NEWS
author wenzelm
Fri Aug 10 00:20:39 2007 +0200 (2007-08-10)
changeset 24210 a865059c4fcb
parent 24187 8bdf5ca5871f
child 24213 71c57c5099d6
permissions -rw-r--r--
* Experimental support for multithreading, using Poly/ML 5.1;
     1 Isabelle NEWS -- history user-relevant changes
     2 ==============================================
     3 
     4 New in this Isabelle version
     5 ----------------------------
     6 
     7 *** General ***
     8 
     9 * More uniform information about legacy features, notably a
    10 warning/error of "Legacy feature: ...", depending on the state of the
    11 tolerate_legacy_features flag (default true). FUTURE INCOMPATIBILITY:
    12 legacy features will disappear eventually.
    13 
    14 * Theory syntax: the header format ``theory A = B + C:'' has been
    15 discontinued in favour of ``theory A imports B C begin''.  Use isatool
    16 fixheaders to convert existing theory files.  INCOMPATIBILITY.
    17 
    18 * Theory syntax: the old non-Isar theory file format has been
    19 discontinued altogether.  Note that ML proof scripts may still be used
    20 with Isar theories; migration is usually quite simple with the ML
    21 function use_legacy_bindings.  INCOMPATIBILITY.
    22 
    23 * Theory syntax: some popular names (e.g. 'class', 'declaration',
    24 'fun', 'help', 'if') are now keywords.  INCOMPATIBILITY, use double
    25 quotes.
    26 
    27 * Theory loader: be more serious about observing the static theory
    28 header specifications (including optional directories), but not the
    29 accidental file locations of previously successful loads.  The strict
    30 update policy of former update_thy is now already performed by
    31 use_thy, so the former has been removed; use_thys updates several
    32 theories simultaneously, just as 'imports' within a theory header
    33 specification, but without merging the results.  Potential
    34 INCOMPATIBILITY: may need to refine theory headers and commands
    35 ROOT.ML which depend on load order.
    36 
    37 * Theory loader: optional support for content-based file
    38 identification, instead of the traditional scheme of full physical
    39 path plus date stamp; configured by the ISABELLE_FILE_IDENT setting
    40 (cf. the system manual).  The new scheme allows to work with
    41 non-finished theories in persistent session images, such that source
    42 files may be moved later on without requiring reloads.
    43 
    44 * Theory loader: old-style ML proof scripts being *attached* to a thy
    45 file (with the same base name as the theory) are considered a legacy
    46 feature, which will disappear eventually. Even now, the theory loader no
    47 longer maintains dependencies on such files.
    48 
    49 * Legacy goal package: reduced interface to the bare minimum required
    50 to keep existing proof scripts running.  Most other user-level
    51 functions are now part of the OldGoals structure, which is *not* open
    52 by default (consider isatool expandshort before open OldGoals).
    53 Removed top_sg, prin, printyp, pprint_term/typ altogether, because
    54 these tend to cause confusion about the actual goal (!) context being
    55 used here, which is not necessarily the same as the_context().
    56 
    57 * Command 'find_theorems': supports "*" wild-card in "name:"
    58 criterion; "with_dups" option.  Certain ProofGeneral versions might
    59 support a specific search form (see ProofGeneral/CHANGES).
    60 
    61 * The ``prems limit'' option (cf. ProofContext.prems_limit) is now -1
    62 by default, which means that "prems" (and also "fixed variables") are
    63 suppressed from proof state output.  Note that the ProofGeneral
    64 settings mechanism allows to change and save options persistently, but
    65 older versions of Isabelle will fail to start up if a negative prems
    66 limit is imposed.
    67 
    68 * Local theory targets may be specified by non-nested blocks of
    69 ``context/locale/class ... begin'' followed by ``end''.  The body may
    70 contain definitions, theorems etc., including any derived mechanism
    71 that has been implemented on top of these primitives.  This concept
    72 generalizes the existing ``theorem (in ...)'' towards more versatility
    73 and scalability.
    74 
    75 * Proof General interface: proper undo of final 'end' command;
    76 discontinued Isabelle/classic mode (ML proof scripts).
    77 
    78 
    79 *** Document preparation ***
    80 
    81 * Added antiquotation @{theory name} which prints the given name,
    82 after checking that it refers to a valid ancestor theory in the
    83 current context.
    84 
    85 * Added antiquotations @{ML_type text} and @{ML_struct text} which
    86 check the given source text as ML type/structure, printing verbatim.
    87 
    88 * Added antiquotation @{abbrev "c args"} which prints the abbreviation
    89 "c args == rhs" given in the current context.  (Any number of
    90 arguments may be given on the LHS.)
    91 
    92 
    93 
    94 *** Pure ***
    95 
    96 * code generator: consts in 'consts_code' Isar commands are now referred
    97   to by usual term syntax (including optional type annotations).
    98 
    99 * code generator: 
   100   - Isar 'definition's, 'constdef's and primitive instance definitions are added
   101     explicitly to the table of defining equations
   102   - primitive definitions are not used as defining equations by default any longer
   103   - defining equations are now definitly restricted to meta "==" and object
   104         equality "="
   105   - HOL theories have been adopted accordingly
   106 
   107 * class_package.ML offers a combination of axclasses and locales to
   108 achieve Haskell-like type classes in Isabelle.  See
   109 HOL/ex/Classpackage.thy for examples.
   110 
   111 * Yet another code generator framework allows to generate executable
   112 code for ML and Haskell (including "class"es).  A short usage sketch:
   113 
   114     internal compilation:
   115         code_gen <list of constants (term syntax)> in SML
   116     writing SML code to a file:
   117         code_gen <list of constants (term syntax)> in SML <filename>
   118     writing OCaml code to a file:
   119         code_gen <list of constants (term syntax)> in OCaml <filename>
   120     writing Haskell code to a bunch of files:
   121         code_gen <list of constants (term syntax)> in Haskell <filename>
   122 
   123 Reasonable default setup of framework in HOL/Main.
   124 
   125 Theorem attributs for selecting and transforming function equations theorems:
   126 
   127     [code fun]:        select a theorem as function equation for a specific constant
   128     [code fun del]:    deselect a theorem as function equation for a specific constant
   129     [code inline]:     select an equation theorem for unfolding (inlining) in place
   130     [code inline del]: deselect an equation theorem for unfolding (inlining) in place
   131 
   132 User-defined serializations (target in {SML, OCaml, Haskell}):
   133 
   134     code_const <and-list of constants (term syntax)>
   135       {(target) <and-list of const target syntax>}+
   136 
   137     code_type <and-list of type constructors>
   138       {(target) <and-list of type target syntax>}+
   139 
   140     code_instance <and-list of instances>
   141       {(target)}+
   142         where instance ::= <type constructor> :: <class>
   143 
   144     code_class <and_list of classes>
   145       {(target) <and-list of class target syntax>}+
   146         where class target syntax ::= <class name> {where {<classop> == <target syntax>}+}?
   147 
   148 code_instance and code_class only apply to target Haskell.
   149 
   150 See HOL theories and HOL/ex/Codegenerator*.thy for usage examples.
   151 Doc/Isar/Advanced/Codegen/ provides a tutorial.
   152 
   153 * Command 'no_translations' removes translation rules from theory
   154 syntax.
   155 
   156 * Overloaded definitions are now actually checked for acyclic
   157 dependencies.  The overloading scheme is slightly more general than
   158 that of Haskell98, although Isabelle does not demand an exact
   159 correspondence to type class and instance declarations.
   160 INCOMPATIBILITY, use ``defs (unchecked overloaded)'' to admit more
   161 exotic versions of overloading -- at the discretion of the user!
   162 
   163 Polymorphic constants are represented via type arguments, i.e. the
   164 instantiation that matches an instance against the most general
   165 declaration given in the signature.  For example, with the declaration
   166 c :: 'a => 'a => 'a, an instance c :: nat => nat => nat is represented
   167 as c(nat).  Overloading is essentially simultaneous structural
   168 recursion over such type arguments.  Incomplete specification patterns
   169 impose global constraints on all occurrences, e.g. c('a * 'a) on the
   170 LHS means that more general c('a * 'b) will be disallowed on any RHS.
   171 Command 'print_theory' outputs the normalized system of recursive
   172 equations, see section "definitions".
   173 
   174 * Configuration options are maintained within the theory or proof
   175 context (with name and type bool/int/string), providing a very simple
   176 interface to a poor-man's version of general context data.  Tools may
   177 declare options in ML (e.g. using Attrib.config_int) and then refer to
   178 these values using Config.get etc.  Users may change options via an
   179 associated attribute of the same name.  This form of context
   180 declaration works particularly well with commands 'declare' or
   181 'using', for example ``declare [[foo = 42]]''.  Thus it has become
   182 very easy to avoid global references, which would not observe Isar
   183 toplevel undo/redo and fail to work with multithreading.
   184 
   185 Various global ML references of Pure and HOL have been turned into
   186 configuration options:
   187 
   188   Unify.search_bound		unify_search_bound
   189   Unify.trace_bound		unify_trace_bound
   190   Unify.trace_simp		unify_trace_simp
   191   Unify.trace_types		unify_trace_types
   192   Simplifier.simp_depth_limit	simp_depth_limit
   193   Blast.depth_limit		blast_depth_limit
   194   DatatypeProp.dtK		datatype_distinctness_limit
   195   fast_arith_neq_limit  	fast_arith_neq_limit
   196   fast_arith_split_limit	fast_arith_split_limit
   197 
   198 * Named collections of theorems may be easily installed as context
   199 data using the functor NamedThmsFun (see
   200 src/Pure/Tools/named_thms.ML).  The user may add or delete facts via
   201 attributes; there is also a toplevel print command.  This facility is
   202 just a common case of general context data, which is the preferred way
   203 for anything more complex than just a list of facts in canonical
   204 order.
   205 
   206 * Isar: command 'declaration' augments a local theory by generic
   207 declaration functions written in ML.  This enables arbitrary content
   208 being added to the context, depending on a morphism that tells the
   209 difference of the original declaration context wrt. the application
   210 context encountered later on.
   211 
   212 * Isar: proper interfaces for simplification procedures.  Command
   213 'simproc_setup' declares named simprocs (with match patterns, and body
   214 text in ML).  Attribute "simproc" adds/deletes simprocs in the current
   215 context.  ML antiquotation @{simproc name} retrieves named simprocs.
   216 
   217 * Isar: an extra pair of brackets around attribute declarations
   218 abbreviates a theorem reference involving an internal dummy fact,
   219 which will be ignored later --- only the effect of the attribute on
   220 the background context will persist.  This form of in-place
   221 declarations is particularly useful with commands like 'declare' and
   222 'using', for example ``have A using [[simproc a]] by simp''.
   223 
   224 * Isar: method "assumption" (and implicit closing of subproofs) now
   225 takes simple non-atomic goal assumptions into account: after applying
   226 an assumption as a rule the resulting subgoals are solved by atomic
   227 assumption steps.  This is particularly useful to finish 'obtain'
   228 goals, such as "!!x. (!!x. P x ==> thesis) ==> P x ==> thesis",
   229 without referring to the original premise "!!x. P x ==> thesis" in the
   230 Isar proof context.  POTENTIAL INCOMPATIBILITY: method "assumption" is
   231 more permissive.
   232 
   233 * Isar: implicit use of prems from the Isar proof context is
   234 considered a legacy feature.  Common applications like ``have A .''
   235 may be replaced by ``have A by fact'' or ``note `A`''.  In general,
   236 referencing facts explicitly here improves readability and
   237 maintainability of proof texts.
   238 
   239 * Isar: improper proof element 'guess' is like 'obtain', but derives
   240 the obtained context from the course of reasoning!  For example:
   241 
   242   assume "EX x y. A x & B y"   -- "any previous fact"
   243   then guess x and y by clarify
   244 
   245 This technique is potentially adventurous, depending on the facts and
   246 proof tools being involved here.
   247 
   248 * Isar: known facts from the proof context may be specified as literal
   249 propositions, using ASCII back-quote syntax.  This works wherever
   250 named facts used to be allowed so far, in proof commands, proof
   251 methods, attributes etc.  Literal facts are retrieved from the context
   252 according to unification of type and term parameters.  For example,
   253 provided that "A" and "A ==> B" and "!!x. P x ==> Q x" are known
   254 theorems in the current context, then these are valid literal facts:
   255 `A` and `A ==> B` and `!!x. P x ==> Q x" as well as `P a ==> Q a` etc.
   256 
   257 There is also a proof method "fact" which does the same composition
   258 for explicit goal states, e.g. the following proof texts coincide with
   259 certain special cases of literal facts:
   260 
   261   have "A" by fact                 ==  note `A`
   262   have "A ==> B" by fact           ==  note `A ==> B`
   263   have "!!x. P x ==> Q x" by fact  ==  note `!!x. P x ==> Q x`
   264   have "P a ==> Q a" by fact       ==  note `P a ==> Q a`
   265 
   266 * Isar: ":" (colon) is no longer a symbolic identifier character in
   267 outer syntax.  Thus symbolic identifiers may be used without
   268 additional white space in declarations like this: ``assume *: A''.
   269 
   270 * Isar: 'print_facts' prints all local facts of the current context,
   271 both named and unnamed ones.
   272 
   273 * Isar: 'def' now admits simultaneous definitions, e.g.:
   274 
   275   def x == "t" and y == "u"
   276 
   277 * Isar: added command 'unfolding', which is structurally similar to
   278 'using', but affects both the goal state and facts by unfolding given
   279 rewrite rules.  Thus many occurrences of the 'unfold' method or
   280 'unfolded' attribute may be replaced by first-class proof text.
   281 
   282 * Isar: methods 'unfold' / 'fold', attributes 'unfolded' / 'folded',
   283 and command 'unfolding' now all support object-level equalities
   284 (potentially conditional).  The underlying notion of rewrite rule is
   285 analogous to the 'rule_format' attribute, but *not* that of the
   286 Simplifier (which is usually more generous).
   287 
   288 * Isar: the goal restriction operator [N] (default N = 1) evaluates a
   289 method expression within a sandbox consisting of the first N
   290 sub-goals, which need to exist.  For example, ``simp_all [3]''
   291 simplifies the first three sub-goals, while (rule foo, simp_all)[]
   292 simplifies all new goals that emerge from applying rule foo to the
   293 originally first one.
   294 
   295 * Isar: schematic goals are no longer restricted to higher-order
   296 patterns; e.g. ``lemma "?P(?x)" by (rule TrueI)'' now works as
   297 expected.
   298 
   299 * Isar: the conclusion of a long theorem statement is now either
   300 'shows' (a simultaneous conjunction, as before), or 'obtains'
   301 (essentially a disjunction of cases with local parameters and
   302 assumptions).  The latter allows to express general elimination rules
   303 adequately; in this notation common elimination rules look like this:
   304 
   305   lemma exE:    -- "EX x. P x ==> (!!x. P x ==> thesis) ==> thesis"
   306     assumes "EX x. P x"
   307     obtains x where "P x"
   308 
   309   lemma conjE:  -- "A & B ==> (A ==> B ==> thesis) ==> thesis"
   310     assumes "A & B"
   311     obtains A and B
   312 
   313   lemma disjE:  -- "A | B ==> (A ==> thesis) ==> (B ==> thesis) ==> thesis"
   314     assumes "A | B"
   315     obtains
   316       A
   317     | B
   318 
   319 The subsequent classical rules even refer to the formal "thesis"
   320 explicitly:
   321 
   322   lemma classical:     -- "(~ thesis ==> thesis) ==> thesis"
   323     obtains "~ thesis"
   324 
   325   lemma Peirce's_Law:  -- "((thesis ==> something) ==> thesis) ==> thesis"
   326     obtains "thesis ==> something"
   327 
   328 The actual proof of an 'obtains' statement is analogous to that of the
   329 Isar proof element 'obtain', only that there may be several cases.
   330 Optional case names may be specified in parentheses; these will be
   331 available both in the present proof and as annotations in the
   332 resulting rule, for later use with the 'cases' method (cf. attribute
   333 case_names).
   334 
   335 * Isar: the assumptions of a long theorem statement are available as
   336 "assms" fact in the proof context.  This is more appropriate than the
   337 (historical) "prems", which refers to all assumptions of the current
   338 context, including those from the target locale, proof body etc.
   339 
   340 * Isar: 'print_statement' prints theorems from the current theory or
   341 proof context in long statement form, according to the syntax of a
   342 top-level lemma.
   343 
   344 * Isar: 'obtain' takes an optional case name for the local context
   345 introduction rule (default "that").
   346 
   347 * Isar: removed obsolete 'concl is' patterns.  INCOMPATIBILITY, use
   348 explicit (is "_ ==> ?foo") in the rare cases where this still happens
   349 to occur.
   350 
   351 * Pure: syntax "CONST name" produces a fully internalized constant
   352 according to the current context.  This is particularly useful for
   353 syntax translations that should refer to internal constant
   354 representations independently of name spaces.
   355 
   356 * Pure: syntax constant for foo (binder "FOO ") is called "foo_binder"
   357 instead of "FOO ". This allows multiple binder declarations to coexist
   358 in the same context.  INCOMPATIBILITY.
