]> matita.cs.unibo.it Git - helm.git/log
helm.git
15 years agounification hint
Enrico Tassi [Fri, 19 Dec 2008 10:07:15 +0000 (10:07 +0000)]
unification hint

15 years agolambda case fixed, used to not properly use the expected type if it was not a Prod
Enrico Tassi [Fri, 19 Dec 2008 10:06:31 +0000 (10:06 +0000)]
lambda case fixed, used to not properly use the expected type if it was not a Prod

15 years agoadded 'unification hint command to add a term to the new unification hint database'
Enrico Tassi [Fri, 19 Dec 2008 10:05:31 +0000 (10:05 +0000)]
added 'unification hint command to add a term to the new unification hint database'

15 years agoidentity coercions are not really inserted, just used as hints for unification
Enrico Tassi [Fri, 19 Dec 2008 10:02:11 +0000 (10:02 +0000)]
identity coercions are not really inserted, just used as hints for unification

15 years ago....
Enrico Tassi [Fri, 19 Dec 2008 09:59:54 +0000 (09:59 +0000)]
....

15 years agoStupid bug fixed: checkin the type in place of the sort during the refinement
Claudio Sacerdoti Coen [Thu, 18 Dec 2008 20:44:31 +0000 (20:44 +0000)]
Stupid bug fixed: checkin the type in place of the sort during the refinement
of axioms.

15 years agoNew disambiguation commented out.
Claudio Sacerdoti Coen [Thu, 18 Dec 2008 20:41:30 +0000 (20:41 +0000)]
New disambiguation commented out.

15 years agoMany axioms are now proved... using many more (but simpler) axioms.
Claudio Sacerdoti Coen [Thu, 18 Dec 2008 20:35:07 +0000 (20:35 +0000)]
Many axioms are now proved... using many more (but simpler) axioms.

15 years agoNow it compiles again.
Claudio Sacerdoti Coen [Thu, 18 Dec 2008 20:32:15 +0000 (20:32 +0000)]
Now it compiles again.

15 years agoRefinement of axioms fixed. We did not check that the declared type must be
Claudio Sacerdoti Coen [Thu, 18 Dec 2008 20:29:33 +0000 (20:29 +0000)]
Refinement of axioms fixed. We did not check that the declared type must be
a type.

15 years ago- improved logging
Ferruccio Guidi [Thu, 18 Dec 2008 15:48:01 +0000 (15:48 +0000)]
- improved logging
- we started to support the coercion for "non"

15 years ago...
Enrico Tassi [Thu, 18 Dec 2008 13:56:15 +0000 (13:56 +0000)]
...

15 years agowe started the support for the coercions "alle" and "alli"
Ferruccio Guidi [Wed, 17 Dec 2008 19:49:35 +0000 (19:49 +0000)]
we started the support for the coercions "alle" and "alli"

15 years agofoo overlap
Enrico Tassi [Wed, 17 Dec 2008 17:52:18 +0000 (17:52 +0000)]
foo overlap

15 years agofoo overlap
Enrico Tassi [Wed, 17 Dec 2008 17:35:00 +0000 (17:35 +0000)]
foo overlap

15 years ago....
Enrico Tassi [Wed, 17 Dec 2008 15:20:44 +0000 (15:20 +0000)]
....

15 years agowe added the implicit coercion for modus tollens
Ferruccio Guidi [Tue, 16 Dec 2008 19:32:12 +0000 (19:32 +0000)]
we added the implicit coercion for modus tollens

15 years agoprevious change was causing divergence
Enrico Tassi [Tue, 16 Dec 2008 19:14:01 +0000 (19:14 +0000)]
previous change was causing divergence

15 years agohints work better now
Enrico Tassi [Tue, 16 Dec 2008 19:13:46 +0000 (19:13 +0000)]
hints work better now

