| Age | Commit message (Collapse) | Author |
|---|
|
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.
|
|
|
|
Objects registered through the callback functions were pushed on the libstack
before the ML-MODULE object itself, leading to anomalies when the corresponding
object was assuming that the ML module was properly defined in any other Coq
module requiring the Declare ML command.
|
|
|
|
When Mltop.add_rec_path was called to add paths to the loadpath, it was
also adding the top directory to the mlpath. In particular, "theories" was
added to the mlpath although it was explicitly marked "~with_ml:false".
The "plugins" and "user-contribs" subdirectories were scanned twice, once
for filling the loadpath and then for filling the mlpath.
This patch adds an argument to Mltop.add_rec_path to explicitly control
which paths from the loadpath are added to the mlpath (none, the top one,
all of them). The "top" option was the single old behavior; the "none"
option is used for "theories"; the "all" option is used to avoid
duplicate scanning for "plugins" and "user-contribs".
|
|
The number of path canonizations was quadratic in the number of potential
plugin directories. This patch makes it linear; on a standard Coq tree,
this brings the number of chdir and getcwd system calls from more than
1,000 down to about 200 at startup.
This also fixes a bug where the Cd vernacular command could cause plugins
to break since relative paths were used to locate them.
|
|
Suggested by @ppedrot
|
|
As noted by @ppedrot, the first is redundant. The patch is basically a renaming.
We didn't make the component optional yet, but this could happen in a
future patch.
|
|
module)
For the moment, there is an Error module in compilers-lib/ocamlbytecomp.cm(x)a
|
|
On the user side, coqtop and coqc take a list of warning names or categories
after -w. No prefix means activate the warning, a "-" prefix means deactivate
it, and "+" means turn the warning into an error. Special categories include
"all", and "default" which contains the warnings enabled by default.
We also provide a vernacular Set Warnings which takes the same flags as argument.
Note that coqc now prints warnings.
The name and category of a warning are printed with the warning itself.
On the developer side, Feedback.msg_warning is still accessible, but the
recommended way to print a warning is in two steps:
1) create it by:
let warn_my_warning =
CWarnings.create ~name:"my-warning" ~category:"my-category"
(fun args -> Pp.strbrk ...)
2) print it by:
warn_my_warning args
|
|
This patch splits pretty printing representation from IO operations.
- `Pp` is kept in charge of the abstract pretty printing representation.
- The `Feedback` module provides interface for doing printing IO.
The patch continues work initiated for 8.5 and has the following effects:
- The following functions in `Pp`: `pp`, `ppnl`, `pperr`, `pperrnl`,
`pperr_flush`, `pp_flush`, `flush_all`, `msg`, `msgnl`, `msgerr`,
`msgerrnl`, `message` are removed. `Feedback.msg_*` functions must be
used instead.
- Feedback provides different backends to handle output, currently,
`stdout`, `emacs` and CoqIDE backends are provided.
- Clients cannot specify flush policy anymore, thus `pp_flush` et al are
gone.
- `Feedback.feedback` takes an `edit_or_state_id` instead of the old
mix.
Lightly tested: Test-suite passes, Proof General and CoqIDE seem to work.
|
|
|
|
|
|
|
|
Since error messages are ultimately passed to Format, which has its own
buffers for concatenating strings, using concatenation for preparing error
messages just doubles the workload and increases memory pressure.
|
|
|
|
subdirectories.
Using Mltop.add_path instead of Mltop.add_rec_path causes the following
kind of behavior:
$ coqtop -nois
Coq < Require Import Coq.Init.Datatypes.
Error: Cannot find a physical path bound to logical path Coq.Init.
The only case where its use is still warranted is the implicit "." path. It
shall not be recursive because Coq might be called from about anywhere.
This patch also removes -I-as since its behavior is no longer the one from
8.4 so it is not worth keeping it around.
|
|
|
|
In particular:
- abstracting the code using calls to Unix opendir, stat, and closedir,
- uniformly using warnings when a directory does not exist (coqtop was
ignoring silently and coqdep was exiting via handle_unix_error),
- uniformly expecting paths in Unix format and warning otherwise.
|
|
(Sorry, was not intended to be pushed)
This reverts commit 5268efdefb396267bfda0c17eb045fa2ed516b3c.
|
|
In particular:
- abstracting the code using calls to Unix opendir, stat, and closedir,
- uniformly using warnings when a directory does not exist (coqtop was
ignoring silently and coqdep was exiting via handle_unix_error).
|
|
This is done by adding a fourth type of loadpath, the ones that are
neither implicit nor root, for the subdirectories of a -Q root.