   359 
   360 * Isar/locales: 'notation' provides a robust interface to the 'syntax'
   361 primitive that also works in a locale context (both for constants and
   362 fixed variables).  Type declaration and internal syntactic
   363 representation of given constants retrieved from the context.
   364 
   365 * Isar/locales: new derived specification elements 'axiomatization',
   366 'definition', 'abbreviation', which support type-inference, admit
   367 object-level specifications (equality, equivalence).  See also the
   368 isar-ref manual.  Examples:
   369 
   370   axiomatization
   371     eq  (infix "===" 50) where
   372     eq_refl: "x === x" and eq_subst: "x === y ==> P x ==> P y"
   373 
   374   definition "f x y = x + y + 1"
   375   definition g where "g x = f x x"
   376 
   377   abbreviation
   378     neq  (infix "=!=" 50) where
   379     "x =!= y == ~ (x === y)"
   380 
   381 These specifications may be also used in a locale context.  Then the
   382 constants being introduced depend on certain fixed parameters, and the
   383 constant name is qualified by the locale base name.  An internal
   384 abbreviation takes care for convenient input and output, making the
   385 parameters implicit and using the original short name.  See also
   386 HOL/ex/Abstract_NAT.thy for an example of deriving polymorphic
   387 entities from a monomorphic theory.
   388 
   389 Presently, abbreviations are only available 'in' a target locale, but
   390 not inherited by general import expressions.  Also note that
   391 'abbreviation' may be used as a type-safe replacement for 'syntax' +
   392 'translations' in common applications.
   393 
   394 Concrete syntax is attached to specified constants in internal form,
   395 independently of name spaces.  The parse tree representation is
   396 slightly different -- use 'notation' instead of raw 'syntax', and
   397 'translations' with explicit "CONST" markup to accommodate this.
   398 
   399 * Pure: command 'print_abbrevs' prints all constant abbreviations of
   400 the current context.  Print mode "no_abbrevs" prevents inversion of
   401 abbreviations on output.
   402 
   403 * Isar/locales: improved parameter handling:
   404 - use of locales "var" and "struct" no longer necessary;
   405 - parameter renamings are no longer required to be injective.
   406   This enables, for example, to define a locale for endomorphisms thus:
   407   locale endom = homom mult mult h.
   408 
   409 * Isar/locales: changed the way locales with predicates are defined.
   410 Instead of accumulating the specification, the imported expression is
   411 now an interpretation.  INCOMPATIBILITY: different normal form of
   412 locale expressions.  In particular, in interpretations of locales with
   413 predicates, goals repesenting already interpreted fragments are not
   414 removed automatically.  Use methods `intro_locales' and
   415 `unfold_locales'; see below.
   416 
   417 * Isar/locales: new methods `intro_locales' and `unfold_locales'
   418 provide backward reasoning on locales predicates.  The methods are
   419 aware of interpretations and discharge corresponding goals.
   420 `intro_locales' is less aggressive then `unfold_locales' and does not
   421 unfold predicates to assumptions.
   422 
   423 * Isar/locales: the order in which locale fragments are accumulated
   424 has changed.  This enables to override declarations from fragments due
   425 to interpretations -- for example, unwanted simp rules.
   426 
   427 * Isar/locales: interpretation in theories and proof contexts has been
   428 extended.  One may now specify (and prove) equations, which are
   429 unfolded in interpreted theorems.  This is useful for replacing
   430 defined concepts (constants depending on locale parameters) by
   431 concepts already existing in the target context.  Example:
   432 
   433   interpretation partial_order ["op <= :: [int, int] => bool"]
   434     where "partial_order.less (op <=) (x::int) y = (x < y)"
   435 
   436 Typically, the constant `partial_order.less' is created by a definition
   437 specification element in the context of locale partial_order.
   438 
   439 * Provers/induct: improved internal context management to support
   440 local fixes and defines on-the-fly.  Thus explicit meta-level
   441 connectives !! and ==> are rarely required anymore in inductive goals
   442 (using object-logic connectives for this purpose has been long
   443 obsolete anyway).  The subsequent proof patterns illustrate advanced
   444 techniques of natural induction; general datatypes and inductive sets
   445 work analogously (see also src/HOL/Lambda for realistic examples).
   446 
   447 (1) This is how to ``strengthen'' an inductive goal wrt. certain
   448 parameters:
   449 
   450   lemma
   451     fixes n :: nat and x :: 'a
   452     assumes a: "A n x"
   453     shows "P n x"
   454     using a                     -- {* make induct insert fact a *}
   455   proof (induct n arbitrary: x) -- {* generalize goal to "!!x. A n x ==> P n x" *}
   456     case 0
   457     show ?case sorry
   458   next
   459     case (Suc n)
   460     note `!!x. A n x ==> P n x` -- {* induction hypothesis, according to induction rule *}
   461     note `A (Suc n) x`          -- {* induction premise, stemming from fact a *}
   462     show ?case sorry
   463   qed
   464 
   465 (2) This is how to perform induction over ``expressions of a certain
   466 form'', using a locally defined inductive parameter n == "a x"
   467 together with strengthening (the latter is usually required to get
   468 sufficiently flexible induction hypotheses):
   469 
   470   lemma
   471     fixes a :: "'a => nat"
   472     assumes a: "A (a x)"
   473     shows "P (a x)"
   474     using a
   475   proof (induct n == "a x" arbitrary: x)
   476     ...
   477 
   478 See also HOL/Isar_examples/Puzzle.thy for an application of the this
   479 particular technique.
   480 
   481 (3) This is how to perform existential reasoning ('obtains' or
   482 'obtain') by induction, while avoiding explicit object-logic
   483 encodings:
   484 
   485   lemma
   486     fixes n :: nat
   487     obtains x :: 'a where "P n x" and "Q n x"
   488   proof (induct n arbitrary: thesis)
   489     case 0
   490     obtain x where "P 0 x" and "Q 0 x" sorry
   491     then show thesis by (rule 0)
   492   next
   493     case (Suc n)
   494     obtain x where "P n x" and "Q n x" by (rule Suc.hyps)
   495     obtain x where "P (Suc n) x" and "Q (Suc n) x" sorry
   496     then show thesis by (rule Suc.prems)
   497   qed
   498 
   499 Here the 'arbitrary: thesis' specification essentially modifies the
   500 scope of the formal thesis parameter, in order to the get the whole
   501 existence statement through the induction as expected.
   502 
   503 * Provers/induct: mutual induction rules are now specified as a list
   504 of rule sharing the same induction cases.  HOL packages usually
   505 provide foo_bar.inducts for mutually defined items foo and bar
   506 (e.g. inductive sets or datatypes).  INCOMPATIBILITY, users need to
   507 specify mutual induction rules differently, i.e. like this:
   508 
   509   (induct rule: foo_bar.inducts)
   510   (induct set: foo bar)
   511   (induct type: foo bar)
   512 
   513 The ML function ProjectRule.projections turns old-style rules into the
   514 new format.
   515 
   516 * Provers/induct: improved handling of simultaneous goals.  Instead of
   517 introducing object-level conjunction, the statement is now split into
   518 several conclusions, while the corresponding symbolic cases are
   519 nested accordingly.  INCOMPATIBILITY, proofs need to be structured
   520 explicitly.  For example:
   521 
   522   lemma
   523     fixes n :: nat
   524     shows "P n" and "Q n"
   525   proof (induct n)
   526     case 0 case 1
   527     show "P 0" sorry
   528   next
   529     case 0 case 2
   530     show "Q 0" sorry
   531   next
   532     case (Suc n) case 1
   533     note `P n` and `Q n`
   534     show "P (Suc n)" sorry
   535   next
   536     case (Suc n) case 2
   537     note `P n` and `Q n`
   538     show "Q (Suc n)" sorry
   539   qed
   540 
   541 The split into subcases may be deferred as follows -- this is
   542 particularly relevant for goal statements with local premises.
   543 
   544   lemma
   545     fixes n :: nat
   546     shows "A n ==> P n" and "B n ==> Q n"
   547   proof (induct n)
   548     case 0
   549     {
   550       case 1
   551       note `A 0`
   552       show "P 0" sorry
   553     next
   554       case 2
   555       note `B 0`
   556       show "Q 0" sorry
   557     }
   558   next
   559     case (Suc n)
   560     note `A n ==> P n` and `B n ==> Q n`
   561     {
   562       case 1
   563       note `A (Suc n)`
   564       show "P (Suc n)" sorry
   565     next
   566       case 2
   567       note `B (Suc n)`
   568       show "Q (Suc n)" sorry
   569     }
   570   qed
   571 
   572 If simultaneous goals are to be used with mutual rules, the statement
   573 needs to be structured carefully as a two-level conjunction, using
   574 lists of propositions separated by 'and':
   575 
   576   lemma
   577     shows "a : A ==> P1 a"
   578           "a : A ==> P2 a"
   579       and "b : B ==> Q1 b"
   580           "b : B ==> Q2 b"
   581           "b : B ==> Q3 b"
   582   proof (induct set: A B)
   583 
   584 * Provers/induct: support coinduction as well.  See
   585 src/HOL/Library/Coinductive_List.thy for various examples.
   586 
   587 * Attribute "symmetric" produces result with standardized schematic
   588 variables (index 0).  Potential INCOMPATIBILITY.
   589 
   590 * Simplifier: by default the simplifier trace only shows top level
   591 rewrites now. That is, trace_simp_depth_limit is set to 1 by
   592 default. Thus there is less danger of being flooded by the trace. The
   593 trace indicates where parts have been suppressed.
   594   
   595 * Provers/classical: removed obsolete classical version of elim_format
   596 attribute; classical elim/dest rules are now treated uniformly when
   597 manipulating the claset.
   598 
   599 * Provers/classical: stricter checks to ensure that supplied intro,
   600 dest and elim rules are well-formed; dest and elim rules must have at
   601 least one premise.
   602 
   603 * Provers/classical: attributes dest/elim/intro take an optional
   604 weight argument for the rule (just as the Pure versions).  Weights are
   605 ignored by automated tools, but determine the search order of single
   606 rule steps.
   607 
   608 * Syntax: input syntax now supports dummy variable binding "%_. b",
   609 where the body does not mention the bound variable.  Note that dummy
   610 patterns implicitly depend on their context of bounds, which makes
   611 "{_. _}" match any set comprehension as expected.  Potential
   612 INCOMPATIBILITY -- parse translations need to cope with syntactic
   613 constant "_idtdummy" in the binding position.
   614 
   615 * Syntax: removed obsolete syntactic constant "_K" and its associated
   616 parse translation.  INCOMPATIBILITY -- use dummy abstraction instead,
   617 for example "A -> B" => "Pi A (%_. B)".
   618 
   619 * Pure: 'class_deps' command visualizes the subclass relation, using
   620 the graph browser tool.
   621 
   622 * Pure: 'print_theory' now suppresses entities with internal name
   623 (trailing "_") by default; use '!' option for full details.
   624 
   625 
   626 *** HOL ***
   627 
   628 * Code generator library theories:
   629   * Pretty_Int represents HOL integers by big integer literals in target
   630     languages.
   631   * Pretty_Char represents HOL characters by character literals in target
   632     languages.
   633   * Pretty_Char_chr like Pretty_Char, but also offers treatment of character
   634     codes; includes Pretty_Int.
   635   * Executable_Set allows to generate code for finite sets using lists.
   636   * Executable_Rat implements rational numbers as triples (sign, enumerator,
   637     denominator).
   638   * Executable_Real implements a subset of real numbers, namly those
   639     representable by rational numbers.
   640   * Efficient_Nat implements natural numbers by integers, which in general will
   641     result in higher efficency; pattern matching with 0/Suc is eliminated;
   642     includes Pretty_Int.
   643   * ML_String provides an additional datatype ml_string; in the HOL default
   644     setup, strings in HOL are mapped to lists of HOL characters in SML; values
   645     of type ml_string are mapped to strings in SML.
   646   * ML_Int provides an additional datatype ml_int which is mapped to to SML
   647     built-in integers.
   648 
   649 * New package for inductive predicates
   650 
   651   An n-ary predicate p with m parameters z_1, ..., z_m can now be defined via
   652 
   653     inductive
   654       p :: "U_1 => ... => U_m => T_1 => ... => T_n => bool"
   655       for z_1 :: U_1 and ... and z_n :: U_m
   656     where
   657       rule_1: "... ==> p z_1 ... z_m t_1_1 ... t_1_n"
   658     | ...
   659 
   660   rather than
   661 
   662     consts s :: "U_1 => ... => U_m => (T_1 * ... * T_n) set"
   663 
   664     abbreviation p :: "U_1 => ... => U_m => T_1 => ... => T_n => bool"
   665     where "p z_1 ... z_m x_1 ... x_n == (x_1, ..., x_n) : s z_1 ... z_m"
   666 
   667     inductive "s z_1 ... z_m"
   668     intros
   669       rule_1: "... ==> (t_1_1, ..., t_1_n) : s z_1 ... z_m"
   670       ...
   671 
   672   For backward compatibility, there is a wrapper allowing inductive
   673   sets to be defined with the new package via
   674 
   675     inductive_set
   676       s :: "U_1 => ... => U_m => (T_1 * ... * T_n) set"
   677       for z_1 :: U_1 and ... and z_n :: U_m
   678     where
   679       rule_1: "... ==> (t_1_1, ..., t_1_n) : s z_1 ... z_m"
   680     | ...
   681 
   682   or
   683 
   684     inductive_set
   685       s :: "U_1 => ... => U_m => (T_1 * ... * T_n) set"
   686       and p :: "U_1 => ... => U_m => T_1 => ... => T_n => bool"
   687       for z_1 :: U_1 and ... and z_n :: U_m
   688     where
   689       "p z_1 ... z_m x_1 ... x_n == (x_1, ..., x_n) : s z_1 ... z_m"
   690     | rule_1: "... ==> p z_1 ... z_m t_1_1 ... t_1_n"
   691     | ...
   692 
   693   if the additional syntax "p ..." is required.
   694 
   695   Many examples can be found in the subdirectories Auth, Bali, Induct,
   696   or MicroJava.
   697 
   698   INCOMPATIBILITIES:
   699 
   700   - Since declaration and definition of inductive sets or predicates
   701     is no longer separated, abbreviations involving the newly introduced
   702     sets or predicates must be specified together with the introduction
   703     rules after the "where" keyword (see example above), rather than before
   704     the actual inductive definition.
   705 
   706   - The variables in induction and elimination rules are now quantified
   707     in the order of their occurrence in the introduction rules, rather than
   708     in alphabetical order. Since this may break some proofs, these proofs
   709     either have to be repaired, e.g. by reordering the variables
   710     a_i_1 ... a_i_{k_i} in Isar "case" statements of the form
   711 
   712       case (rule_i a_i_1 ... a_i_{k_i})
   713 
   714     or the old order of quantification has to be restored by explicitly adding
   715     meta-level quantifiers in the introduction rules, i.e.
   716 
   717       | rule_i: "!!a_i_1 ... a_i_{k_i}. ... ==> p z_1 ... z_m t_i_1 ... t_i_n"
   718 
   719   - The format of the elimination rules is now
   720 
   721       p z_1 ... z_m x_1 ... x_n ==>
   722         (!!a_1_1 ... a_1_{k_1}. x_1 = t_1_1 ==> ... ==> x_n = t_1_n ==> ... ==> P)
   723         ==> ... ==> P
   724 
   725     for predicates and
   726 
   727       (x_1, ..., x_n) : s z_1 ... z_m ==>
   728         (!!a_1_1 ... a_1_{k_1}. x_1 = t_1_1 ==> ... ==> x_n = t_1_n ==> ... ==> P)
   729         ==> ... ==> P
   730 
   731     for sets rather than
   732 
   733       x : s z_1 ... z_m ==>
   734         (!!a_1_1 ... a_1_{k_1}. x = (t_1_1, ..., t_1_n) ==> ... ==> P)
   735         ==> ... ==> P
   736 
   737     This may require terms in goals to be expanded to n-tuples (e.g. using case_tac
   738     or simplification with the split_paired_all rule) before the above elimination
   739     rule is applicable.
   740 
   741   - The elimination or case analysis rules for (mutually) inductive sets or
   742     predicates are now called "p_1.cases" ... "p_k.cases". The list of rules
   743     "p_1_..._p_k.elims" is no longer available.
   744 
   745 * Method "metis" proves goals by applying the Metis general-purpose
   746 resolution prover.  Examples are in the directory MetisExamples.  See
   747 also http://gilith.com/software/metis/
   748   
   749 * Command 'sledgehammer' invokes external automatic theorem provers as
   750 background processes.  It generates calls to the "metis" method if
   751 successful. These can be pasted into the proof.  Users do not have to
   752 wait for the automatic provers to return.
   753 
   754 * Case-expressions allow arbitrary constructor-patterns (including "_") and
   755   take their order into account, like in functional programming.
   756   Internally, this is translated into nested case-expressions; missing cases
   757   are added and mapped to the predefined constant "undefined". In complicated
   758   cases printing may no longer show the original input but the internal
   759   form. Lambda-abstractions allow the same form of pattern matching:
   760   "% pat1 => e1 | ..." is an abbreviation for
   761   "%x. case x of pat1 => e1 | ..." where x is a new variable.
