]>
matita.cs.unibo.it Git - helm.git/log
Claudio Sacerdoti Coen [Mon, 12 Sep 2005 10:44:12 +0000 (10:44 +0000)]
...
Claudio Sacerdoti Coen [Mon, 12 Sep 2005 10:26:36 +0000 (10:26 +0000)]
...
Stefano Zacchiroli [Mon, 12 Sep 2005 09:19:28 +0000 (09:19 +0000)]
- exported CPS iterator visit_magic
- added CPS iterator vivist_variable
- added freshen_term/freshen_obj which freshen instance numbers on ASTs
Stefano Zacchiroli [Mon, 12 Sep 2005 09:18:15 +0000 (09:18 +0000)]
Filled pre-generated notation levels with productions on dummy tokens (never generated by the lexer). This hack works around a bug in camlp4 which on delete_rule remove an entire precedence level if it gets empty after the delete_rule invocation.
Stefano Zacchiroli [Mon, 12 Sep 2005 09:16:56 +0000 (09:16 +0000)]
done some items
Stefano Zacchiroli [Mon, 12 Sep 2005 09:16:34 +0000 (09:16 +0000)]
commented Record type constructor
Claudio Sacerdoti Coen [Mon, 12 Sep 2005 08:52:13 +0000 (08:52 +0000)]
An old note (that goes back to July) integrated in matita.txt.
Enrico Tassi [Fri, 9 Sep 2005 13:03:25 +0000 (13:03 +0000)]
the case Appl Meta vs t was not executed in case t was a Constant,Mutcase,Cofix,Fix,.... (I guess only if it was a Rel/Meta).
Enrico Tassi [Fri, 9 Sep 2005 09:24:53 +0000 (09:24 +0000)]
workaround for sequent viewer flickering
Enrico Tassi [Fri, 9 Sep 2005 09:06:13 +0000 (09:06 +0000)]
added left/right
Claudio Sacerdoti Coen [Thu, 8 Sep 2005 17:13:46 +0000 (17:13 +0000)]
...
Stefano Zacchiroli [Thu, 8 Sep 2005 12:49:43 +0000 (12:49 +0000)]
removed some debugging prints
Stefano Zacchiroli [Thu, 8 Sep 2005 11:55:16 +0000 (11:55 +0000)]
items related to lazy parsing
Stefano Zacchiroli [Thu, 8 Sep 2005 10:42:03 +0000 (10:42 +0000)]
implemented lazy disambiguation of tactics arguments, when those
arguments have to be parsed in contexts extracted by a pattern
Stefano Zacchiroli [Thu, 8 Sep 2005 10:40:16 +0000 (10:40 +0000)]
added set_metasenv on proof_status
Stefano Zacchiroli [Thu, 8 Sep 2005 10:39:49 +0000 (10:39 +0000)]
- uses runtime base dir to reference logo with qed
- removed some dead code
Stefano Zacchiroli [Thu, 8 Sep 2005 10:38:05 +0000 (10:38 +0000)]
removed some dead code
Stefano Zacchiroli [Thu, 8 Sep 2005 10:37:34 +0000 (10:37 +0000)]
tests changing under a binder
Stefano Zacchiroli [Thu, 8 Sep 2005 10:36:40 +0000 (10:36 +0000)]
uses Map.equal to compare universes
Stefano Zacchiroli [Thu, 8 Sep 2005 09:42:30 +0000 (09:42 +0000)]
completed test
Claudio Sacerdoti Coen [Wed, 7 Sep 2005 17:15:09 +0000 (17:15 +0000)]
Unsharing bugs fixed.
Claudio Sacerdoti Coen [Wed, 7 Sep 2005 16:56:30 +0000 (16:56 +0000)]
Unsharing bug due to a very stupid typo fixed.
Claudio Sacerdoti Coen [Wed, 7 Sep 2005 16:20:03 +0000 (16:20 +0000)]
...