Note: this means that scanning for available directories is no longer done
on the fly for -Q, but once and for all, as with -R.
|
|
|
|
Instead of modifying exceptions to wear additional information, we instead use
a dedicated type now. All exception-using functions were modified to support
this new type, in particular Future's fix_exn-s and the tactic monad.
To solve the problem of enriching exceptions at raise time and recover this
data in the try-with handler, we use a global datastructure recording the
given piece of data imperatively that we retrieve in the try-with handler.
We ensure that such instrumented try-with destroy the data so that there
may not be confusion with another exception. To further harden the correction
of this structure, we also check for pointer equality with the last raised
exception.
The global data structure is not thread-safe for now, which is incorrect as
the STM uses threads and enriched exceptions. Yet, we splitted the patch in
two parts, so that we do not introduce dependencies to the Thread library
immediatly. This will allow to revert only the second patch if ever we
switch to OCaml-coded lightweight threads.
|
|
|
|
This fixes the following bug related to stm: if one passes -I to coqide, then
such flag is passes to the workers; but if one uses "Add ML LoadPath" to
extend the paths in which coq looks for plugins, this extra path was
no passed to the slaves (via the command line) nor store in the system
state and hence sent to the slaves.
With this patch, when a cmxs is loaded, its full path is stored in the
summary and hence sent to the workers as one may expect.
|
|
E.g.
Coq options are:
-I dir look for ML files in dir
-include dir (idem)
[...]
-h, --help print this list of options
With the flag '-toploop coqidetop' these extra option are also available:
--help-XML-protocol print the documentation of the XML protocol used by CoqIDE
|
|
|
|
User interface writers can drop a footop.cmxs in $(COQLIB)/toploop/
and pass -toploop footop to customize the coq main loop.
A toploop must set Coqtop.toploop_init and Coqtop.toploop_run to
functions respectively initializing the toploop (and parsing toploop
specific arguments) and running the main loop itself.
For backward compatibility -ideslave and -async-proofs worker do
set the toploop to coqidetop and stmworkertop respectively.
|
|
caching time, i.e. when the Declare ML Module command is evaluated. This can
be used by both static and dynamic plugins.
|
|
with -R. (Fix for Rocq/Rational.)"
This reverts commit 7d3ce4012a53b123dac95381bf46aac65f865d69.
Conflicts:
CHANGES
|
|
This option complements -I-as and -R. As the two other options, it adds a
new loadpath, but contrarily to them, files are not looked into the
directory unless fully qualified.
|
|
(Fix for Rocq/Rational.)
|
|
- Option -I no longer handles loadpath, only mlpath. This is the same
behavior as coq_makefile. Option -I-as is unchanged.
- Option -R no longer recursively adds to mlpath; only the root directory
is added.
- user-contrib/ and xdg directories are no longer recursively added to
the loadpath.
- theories/ and plugins/ are no longer recursively added to the loadpath
when option -nois is passed.
- All the preconfigured directories are still recursively added to the
mlpath though.
|
|
|
|
The exact filename has to be written. This is coherent with the RefMan.
|
|
This summary entry is special because its unfreeze may load ML
code that in turns adds summary entries. Hence it is the first
summary piece to be unfreezed, otherwise some summaries may get
lost.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16752 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16652 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
the OCaml GC, as indicated by [Gc.stat].
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16651 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
it into the standard logger instead.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16491 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
States.freeze takes ~marshallable:bool, so that (only) when we want to
marshal data to disk/network we can ask the freeze functions of the
summary to force lazy values. The flag is propagated to Lib and Summary.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16478 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
the interface.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16372 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16290 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16260 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
Ok, this is merely a matter of taste, but up to now the usage
in Coq is rather to use capital letters instead of _ in the
names of inner modules.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16221 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
exception information addition facility.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16213 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16211 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
I hope I did not forget some [with] clauses. Otherwise, some
stack frame will be missing from the debug.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16167 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16161 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
native OCaml code.
Warning: the "retroknowledge" mechanism has not been ported to the native
compiler, because integers and persistent arrays will ultimately be defined as
primitive constructions. Until then, computation on numbers may be faster using
the VM, since it takes advantage of machine integers.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16136 85f007b7-540e-0410-9357-904b9bb8a0f7
|