   762 
   763 * IntDef: The constant "int :: nat => int" has been removed; now "int"
   764   is an abbreviation for "of_nat :: nat => int". The simplification rules
   765   for "of_nat" have been changed to work like "int" did previously.
   766   (potential INCOMPATIBILITY)
   767   - "of_nat (Suc m)" simplifies to "1 + of_nat m" instead of "of_nat m + 1"
   768   - of_nat_diff and of_nat_mult are no longer default simp rules
   769 
   770 * Method "algebra" solves polynomial equations over (semi)rings using
   771   Groebner bases. The (semi)ring structure is defined by locales and
   772   the tool setup depends on that generic context. Installing the
   773   method for a specific type involves instantiating the locale and
   774   possibly adding declarations for computation on the coefficients.
   775   The method is already instantiated for natural numbers and for the
   776   axiomatic class of idoms with numerals.  See also the paper by
   777   Chaieb and Wenzel at CALCULEMUS 2007 for the general principles
   778   underlying this architecture of context-aware proof-tools.
   779 
   780 * constant "List.op @" now named "List.append".  Use ML antiquotations
   781 @{const_name List.append} or @{term " ... @ ... "} to circumvent
   782 possible incompatibilities when working on ML level.
   783 
   784 * Constant renames due to introduction of canonical name prefixing for
   785   class package:
   786 
   787     HOL.abs ~> HOL.minus_class.abs
   788     HOL.divide ~> HOL.divide_class.divide
   789     Nat.power ~> Nat.power_class.power
   790     Nat.size ~> Nat.size_class.size
   791     Numeral.number_of ~> Numeral.number_class.number_of
   792     FixedPoint.Inf ~> FixedPoint.complete_lattice_class.Inf
   793 
   794 * Rudimentary class target mechanism involves constant renames:
   795 
   796     Orderings.min ~> Orderings.ord_class.min
   797     Orderings.max ~> Orderings.ord_class.max
   798     FixedPoint.Sup ~> FixedPoint.complete_lattice_class.Sup
   799 
   800 * primrec: missing cases mapped to "undefined" instead of "arbitrary"
   801 
   802 * new constant "undefined" with axiom "undefined x = undefined"
   803 
   804 * new class "default" with associated constant "default"
   805 
   806 * new function listsum :: 'a list => 'a for arbitrary monoids.
   807   Special syntax: "SUM x <- xs. f x" (and latex variants)
   808 
   809 * new (input only) syntax for Haskell-like list comprehension, eg
   810   [(x,y). x <- xs, y <- ys, x ~= y]
   811   For details see List.thy.
   812 
   813 * The special syntax for function "filter" has changed from [x : xs. P] to
   814   [x <- xs. P] to avoid an ambiguity caused by list comprehension syntax,
   815   and for uniformity. INCOMPATIBILITY
   816 
   817 * Lemma "set_take_whileD" renamed to "set_takeWhileD"
   818 
   819 * New lemma collection field_simps (an extension of ring_simps)
   820   for manipulating (in)equations involving division. Multiplies
   821   with all denominators that can be proved to be non-zero (in equations)
   822   or positive/negative (in inequations).
   823 
   824 * Lemma collections ring_eq_simps, group_eq_simps and ring_distrib
   825   have been improved and renamed to ring_simps, group_simps and ring_distribs.
   826   Removed lemmas field_xyz in Ring_and_Field
   827   because they were subsumed by lemmas xyz.
   828 INCOMPATIBILITY.
   829 
   830 * Library/Pretty_Int.thy: maps HOL numerals on target language integer literals
   831   when generating code.
   832 
   833 * Library/Pretty_Char.thy: maps HOL characters on target language character literals
   834   when generating code.
   835 
   836 * Library/Commutative_Ring.thy: switched from recdef to function package;
   837   constants add, mul, pow now curried.  Infix syntax for algebraic operations.
   838 
   839 * Some steps towards more uniform lattice theory development in HOL.
   840 
   841     constants "meet" and "join" now named "inf" and "sup"
   842     constant "Meet" now named "Inf"
   843 
   844     classes "meet_semilorder" and "join_semilorder" now named
   845       "lower_semilattice" and "upper_semilattice"
   846     class "lorder" now named "lattice"
   847     class "comp_lat" now named "complete_lattice"
   848 
   849     Instantiation of lattice classes allows explicit definitions
   850     for "inf" and "sup" operations.
   851 
   852   INCOMPATIBILITY.  Theorem renames:
   853 
   854     meet_left_le            ~> inf_le1
   855     meet_right_le           ~> inf_le2
   856     join_left_le            ~> sup_ge1
   857     join_right_le           ~> sup_ge2
   858     meet_join_le            ~> inf_sup_ord
   859     le_meetI                ~> le_infI
   860     join_leI                ~> le_supI
   861     le_meet                 ~> le_inf_iff
   862     le_join                 ~> ge_sup_conv
   863     meet_idempotent         ~> inf_idem
   864     join_idempotent         ~> sup_idem
   865     meet_comm               ~> inf_commute
   866     join_comm               ~> sup_commute
   867     meet_leI1               ~> le_infI1
   868     meet_leI2               ~> le_infI2
   869     le_joinI1               ~> le_supI1
   870     le_joinI2               ~> le_supI2
   871     meet_assoc              ~> inf_assoc
   872     join_assoc              ~> sup_assoc
   873     meet_left_comm          ~> inf_left_commute
   874     meet_left_idempotent    ~> inf_left_idem
   875     join_left_comm          ~> sup_left_commute
   876     join_left_idempotent    ~> sup_left_idem
   877     meet_aci                ~> inf_aci
   878     join_aci                ~> sup_aci
   879     le_def_meet             ~> le_iff_inf
   880     le_def_join             ~> le_iff_sup
   881     join_absorp2            ~> sup_absorb2
   882     join_absorp1            ~> sup_absorb1
   883     meet_absorp1            ~> inf_absorb1
   884     meet_absorp2            ~> inf_absorb2
   885     meet_join_absorp        ~> inf_sup_absorb
   886     join_meet_absorp        ~> sup_inf_absorb
   887     distrib_join_le         ~> distrib_sup_le
   888     distrib_meet_le         ~> distrib_inf_le
   889 
   890     add_meet_distrib_left   ~> add_inf_distrib_left
   891     add_join_distrib_left   ~> add_sup_distrib_left
   892     is_join_neg_meet        ~> is_join_neg_inf
   893     is_meet_neg_join        ~> is_meet_neg_sup
   894     add_meet_distrib_right  ~> add_inf_distrib_right
   895     add_join_distrib_right  ~> add_sup_distrib_right
   896     add_meet_join_distribs  ~> add_sup_inf_distribs
   897     join_eq_neg_meet        ~> sup_eq_neg_inf
   898     meet_eq_neg_join        ~> inf_eq_neg_sup
   899     add_eq_meet_join        ~> add_eq_inf_sup
   900     meet_0_imp_0            ~> inf_0_imp_0
   901     join_0_imp_0            ~> sup_0_imp_0
   902     meet_0_eq_0             ~> inf_0_eq_0
   903     join_0_eq_0             ~> sup_0_eq_0
   904     neg_meet_eq_join        ~> neg_inf_eq_sup
   905     neg_join_eq_meet        ~> neg_sup_eq_inf
   906     join_eq_if              ~> sup_eq_if
   907 
   908     mono_meet               ~> mono_inf
   909     mono_join               ~> mono_sup
   910     meet_bool_eq            ~> inf_bool_eq
   911     join_bool_eq            ~> sup_bool_eq
   912     meet_fun_eq             ~> inf_fun_eq
   913     join_fun_eq             ~> sup_fun_eq
   914     meet_set_eq             ~> inf_set_eq
   915     join_set_eq             ~> sup_set_eq
   916     meet1_iff               ~> inf1_iff
   917     meet2_iff               ~> inf2_iff
   918     meet1I                  ~> inf1I
   919     meet2I                  ~> inf2I
   920     meet1D1                 ~> inf1D1
   921     meet2D1                 ~> inf2D1
   922     meet1D2                 ~> inf1D2
   923     meet2D2                 ~> inf2D2
   924     meet1E                  ~> inf1E
   925     meet2E                  ~> inf2E
   926     join1_iff               ~> sup1_iff
   927     join2_iff               ~> sup2_iff
   928     join1I1                 ~> sup1I1
   929     join2I1                 ~> sup2I1
   930     join1I1                 ~> sup1I1
   931     join2I2                 ~> sup1I2
   932     join1CI                 ~> sup1CI
   933     join2CI                 ~> sup2CI
   934     join1E                  ~> sup1E
   935     join2E                  ~> sup2E
   936 
   937     is_meet_Meet            ~> is_meet_Inf
   938     Meet_bool_def           ~> Inf_bool_def
   939     Meet_fun_def            ~> Inf_fun_def
   940     Meet_greatest           ~> Inf_greatest
   941     Meet_lower              ~> Inf_lower
   942     Meet_set_def            ~> Inf_set_def
   943 
   944     listsp_meetI            ~> listsp_infI
   945     listsp_meet_eq          ~> listsp_inf_eq
   946 
   947     meet_min                ~> inf_min
   948     join_max                ~> sup_max
   949 
   950 * Classes "order" and "linorder": facts "refl", "trans" and
   951 "cases" renamed ro "order_refl", "order_trans" and "linorder_cases", to
   952 avoid clashes with HOL "refl" and "trans". INCOMPATIBILITY.
   953 
   954 * Classes "order" and "linorder": 
   955 potential INCOMPATIBILITY: order of proof goals in order/linorder instance
   956 proofs changed.
   957 
   958 * Dropped lemma duplicate def_imp_eq in favor of meta_eq_to_obj_eq.
   959 INCOMPATIBILITY.
   960 
   961 * Dropped lemma duplicate if_def2 in favor of if_bool_eq_conj.
   962 INCOMPATIBILITY.
   963 
   964 * Added syntactic class "size"; overloaded constant "size" now has
   965 type "'a::size ==> bool"
   966 
   967 * Renamed constants "Divides.op div", "Divides.op mod" and "Divides.op
   968 dvd" to "Divides.div_class.div", "Divides.div_class.mod" and "Divides.dvd". INCOMPATIBILITY.
   969 
   970 * Added method "lexicographic_order" automatically synthesizes
   971 termination relations as lexicographic combinations of size measures
   972 -- 'function' package.
   973 
   974 * HOL/records: generalised field-update to take a function on the
   975 field rather than the new value: r(|A := x|) is translated to A_update
   976 (K x) r The K-combinator that is internally used is called K_record.
   977 INCOMPATIBILITY: Usage of the plain update functions has to be
   978 adapted.
   979  
   980 * axclass "semiring_0" now contains annihilation axioms x * 0 = 0 and
   981 0 * x = 0, which are required for a semiring.  Richer structures do
   982 not inherit from semiring_0 anymore, because this property is a
   983 theorem there, not an axiom.  INCOMPATIBILITY: In instances of
   984 semiring_0, there is more to prove, but this is mostly trivial.
   985 
   986 * axclass "recpower" was generalized to arbitrary monoids, not just
   987 commutative semirings.  INCOMPATIBILITY: If you use recpower and need
   988 commutativity or a semiring property, add the corresponding classes.
   989 
   990 * Unified locale partial_order with class definition (cf. theory
   991 Orderings), added parameter ``less''.  INCOMPATIBILITY.
   992 
   993 * Constant "List.list_all2" in List.thy now uses authentic syntax.
   994 INCOMPATIBILITY: translations containing list_all2 may go wrong.  On
   995 Isar level, use abbreviations instead.
   996 
   997 * Renamed constant "List.op mem" to "List.memberl" INCOMPATIBILITY:
   998 rarely occuring name references (e.g. ``List.op mem.simps'') require
   999 renaming (e.g. ``List.memberl.simps'').
  1000 
  1001 * Renamed constants "0" to "HOL.zero_class.zero" and "1" to "HOL.one_class.one".
  1002 INCOMPATIBILITY.
  1003 
  1004 * Added class "HOL.eq", allowing for code generation with polymorphic equality.
  1005 
  1006 * Numeral syntax: type 'bin' which was a mere type copy of 'int' has
  1007 been abandoned in favour of plain 'int'. INCOMPATIBILITY --
  1008 significant changes for setting up numeral syntax for types:
  1009 
  1010   - new constants Numeral.pred and Numeral.succ instead
  1011       of former Numeral.bin_pred and Numeral.bin_succ.
  1012   - Use integer operations instead of bin_add, bin_mult and so on.
  1013   - Numeral simplification theorems named Numeral.numeral_simps instead of Bin_simps.
  1014   - ML structure Bin_Simprocs now named Int_Numeral_Base_Simprocs.
  1015 
  1016 See HOL/Integ/IntArith.thy for an example setup.
  1017 
  1018 * New top level command 'normal_form' computes the normal form of a
  1019 term that may contain free variables. For example ``normal_form
  1020 "rev[a,b,c]"'' produces ``[b,c,a]'' (without proof).  This command is
  1021 suitable for heavy-duty computations because the functions are
  1022 compiled to ML first.
  1023 
  1024 * Alternative iff syntax "A <-> B" for equality on bool (with priority
  1025 25 like -->); output depends on the "iff" print_mode, the default is
  1026 "A = B" (with priority 50).
  1027 
  1028 * Renamed constants in HOL.thy and Orderings.thy:
  1029     op +   ~> HOL.plus_class.plus
  1030     op -   ~> HOL.minus_class.minus
  1031     uminus ~> HOL.minus_class.uminus
  1032     abs    ~> HOL.abs_class.abs
  1033     op *   ~> HOL.times_class.times
  1034     op <   ~> HOL.ord_class.less
  1035     op <=  ~> HOL.ord_class.less_eq
  1036 
  1037 Adaptions may be required in the following cases:
  1038 
  1039 a) User-defined constants using any of the names "plus", "minus", "times",
  1040 "less" or "less_eq". The standard syntax translations for "+", "-" and "*"
  1041 may go wrong.
  1042 INCOMPATIBILITY: use more specific names.
  1043 
  1044 b) Variables named "plus", "minus", "times", "less", "less_eq"
  1045 INCOMPATIBILITY: use more specific names.
  1046 
  1047 c) Permutative equations (e.g. "a + b = b + a")
  1048 Since the change of names also changes the order of terms, permutative
  1049 rewrite rules may get applied in a different order. Experience shows that
  1050 this is rarely the case (only two adaptions in the whole Isabelle
  1051 distribution).
  1052 INCOMPATIBILITY: rewrite proofs
  1053 
  1054 d) ML code directly refering to constant names
  1055 This in general only affects hand-written proof tactics, simprocs and so on.
  1056 INCOMPATIBILITY: grep your sourcecode and replace names.  Consider use
  1057 of const_name ML antiquotations.
  1058 
  1059 * Relations less (<) and less_eq (<=) are also available on type bool.
  1060 Modified syntax to disallow nesting without explicit parentheses,
  1061 e.g. "(x < y) < z" or "x < (y < z)", but NOT "x < y < z".
  1062 
  1063 * "LEAST x:A. P" expands to "LEAST x. x:A & P" (input only).
  1064 
  1065 * Relation composition operator "op O" now has precedence 75 and binds
  1066 stronger than union and intersection. INCOMPATIBILITY.
  1067 
  1068 * The old set interval syntax "{m..n(}" (and relatives) has been
  1069 removed.  Use "{m..<n}" (and relatives) instead.
  1070 
  1071 * In the context of the assumption "~(s = t)" the Simplifier rewrites
  1072 "t = s" to False (by simproc "neq_simproc").  For backward
  1073 compatibility this can be disabled by ML "reset use_neq_simproc".
  1074 
  1075 * "m dvd n" where m and n are numbers is evaluated to True/False by
  1076 simp.
  1077 
  1078 * Theorem Cons_eq_map_conv no longer declared as ``simp''.
  1079 
  1080 * Theorem setsum_mult renamed to setsum_right_distrib.
  1081 
  1082 * Prefer ex1I over ex_ex1I in single-step reasoning, e.g. by the
  1083 ``rule'' method.
  1084 
  1085 * Reimplemented methods ``sat'' and ``satx'', with several
  1086 improvements: goals no longer need to be stated as "<prems> ==>
  1087 False", equivalences (i.e. "=" on type bool) are handled, variable
  1088 names of the form "lit_<n>" are no longer reserved, significant
  1089 speedup.
  1090 
  1091 * Methods ``sat'' and ``satx'' can now replay MiniSat proof traces.
  1092 zChaff is still supported as well.
  1093 
  1094 * 'inductive' and 'datatype': provide projections of mutual rules,
  1095 bundled as foo_bar.inducts;
  1096 
  1097 * Library: moved theories Parity, GCD, Binomial, Infinite_Set to
  1098 Library.
  1099 
  1100 * Library: moved theory Accessible_Part to main HOL.
  1101 
  1102 * Library: added theory Coinductive_List of potentially infinite lists
  1103 as greatest fixed-point.
  1104 
  1105 * Library: added theory AssocList which implements (finite) maps as
  1106 association lists.