15 years agohints attached
Enrico Tassi [Tue, 16 Dec 2008 13:58:09 +0000 (13:58 +0000)]
hints attached

15 years agoremoved debug code
Enrico Tassi [Tue, 16 Dec 2008 13:57:22 +0000 (13:57 +0000)]
removed debug code

15 years agofixed a bug, it used to report o wrong is_normal bit in case of match
Enrico Tassi [Tue, 16 Dec 2008 13:54:37 +0000 (13:54 +0000)]
fixed a bug, it used to report o wrong is_normal bit in case of match

15 years agomake it compile again
Enrico Tassi [Tue, 16 Dec 2008 13:53:16 +0000 (13:53 +0000)]
make it compile again

15 years agowrap object_not_found
Enrico Tassi [Tue, 16 Dec 2008 13:53:03 +0000 (13:53 +0000)]
wrap object_not_found

15 years agomove the printing in the right place
Enrico Tassi [Tue, 16 Dec 2008 13:50:55 +0000 (13:50 +0000)]
move the printing in the right place

15 years agoauto expansion of \tex macros added as a switch in the edit menu
Enrico Tassi [Tue, 16 Dec 2008 09:00:31 +0000 (09:00 +0000)]
auto expansion of \tex macros added as a switch in the edit menu

15 years agoadded one asser
Enrico Tassi [Tue, 16 Dec 2008 09:00:17 +0000 (09:00 +0000)]
added one asser

15 years agodep.opt regenerated
Enrico Tassi [Mon, 15 Dec 2008 22:55:17 +0000 (22:55 +0000)]
dep.opt regenerated

15 years agoadded unification hints
Enrico Tassi [Mon, 15 Dec 2008 22:49:47 +0000 (22:49 +0000)]
added unification hints

15 years agoSome changes to the pullback test, for debugging
Wilmer Ricciotti [Mon, 15 Dec 2008 19:20:08 +0000 (19:20 +0000)]
Some changes to the pullback test, for debugging

15 years agoFirst attempt to implement unification hints.
Wilmer Ricciotti [Mon, 15 Dec 2008 19:16:46 +0000 (19:16 +0000)]
First attempt to implement unification hints.

15 years agouse named types to force some constraints asap
Enrico Tassi [Mon, 15 Dec 2008 13:50:17 +0000 (13:50 +0000)]
use named types to force some constraints asap

15 years agoTo check if a term is type, do a whd of its sort before matching it.
Enrico Tassi [Mon, 15 Dec 2008 13:26:26 +0000 (13:26 +0000)]
To check if a term is type, do a whd of its sort before matching it.

15 years ago...
Enrico Tassi [Mon, 15 Dec 2008 13:13:33 +0000 (13:13 +0000)]
...

15 years agoaded some whd, you should be able to declare something like:
Enrico Tassi [Mon, 15 Dec 2008 13:12:28 +0000 (13:12 +0000)]
aded some whd, you should be able to declare something like:
 inductive A : FOO :=
where FOO is defined as a sort (or unfolds to somthing ending with a sort)

15 years agoThe library is no longer automatically used during disambiguation.
Claudio Sacerdoti Coen [Sun, 14 Dec 2008 23:03:02 +0000 (23:03 +0000)]
The library is no longer automatically used during disambiguation.
It is used only when the user press the More button or when he selects
the library from the Debug (?????) menu!

Moreover, using the library or not using it has now a different behaviour:
1. when the library is NOT used, symbols with no preferences are no
   longer automatically looked for in the library
2. when the library IS used, only passes 5 and 6 are tried.

CONS:
1. the Debug menu voice is now necessary: move it elsewhere? always open the
   disambiguation window making it non-modal? put an hyperlink in the error
   message window?
2. when the library is used, all previous preferences are completely ignored
   for all symbols. This makes the system much more stupid than what it used
   to be.