Stefano Zacchiroli [Tue, 6 Sep 2005 16:50:33 +0000 (16:50 +0000)]
bugfix: avoid losing attributes on boxes which have a single children
this used to cause impossibility of selecting a let .. in term
Stefano Zacchiroli [Tue, 6 Sep 2005 16:49:07 +0000 (16:49 +0000)]
removed dead code
Stefano Zacchiroli [Tue, 6 Sep 2005 16:48:21 +0000 (16:48 +0000)]
added {get,set}_attr
Stefano Zacchiroli [Tue, 6 Sep 2005 16:48:04 +0000 (16:48 +0000)]
added {get,set,pp}_attr
Stefano Zacchiroli [Tue, 6 Sep 2005 16:47:29 +0000 (16:47 +0000)]
uses \def symbol for definitions in context
Stefano Zacchiroli [Tue, 6 Sep 2005 16:46:03 +0000 (16:46 +0000)]
done 1 item
Stefano Zacchiroli [Tue, 6 Sep 2005 14:39:44 +0000 (14:39 +0000)]
updated
Stefano Zacchiroli [Tue, 6 Sep 2005 14:37:42 +0000 (14:37 +0000)]
misc fixes in cic browser queries (pretty printing, url bar, ...)
Alberto Griggio [Tue, 6 Sep 2005 14:18:12 +0000 (14:18 +0000)]
added dirty hack to blacklist mult_n_2, which causes troubles... :-(
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 14:14:13 +0000 (14:14 +0000)]
...
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 12:21:41 +0000 (12:21 +0000)]
...
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 12:12:44 +0000 (12:12 +0000)]
...
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 11:59:41 +0000 (11:59 +0000)]
Refiner substituted with the type-checker in a case that is known to be already
well-typed.
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 11:38:00 +0000 (11:38 +0000)]
...
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 10:25:15 +0000 (10:25 +0000)]
New bug exposed.
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 10:24:58 +0000 (10:24 +0000)]
...
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 10:12:12 +0000 (10:12 +0000)]
Dead code/files removed.
Stefano Zacchiroli [Tue, 6 Sep 2005 09:13:04 +0000 (09:13 +0000)]
removed debugging prints and better sample
Stefano Zacchiroli [Tue, 6 Sep 2005 09:06:38 +0000 (09:06 +0000)]
added <include href="foo/bar/baz.xml" /> support
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 08:59:09 +0000 (08:59 +0000)]
Bug fixed: the small logo was searched in the wrong path.
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 08:43:30 +0000 (08:43 +0000)]
Serious bug fixed: unsharing was not performed over sequents.
Claudio Sacerdoti Coen [Tue, 6 Sep 2005 08:35:34 +0000 (08:35 +0000)]
Dead code removed.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 17:14:12 +0000 (17:14 +0000)]
Added aliases and notation.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 16:59:40 +0000 (16:59 +0000)]
...
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 16:48:46 +0000 (16:48 +0000)]
Unfold tactic generalized to perform zeta-reduction.
WARNING: the tactic is bugged since it tries to zeta-expand a term that lives
in the context of the goal, not in the context of the pattern!
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 16:47:07 +0000 (16:47 +0000)]
locate_in_* functions generalized to handle equalities in a context.
Usually only the context length is used (to compute some lifting factor).
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 16:36:21 +0000 (16:36 +0000)]
New change in patterns: the pattern "in H" is now interpreted as
"in H \vdash ?". It used to be interpreted as "in H \vdash %".
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 16:14:57 +0000 (16:14 +0000)]
The refined form of a reference to a let-in bound variable in the context
used to be the zeta-expanded form. It is now the reference.
Stefano Zacchiroli [Mon, 5 Sep 2005 15:55:46 +0000 (15:55 +0000)]
use uniform naming for referencing cicNotation* modules
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 15:48:04 +0000 (15:48 +0000)]
...
Stefano Zacchiroli [Mon, 5 Sep 2005 15:29:34 +0000 (15:29 +0000)]
removed tedious debugging message
Stefano Zacchiroli [Mon, 5 Sep 2005 15:28:12 +0000 (15:28 +0000)]
avoid generating multiple times the same xref/href
Enrico Tassi [Mon, 5 Sep 2005 15:25:56 +0000 (15:25 +0000)]
fix generation of applications of applications.