  1107 
  1108 * Added proof method ``evaluation'' for efficiently solving a goal
  1109 (i.e. a boolean expression) by compiling it to ML. The goal is
  1110 "proved" (via an oracle) if it evaluates to True.
  1111 
  1112 * Linear arithmetic now splits certain operators (e.g. min, max, abs)
  1113 also when invoked by the simplifier.  This results in the simplifier
  1114 being more powerful on arithmetic goals.  INCOMPATIBILITY.  Set
  1115 fast_arith_split_limit to 0 to obtain the old behavior.
  1116 
  1117 * Support for hex (0x20) and binary (0b1001) numerals.
  1118 
  1119 * New method: reify eqs (t), where eqs are equations for an
  1120 interpretation I :: 'a list => 'b => 'c and t::'c is an optional
  1121 parameter, computes a term s::'b and a list xs::'a list and proves the
  1122 theorem I xs s = t. This is also known as reification or quoting. The
  1123 resulting theorem is applied to the subgoal to substitute t with I xs
  1124 s.  If t is omitted, the subgoal itself is reified.
  1125 
  1126 * New method: reflection corr_thm eqs (t). The parameters eqs and (t)
  1127 are as explained above. corr_thm is a theorem for I vs (f t) = I vs t,
  1128 where f is supposed to be a computable function (in the sense of code
  1129 generattion). The method uses reify to compute s and xs as above then
  1130 applies corr_thm and uses normalization by evaluation to "prove" f s =
  1131 r and finally gets the theorem t = r, which is again applied to the
  1132 subgoal. An Example is available in HOL/ex/ReflectionEx.thy.
  1133 
  1134 * Reflection: Automatic reification now handels binding, an example
  1135 is available in HOL/ex/ReflectionEx.thy
  1136 
  1137 
  1138 *** HOL-Algebra ***
  1139 
  1140 * Formalisation of ideals and the quotient construction over rings.
  1141 
  1142 * Order and lattice theory no longer based on records.
  1143 INCOMPATIBILITY.
  1144 
  1145 * Renamed lemmas least_carrier -> least_closed and greatest_carrier ->
  1146 greatest_closed.  INCOMPATIBILITY.
  1147 
  1148 * Method algebra is now set up via an attribute.  For examples see
  1149 Ring.thy.  INCOMPATIBILITY: the method is now weaker on combinations
  1150 of algebraic structures.
  1151 
  1152 * Renamed theory CRing to Ring.
  1153 
  1154 
  1155 *** HOL-Complex ***
  1156 
  1157 * Theory Real: new method ferrack implements quantifier elimination
  1158 for linear arithmetic over the reals. The quantifier elimination
  1159 feature is used only for decision, for compatibility with arith. This
  1160 means a goal is either solved or left unchanged, no simplification.
  1161 
  1162 * Hyperreal: Functions root and sqrt are now defined on negative real
  1163 inputs so that root n (- x) = - root n x and sqrt (- x) = - sqrt x.
  1164 Nonnegativity side conditions have been removed from many lemmas, so
  1165 that more subgoals may now be solved by simplification; potential
  1166 INCOMPATIBILITY.
  1167 
  1168 * Real: New axiomatic classes formalize real normed vector spaces and
  1169 algebras, using new overloaded constants scaleR :: real => 'a => 'a
  1170 and norm :: 'a => real.
  1171 
  1172 * Real: New constant of_real :: real => 'a::real_algebra_1 injects
  1173 from reals into other types. The overloaded constant Reals :: 'a set
  1174 is now defined as range of_real; potential INCOMPATIBILITY.
  1175 
  1176 * Real: ML code generation is supported now and hence also quickcheck.
  1177 Reals are implemented as arbitrary precision rationals.
  1178 
  1179 * Hyperreal: Several constants that previously worked only for the
  1180 reals have been generalized, so they now work over arbitrary vector
  1181 spaces. Type annotations may need to be added in some cases; potential
  1182 INCOMPATIBILITY.
  1183 
  1184   Infinitesimal  :: ('a::real_normed_vector) star set
  1185   HFinite        :: ('a::real_normed_vector) star set
  1186   HInfinite      :: ('a::real_normed_vector) star set
  1187   approx         :: ('a::real_normed_vector) star => 'a star => bool
  1188   monad          :: ('a::real_normed_vector) star => 'a star set
  1189   galaxy         :: ('a::real_normed_vector) star => 'a star set
  1190   (NS)LIMSEQ     :: [nat => 'a::real_normed_vector, 'a] => bool
  1191   (NS)convergent :: (nat => 'a::real_normed_vector) => bool
  1192   (NS)Bseq       :: (nat => 'a::real_normed_vector) => bool
  1193   (NS)Cauchy     :: (nat => 'a::real_normed_vector) => bool
  1194   (NS)LIM        :: ['a::real_normed_vector => 'b::real_normed_vector, 'a, 'b] => bool
  1195   is(NS)Cont     :: ['a::real_normed_vector => 'b::real_normed_vector, 'a] => bool
  1196   deriv          :: ['a::real_normed_field => 'a, 'a, 'a] => bool
  1197   sgn            :: 'a::real_normed_vector => 'a
  1198   exp            :: 'a::{recpower,real_normed_field,banach} => 'a
  1199 
  1200 * Complex: Some complex-specific constants are now abbreviations for
  1201 overloaded ones: complex_of_real = of_real, cmod = norm, hcmod =
  1202 hnorm.  Other constants have been entirely removed in favor of the
  1203 polymorphic versions (INCOMPATIBILITY):
  1204 
  1205   approx        <-- capprox
  1206   HFinite       <-- CFinite
  1207   HInfinite     <-- CInfinite
  1208   Infinitesimal <-- CInfinitesimal
  1209   monad         <-- cmonad
  1210   galaxy        <-- cgalaxy
  1211   (NS)LIM       <-- (NS)CLIM, (NS)CRLIM
  1212   is(NS)Cont    <-- is(NS)Contc, is(NS)contCR
  1213   (ns)deriv     <-- (ns)cderiv
  1214 
  1215 
  1216 *** ML ***
  1217 
  1218 * Generic arithmetic modules: Tools/integer.ML, Tools/rat.ML, Tools/float.ML
  1219 
  1220 * Context data interfaces (Theory/Proof/GenericDataFun): removed
  1221 name/print, uninitialized data defaults to ad-hoc copy of empty value,
  1222 init only required for impure data.  INCOMPATIBILITY: empty really
  1223 need to be empty (no dependencies on theory content!)
  1224 
  1225 * ML within Isar: antiquotations allow to embed statically-checked
  1226 formal entities in the source, referring to the context available at
  1227 compile-time.  For example:
  1228 
  1229 ML {* @{typ "'a => 'b"} *}
  1230 ML {* @{term "%x. x"} *}
  1231 ML {* @{prop "x == y"} *}
  1232 ML {* @{ctyp "'a => 'b"} *}
  1233 ML {* @{cterm "%x. x"} *}
  1234 ML {* @{cprop "x == y"} *}
  1235 ML {* @{thm asm_rl} *}
  1236 ML {* @{thms asm_rl} *}
  1237 ML {* @{const_name c} *}
  1238 ML {* @{const_syntax c} *}
  1239 ML {* @{context} *}
  1240 ML {* @{theory} *}
  1241 ML {* @{theory Pure} *}
  1242 ML {* @{simpset} *}
  1243 ML {* @{claset} *}
  1244 ML {* @{clasimpset} *}
  1245 
  1246 The same works for sources being ``used'' within an Isar context.
  1247 
  1248 * ML in Isar: improved error reporting; extra verbosity with
  1249 Toplevel.debug enabled.
  1250 
  1251 * Pure/library:
  1252 
  1253   val burrow: ('a list -> 'b list) -> 'a list list -> 'b list list
  1254   val fold_burrow: ('a list -> 'c -> 'b list * 'd) -> 'a list list -> 'c -> 'b list list * 'd
  1255 
  1256 The semantics of "burrow" is: "take a function with *simulatanously*
  1257 transforms a list of value, and apply it *simulatanously* to a list of
  1258 list of values of the appropriate type". Compare this with "map" which
  1259 would *not* apply its argument function simulatanously but in
  1260 sequence; "fold_burrow" has an additional context.
  1261 
  1262 * Pure/library: functions map2 and fold2 with curried syntax for
  1263 simultanous mapping and folding:
  1264 
  1265     val map2: ('a -> 'b -> 'c) -> 'a list -> 'b list -> 'c list
  1266     val fold2: ('a -> 'b -> 'c -> 'c) -> 'a list -> 'b list -> 'c -> 'c
  1267 
  1268 * Pure/library: indexed lists - some functions in the Isabelle library
  1269 treating lists over 'a as finite mappings from [0...n] to 'a have been
  1270 given more convenient names and signatures reminiscent of similar
  1271 functions for alists, tables, etc:
  1272 
  1273   val nth: 'a list -> int -> 'a 
  1274   val nth_map: int -> ('a -> 'a) -> 'a list -> 'a list
  1275   val fold_index: (int * 'a -> 'b -> 'b) -> 'a list -> 'b -> 'b
  1276 
  1277 Note that fold_index starts counting at index 0, not 1 like foldln
  1278 used to.
  1279 
  1280 * Pure/library: added general ``divide_and_conquer'' combinator on
  1281 lists.
  1282 
  1283 * Pure/General/table.ML: the join operations now works via exceptions
  1284 DUP/SAME instead of type option.  This is simpler in simple cases, and
  1285 admits slightly more efficient complex applications.
  1286 
  1287 * Pure: datatype Context.generic joins theory/Proof.context and
  1288 provides some facilities for code that works in either kind of
  1289 context, notably GenericDataFun for uniform theory and proof data.
  1290 
  1291 * Pure: 'advanced' translation functions (parse_translation etc.) now
  1292 use Context.generic instead of just theory.
  1293 
  1294 * Pure: simplified internal attribute type, which is now always
  1295 Context.generic * thm -> Context.generic * thm.  Global (theory)
  1296 vs. local (Proof.context) attributes have been discontinued, while
  1297 minimizing code duplication.  Thm.rule_attribute and
  1298 Thm.declaration_attribute build canonical attributes; see also
  1299 structure Context for further operations on Context.generic, notably
  1300 GenericDataFun.  INCOMPATIBILITY, need to adapt attribute type
  1301 declarations and definitions.
  1302 
  1303 * Pure/kernel: consts certification ignores sort constraints given in
  1304 signature declarations.  (This information is not relevant to the
  1305 logic, but only for type inference.)  IMPORTANT INTERNAL CHANGE,
  1306 potential INCOMPATIBILITY.
  1307 
  1308 * Pure: axiomatic type classes are now purely definitional, with
  1309 explicit proofs of class axioms and super class relations performed
  1310 internally.  See Pure/axclass.ML for the main internal interfaces --
  1311 notably AxClass.define_class supercedes AxClass.add_axclass, and
  1312 AxClass.axiomatize_class/classrel/arity supercede
  1313 Sign.add_classes/classrel/arities.
  1314 
  1315 * Pure/Isar: Args/Attrib parsers operate on Context.generic --
  1316 global/local versions on theory vs. Proof.context have been
  1317 discontinued; Attrib.syntax and Method.syntax have been adapted
  1318 accordingly.  INCOMPATIBILITY, need to adapt parser expressions for
  1319 attributes, methods, etc.
  1320 
  1321 * Pure: several functions of signature "... -> theory -> theory * ..."
  1322 have been reoriented to "... -> theory -> ... * theory" in order to
  1323 allow natural usage in combination with the ||>, ||>>, |-> and
  1324 fold_map combinators.
  1325 
  1326 * Pure: official theorem names (closed derivations) and additional
  1327 comments (tags) are now strictly separate.  Name hints -- which are
  1328 maintained as tags -- may be attached any time without affecting the
  1329 derivation.
  1330 
  1331 * Pure: primitive rule lift_rule now takes goal cterm instead of an
  1332 actual goal state (thm).  Use Thm.lift_rule (Thm.cprem_of st i) to
  1333 achieve the old behaviour.
  1334 
  1335 * Pure: the "Goal" constant is now called "prop", supporting a
  1336 slightly more general idea of ``protecting'' meta-level rule
  1337 statements.
  1338 
  1339 * Pure: Logic.(un)varify only works in a global context, which is now
  1340 enforced instead of silently assumed.  INCOMPATIBILITY, may use
  1341 Logic.legacy_(un)varify as temporary workaround.
  1342 
  1343 * Pure: structure Name provides scalable operations for generating
  1344 internal variable names, notably Name.variants etc.  This replaces
  1345 some popular functions from term.ML:
  1346 
  1347   Term.variant		->  Name.variant
  1348   Term.variantlist	->  Name.variant_list  (*canonical argument order*)
  1349   Term.invent_names	->  Name.invent_list
  1350 
  1351 Note that low-level renaming rarely occurs in new code -- operations
  1352 from structure Variable are used instead (see below).
  1353 
  1354 * Pure: structure Variable provides fundamental operations for proper
  1355 treatment of fixed/schematic variables in a context.  For example,
  1356 Variable.import introduces fixes for schematics of given facts and
  1357 Variable.export reverses the effect (up to renaming) -- this replaces
  1358 various freeze_thaw operations.
  1359 
  1360 * Pure: structure Goal provides simple interfaces for
  1361 init/conclude/finish and tactical prove operations (replacing former
  1362 Tactic.prove).  Goal.prove is the canonical way to prove results
  1363 within a given context; Goal.prove_global is a degraded version for
  1364 theory level goals, including a global Drule.standard.  Note that
  1365 OldGoals.prove_goalw_cterm has long been obsolete, since it is
  1366 ill-behaved in a local proof context (e.g. with local fixes/assumes or
  1367 in a locale context).
  1368 
  1369 * Isar: simplified treatment of user-level errors, using exception
  1370 ERROR of string uniformly.  Function error now merely raises ERROR,
  1371 without any side effect on output channels.  The Isar toplevel takes
  1372 care of proper display of ERROR exceptions.  ML code may use plain
  1373 handle/can/try; cat_error may be used to concatenate errors like this:
  1374 
  1375   ... handle ERROR msg => cat_error msg "..."
  1376 
  1377 Toplevel ML code (run directly or through the Isar toplevel) may be
  1378 embedded into the Isar toplevel with exception display/debug like
  1379 this:
  1380 
  1381   Isar.toplevel (fn () => ...)
  1382 
  1383 INCOMPATIBILITY, removed special transform_error facilities, removed
  1384 obsolete variants of user-level exceptions (ERROR_MESSAGE,
  1385 Context.PROOF, ProofContext.CONTEXT, Proof.STATE, ProofHistory.FAIL)
  1386 -- use plain ERROR instead.
  1387 
  1388 * Isar: theory setup now has type (theory -> theory), instead of a
  1389 list.  INCOMPATIBILITY, may use #> to compose setup functions.
  1390 
  1391 * Isar: installed ML toplevel pretty printer for type Proof.context,
  1392 subject to ProofContext.debug/verbose flags.
  1393 
  1394 * Isar: Toplevel.theory_to_proof admits transactions that modify the
  1395 theory before entering a proof state.  Transactions now always see a
  1396 quasi-functional intermediate checkpoint, both in interactive and
  1397 batch mode.
  1398 
  1399 * Simplifier: the simpset of a running simplification process now
  1400 contains a proof context (cf. Simplifier.the_context), which is the
  1401 very context that the initial simpset has been retrieved from (by
  1402 simpset_of/local_simpset_of).  Consequently, all plug-in components
  1403 (solver, looper etc.) may depend on arbitrary proof data.
  1404 
  1405 * Simplifier.inherit_context inherits the proof context (plus the
  1406 local bounds) of the current simplification process; any simproc
  1407 etc. that calls the Simplifier recursively should do this!  Removed
  1408 former Simplifier.inherit_bounds, which is already included here --
  1409 INCOMPATIBILITY.  Tools based on low-level rewriting may even have to
  1410 specify an explicit context using Simplifier.context/theory_context.
  1411 
  1412 * Simplifier/Classical Reasoner: more abstract interfaces
  1413 change_simpset/claset for modifying the simpset/claset reference of a
  1414 theory; raw versions simpset/claset_ref etc. have been discontinued --
  1415 INCOMPATIBILITY.
  1416 
  1417 * Provers: more generic wrt. syntax of object-logics, avoid hardwired
  1418 "Trueprop" etc.
  1419 
  1420 
  1421 *** System ***
  1422 
  1423 * settings: ML_IDENTIFIER -- which is appended to user specific heap
  1424 locations -- now includes the Isabelle version identifier as well.
  1425 This simplifies use of multiple Isabelle installations.
  1426 
  1427 * isabelle-process: option -S (secure mode) disables some critical
  1428 operations, notably runtime compilation and evaluation of ML source
  1429 code.