PROS:
disambiguation is now much faster in case of error

15 years ago- improved logging
Ferruccio Guidi [Sun, 14 Dec 2008 22:22:38 +0000 (22:22 +0000)]
- improved logging
- optimization in subst

15 years agonew kernel basic_ag (with absolute local references)
Ferruccio Guidi [Sun, 14 Dec 2008 19:34:19 +0000 (19:34 +0000)]
new kernel basic_ag (with absolute local references)

15 years agowe are changing the kernel version from basic_rg to basic_ag
Ferruccio Guidi [Sun, 14 Dec 2008 15:32:38 +0000 (15:32 +0000)]
we are changing the kernel version from basic_rg to basic_ag

15 years agoautItem : the uris of the objects involved in the implicit coercions
Ferruccio Guidi [Sun, 14 Dec 2008 15:23:54 +0000 (15:23 +0000)]
autItem : the uris of the objects involved in the implicit coercions
log     : improved log output
basic_rg: improved logging

15 years ago...
Enrico Tassi [Sun, 14 Dec 2008 11:37:50 +0000 (11:37 +0000)]
...

15 years agoeureka! match x with [ C gfds fsd fdg fdg fdg fd gfdg fd => c ] is
Enrico Tassi [Sun, 14 Dec 2008 11:36:02 +0000 (11:36 +0000)]
eureka! match x with [ C gfds fsd fdg fdg fdg fd gfdg fd => c ] is
now correctly broken if too long

15 years ago...
Enrico Tassi [Sun, 14 Dec 2008 10:56:38 +0000 (10:56 +0000)]
...

15 years agoimproved type hierarchy management
Ferruccio Guidi [Fri, 12 Dec 2008 21:26:57 +0000 (21:26 +0000)]
improved type hierarchy management

15 years ago...
Claudio Sacerdoti Coen [Fri, 12 Dec 2008 14:59:39 +0000 (14:59 +0000)]
...

15 years agoA parser (and a scanner) to import "~C" files into Matita.
Claudio Sacerdoti Coen [Fri, 12 Dec 2008 14:41:18 +0000 (14:41 +0000)]
A parser (and a scanner) to import "~C" files into Matita.

15 years ago...
Claudio Sacerdoti Coen [Fri, 12 Dec 2008 14:32:56 +0000 (14:32 +0000)]
...

15 years ago1. new expressions AND, OR, XOR
Claudio Sacerdoti Coen [Fri, 12 Dec 2008 14:32:23 +0000 (14:32 +0000)]
1. new expressions AND, OR, XOR
2. constants must be initialized
3. bug fixed: the size n of an array means that the array
   has (n+1) elements. Thus the last valid index is n.

15 years ago- basic_rg: architectural bug fix
Ferruccio Guidi [Fri, 12 Dec 2008 12:45:42 +0000 (12:45 +0000)]
- basic_rg: architectural bug fix
- toplevel: more help on command line options

15 years ago...
Enrico Tassi [Fri, 12 Dec 2008 12:14:31 +0000 (12:14 +0000)]
...

15 years agoadded some hardcoded universes, needed to typecheck the librarya
Enrico Tassi [Fri, 12 Dec 2008 12:11:54 +0000 (12:11 +0000)]
added some hardcoded universes, needed to typecheck the librarya

15 years agoadded some messages
Enrico Tassi [Fri, 12 Dec 2008 12:11:12 +0000 (12:11 +0000)]
added some messages

15 years agoadded empty_db, usefull to avoid translating all old coercions to obtain a new coerci...
Enrico Tassi [Fri, 12 Dec 2008 12:10:40 +0000 (12:10 +0000)]
added empty_db, usefull to avoid translating all old coercions to obtain a new coercions database

15 years agouse new function to clear caches so that objects are translated correctly
Enrico Tassi [Fri, 12 Dec 2008 12:09:25 +0000 (12:09 +0000)]
use new function to clear caches so that objects are translated correctly

15 years agobetter error message, functions to clear various caches used during translation of...
Enrico Tassi [Fri, 12 Dec 2008 12:08:16 +0000 (12:08 +0000)]
better error message, functions to clear various caches used during translation of old2new cic, logger functions implemented inside the typechecker so that
uncommenting one line makes the type checker more verbose when recursively typing stuff

15 years agothanks to the new fixes to notation, I can define \land for both pairs and
Enrico Tassi [Fri, 12 Dec 2008 11:59:32 +0000 (11:59 +0000)]
thanks to the new fixes to notation, I can define \land for both pairs and
families!