Enrico Tassi [Mon, 5 Sep 2005 15:25:25 +0000 (15:25 +0000)]
added @raise in comment (and source)
Enrico Tassi [Mon, 5 Sep 2005 15:24:21 +0000 (15:24 +0000)]
simplify and let-in
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 14:42:52 +0000 (14:42 +0000)]
...
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 14:41:47 +0000 (14:41 +0000)]
Assert false (for imbricated theorems) changed to a nice error message.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 14:27:39 +0000 (14:27 +0000)]
New strategy for let-in unfolding (aka zeta reduction): a reference to a
let-bound variable is expanded if and only if the simplified form of its definiendum (applied to the actual arguments) is different from the
non-simplified form.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 12:42:57 +0000 (12:42 +0000)]
Critical bug fixed: the get_cooked_obj was called on an object that was not
in cache. The bug did not manifest when the environment was trusted.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 12:04:50 +0000 (12:04 +0000)]
Typing errors fixed.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 11:47:49 +0000 (11:47 +0000)]
The logo was searched in the wrong directory.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 11:44:28 +0000 (11:44 +0000)]
Bug fixed: Invalid_argument was raised by List.combine in place of a typing
error. A similar thing occurred during pretty-printing.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 10:35:41 +0000 (10:35 +0000)]
LAMBDA-TYPES moved under contrib, fixed (to use the library of Coq even if
the library of matita is already compiled) and added to the daily benchmark.
Stefano Zacchiroli [Mon, 5 Sep 2005 10:26:09 +0000 (10:26 +0000)]
bug fix for IDA: uses remote names (i.e. http://..) when patching xml bases so that external links like images gets valid URIs instead of invalid local ones like /tmp/...
Ferruccio Guidi [Mon, 5 Sep 2005 10:15:21 +0000 (10:15 +0000)]
LAMBDA-TYPES moved under contribs
Enrico Tassi [Mon, 5 Sep 2005 10:04:00 +0000 (10:04 +0000)]
The popup that asks to generate .moo for a .ma shows only the basename and not the full path of the file.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 09:46:26 +0000 (09:46 +0000)]
Bug fixed: matitac used to stop too early when an ambigous identifier was met
and auto_disambiguation was true.
Claudio Sacerdoti Coen [Mon, 5 Sep 2005 09:24:51 +0000 (09:24 +0000)]
create_owner_environment missing from matitatop initialization
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 13:02:44 +0000 (13:02 +0000)]
...
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 13:01:37 +0000 (13:01 +0000)]
Unsharing removed since it is now used in Cic2acic.
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 12:58:23 +0000 (12:58 +0000)]
Unsharing finally introduced (but just for object processing, not yet for terms
and sequents).
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 12:40:21 +0000 (12:40 +0000)]
Bug fixed: sorts and implicits were not unshared correctly.
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 12:08:58 +0000 (12:08 +0000)]
Assert added to check whether an unsharing problem is met!
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 10:09:38 +0000 (10:09 +0000)]
...
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 10:08:37 +0000 (10:08 +0000)]
...
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 10:01:55 +0000 (10:01 +0000)]
...
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 09:58:02 +0000 (09:58 +0000)]
Serious bug fixed previously introduced by me in the kernel fixed: in some cases
an application with no arguments was created.
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 09:42:34 +0000 (09:42 +0000)]
Stupid typo fixed.
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 09:42:01 +0000 (09:42 +0000)]
Missing initialization of the trusting function (for the kernel).
Claudio Sacerdoti Coen [Fri, 2 Sep 2005 09:11:58 +0000 (09:11 +0000)]
Unused index (because of UNIQUE field count) dropped.
Claudio Sacerdoti Coen [Thu, 1 Sep 2005 11:35:22 +0000 (11:35 +0000)]
...
Claudio Sacerdoti Coen [Thu, 1 Sep 2005 11:34:04 +0000 (11:34 +0000)]
...
Claudio Sacerdoti Coen [Thu, 1 Sep 2005 10:28:28 +0000 (10:28 +0000)]
New index refObj_occurrence on refObj. It is required by rdfly and by the graph
generator to efficiently produce backward dependencies graphs.
Alberto Griggio [Thu, 1 Sep 2005 09:43:23 +0000 (09:43 +0000)]
changed default parameter values...