  1430 
  1431 * Experimental support for multithreading, using Poly/ML 5.1 (internal
  1432 version from CVS). The maximum number of threads is specified via usedir
  1433 option -M or the "max-threads" setting in Proof General. User-code needs
  1434 to observe certain guidelines for thread-safe programming, see appendix
  1435 A in the Isar Implementation manual.
  1436 
  1437 
  1438 New in Isabelle2005 (October 2005)
  1439 ----------------------------------
  1440 
  1441 *** General ***
  1442 
  1443 * Theory headers: the new header syntax for Isar theories is
  1444 
  1445   theory <name>
  1446   imports <theory1> ... <theoryN>
  1447   uses <file1> ... <fileM>
  1448   begin
  1449 
  1450 where the 'uses' part is optional.  The previous syntax
  1451 
  1452   theory <name> = <theory1> + ... + <theoryN>:
  1453 
  1454 will disappear in the next release.  Use isatool fixheaders to convert
  1455 existing theory files.  Note that there is no change in ancient
  1456 non-Isar theories now, but these will disappear soon.
  1457 
  1458 * Theory loader: parent theories can now also be referred to via
  1459 relative and absolute paths.
  1460 
  1461 * Command 'find_theorems' searches for a list of criteria instead of a
  1462 list of constants. Known criteria are: intro, elim, dest, name:string,
  1463 simp:term, and any term. Criteria can be preceded by '-' to select
  1464 theorems that do not match. Intro, elim, dest select theorems that
  1465 match the current goal, name:s selects theorems whose fully qualified
  1466 name contain s, and simp:term selects all simplification rules whose
  1467 lhs match term.  Any other term is interpreted as pattern and selects
  1468 all theorems matching the pattern. Available in ProofGeneral under
  1469 'ProofGeneral -> Find Theorems' or C-c C-f.  Example:
  1470 
  1471   C-c C-f (100) "(_::nat) + _ + _" intro -name: "HOL."
  1472 
  1473 prints the last 100 theorems matching the pattern "(_::nat) + _ + _",
  1474 matching the current goal as introduction rule and not having "HOL."
  1475 in their name (i.e. not being defined in theory HOL).
  1476 
  1477 * Command 'thms_containing' has been discontinued in favour of
  1478 'find_theorems'; INCOMPATIBILITY.
  1479 
  1480 * Communication with Proof General is now 8bit clean, which means that
  1481 Unicode text in UTF-8 encoding may be used within theory texts (both
  1482 formal and informal parts).  Cf. option -U of the Isabelle Proof
  1483 General interface.  Here are some simple examples (cf. src/HOL/ex):
  1484 
  1485   http://isabelle.in.tum.de/library/HOL/ex/Hebrew.html
  1486   http://isabelle.in.tum.de/library/HOL/ex/Chinese.html
  1487 
  1488 * Improved efficiency of the Simplifier and, to a lesser degree, the
  1489 Classical Reasoner.  Typical big applications run around 2 times
  1490 faster.
  1491 
  1492 
  1493 *** Document preparation ***
  1494 
  1495 * Commands 'display_drafts' and 'print_drafts' perform simple output
  1496 of raw sources.  Only those symbols that do not require additional
  1497 LaTeX packages (depending on comments in isabellesym.sty) are
  1498 displayed properly, everything else is left verbatim.  isatool display
  1499 and isatool print are used as front ends (these are subject to the
  1500 DVI/PDF_VIEWER and PRINT_COMMAND settings, respectively).
  1501 
  1502 * Command tags control specific markup of certain regions of text,
  1503 notably folding and hiding.  Predefined tags include "theory" (for
  1504 theory begin and end), "proof" for proof commands, and "ML" for
  1505 commands involving ML code; the additional tags "visible" and
  1506 "invisible" are unused by default.  Users may give explicit tag
  1507 specifications in the text, e.g. ''by %invisible (auto)''.  The
  1508 interpretation of tags is determined by the LaTeX job during document
  1509 preparation: see option -V of isatool usedir, or options -n and -t of
  1510 isatool document, or even the LaTeX macros \isakeeptag, \isafoldtag,
  1511 \isadroptag.
  1512 
  1513 Several document versions may be produced at the same time via isatool
  1514 usedir (the generated index.html will link all of them).  Typical
  1515 specifications include ''-V document=theory,proof,ML'' to present
  1516 theory/proof/ML parts faithfully, ''-V outline=/proof,/ML'' to fold
  1517 proof and ML commands, and ''-V mutilated=-theory,-proof,-ML'' to omit
  1518 these parts without any formal replacement text.  The Isabelle site
  1519 default settings produce ''document'' and ''outline'' versions as
  1520 specified above.
  1521 
  1522 * Several new antiquotations:
  1523 
  1524   @{term_type term} prints a term with its type annotated;
  1525 
  1526   @{typeof term} prints the type of a term;
  1527 
  1528   @{const const} is the same as @{term const}, but checks that the
  1529   argument is a known logical constant;
  1530 
  1531   @{term_style style term} and @{thm_style style thm} print a term or
  1532   theorem applying a "style" to it
  1533 
  1534   @{ML text}
  1535 
  1536 Predefined styles are 'lhs' and 'rhs' printing the lhs/rhs of
  1537 definitions, equations, inequations etc., 'concl' printing only the
  1538 conclusion of a meta-logical statement theorem, and 'prem1' .. 'prem19'
  1539 to print the specified premise.  TermStyle.add_style provides an ML
  1540 interface for introducing further styles.  See also the "LaTeX Sugar"
  1541 document practical applications.  The ML antiquotation prints
  1542 type-checked ML expressions verbatim.
  1543 
  1544 * Markup commands 'chapter', 'section', 'subsection', 'subsubsection',
  1545 and 'text' support optional locale specification '(in loc)', which
  1546 specifies the default context for interpreting antiquotations.  For
  1547 example: 'text (in lattice) {* @{thm inf_assoc}*}'.
  1548 
  1549 * Option 'locale=NAME' of antiquotations specifies an alternative
  1550 context interpreting the subsequent argument.  For example: @{thm
  1551 [locale=lattice] inf_assoc}.
  1552 
  1553 * Proper output of proof terms (@{prf ...} and @{full_prf ...}) within
  1554 a proof context.
  1555 
  1556 * Proper output of antiquotations for theory commands involving a
  1557 proof context (such as 'locale' or 'theorem (in loc) ...').
  1558 
  1559 * Delimiters of outer tokens (string etc.) now produce separate LaTeX
  1560 macros (\isachardoublequoteopen, isachardoublequoteclose etc.).
  1561 
  1562 * isatool usedir: new option -C (default true) controls whether option
  1563 -D should include a copy of the original document directory; -C false
  1564 prevents unwanted effects such as copying of administrative CVS data.
  1565 
  1566 
  1567 *** Pure ***
  1568 
  1569 * Considerably improved version of 'constdefs' command.  Now performs
  1570 automatic type-inference of declared constants; additional support for
  1571 local structure declarations (cf. locales and HOL records), see also
  1572 isar-ref manual.  Potential INCOMPATIBILITY: need to observe strictly
  1573 sequential dependencies of definitions within a single 'constdefs'
  1574 section; moreover, the declared name needs to be an identifier.  If
  1575 all fails, consider to fall back on 'consts' and 'defs' separately.
  1576 
  1577 * Improved indexed syntax and implicit structures.  First of all,
  1578 indexed syntax provides a notational device for subscripted
  1579 application, using the new syntax \<^bsub>term\<^esub> for arbitrary
  1580 expressions.  Secondly, in a local context with structure
  1581 declarations, number indexes \<^sub>n or the empty index (default
  1582 number 1) refer to a certain fixed variable implicitly; option
  1583 show_structs controls printing of implicit structures.  Typical
  1584 applications of these concepts involve record types and locales.
  1585 
  1586 * New command 'no_syntax' removes grammar declarations (and
  1587 translations) resulting from the given syntax specification, which is
  1588 interpreted in the same manner as for the 'syntax' command.
  1589 
  1590 * 'Advanced' translation functions (parse_translation etc.) may depend
  1591 on the signature of the theory context being presently used for
  1592 parsing/printing, see also isar-ref manual.
  1593 
  1594 * Improved 'oracle' command provides a type-safe interface to turn an
  1595 ML expression of type theory -> T -> term into a primitive rule of
  1596 type theory -> T -> thm (i.e. the functionality of Thm.invoke_oracle
  1597 is already included here); see also FOL/ex/IffExample.thy;
  1598 INCOMPATIBILITY.
  1599 
  1600 * axclass: name space prefix for class "c" is now "c_class" (was "c"
  1601 before); "cI" is no longer bound, use "c.intro" instead.
  1602 INCOMPATIBILITY.  This change avoids clashes of fact bindings for
  1603 axclasses vs. locales.
  1604 
  1605 * Improved internal renaming of symbolic identifiers -- attach primes
  1606 instead of base 26 numbers.
  1607 
  1608 * New flag show_question_marks controls printing of leading question
  1609 marks in schematic variable names.
  1610 
  1611 * In schematic variable names, *any* symbol following \<^isub> or
  1612 \<^isup> is now treated as part of the base name.  For example, the
  1613 following works without printing of awkward ".0" indexes:
  1614 
  1615   lemma "x\<^isub>1 = x\<^isub>2 ==> x\<^isub>2 = x\<^isub>1"
  1616     by simp
  1617 
  1618 * Inner syntax includes (*(*nested*) comments*).
  1619 
  1620 * Pretty printer now supports unbreakable blocks, specified in mixfix
  1621 annotations as "(00...)".
  1622 
  1623 * Clear separation of logical types and nonterminals, where the latter
  1624 may only occur in 'syntax' specifications or type abbreviations.
  1625 Before that distinction was only partially implemented via type class
  1626 "logic" vs. "{}".  Potential INCOMPATIBILITY in rare cases of improper
  1627 use of 'types'/'consts' instead of 'nonterminals'/'syntax'.  Some very
  1628 exotic syntax specifications may require further adaption
  1629 (e.g. Cube/Cube.thy).
  1630 
  1631 * Removed obsolete type class "logic", use the top sort {} instead.
  1632 Note that non-logical types should be declared as 'nonterminals'
  1633 rather than 'types'.  INCOMPATIBILITY for new object-logic
  1634 specifications.
  1635 
  1636 * Attributes 'induct' and 'cases': type or set names may now be
  1637 locally fixed variables as well.
  1638 
  1639 * Simplifier: can now control the depth to which conditional rewriting
  1640 is traced via the PG menu Isabelle -> Settings -> Trace Simp Depth
  1641 Limit.
  1642 
  1643 * Simplifier: simplification procedures may now take the current
  1644 simpset into account (cf. Simplifier.simproc(_i) / mk_simproc
  1645 interface), which is very useful for calling the Simplifier
  1646 recursively.  Minor INCOMPATIBILITY: the 'prems' argument of simprocs
  1647 is gone -- use prems_of_ss on the simpset instead.  Moreover, the
  1648 low-level mk_simproc no longer applies Logic.varify internally, to
  1649 allow for use in a context of fixed variables.
  1650 
  1651 * thin_tac now works even if the assumption being deleted contains !!
  1652 or ==>.  More generally, erule now works even if the major premise of
  1653 the elimination rule contains !! or ==>.
  1654 
  1655 * Method 'rules' has been renamed to 'iprover'. INCOMPATIBILITY.
  1656 
  1657 * Reorganized bootstrapping of the Pure theories; CPure is now derived
  1658 from Pure, which contains all common declarations already.  Both
  1659 theories are defined via plain Isabelle/Isar .thy files.
  1660 INCOMPATIBILITY: elements of CPure (such as the CPure.intro /
  1661 CPure.elim / CPure.dest attributes) now appear in the Pure name space;
  1662 use isatool fixcpure to adapt your theory and ML sources.
  1663 
  1664 * New syntax 'name(i-j, i-, i, ...)' for referring to specific
  1665 selections of theorems in named facts via index ranges.
  1666 
  1667 * 'print_theorems': in theory mode, really print the difference
  1668 wrt. the last state (works for interactive theory development only),
  1669 in proof mode print all local facts (cf. 'print_facts');
  1670 
  1671 * 'hide': option '(open)' hides only base names.
  1672 
  1673 * More efficient treatment of intermediate checkpoints in interactive
  1674 theory development.
  1675 
  1676 * Code generator is now invoked via code_module (incremental code
  1677 generation) and code_library (modular code generation, ML structures
  1678 for each theory).  INCOMPATIBILITY: new keywords 'file' and 'contains'
  1679 must be quoted when used as identifiers.
  1680 
  1681 * New 'value' command for reading, evaluating and printing terms using
  1682 the code generator.  INCOMPATIBILITY: command keyword 'value' must be
  1683 quoted when used as identifier.
  1684 
  1685 
  1686 *** Locales ***
  1687 
  1688 * New commands for the interpretation of locale expressions in
  1689 theories (1), locales (2) and proof contexts (3).  These generate
  1690 proof obligations from the expression specification.  After the
  1691 obligations have been discharged, theorems of the expression are added
  1692 to the theory, target locale or proof context.  The synopsis of the
  1693 commands is a follows:
  1694 
  1695   (1) interpretation expr inst
  1696   (2) interpretation target < expr
  1697   (3) interpret expr inst
  1698 
  1699 Interpretation in theories and proof contexts require a parameter
  1700 instantiation of terms from the current context.  This is applied to
  1701 specifications and theorems of the interpreted expression.
  1702 Interpretation in locales only permits parameter renaming through the
  1703 locale expression.  Interpretation is smart in that interpretations
  1704 that are active already do not occur in proof obligations, neither are
  1705 instantiated theorems stored in duplicate.  Use 'print_interps' to
  1706 inspect active interpretations of a particular locale.  For details,
  1707 see the Isar Reference manual.  Examples can be found in
  1708 HOL/Finite_Set.thy and HOL/Algebra/UnivPoly.thy.
  1709 
  1710 INCOMPATIBILITY: former 'instantiate' has been withdrawn, use
  1711 'interpret' instead.
  1712 
  1713 * New context element 'constrains' for adding type constraints to
  1714 parameters.
  1715 
  1716 * Context expressions: renaming of parameters with syntax
  1717 redeclaration.
  1718 
  1719 * Locale declaration: 'includes' disallowed.
  1720 
  1721 * Proper static binding of attribute syntax -- i.e. types / terms /
  1722 facts mentioned as arguments are always those of the locale definition
  1723 context, independently of the context of later invocations.  Moreover,
  1724 locale operations (renaming and type / term instantiation) are applied
  1725 to attribute arguments as expected.
  1726 
  1727 INCOMPATIBILITY of the ML interface: always pass Attrib.src instead of
  1728 actual attributes; rare situations may require Attrib.attribute to
  1729 embed those attributes into Attrib.src that lack concrete syntax.
  1730 Attribute implementations need to cooperate properly with the static
  1731 binding mechanism.  Basic parsers Args.XXX_typ/term/prop and
  1732 Attrib.XXX_thm etc. already do the right thing without further
  1733 intervention.  Only unusual applications -- such as "where" or "of"
  1734 (cf. src/Pure/Isar/attrib.ML), which process arguments depending both
  1735 on the context and the facts involved -- may have to assign parsed
  1736 values to argument tokens explicitly.
  1737 
  1738 * Changed parameter management in theorem generation for long goal
  1739 statements with 'includes'.  INCOMPATIBILITY: produces a different
  1740 theorem statement in rare situations.
  1741 
  1742 * Locale inspection command 'print_locale' omits notes elements.  Use
  1743 'print_locale!' to have them included in the output.
  1744 
  1745 
  1746 *** Provers ***
  1747 
  1748 * Provers/hypsubst.ML: improved version of the subst method, for
  1749 single-step rewriting: it now works in bound variable contexts. New is
  1750 'subst (asm)', for rewriting an assumption.  INCOMPATIBILITY: may
  1751 rewrite a different subterm than the original subst method, which is
  1752 still available as 'simplesubst'.
  1753 
  1754 * Provers/quasi.ML: new transitivity reasoners for transitivity only
  1755 and quasi orders.
  1756 
  1757 * Provers/trancl.ML: new transitivity reasoner for transitive and
  1758 reflexive-transitive closure of relations.
  1759 
  1760 * Provers/blast.ML: new reference depth_limit to make blast's depth
  1761 limit (previously hard-coded with a value of 20) user-definable.
  1762 
  1763 * Provers/simplifier.ML has been moved to Pure, where Simplifier.setup
  1764 is peformed already.  Object-logics merely need to finish their
  1765 initial simpset configuration as before.  INCOMPATIBILITY.
  1766 
  1767 
  1768 *** HOL ***
  1769 
  1770 * Symbolic syntax of Hilbert Choice Operator is now as follows:
  1771 
  1772   syntax (epsilon)
  1773     "_Eps" :: "[pttrn, bool] => 'a"    ("(3\<some>_./ _)" [0, 10] 10)
  1774 
  1775 The symbol \<some> is displayed as the alternative epsilon of LaTeX
  1776 and x-symbol; use option '-m epsilon' to get it actually printed.
  1777 Moreover, the mathematically important symbolic identifier \<epsilon>
  1778 becomes available as variable, constant etc.  INCOMPATIBILITY,
  1779 
  1780 * "x > y" abbreviates "y < x" and "x >= y" abbreviates "y <= x".
  1781 Similarly for all quantifiers: "ALL x > y" etc.  The x-symbol for >=
  1782 is \<ge>. New transitivity rules have been added to HOL/Orderings.thy to
  1783 support corresponding Isar calculations.