15 years agopattern matching over Ast.Case simplified:
Enrico Tassi [Fri, 12 Dec 2008 11:43:39 +0000 (11:43 +0000)]
pattern matching over Ast.Case simplified:
- it is not possible to match the `in intype` foo part
- it is not possible to match the outtype
eta expansion still done on Ast (hard to do it in Cic. The abstracted variable
does not have a type :-( thus you may have to double your notations. Name for fresh variables changed to \etaX instead of freshX (much shorter in terms of screen pixels)

15 years agoApplications are now processed from left to right.
Claudio Sacerdoti Coen [Thu, 11 Dec 2008 22:39:49 +0000 (22:39 +0000)]
Applications are now processed from left to right.

15 years agosupport for mathml mpadded tag added (allows to overlap two subsequent symbols);...
Enrico Tassi [Thu, 11 Dec 2008 14:16:47 +0000 (14:16 +0000)]
support for mathml mpadded tag added (allows to overlap two subsequent symbols); you library needs to be recompiled after the update

15 years ago- new semantic log system
Ferruccio Guidi [Wed, 10 Dec 2008 21:25:21 +0000 (21:25 +0000)]
- new semantic log system
- new brg terms
- bug fix im MetaOutput and MetaBrg

15 years agofirst version of kernel "basic_rg"
Ferruccio Guidi [Tue, 9 Dec 2008 19:26:31 +0000 (19:26 +0000)]
first version of kernel "basic_rg"
the grundlagen does not typecheck because some of its lambdas are actually pis
(this is known)

15 years agonew disambiguatio attached with the right universe graph
Enrico Tassi [Tue, 9 Dec 2008 18:30:47 +0000 (18:30 +0000)]
new disambiguatio attached with the right universe graph

15 years agoraise uncertain when a sort is not a sort but may be, use max for mixing universes...
Enrico Tassi [Tue, 9 Dec 2008 18:29:37 +0000 (18:29 +0000)]
raise uncertain when a sort is not a sort but may be, use max for mixing universes, coercions to sort in the prod case are triggered before putting the source type in the context

15 years agobetter coercions indexing and lookup
Enrico Tassi [Tue, 9 Dec 2008 18:27:31 +0000 (18:27 +0000)]
better coercions indexing and lookup

15 years agobetter max function (instead of @) for combining universes
Enrico Tassi [Tue, 9 Dec 2008 18:26:18 +0000 (18:26 +0000)]
better max function (instead of @) for combining universes

15 years agoadded an exception
Enrico Tassi [Tue, 9 Dec 2008 18:25:04 +0000 (18:25 +0000)]
added an exception

15 years agosaturate was not returning the correct (saturated) type (but a correct number
Enrico Tassi [Tue, 9 Dec 2008 16:04:20 +0000 (16:04 +0000)]
saturate was not returning the correct (saturated) type (but a correct number
of arguments.

15 years agofixed notation
Enrico Tassi [Tue, 9 Dec 2008 12:18:42 +0000 (12:18 +0000)]
fixed notation

15 years ago...
Enrico Tassi [Tue, 9 Dec 2008 09:02:08 +0000 (09:02 +0000)]
...

