not
be there.
We move it to `Declare`, waiting for more investigation.
|
|
We place the check for unhandled exceptions in the `is_anomaly`
function, and consider all the exceptions non-handled by the printers
always anomalies.
This reworks the solution implemented in
ea3909466eaaf86ff212c0a002e5df11e4a979f5 , in particular
`allow_uncaught` cannot be used anymore, all exceptions must install a
printer.
In order to pass the test-suite CI we also had to register some
printers, that were not registered for no reason, forcing clients to
call a post-processing step on errors.
|
|
|
|
This replaces the mismatched context error, which occurred when
Instance := {} was used with strictly more fields than declared.
Since we later check that field names match those declared for the
instance, now that we reject duplicates we know that there are no
extra fields.
|
|
|
|
|
|
This was dead code, it was never raised ever.
|
|
|
|
As per https://github.com/coq/coq/pull/8965#issuecomment-441440779
|
|
Users can now register string notations for custom inductives.
Much of the code and documentation was copied from numeral notations.
I chose to use a 256-constructor inductive for primitive string syntax
because (a) it is easy to convert between character codes and
constructors, and (b) it is more efficient than the existing `ascii`
type.
Some choices about proofs of the new `byte` type were made based on
efficiency. For example, https://github.com/coq/coq/issues/8517 means
that we cannot simply use `Scheme Equality` for this type, and I have
taken some care to ensure that the proofs of decidable equality and
conversion are fast. (Unfortunately, the `Init/Byte.v` file is the
slowest one in the prelude (it takes a couple of seconds to build), and
I'm not sure where the slowness is.)
In String.v, some uses of `0` as a `nat` were replaced by `O`, because
the file initially refused to check interactively otherwise (it
complained that `0` could not be interpreted in `string_scope` before
loading `Coq.Strings.String`).
There is unfortunately a decent amount of code duplication between
numeral notations and string notations.
I have not put too much thought into chosing names; most names have been
chosen to be similar to numeral notations, though I chose the name
`byte` from
https://github.com/coq/coq/issues/8483#issuecomment-421671785.
Unfortunately, this feature does not support declaring string syntax for
`list ascii`, unless that type is wrapped in a record or other inductive
type. This is not a fundamental limitation; it should be relatively
easy for someone who knows the API of the reduction machinery in Coq to
extend both this and numeral notations to support any type whose hnf
starts with an inductive type. (The reason for needing an inductive
type to bottom out at is that this is how the plugin determines what
constructors are the entry points for printing the given notation.
However, see also https://github.com/coq/coq/issues/8964 for
complications that are more likely to arise if inductive type families
are supported.)
N.B. I generated the long lists of constructors for the `byte` type with
short python scripts.
Closes #8853
|
|
This makes the make-based build system stop linking to Camlp5's
gramlib and instead links to our own gramlib.
We use the style done in the packing of `Stdlib` in OCaml 4.07.
As to introduce a minimal amount of noise in history we use an
autogenerated `gramlib__pack` directory.
Co-authored-by: Gaëtan Gilbert <gaetan.gilbert@skyskimmer.net>
|
|
|
|
This shall eventually allow to call Himsg at any point of the
execution, independently of the exact current global environment.
|
|
Switch to using exceptions rather than user errors. This will be
required because the machinery for printing constrs is not available in
notation.ml, so we move the error message printing to himsg.ml instead.
This is commit 2/4 in the fix for #8401.
|
|
Continuing the interface cleanup we place `Constrexpr` in the
internalization module, which is the one that eliminates it.
This slims down `pretyping` considerably, including removing the
`Univdecls` module which existed only due to bad dependency ordering
in the first place. Thanks to @ Skyskimmer we also remove a duplicate
`univ_decl` definition among `Misctypes` and `UState`.
This is mostly a proof of concept yet as it depends on quite a few
patches of the tree. For sure some tweaks will be necessary, but it
should be good for review now.
IMO the tree is now in a state where we can could easy eliminate more
than 10 modules without any impact, IMHO this is a net saving API-wise
and would help people to understand the structure of the code better.
|
|
This API is a bit strange, I expect it will change at some point.
|
|
|
|
Instead, we properly register a printer for such exception and update
the code.
|
|
The exception needs to carry aroud a pair of `env, sigma` so printing
is correct. This gets rid of a few global calls, and it is IMO the
right thing to do.
While we are at it, we incorporate some fixes to a couple of
additional printing functions missing the `env, sigma` pair.
|
|
We'd like to handle proofs functionally we thus recommend not to use
printing functions without an explicit context.
We also adapt most of the code, making more explicit where the
printing environment is coming from.