  1784 
  1785 * "{x:A. P}" abbreviates "{x. x:A & P}", and similarly for "\<in>"
  1786 instead of ":".
  1787 
  1788 * theory SetInterval: changed the syntax for open intervals:
  1789 
  1790   Old       New
  1791   {..n(}    {..<n}
  1792   {)n..}    {n<..}
  1793   {m..n(}   {m..<n}
  1794   {)m..n}   {m<..n}
  1795   {)m..n(}  {m<..<n}
  1796 
  1797 The old syntax is still supported but will disappear in the next
  1798 release.  For conversion use the following Emacs search and replace
  1799 patterns (these are not perfect but work quite well):
  1800 
  1801   {)\([^\.]*\)\.\.  ->  {\1<\.\.}
  1802   \.\.\([^(}]*\)(}  ->  \.\.<\1}
  1803 
  1804 * Theory Commutative_Ring (in Library): method comm_ring for proving
  1805 equalities in commutative rings; method 'algebra' provides a generic
  1806 interface.
  1807 
  1808 * Theory Finite_Set: changed the syntax for 'setsum', summation over
  1809 finite sets: "setsum (%x. e) A", which used to be "\<Sum>x:A. e", is
  1810 now either "SUM x:A. e" or "\<Sum>x \<in> A. e". The bound variable can
  1811 be a tuple pattern.
  1812 
  1813 Some new syntax forms are available:
  1814 
  1815   "\<Sum>x | P. e"      for     "setsum (%x. e) {x. P}"
  1816   "\<Sum>x = a..b. e"   for     "setsum (%x. e) {a..b}"
  1817   "\<Sum>x = a..<b. e"  for     "setsum (%x. e) {a..<b}"
  1818   "\<Sum>x < k. e"      for     "setsum (%x. e) {..<k}"
  1819 
  1820 The latter form "\<Sum>x < k. e" used to be based on a separate
  1821 function "Summation", which has been discontinued.
  1822 
  1823 * theory Finite_Set: in structured induction proofs, the insert case
  1824 is now 'case (insert x F)' instead of the old counterintuitive 'case
  1825 (insert F x)'.
  1826 
  1827 * The 'refute' command has been extended to support a much larger
  1828 fragment of HOL, including axiomatic type classes, constdefs and
  1829 typedefs, inductive datatypes and recursion.
  1830 
  1831 * New tactics 'sat' and 'satx' to prove propositional tautologies.
  1832 Requires zChaff with proof generation to be installed.  See
  1833 HOL/ex/SAT_Examples.thy for examples.
  1834 
  1835 * Datatype induction via method 'induct' now preserves the name of the
  1836 induction variable. For example, when proving P(xs::'a list) by
  1837 induction on xs, the induction step is now P(xs) ==> P(a#xs) rather
  1838 than P(list) ==> P(a#list) as previously.  Potential INCOMPATIBILITY
  1839 in unstructured proof scripts.
  1840 
  1841 * Reworked implementation of records.  Improved scalability for
  1842 records with many fields, avoiding performance problems for type
  1843 inference. Records are no longer composed of nested field types, but
  1844 of nested extension types. Therefore the record type only grows linear
  1845 in the number of extensions and not in the number of fields.  The
  1846 top-level (users) view on records is preserved.  Potential
  1847 INCOMPATIBILITY only in strange cases, where the theory depends on the
  1848 old record representation. The type generated for a record is called
  1849 <record_name>_ext_type.
  1850 
  1851 Flag record_quick_and_dirty_sensitive can be enabled to skip the
  1852 proofs triggered by a record definition or a simproc (if
  1853 quick_and_dirty is enabled).  Definitions of large records can take
  1854 quite long.
  1855 
  1856 New simproc record_upd_simproc for simplification of multiple record
  1857 updates enabled by default.  Moreover, trivial updates are also
  1858 removed: r(|x := x r|) = r.  INCOMPATIBILITY: old proofs break
  1859 occasionally, since simplification is more powerful by default.
  1860 
  1861 * typedef: proper support for polymorphic sets, which contain extra
  1862 type-variables in the term.
  1863 
  1864 * Simplifier: automatically reasons about transitivity chains
  1865 involving "trancl" (r^+) and "rtrancl" (r^*) by setting up tactics
  1866 provided by Provers/trancl.ML as additional solvers.  INCOMPATIBILITY:
  1867 old proofs break occasionally as simplification may now solve more
  1868 goals than previously.
  1869 
  1870 * Simplifier: converts x <= y into x = y if assumption y <= x is
  1871 present.  Works for all partial orders (class "order"), in particular
  1872 numbers and sets.  For linear orders (e.g. numbers) it treats ~ x < y
  1873 just like y <= x.
  1874 
  1875 * Simplifier: new simproc for "let x = a in f x".  If a is a free or
  1876 bound variable or a constant then the let is unfolded.  Otherwise
  1877 first a is simplified to b, and then f b is simplified to g. If
  1878 possible we abstract b from g arriving at "let x = b in h x",
  1879 otherwise we unfold the let and arrive at g.  The simproc can be
  1880 enabled/disabled by the reference use_let_simproc.  Potential
  1881 INCOMPATIBILITY since simplification is more powerful by default.
  1882 
  1883 * Classical reasoning: the meson method now accepts theorems as arguments.
  1884 
  1885 * Prover support: pre-release of the Isabelle-ATP linkup, which runs background
  1886 jobs to provide advice on the provability of subgoals.
  1887 
  1888 * Theory OrderedGroup and Ring_and_Field: various additions and
  1889 improvements to faciliate calculations involving equalities and
  1890 inequalities.
  1891 
  1892 The following theorems have been eliminated or modified
  1893 (INCOMPATIBILITY):
  1894 
  1895   abs_eq             now named abs_of_nonneg
  1896   abs_of_ge_0        now named abs_of_nonneg
  1897   abs_minus_eq       now named abs_of_nonpos
  1898   imp_abs_id         now named abs_of_nonneg
  1899   imp_abs_neg_id     now named abs_of_nonpos
  1900   mult_pos           now named mult_pos_pos
  1901   mult_pos_le        now named mult_nonneg_nonneg
  1902   mult_pos_neg_le    now named mult_nonneg_nonpos
  1903   mult_pos_neg2_le   now named mult_nonneg_nonpos2
  1904   mult_neg           now named mult_neg_neg
  1905   mult_neg_le        now named mult_nonpos_nonpos
  1906 
  1907 * The following lemmas in Ring_and_Field have been added to the simplifier:
  1908      
  1909      zero_le_square
  1910      not_square_less_zero 
  1911 
  1912   The following lemmas have been deleted from Real/RealPow:
  1913   
  1914      realpow_zero_zero
  1915      realpow_two
  1916      realpow_less
  1917      zero_le_power
  1918      realpow_two_le
  1919      abs_realpow_two
  1920      realpow_two_abs     
  1921 
  1922 * Theory Parity: added rules for simplifying exponents.
  1923 
  1924 * Theory List:
  1925 
  1926 The following theorems have been eliminated or modified
  1927 (INCOMPATIBILITY):
  1928 
  1929   list_all_Nil       now named list_all.simps(1)
  1930   list_all_Cons      now named list_all.simps(2)
  1931   list_all_conv      now named list_all_iff
  1932   set_mem_eq         now named mem_iff
  1933 
  1934 * Theories SetsAndFunctions and BigO (see HOL/Library) support
  1935 asymptotic "big O" calculations.  See the notes in BigO.thy.
  1936 
  1937 
  1938 *** HOL-Complex ***
  1939 
  1940 * Theory RealDef: better support for embedding natural numbers and
  1941 integers in the reals.
  1942 
  1943 The following theorems have been eliminated or modified
  1944 (INCOMPATIBILITY):
  1945 
  1946   exp_ge_add_one_self  now requires no hypotheses
  1947   real_of_int_add      reversed direction of equality (use [symmetric])
  1948   real_of_int_minus    reversed direction of equality (use [symmetric])
  1949   real_of_int_diff     reversed direction of equality (use [symmetric])
  1950   real_of_int_mult     reversed direction of equality (use [symmetric])
  1951 
  1952 * Theory RComplete: expanded support for floor and ceiling functions.
  1953 
  1954 * Theory Ln is new, with properties of the natural logarithm
  1955 
  1956 * Hyperreal: There is a new type constructor "star" for making
  1957 nonstandard types.  The old type names are now type synonyms:
  1958 
  1959   hypreal = real star
  1960   hypnat = nat star
  1961   hcomplex = complex star
  1962 
  1963 * Hyperreal: Many groups of similarly-defined constants have been
  1964 replaced by polymorphic versions (INCOMPATIBILITY):
  1965 
  1966   star_of <-- hypreal_of_real, hypnat_of_nat, hcomplex_of_complex
  1967 
  1968   starset      <-- starsetNat, starsetC
  1969   *s*          <-- *sNat*, *sc*
  1970   starset_n    <-- starsetNat_n, starsetC_n
  1971   *sn*         <-- *sNatn*, *scn*
  1972   InternalSets <-- InternalNatSets, InternalCSets
  1973 
  1974   starfun      <-- starfun{Nat,Nat2,C,RC,CR}
  1975   *f*          <-- *fNat*, *fNat2*, *fc*, *fRc*, *fcR*
  1976   starfun_n    <-- starfun{Nat,Nat2,C,RC,CR}_n
  1977   *fn*         <-- *fNatn*, *fNat2n*, *fcn*, *fRcn*, *fcRn*
  1978   InternalFuns <-- InternalNatFuns, InternalNatFuns2, Internal{C,RC,CR}Funs
  1979 
  1980 * Hyperreal: Many type-specific theorems have been removed in favor of
  1981 theorems specific to various axiomatic type classes (INCOMPATIBILITY):
  1982 
  1983   add_commute <-- {hypreal,hypnat,hcomplex}_add_commute
  1984   add_assoc   <-- {hypreal,hypnat,hcomplex}_add_assocs
  1985   OrderedGroup.add_0 <-- {hypreal,hypnat,hcomplex}_add_zero_left
  1986   OrderedGroup.add_0_right <-- {hypreal,hcomplex}_add_zero_right
  1987   right_minus <-- hypreal_add_minus
  1988   left_minus <-- {hypreal,hcomplex}_add_minus_left
  1989   mult_commute <-- {hypreal,hypnat,hcomplex}_mult_commute
  1990   mult_assoc <-- {hypreal,hypnat,hcomplex}_mult_assoc
  1991   mult_1_left <-- {hypreal,hypnat}_mult_1, hcomplex_mult_one_left
  1992   mult_1_right <-- hcomplex_mult_one_right
  1993   mult_zero_left <-- hcomplex_mult_zero_left
  1994   left_distrib <-- {hypreal,hypnat,hcomplex}_add_mult_distrib
  1995   right_distrib <-- hypnat_add_mult_distrib2
  1996   zero_neq_one <-- {hypreal,hypnat,hcomplex}_zero_not_eq_one
  1997   right_inverse <-- hypreal_mult_inverse
  1998   left_inverse <-- hypreal_mult_inverse_left, hcomplex_mult_inv_left
  1999   order_refl <-- {hypreal,hypnat}_le_refl
  2000   order_trans <-- {hypreal,hypnat}_le_trans
  2001   order_antisym <-- {hypreal,hypnat}_le_anti_sym
  2002   order_less_le <-- {hypreal,hypnat}_less_le
  2003   linorder_linear <-- {hypreal,hypnat}_le_linear
  2004   add_left_mono <-- {hypreal,hypnat}_add_left_mono
  2005   mult_strict_left_mono <-- {hypreal,hypnat}_mult_less_mono2
  2006   add_nonneg_nonneg <-- hypreal_le_add_order
  2007 
  2008 * Hyperreal: Separate theorems having to do with type-specific
  2009 versions of constants have been merged into theorems that apply to the
  2010 new polymorphic constants (INCOMPATIBILITY):
  2011 
  2012   STAR_UNIV_set <-- {STAR_real,NatStar_real,STARC_complex}_set
  2013   STAR_empty_set <-- {STAR,NatStar,STARC}_empty_set
  2014   STAR_Un <-- {STAR,NatStar,STARC}_Un
  2015   STAR_Int <-- {STAR,NatStar,STARC}_Int
  2016   STAR_Compl <-- {STAR,NatStar,STARC}_Compl
  2017   STAR_subset <-- {STAR,NatStar,STARC}_subset
  2018   STAR_mem <-- {STAR,NatStar,STARC}_mem
  2019   STAR_mem_Compl <-- {STAR,STARC}_mem_Compl
  2020   STAR_diff <-- {STAR,STARC}_diff
  2021   STAR_star_of_image_subset <-- {STAR_hypreal_of_real, NatStar_hypreal_of_real,
  2022     STARC_hcomplex_of_complex}_image_subset
  2023   starset_n_Un <-- starset{Nat,C}_n_Un
  2024   starset_n_Int <-- starset{Nat,C}_n_Int
  2025   starset_n_Compl <-- starset{Nat,C}_n_Compl
  2026   starset_n_diff <-- starset{Nat,C}_n_diff
  2027   InternalSets_Un <-- Internal{Nat,C}Sets_Un
  2028   InternalSets_Int <-- Internal{Nat,C}Sets_Int
  2029   InternalSets_Compl <-- Internal{Nat,C}Sets_Compl
  2030   InternalSets_diff <-- Internal{Nat,C}Sets_diff
  2031   InternalSets_UNIV_diff <-- Internal{Nat,C}Sets_UNIV_diff
  2032   InternalSets_starset_n <-- Internal{Nat,C}Sets_starset{Nat,C}_n
  2033   starset_starset_n_eq <-- starset{Nat,C}_starset{Nat,C}_n_eq
  2034   starset_n_starset <-- starset{Nat,C}_n_starset{Nat,C}
  2035   starfun_n_starfun <-- starfun{Nat,Nat2,C,RC,CR}_n_starfun{Nat,Nat2,C,RC,CR}
  2036   starfun <-- starfun{Nat,Nat2,C,RC,CR}
  2037   starfun_mult <-- starfun{Nat,Nat2,C,RC,CR}_mult
  2038   starfun_add <-- starfun{Nat,Nat2,C,RC,CR}_add
  2039   starfun_minus <-- starfun{Nat,Nat2,C,RC,CR}_minus
  2040   starfun_diff <-- starfun{C,RC,CR}_diff
  2041   starfun_o <-- starfun{NatNat2,Nat2,_stafunNat,C,C_starfunRC,_starfunCR}_o
  2042   starfun_o2 <-- starfun{NatNat2,_stafunNat,C,C_starfunRC,_starfunCR}_o2
  2043   starfun_const_fun <-- starfun{Nat,Nat2,C,RC,CR}_const_fun
  2044   starfun_inverse <-- starfun{Nat,C,RC,CR}_inverse
  2045   starfun_eq <-- starfun{Nat,Nat2,C,RC,CR}_eq
  2046   starfun_eq_iff <-- starfun{C,RC,CR}_eq_iff
  2047   starfun_Id <-- starfunC_Id
  2048   starfun_approx <-- starfun{Nat,CR}_approx
  2049   starfun_capprox <-- starfun{C,RC}_capprox
  2050   starfun_abs <-- starfunNat_rabs
  2051   starfun_lambda_cancel <-- starfun{C,CR,RC}_lambda_cancel
  2052   starfun_lambda_cancel2 <-- starfun{C,CR,RC}_lambda_cancel2
  2053   starfun_mult_HFinite_approx <-- starfunCR_mult_HFinite_capprox
  2054   starfun_mult_CFinite_capprox <-- starfun{C,RC}_mult_CFinite_capprox
  2055   starfun_add_capprox <-- starfun{C,RC}_add_capprox
  2056   starfun_add_approx <-- starfunCR_add_approx
  2057   starfun_inverse_inverse <-- starfunC_inverse_inverse
  2058   starfun_divide <-- starfun{C,CR,RC}_divide
  2059   starfun_n <-- starfun{Nat,C}_n
  2060   starfun_n_mult <-- starfun{Nat,C}_n_mult
  2061   starfun_n_add <-- starfun{Nat,C}_n_add
  2062   starfun_n_add_minus <-- starfunNat_n_add_minus
  2063   starfun_n_const_fun <-- starfun{Nat,C}_n_const_fun
  2064   starfun_n_minus <-- starfun{Nat,C}_n_minus
  2065   starfun_n_eq <-- starfun{Nat,C}_n_eq
  2066 
  2067   star_n_add <-- {hypreal,hypnat,hcomplex}_add
  2068   star_n_minus <-- {hypreal,hcomplex}_minus
  2069   star_n_diff <-- {hypreal,hcomplex}_diff
  2070   star_n_mult <-- {hypreal,hcomplex}_mult
  2071   star_n_inverse <-- {hypreal,hcomplex}_inverse
  2072   star_n_le <-- {hypreal,hypnat}_le
  2073   star_n_less <-- {hypreal,hypnat}_less
  2074   star_n_zero_num <-- {hypreal,hypnat,hcomplex}_zero_num
  2075   star_n_one_num <-- {hypreal,hypnat,hcomplex}_one_num
  2076   star_n_abs <-- hypreal_hrabs
  2077   star_n_divide <-- hcomplex_divide
  2078 
  2079   star_of_add <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_add
  2080   star_of_minus <-- {hypreal_of_real,hcomplex_of_complex}_minus
  2081   star_of_diff <-- hypreal_of_real_diff
  2082   star_of_mult <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_mult
  2083   star_of_one <-- {hypreal_of_real,hcomplex_of_complex}_one
  2084   star_of_zero <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_zero
  2085   star_of_le <-- {hypreal_of_real,hypnat_of_nat}_le_iff
  2086   star_of_less <-- {hypreal_of_real,hypnat_of_nat}_less_iff
  2087   star_of_eq <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_eq_iff
  2088   star_of_inverse <-- {hypreal_of_real,hcomplex_of_complex}_inverse
  2089   star_of_divide <-- {hypreal_of_real,hcomplex_of_complex}_divide
  2090   star_of_of_nat <-- {hypreal_of_real,hcomplex_of_complex}_of_nat
  2091   star_of_of_int <-- {hypreal_of_real,hcomplex_of_complex}_of_int
  2092   star_of_number_of <-- {hypreal,hcomplex}_number_of
  2093   star_of_number_less <-- number_of_less_hypreal_of_real_iff
  2094   star_of_number_le <-- number_of_le_hypreal_of_real_iff
  2095   star_of_eq_number <-- hypreal_of_real_eq_number_of_iff
  2096   star_of_less_number <-- hypreal_of_real_less_number_of_iff
  2097   star_of_le_number <-- hypreal_of_real_le_number_of_iff
  2098   star_of_power <-- hypreal_of_real_power
  2099   star_of_eq_0 <-- hcomplex_of_complex_zero_iff
  2100 
  2101 * Hyperreal: new method "transfer" that implements the transfer
  2102 principle of nonstandard analysis. With a subgoal that mentions
  2103 nonstandard types like "'a star", the command "apply transfer"
  2104 replaces it with an equivalent one that mentions only standard types.