15 years agooption to collapse all tex macros implemented
Enrico Tassi [Tue, 9 Dec 2008 08:57:39 +0000 (08:57 +0000)]
option to collapse all tex macros implemented

15 years agoA (boring and long) once-in-a-life exercise on proving a trivial property
Claudio Sacerdoti Coen [Mon, 8 Dec 2008 23:56:26 +0000 (23:56 +0000)]
A (boring and long) once-in-a-life exercise on proving a trivial property
on natural numbers in ND.

15 years agonon active but almost working implementation of \TeX macro substitution while colouring
Enrico Tassi [Mon, 8 Dec 2008 17:07:45 +0000 (17:07 +0000)]
non active but almost working implementation of \TeX macro substitution while colouring

15 years agoalt-l for not working nymore for \fox where x was in an eq class
Enrico Tassi [Mon, 8 Dec 2008 14:02:53 +0000 (14:02 +0000)]
alt-l for not working nymore for \fox where x was in an eq class

15 years ago...
Enrico Tassi [Mon, 8 Dec 2008 12:50:12 +0000 (12:50 +0000)]
...

15 years agobetter replacement for \\def
Enrico Tassi [Mon, 8 Dec 2008 09:34:40 +0000 (09:34 +0000)]
better replacement for \\def

15 years ago3.5 -> 35 to help crappy latex
Enrico Tassi [Mon, 8 Dec 2008 00:39:33 +0000 (00:39 +0000)]
3.5 -> 35 to help crappy latex

15 years agoNew exception considered.
Claudio Sacerdoti Coen [Sun, 7 Dec 2008 19:05:29 +0000 (19:05 +0000)]
New exception considered.

15 years agoBug fixed: every time we form a Prod, we must typecheck it before trying
Claudio Sacerdoti Coen [Sun, 7 Dec 2008 18:57:15 +0000 (18:57 +0000)]
Bug fixed: every time we form a Prod, we must typecheck it before trying
unification.

15 years agobetter images
Enrico Tassi [Sun, 7 Dec 2008 18:05:18 +0000 (18:05 +0000)]
better images

15 years agonew concept of virtuals, defined only in the gui that behave as the old (still
Enrico Tassi [Sat, 6 Dec 2008 18:12:03 +0000 (18:12 +0000)]
new concept of virtuals, defined only in the gui that behave as the old (still
present and used) ut8tables and ligatures mapping \foo or => to unicode symbols.

support in the gui for utomatical sumstitution of a virtual with its utf8
counterpart (disabled now, grep for if false && str = " ").

support for uf8 equivalnce classes, names simalrsymbols, activated by alt-l.

alt-l is now overloaded, can expand a \foo or => to a unicode symbol
and cycle on unicode symbols in the same eq class of the one just before the
cursor.

classes are already defined for letters mapping them to variants (other
alfabets and fonts) and on arrows and <.

15 years agometaAut: now we use hash tables properly (processing time: 2s)
Ferruccio Guidi [Sat, 6 Dec 2008 14:42:09 +0000 (14:42 +0000)]
metaAut: now we use hash tables properly (processing time: 2s)
brgReduction: started

15 years agonew exception captured
Claudio Sacerdoti Coen [Fri, 5 Dec 2008 23:27:42 +0000 (23:27 +0000)]
new exception captured

15 years agoAdded new syntax Type[n] where n is a number.
Claudio Sacerdoti Coen [Fri, 5 Dec 2008 23:12:39 +0000 (23:12 +0000)]
Added new syntax  Type[n]  where n is a number.
The old kernel interpretes it simply as Type.
The new kernel as Typen.

15 years agoDebugging code removed.
Claudio Sacerdoti Coen [Fri, 5 Dec 2008 23:03:13 +0000 (23:03 +0000)]
Debugging code removed.

15 years agoDebugging code removed
Claudio Sacerdoti Coen [Fri, 5 Dec 2008 23:02:44 +0000 (23:02 +0000)]
Debugging code removed

15 years agoDebugging instructions removed.
Claudio Sacerdoti Coen [Fri, 5 Dec 2008 23:02:10 +0000 (23:02 +0000)]
Debugging instructions removed.