Claudio Sacerdoti Coen [Thu, 1 Sep 2005 08:42:30 +0000 (08:42 +0000)]
Comment added.
Claudio Sacerdoti Coen [Thu, 1 Sep 2005 08:42:02 +0000 (08:42 +0000)]
...
Claudio Sacerdoti Coen [Thu, 1 Sep 2005 08:40:28 +0000 (08:40 +0000)]
...
Alberto Griggio [Wed, 31 Aug 2005 23:08:10 +0000 (23:08 +0000)]
fixed bug in compute_equality_weight caused by wrong assumption on function
arguments evaluation order
Claudio Sacerdoti Coen [Wed, 31 Aug 2005 10:53:16 +0000 (10:53 +0000)]
Since several weeks whelp did not compile any longer. Fixed:
1. the new getter implementation by Zack does not implement yet remote
calls. I have changed the (sample) configuration file to use an internal
getter (maybe also speeding up the tool?)
2. the new parser implementation does not have hard-coded notation. The
notation is now loaded from two files on disk (core_notation.moo and
coq.moo). I have added to entries in the configuration file to hold the
path to the two moo files.
3. Whelp needs to parse and pretty-print the alias definitions. The old
parser and pretty-printer was removed from CVS. I have re-used the new
pretty-printer, but no parser was available. I have reimplemented it
(in DisambiguatePp :-). The file should probably be renamed. Moreover,
the code implemented is very redundant with the corresponding one in
MatitaEngine
Claudio Sacerdoti Coen [Wed, 31 Aug 2005 09:35:52 +0000 (09:35 +0000)]
...
Claudio Sacerdoti Coen [Wed, 31 Aug 2005 08:28:29 +0000 (08:28 +0000)]
(** xxx **) ==> (** xxx *)
Claudio Sacerdoti Coen [Tue, 30 Aug 2005 10:39:57 +0000 (10:39 +0000)]
...
Claudio Sacerdoti Coen [Tue, 30 Aug 2005 10:15:44 +0000 (10:15 +0000)]
Bug fixed: "cic:/dummy_i" is an invalid URI (that used to be erroneously accepted by the getter). Changed to "cic:/dummy_i.con".
Claudio Sacerdoti Coen [Tue, 30 Aug 2005 09:39:12 +0000 (09:39 +0000)]
A parser for aliases implemented (required by the Whelp).
Claudio Sacerdoti Coen [Tue, 30 Aug 2005 08:51:02 +0000 (08:51 +0000)]
CicNotationPres self-reference (misteriously accepted by ocaml!!!)
Claudio Sacerdoti Coen [Mon, 29 Aug 2005 11:50:18 +0000 (11:50 +0000)]
WARNING: this commit changes the DB representation; you need to alter the
DB by hand after this commit
- h_position: varchar(255) -> varchar(62)
- h_sort: varchar(255) -> varchar(6)
- several indexes have been redesigned from scratch.
Motivation: MySql is not able to exploit two indexes (on columns c1 and c2)
in a query of type "select ... from table where c1 = x and c2 = y".
Only one of the two indexes is exploited; the other test is done on each line. My commit introduces a single index on c1 AND c2 in place of two indexes.
WARNING: it is sure that a few queries (e.g. match) are greatly optimized by
this commit. On the contrary I ignore if the performance of other queries drops.Moreover, I have not optimized yet the indexes over the hits and count tables.
Claudio Sacerdoti Coen [Mon, 29 Aug 2005 11:46:28 +0000 (11:46 +0000)]
WARNING: this commit changes the DB representation; you need to alter the
DB by hand after this commit
- h_position: varchar(255) -> varchar(62)
- h_sort: varchar(255) -> varchar(6)
- several indexes have been redesigned from scratch.
Motivation: MySql is not able to exploit two indexes (on columns c1 and c2)
in a query of type "select ... from table where c1 = x and c2 = y".
Only one of the two indexes is exploited; the other test is done on each line.
My commit introduces a single index on c1 AND c2 in place of two indexes.
WARNING: it is sure that a few queries (e.g. match) are greatly optimized by
this commit. On the contrary I ignore if the performance of other queries drops.
Moreover, I have not optimized yet the indexes over the hits and count tables.