  2105 To be successful, all free variables must have standard types; non-
  2106 standard variables must have explicit universal quantifiers.
  2107 
  2108 * Hyperreal: A theory of Taylor series.
  2109 
  2110 
  2111 *** HOLCF ***
  2112 
  2113 * Discontinued special version of 'constdefs' (which used to support
  2114 continuous functions) in favor of the general Pure one with full
  2115 type-inference.
  2116 
  2117 * New simplification procedure for solving continuity conditions; it
  2118 is much faster on terms with many nested lambda abstractions (cubic
  2119 instead of exponential time).
  2120 
  2121 * New syntax for domain package: selector names are now optional.
  2122 Parentheses should be omitted unless argument is lazy, for example:
  2123 
  2124   domain 'a stream = cons "'a" (lazy "'a stream")
  2125 
  2126 * New command 'fixrec' for defining recursive functions with pattern
  2127 matching; defining multiple functions with mutual recursion is also
  2128 supported.  Patterns may include the constants cpair, spair, up, sinl,
  2129 sinr, or any data constructor defined by the domain package. The given
  2130 equations are proven as rewrite rules. See HOLCF/ex/Fixrec_ex.thy for
  2131 syntax and examples.
  2132 
  2133 * New commands 'cpodef' and 'pcpodef' for defining predicate subtypes
  2134 of cpo and pcpo types. Syntax is exactly like the 'typedef' command,
  2135 but the proof obligation additionally includes an admissibility
  2136 requirement. The packages generate instances of class cpo or pcpo,
  2137 with continuity and strictness theorems for Rep and Abs.
  2138 
  2139 * HOLCF: Many theorems have been renamed according to a more standard naming
  2140 scheme (INCOMPATIBILITY):
  2141 
  2142   foo_inject:  "foo$x = foo$y ==> x = y"
  2143   foo_eq:      "(foo$x = foo$y) = (x = y)"
  2144   foo_less:    "(foo$x << foo$y) = (x << y)"
  2145   foo_strict:  "foo$UU = UU"
  2146   foo_defined: "... ==> foo$x ~= UU"
  2147   foo_defined_iff: "(foo$x = UU) = (x = UU)"
  2148 
  2149 
  2150 *** ZF ***
  2151 
  2152 * ZF/ex: theories Group and Ring provide examples in abstract algebra,
  2153 including the First Isomorphism Theorem (on quotienting by the kernel
  2154 of a homomorphism).
  2155 
  2156 * ZF/Simplifier: install second copy of type solver that actually
  2157 makes use of TC rules declared to Isar proof contexts (or locales);
  2158 the old version is still required for ML proof scripts.
  2159 
  2160 
  2161 *** Cube ***
  2162 
  2163 * Converted to Isar theory format; use locales instead of axiomatic
  2164 theories.
  2165 
  2166 
  2167 *** ML ***
  2168 
  2169 * Pure/library.ML: added ##>, ##>>, #>> -- higher-order counterparts
  2170 for ||>, ||>>, |>>,
  2171 
  2172 * Pure/library.ML no longer defines its own option datatype, but uses
  2173 that of the SML basis, which has constructors NONE and SOME instead of
  2174 None and Some, as well as exception Option.Option instead of OPTION.
  2175 The functions the, if_none, is_some, is_none have been adapted
  2176 accordingly, while Option.map replaces apsome.
  2177 
  2178 * Pure/library.ML: the exception LIST has been given up in favour of
  2179 the standard exceptions Empty and Subscript, as well as
  2180 Library.UnequalLengths.  Function like Library.hd and Library.tl are
  2181 superceded by the standard hd and tl functions etc.
  2182 
  2183 A number of basic list functions are no longer exported to the ML
  2184 toplevel, as they are variants of predefined functions.  The following
  2185 suggests how one can translate existing code:
  2186 
  2187     rev_append xs ys = List.revAppend (xs, ys)
  2188     nth_elem (i, xs) = List.nth (xs, i)
  2189     last_elem xs = List.last xs
  2190     flat xss = List.concat xss
  2191     seq fs = List.app fs
  2192     partition P xs = List.partition P xs
  2193     mapfilter f xs = List.mapPartial f xs
  2194 
  2195 * Pure/library.ML: several combinators for linear functional
  2196 transformations, notably reverse application and composition:
  2197 
  2198   x |> f                f #> g
  2199   (x, y) |-> f          f #-> g
  2200 
  2201 * Pure/library.ML: introduced/changed precedence of infix operators:
  2202 
  2203   infix 1 |> |-> ||> ||>> |>> |>>> #> #->;
  2204   infix 2 ?;
  2205   infix 3 o oo ooo oooo;
  2206   infix 4 ~~ upto downto;
  2207 
  2208 Maybe INCOMPATIBILITY when any of those is used in conjunction with other
  2209 infix operators.
  2210 
  2211 * Pure/library.ML: natural list combinators fold, fold_rev, and
  2212 fold_map support linear functional transformations and nesting.  For
  2213 example:
  2214 
  2215   fold f [x1, ..., xN] y =
  2216     y |> f x1 |> ... |> f xN
  2217 
  2218   (fold o fold) f [xs1, ..., xsN] y =
  2219     y |> fold f xs1 |> ... |> fold f xsN
  2220 
  2221   fold f [x1, ..., xN] =
  2222     f x1 #> ... #> f xN
  2223 
  2224   (fold o fold) f [xs1, ..., xsN] =
  2225     fold f xs1 #> ... #> fold f xsN
  2226 
  2227 * Pure/library.ML: the following selectors on type 'a option are
  2228 available:
  2229 
  2230   the:               'a option -> 'a  (*partial*)
  2231   these:             'a option -> 'a  where 'a = 'b list
  2232   the_default: 'a -> 'a option -> 'a
  2233   the_list:          'a option -> 'a list
  2234 
  2235 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  2236 basic operations for association lists, following natural argument
  2237 order; moreover the explicit equality predicate passed here avoids
  2238 potentially expensive polymorphic runtime equality checks.
  2239 The old functions may be expressed as follows:
  2240 
  2241   assoc = uncurry (AList.lookup (op =))
  2242   assocs = these oo AList.lookup (op =)
  2243   overwrite = uncurry (AList.update (op =)) o swap
  2244 
  2245 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  2246 
  2247   val make: ('a -> 'b) -> 'a list -> ('a * 'b) list
  2248   val find: ('a * 'b -> bool) -> ('c * 'b) list -> 'a -> 'c list
  2249 
  2250 replacing make_keylist and keyfilter (occassionally used)
  2251 Naive rewrites:
  2252 
  2253   make_keylist = AList.make
  2254   keyfilter = AList.find (op =)
  2255 
  2256 * eq_fst and eq_snd now take explicit equality parameter, thus
  2257   avoiding eqtypes. Naive rewrites:
  2258 
  2259     eq_fst = eq_fst (op =)
  2260     eq_snd = eq_snd (op =)
  2261 
  2262 * Removed deprecated apl and apr (rarely used).
  2263   Naive rewrites:
  2264 
  2265     apl (n, op) =>>= curry op n
  2266     apr (op, m) =>>= fn n => op (n, m)
  2267 
  2268 * Pure/General: structure OrdList (cf. Pure/General/ord_list.ML)
  2269 provides a reasonably efficient light-weight implementation of sets as
  2270 lists.
  2271 
  2272 * Pure/General: generic tables (cf. Pure/General/table.ML) provide a
  2273 few new operations; existing lookup and update are now curried to
  2274 follow natural argument order (for use with fold etc.);
  2275 INCOMPATIBILITY, use (uncurry Symtab.lookup) etc. as last resort.
  2276 
  2277 * Pure/General: output via the Isabelle channels of
  2278 writeln/warning/error etc. is now passed through Output.output, with a
  2279 hook for arbitrary transformations depending on the print_mode
  2280 (cf. Output.add_mode -- the first active mode that provides a output
  2281 function wins).  Already formatted output may be embedded into further
  2282 text via Output.raw; the result of Pretty.string_of/str_of and derived
  2283 functions (string_of_term/cterm/thm etc.) is already marked raw to
  2284 accommodate easy composition of diagnostic messages etc.  Programmers
  2285 rarely need to care about Output.output or Output.raw at all, with
  2286 some notable exceptions: Output.output is required when bypassing the
  2287 standard channels (writeln etc.), or in token translations to produce
  2288 properly formatted results; Output.raw is required when capturing
  2289 already output material that will eventually be presented to the user
  2290 a second time.  For the default print mode, both Output.output and
  2291 Output.raw have no effect.
  2292 
  2293 * Pure/General: Output.time_accumulator NAME creates an operator ('a
  2294 -> 'b) -> 'a -> 'b to measure runtime and count invocations; the
  2295 cumulative results are displayed at the end of a batch session.
  2296 
  2297 * Pure/General: File.sysify_path and File.quote_sysify path have been
  2298 replaced by File.platform_path and File.shell_path (with appropriate
  2299 hooks).  This provides a clean interface for unusual systems where the
  2300 internal and external process view of file names are different.
  2301 
  2302 * Pure: more efficient orders for basic syntactic entities: added
  2303 fast_string_ord, fast_indexname_ord, fast_term_ord; changed sort_ord
  2304 and typ_ord to use fast_string_ord and fast_indexname_ord (term_ord is
  2305 NOT affected); structures Symtab, Vartab, Typtab, Termtab use the fast
  2306 orders now -- potential INCOMPATIBILITY for code that depends on a
  2307 particular order for Symtab.keys, Symtab.dest, etc. (consider using
  2308 Library.sort_strings on result).
  2309 
  2310 * Pure/term.ML: combinators fold_atyps, fold_aterms, fold_term_types,
  2311 fold_types traverse types/terms from left to right, observing natural
  2312 argument order.  Supercedes previous foldl_XXX versions, add_frees,
  2313 add_vars etc. have been adapted as well: INCOMPATIBILITY.
  2314 
  2315 * Pure: name spaces have been refined, with significant changes of the
  2316 internal interfaces -- INCOMPATIBILITY.  Renamed cond_extern(_table)
  2317 to extern(_table).  The plain name entry path is superceded by a
  2318 general 'naming' context, which also includes the 'policy' to produce
  2319 a fully qualified name and external accesses of a fully qualified
  2320 name; NameSpace.extend is superceded by context dependent
  2321 Sign.declare_name.  Several theory and proof context operations modify
  2322 the naming context.  Especially note Theory.restore_naming and
  2323 ProofContext.restore_naming to get back to a sane state; note that
  2324 Theory.add_path is no longer sufficient to recover from
  2325 Theory.absolute_path in particular.
  2326 
  2327 * Pure: new flags short_names (default false) and unique_names
  2328 (default true) for controlling output of qualified names.  If
  2329 short_names is set, names are printed unqualified.  If unique_names is
  2330 reset, the name prefix is reduced to the minimum required to achieve
  2331 the original result when interning again, even if there is an overlap
  2332 with earlier declarations.
  2333 
  2334 * Pure/TheoryDataFun: change of the argument structure; 'prep_ext' is
  2335 now 'extend', and 'merge' gets an additional Pretty.pp argument
  2336 (useful for printing error messages).  INCOMPATIBILITY.
  2337 
  2338 * Pure: major reorganization of the theory context.  Type Sign.sg and
  2339 Theory.theory are now identified, referring to the universal
  2340 Context.theory (see Pure/context.ML).  Actual signature and theory
  2341 content is managed as theory data.  The old code and interfaces were
  2342 spread over many files and structures; the new arrangement introduces
  2343 considerable INCOMPATIBILITY to gain more clarity:
  2344 
  2345   Context -- theory management operations (name, identity, inclusion,
  2346     parents, ancestors, merge, etc.), plus generic theory data;
  2347 
  2348   Sign -- logical signature and syntax operations (declaring consts,
  2349     types, etc.), plus certify/read for common entities;
  2350 
  2351   Theory -- logical theory operations (stating axioms, definitions,
  2352     oracles), plus a copy of logical signature operations (consts,
  2353     types, etc.); also a few basic management operations (Theory.copy,
  2354     Theory.merge, etc.)
  2355 
  2356 The most basic sign_of operations (Theory.sign_of, Thm.sign_of_thm
  2357 etc.) as well as the sign field in Thm.rep_thm etc. have been retained
  2358 for convenience -- they merely return the theory.
  2359 
  2360 * Pure: type Type.tsig is superceded by theory in most interfaces.
  2361 
  2362 * Pure: the Isar proof context type is already defined early in Pure
  2363 as Context.proof (note that ProofContext.context and Proof.context are
  2364 aliases, where the latter is the preferred name).  This enables other
  2365 Isabelle components to refer to that type even before Isar is present.
  2366 
  2367 * Pure/sign/theory: discontinued named name spaces (i.e. classK,
  2368 typeK, constK, axiomK, oracleK), but provide explicit operations for
  2369 any of these kinds.  For example, Sign.intern typeK is now
  2370 Sign.intern_type, Theory.hide_space Sign.typeK is now
  2371 Theory.hide_types.  Also note that former
  2372 Theory.hide_classes/types/consts are now
  2373 Theory.hide_classes_i/types_i/consts_i, while the non '_i' versions
  2374 internalize their arguments!  INCOMPATIBILITY.
  2375 
  2376 * Pure: get_thm interface (of PureThy and ProofContext) expects
  2377 datatype thmref (with constructors Name and NameSelection) instead of
  2378 plain string -- INCOMPATIBILITY;
  2379 
  2380 * Pure: cases produced by proof methods specify options, where NONE
  2381 means to remove case bindings -- INCOMPATIBILITY in
  2382 (RAW_)METHOD_CASES.
  2383 
  2384 * Pure: the following operations retrieve axioms or theorems from a
  2385 theory node or theory hierarchy, respectively:
  2386 
  2387   Theory.axioms_of: theory -> (string * term) list
  2388   Theory.all_axioms_of: theory -> (string * term) list
  2389   PureThy.thms_of: theory -> (string * thm) list
  2390   PureThy.all_thms_of: theory -> (string * thm) list
  2391 
  2392 * Pure: print_tac now outputs the goal through the trace channel.
  2393 
  2394 * Isar toplevel: improved diagnostics, mostly for Poly/ML only.
  2395 Reference Toplevel.debug (default false) controls detailed printing
  2396 and tracing of low-level exceptions; Toplevel.profiling (default 0)
  2397 controls execution profiling -- set to 1 for time and 2 for space
  2398 (both increase the runtime).
  2399 
  2400 * Isar session: The initial use of ROOT.ML is now always timed,
  2401 i.e. the log will show the actual process times, in contrast to the
  2402 elapsed wall-clock time that the outer shell wrapper produces.
  2403 
  2404 * Simplifier: improved handling of bound variables (nameless
  2405 representation, avoid allocating new strings).  Simprocs that invoke
  2406 the Simplifier recursively should use Simplifier.inherit_bounds to
  2407 avoid local name clashes.  Failure to do so produces warnings
  2408 "Simplifier: renamed bound variable ..."; set Simplifier.debug_bounds
  2409 for further details.
  2410 
  2411 * ML functions legacy_bindings and use_legacy_bindings produce ML fact
  2412 bindings for all theorems stored within a given theory; this may help
  2413 in porting non-Isar theories to Isar ones, while keeping ML proof
  2414 scripts for the time being.