An open task is to refactor some code so we gradually make the
`Pfedit.get_current_context ()` disappear.
|
|
|
|
|
|
As per https://github.com/coq/coq/pull/716#issuecomment-305140839
Partially using
```bash
git grep --name-only 'anomaly\s*\(~label:"[^"]*"\s*\)\?\(Pp.\)\?(\(\(Pp.\)\?str\)\?\s*".*[^\.!]")' | xargs sed s'/\(anomaly\s*\(~label:"[^"]*"\s*\)\?\(Pp.\)\?(\(\(Pp.\)\?str\)\?\s*".*\s*[^\.! ]\)\s*")/\1.")/g' -i
```
and
```bash
git grep --name-only ' !"' | xargs sed s'/ !"/!"/g' -i
```
The rest were manually edited by looking at the results of
```bash
git grep anomaly | grep '\.ml' | grep -v 'anomaly\s*\(~label:"[^"]*"\s*\)\?\(Pp\.\)\?(\(\(Pp.\)\?str\)\?\s*".*\(\.\|!\)")' | grep 'anomaly\($\|[^_]\)' | less
```
|
|
This completes the Loc.ghost removal, the idea is to gear the API
towards optional, but uniform, location handling.
We don't print <unknown> anymore in the case there is no location.
This is what the test suite expects.
The old printing logic for located items was a bit inconsistent as
it sometimes printed <unknown> and other times it printed nothing as
the caller checked for `is_ghost` upstream.
|
|
Now it is a private field, locations are optional.
|
|
|
|
We remove the camlp4 compatibility layer, and try to clean up
most structures. `parsing/compat` is gone.
We added some documentation to the lexer/parser interfaces that are
often obscured by module includes.
|
|
|
|
Previously to this patch, Coq featured to distinct logging paths: the
console legacy one, based on `Pp.std_ppcmds` and Ocaml's `Format`
module, and the `Feedback` one, intended to encapsulate message inside a
more general, GUI-based feedback protocol.
This patch removes the legacy logging path and makes feedback
canonical. Thus, the core of Coq has no dependency on console code
anymore.
Additionally, this patch resolves the duplication of "document" formats
present in the same situation. The original console-based printing code
relied on an opaque datatype `std_ppcmds`, (mostly a reification of
`Format`'s format strings) that could be then rendered to the console.
However, the feedback path couldn't reuse this type due to its opaque
nature. The first versions just embedded rending of `std_ppcmds` to a
string, however in 8.5 a new "rich printing" type, `Richpp.richpp` was
introduced.
The idea for this type was to be serializable, however it brought
several problems: it didn't have proper document manipulation
operations, its format was overly verbose and didn't preserve the full
layout, and it still relied on `Format` for generation, making
client-side rendering difficult.
We thus follow the plan outlined in CEP#9, that is to say, we take a
public and refactored version of `std_ppcmds` as the canonical "document
type", and move feedback to be over there. The toplevel now is
implemented as a feedback listener and has ownership of the console.
`richpp` is now IDE-specific, and only used for legacy rendering. It
could go away in future versions. `std_ppcmds` carries strictly more
information and is friendlier to client-side rendering and display
control.
Thus, the new panorama is:
- `Feedback` has become a very module for event dispatching.
- `Pp` contains a target-independent box-based document format.
It also contains the `Format`-based renderer.
- All console access lives in `toplevel`, with console handlers private
to coqtop.
_NOTE_: After this patch, many printing parameters such as printing
width or depth should be set client-side. This works better IMO,
clients don't need to notify Coq about resizing anywmore. Indeed, for
box-based capable backends such as HTML or LaTeX, the UI can directly
render and let the engine perform the word breaking work.
_NOTE_: Many messages could benefit from new features of the output
format, however we have chosen not to alter them to preserve output.
A Future commits will move console tag handling in `Pp_style` to
`toplevel/`, where it logically belongs.
The only change with regards to printing is that the "Error:" header was
added to console output in several different positions, we have removed
some of this duplication, now error messages should be a bit more
consistent.
|
|
We also remove flushing operations `msg_with`, now the flushing
responsibility belong to the owner of the formatter.
|
|
This is what has always been used, so it doesn't represent a functional
change.
This is just a preliminary patch, but many more possibilities could be
done wrt tags.
|
|
Currently, the STM, vernac interpretation, and the toplevel are
intertwined in a mutual dependency that needs to be resolved using
imperative callbacks.
This is problematic for a few reasons, in particular it makes the
interpretation of commands that affect the document quite intricate.
As a first step, we split the `toplevel/` directory into two: "pure"
vernac interpretation is moved to the `vernac/` directory, on which
the STM relies.
Test suite passes, and only one command seems to be disabled with this
approach, "Show Script" which is to my understanding
obsolete. Subsequent commits will fix this and refine some of the
invariants that are not needed anymore.
|