15 years agoUseless code removed.
Claudio Sacerdoti Coen [Fri, 5 Dec 2008 21:58:39 +0000 (21:58 +0000)]
Useless code removed.

15 years agoUseless code removed.
Claudio Sacerdoti Coen [Fri, 5 Dec 2008 21:42:51 +0000 (21:42 +0000)]
Useless code removed.

15 years agoUseless code removed.
Claudio Sacerdoti Coen [Fri, 5 Dec 2008 21:35:37 +0000 (21:35 +0000)]
Useless code removed.

15 years agoCase-sensitive fixes.
Claudio Sacerdoti Coen [Fri, 5 Dec 2008 20:34:51 +0000 (20:34 +0000)]
Case-sensitive fixes.

15 years agoa few missing ~subst added to whd
Enrico Tassi [Fri, 5 Dec 2008 18:16:01 +0000 (18:16 +0000)]
a few missing ~subst added to whd

15 years agocoercions are there, but not heavily tested
Enrico Tassi [Fri, 5 Dec 2008 18:10:04 +0000 (18:10 +0000)]
coercions are there, but not heavily tested

15 years agoif todo_dom was [] disambiguation was performed twice
Enrico Tassi [Fri, 5 Dec 2008 18:07:54 +0000 (18:07 +0000)]
if todo_dom was [] disambiguation was performed twice

15 years agostill commented, but benchmarks the new/old disambigution process
Enrico Tassi [Fri, 5 Dec 2008 11:29:39 +0000 (11:29 +0000)]
still commented, but benchmarks the new/old disambigution process

15 years agodisambiguation takes ~mk_localization_tbl and not ~localization_tbl, thus can
Enrico Tassi [Fri, 5 Dec 2008 11:26:55 +0000 (11:26 +0000)]
disambiguation takes ~mk_localization_tbl and not ~localization_tbl, thus can
create the hashtable when needed (after every interpretation)

15 years agoraise failure instead of uncertain if two terms are meta closed
Enrico Tassi [Fri, 5 Dec 2008 11:23:48 +0000 (11:23 +0000)]
raise failure instead of uncertain if two terms are meta closed

15 years agoFixes:
Enrico Tassi [Thu, 4 Dec 2008 19:36:50 +0000 (19:36 +0000)]
Fixes:
 - new letin interpretation in nCicDisambiguation fixed to not swap arguments
 - new refiner raises good exception in case a sort in not such
 - reference raise good exception instead of assert false
Improvements:
 - new disambiguation attached
 - constructor -> indty function in reference

15 years agohousekeeping:
Enrico Tassi [Thu, 4 Dec 2008 13:06:29 +0000 (13:06 +0000)]
housekeeping:
- unused functions removed from Enviconment
- unused file disambiguatePp removed
- disambiguation callbacks pushed down to Disambiguate
  from MultiPassDisambiguator

15 years agoBug fixed: pretty-printing of aliases when the OK button is pressed in the
Enrico Tassi [Thu, 4 Dec 2008 12:36:44 +0000 (12:36 +0000)]
Bug fixed: pretty-printing of aliases when the OK button is pressed in the
ambiguous error window.

15 years agoThe aliases and multi_aliases in the lexicon status are now
Claudio Sacerdoti Coen [Wed, 3 Dec 2008 22:53:01 +0000 (22:53 +0000)]
The aliases and multi_aliases in the lexicon status are now
LexiconAst.alias_spec (they used to be DisambiguateTypes.codomain_item).
Benefit: it is now possible to use the very same set of aliases both for the
new and old terms. Moreover, this commits simplifies quite a bit of code.

Known bugs: when clicking the "OK" button in the error message disambiguation
window, an assertion is often raised. When the assertion is not raised, the
set of aliases that are printed are not syntactically/semantically correct.