  2415 
  2416 * ML operator HTML.with_charset specifies the charset begin used for
  2417 generated HTML files.  For example:
  2418 
  2419   HTML.with_charset "utf-8" use_thy "Hebrew";
  2420   HTML.with_charset "utf-8" use_thy "Chinese";
  2421 
  2422 
  2423 *** System ***
  2424 
  2425 * Allow symlinks to all proper Isabelle executables (Isabelle,
  2426 isabelle, isatool etc.).
  2427 
  2428 * ISABELLE_DOC_FORMAT setting specifies preferred document format (for
  2429 isatool doc, isatool mkdir, display_drafts etc.).
  2430 
  2431 * isatool usedir: option -f allows specification of the ML file to be
  2432 used by Isabelle; default is ROOT.ML.
  2433 
  2434 * New isatool version outputs the version identifier of the Isabelle
  2435 distribution being used.
  2436 
  2437 * HOL: new isatool dimacs2hol converts files in DIMACS CNF format
  2438 (containing Boolean satisfiability problems) into Isabelle/HOL
  2439 theories.
  2440 
  2441 
  2442 
  2443 New in Isabelle2004 (April 2004)
  2444 --------------------------------
  2445 
  2446 *** General ***
  2447 
  2448 * Provers/order.ML:  new efficient reasoner for partial and linear orders.
  2449   Replaces linorder.ML.
  2450 
  2451 * Pure: Greek letters (except small lambda, \<lambda>), as well as Gothic
  2452   (\<aa>...\<zz>\<AA>...\<ZZ>), calligraphic (\<A>...\<Z>), and Euler
  2453   (\<a>...\<z>), are now considered normal letters, and can therefore
  2454   be used anywhere where an ASCII letter (a...zA...Z) has until
  2455   now. COMPATIBILITY: This obviously changes the parsing of some
  2456   terms, especially where a symbol has been used as a binder, say
  2457   '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed
  2458   as an identifier.  Fix it by inserting a space around former
  2459   symbols.  Call 'isatool fixgreek' to try to fix parsing errors in
  2460   existing theory and ML files.
  2461 
  2462 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.
  2463 
  2464 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now
  2465   allowed in identifiers. Similar to Greek letters \<^isub> is now considered
  2466   a normal (but invisible) letter. For multiple letter subscripts repeat
  2467   \<^isub> like this: x\<^isub>1\<^isub>2.
  2468 
  2469 * Pure: There are now sub-/superscripts that can span more than one
  2470   character. Text between \<^bsub> and \<^esub> is set in subscript in
  2471   ProofGeneral and LaTeX, text between \<^bsup> and \<^esup> in
  2472   superscript. The new control characters are not identifier parts.
  2473 
  2474 * Pure: Control-symbols of the form \<^raw:...> will literally print the
  2475   content of "..." to the latex file instead of \isacntrl... . The "..."
  2476   may consist of any printable characters excluding the end bracket >.
  2477 
  2478 * Pure: Using new Isar command "finalconsts" (or the ML functions
  2479   Theory.add_finals or Theory.add_finals_i) it is now possible to
  2480   declare constants "final", which prevents their being given a definition
  2481   later.  It is useful for constants whose behaviour is fixed axiomatically
  2482   rather than definitionally, such as the meta-logic connectives.
  2483 
  2484 * Pure: 'instance' now handles general arities with general sorts
  2485   (i.e. intersections of classes),
  2486 
  2487 * Presentation: generated HTML now uses a CSS style sheet to make layout
  2488   (somewhat) independent of content. It is copied from lib/html/isabelle.css.
  2489   It can be changed to alter the colors/layout of generated pages.
  2490 
  2491 
  2492 *** Isar ***
  2493 
  2494 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,
  2495   cut_tac, subgoal_tac and thin_tac:
  2496   - Now understand static (Isar) contexts.  As a consequence, users of Isar
  2497     locales are no longer forced to write Isar proof scripts.
  2498     For details see Isar Reference Manual, paragraph 4.3.2: Further tactic
  2499     emulations.
  2500   - INCOMPATIBILITY: names of variables to be instantiated may no
  2501     longer be enclosed in quotes.  Instead, precede variable name with `?'.
  2502     This is consistent with the instantiation attribute "where".
  2503 
  2504 * Attributes "where" and "of":
  2505   - Now take type variables of instantiated theorem into account when reading
  2506     the instantiation string.  This fixes a bug that caused instantiated
  2507     theorems to have too special types in some circumstances.
  2508   - "where" permits explicit instantiations of type variables.
  2509 
  2510 * Calculation commands "moreover" and "also" no longer interfere with
  2511   current facts ("this"), admitting arbitrary combinations with "then"
  2512   and derived forms.
  2513 
  2514 * Locales:
  2515   - Goal statements involving the context element "includes" no longer
  2516     generate theorems with internal delta predicates (those ending on
  2517     "_axioms") in the premise.
  2518     Resolve particular premise with <locale>.intro to obtain old form.
  2519   - Fixed bug in type inference ("unify_frozen") that prevented mix of target
  2520     specification and "includes" elements in goal statement.
  2521   - Rule sets <locale>.intro and <locale>.axioms no longer declared as
  2522     [intro?] and [elim?] (respectively) by default.
  2523   - Experimental command for instantiation of locales in proof contexts:
  2524         instantiate <label>[<attrs>]: <loc>
  2525     Instantiates locale <loc> and adds all its theorems to the current context
  2526     taking into account their attributes.  Label and attrs are optional
  2527     modifiers, like in theorem declarations.  If present, names of
  2528     instantiated theorems are qualified with <label>, and the attributes
  2529     <attrs> are applied after any attributes these theorems might have already.
  2530       If the locale has assumptions, a chained fact of the form
  2531     "<loc> t1 ... tn" is expected from which instantiations of the parameters
  2532     are derived.  The command does not support old-style locales declared
  2533     with "locale (open)".
  2534       A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.
  2535 
  2536 * HOL: Tactic emulation methods induct_tac and case_tac understand static
  2537   (Isar) contexts.
  2538 
  2539 
  2540 *** HOL ***
  2541 
  2542 * Proof import: new image HOL4 contains the imported library from
  2543   the HOL4 system with about 2500 theorems. It is imported by
  2544   replaying proof terms produced by HOL4 in Isabelle. The HOL4 image
  2545   can be used like any other Isabelle image.  See
  2546   HOL/Import/HOL/README for more information.
  2547 
  2548 * Simplifier:
  2549   - Much improved handling of linear and partial orders.
  2550     Reasoners for linear and partial orders are set up for type classes
  2551     "linorder" and "order" respectively, and are added to the default simpset
  2552     as solvers.  This means that the simplifier can build transitivity chains
  2553     to solve goals from the assumptions.
  2554   - INCOMPATIBILITY: old proofs break occasionally.  Typically, applications
  2555     of blast or auto after simplification become unnecessary because the goal
  2556     is solved by simplification already.
  2557 
  2558 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,
  2559     all proved in axiomatic type classes for semirings, rings and fields.
  2560 
  2561 * Numerics:
  2562   - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are
  2563     now formalized using the Ring_and_Field theory mentioned above.
  2564   - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently
  2565     than before, because now they are set up once in a generic manner.
  2566   - INCOMPATIBILITY: many type-specific arithmetic laws have gone.
  2567     Look for the general versions in Ring_and_Field (and Power if they concern
  2568     exponentiation).
  2569 
  2570 * Type "rat" of the rational numbers is now available in HOL-Complex.
  2571 
  2572 * Records:
  2573   - Record types are now by default printed with their type abbreviation
  2574     instead of the list of all field types. This can be configured via
  2575     the reference "print_record_type_abbr".
  2576   - Simproc "record_upd_simproc" for simplification of multiple updates added
  2577     (not enabled by default).
  2578   - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.
  2579     EX x. x = sel r to True (not enabled by default).
  2580   - Tactic "record_split_simp_tac" to split and simplify records added.
  2581 
  2582 * 'specification' command added, allowing for definition by
  2583   specification.  There is also an 'ax_specification' command that
  2584   introduces the new constants axiomatically.
  2585 
  2586 * arith(_tac) is now able to generate counterexamples for reals as well.
  2587 
  2588 * HOL-Algebra: new locale "ring" for non-commutative rings.
  2589 
  2590 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function
  2591   definitions, thanks to Sava Krsti\'{c} and John Matthews.
  2592 
  2593 * HOL-Matrix: a first theory for matrices in HOL with an application of
  2594   matrix theory to linear programming.
  2595 
  2596 * Unions and Intersections:
  2597   The latex output syntax of UN and INT has been changed
  2598   from "\Union x \in A. B" to "\Union_{x \in A} B"
  2599   i.e. the index formulae has become a subscript.
  2600   Similarly for "\Union x. B", and for \Inter instead of \Union.
  2601 
  2602 * Unions and Intersections over Intervals:
  2603   There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is
  2604   also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"
  2605   like in normal math, and corresponding versions for < and for intersection.
  2606 
  2607 * HOL/List: Ordering "lexico" is renamed "lenlex" and the standard
  2608   lexicographic dictonary ordering has been added as "lexord".
  2609 
  2610 * ML: the legacy theory structures Int and List have been removed. They had
  2611   conflicted with ML Basis Library structures having the same names.
  2612 
  2613 * 'refute' command added to search for (finite) countermodels.  Only works
  2614   for a fragment of HOL.  The installation of an external SAT solver is
  2615   highly recommended.  See "HOL/Refute.thy" for details.
  2616 
  2617 * 'quickcheck' command: Allows to find counterexamples by evaluating
  2618   formulae under an assignment of free variables to random values.
  2619   In contrast to 'refute', it can deal with inductive datatypes,
  2620   but cannot handle quantifiers. See "HOL/ex/Quickcheck_Examples.thy"
  2621   for examples.
  2622 
  2623 
  2624 *** HOLCF ***
  2625 
  2626 * Streams now come with concatenation and are part of the HOLCF image
  2627 
  2628 
  2629 
  2630 New in Isabelle2003 (May 2003)
  2631 ------------------------------
  2632 
  2633 *** General ***
  2634 
  2635 * Provers/simplifier:
  2636 
  2637   - Completely reimplemented method simp (ML: Asm_full_simp_tac):
  2638     Assumptions are now subject to complete mutual simplification,
  2639     not just from left to right. The simplifier now preserves
  2640     the order of assumptions.
  2641 
  2642     Potential INCOMPATIBILITY:
  2643 
  2644     -- simp sometimes diverges where the old version did
  2645        not, e.g. invoking simp on the goal
  2646 
  2647         [| P (f x); y = x; f x = f y |] ==> Q
  2648 
  2649        now gives rise to the infinite reduction sequence
  2650 
  2651         P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
  2652 
  2653        Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
  2654        kind of problem.
  2655 
  2656     -- Tactics combining classical reasoner and simplification (such as auto)
  2657        are also affected by this change, because many of them rely on
  2658        simp. They may sometimes diverge as well or yield a different numbers
  2659        of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
  2660        in case of problems. Sometimes subsequent calls to the classical
  2661        reasoner will fail because a preceeding call to the simplifier too
  2662        eagerly simplified the goal, e.g. deleted redundant premises.
  2663 
  2664   - The simplifier trace now shows the names of the applied rewrite rules
  2665 
  2666   - You can limit the number of recursive invocations of the simplifier
  2667     during conditional rewriting (where the simplifie tries to solve the
  2668     conditions before applying the rewrite rule):
  2669     ML "simp_depth_limit := n"
  2670     where n is an integer. Thus you can force termination where previously
  2671     the simplifier would diverge.
  2672 
  2673   - Accepts free variables as head terms in congruence rules.  Useful in Isar.
  2674 
  2675   - No longer aborts on failed congruence proof.  Instead, the
  2676     congruence is ignored.
  2677 
  2678 * Pure: New generic framework for extracting programs from constructive
  2679   proofs. See HOL/Extraction.thy for an example instantiation, as well
  2680   as HOL/Extraction for some case studies.
  2681 
  2682 * Pure: The main goal of the proof state is no longer shown by default, only
  2683 the subgoals. This behaviour is controlled by a new flag.
  2684    PG menu: Isabelle/Isar -> Settings -> Show Main Goal
  2685 (ML: Proof.show_main_goal).
  2686 
  2687 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
  2688 rules whose conclusion matches subgoal 1:
  2689       PG menu: Isabelle/Isar -> Show me -> matching rules
  2690 The rules are ordered by how closely they match the subgoal.
  2691 In particular, rules that solve a subgoal outright are displayed first
  2692 (or rather last, the way they are printed).
  2693 (ML: ProofGeneral.print_intros())
  2694 
  2695 * Pure: New flag trace_unify_fail causes unification to print
  2696 diagnostic information (PG: in trace buffer) when it fails. This is
  2697 useful for figuring out why single step proofs like rule, erule or
  2698 assumption failed.
  2699 
  2700 * Pure: Locale specifications now produce predicate definitions
  2701 according to the body of text (covering assumptions modulo local
  2702 definitions); predicate "loc_axioms" covers newly introduced text,
  2703 while "loc" is cumulative wrt. all included locale expressions; the
  2704 latter view is presented only on export into the global theory
  2705 context; potential INCOMPATIBILITY, use "(open)" option to fall back
  2706 on the old view without predicates;
  2707 
  2708 * Pure: predefined locales "var" and "struct" are useful for sharing
  2709 parameters (as in CASL, for example); just specify something like
  2710 ``var x + var y + struct M'' as import;
  2711 
  2712 * Pure: improved thms_containing: proper indexing of facts instead of
  2713 raw theorems; check validity of results wrt. current name space;
  2714 include local facts of proof configuration (also covers active
  2715 locales), cover fixed variables in index; may use "_" in term
  2716 specification; an optional limit for the number of printed facts may
  2717 be given (the default is 40);
  2718 
  2719 * Pure: disallow duplicate fact bindings within new-style theory files
  2720 (batch-mode only);
  2721 
  2722 * Provers: improved induct method: assumptions introduced by case
  2723 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
  2724 the goal statement); "foo" still refers to all facts collectively;
  2725 
  2726 * Provers: the function blast.overloaded has been removed: all constants
  2727 are regarded as potentially overloaded, which improves robustness in exchange
  2728 for slight decrease in efficiency;
  2729 
  2730 * Provers/linorder: New generic prover for transitivity reasoning over
  2731 linear orders.  Note: this prover is not efficient!
  2732 
  2733 * Isar: preview of problems to finish 'show' now produce an error
  2734 rather than just a warning (in interactive mode);
  2735 
  2736 
  2737 *** HOL ***
  2738 
  2739 * arith(_tac)
  2740 
  2741  - Produces a counter example if it cannot prove a goal.
  2742    Note that the counter example may be spurious if the goal is not a formula
  2743    of quantifier-free linear arithmetic.
  2744    In ProofGeneral the counter example appears in the trace buffer.
  2745 
  2746  - Knows about div k and mod k where k is a numeral of type nat or int.
  2747 
  2748  - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
  2749    linear arithmetic fails. This takes account of quantifiers and divisibility.
  2750    Presburger arithmetic can also be called explicitly via presburger(_tac).
  2751 
  2752 * simp's arithmetic capabilities have been enhanced a bit: it now
  2753 takes ~= in premises into account (by performing a case split);
  2754 
  2755 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
  2756 are distributed over a sum of terms;
  2757 
  2758 * New tactic "trans_tac" and method "trans" instantiate
  2759 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
  2760 "<=", "<" and "=").
  2761 
  2762 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main
  2763 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
  2764 
  2765 * 'typedef' command has new option "open" to suppress the set
  2766 definition;
  2767 
  2768 * functions Min and Max on finite sets have been introduced (theory
  2769 Finite_Set);
  2770 
  2771 * attribute [symmetric] now works for relations as well; it turns
  2772 (x,y) : R^-1 into (y,x) : R, and vice versa;
  2773 
  2774 * induct over a !!-quantified statement (say !!x1..xn):
  2775   each "case" automatically performs "fix x1 .. xn" with exactly those names.
  2776 
  2777 * Map: `empty' is no longer a constant but a syntactic abbreviation for
  2778 %x. None. Warning: empty_def now refers to the previously hidden definition
  2779 of the empty set.
  2780 
  2781 * Algebra: formalization of classical algebra.  Intended as base for
  2782 any algebraic development in Isabelle.  Currently covers group theory
  2783 (up to Sylow's theorem) and ring theory (Universal Property of
  2784 Univariate Polynomials).  Contributions welcome;
  2785 
  2786 * GroupTheory: deleted, since its material has been moved to Algebra;
  2787 
  2788 * Complex: new directory of the complex numbers with numeric constants,
  2789 nonstandard complex numbers, and some complex analysis, standard and
  2790 nonstandard (Jacques Fleuriot);
  2791 
  2792 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
  2793 
  2794 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques
  2795 Fleuriot);
  2796 
  2797 * Real/HahnBanach: updated and adapted to locales;
  2798 
  2799 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
  2800 Gray and Kramer);
  2801 
  2802 * UNITY: added the Meier-Sanders theory of progress sets;
  2803 
  2804 * MicroJava: bytecode verifier and lightweight bytecode verifier
  2805 as abstract algorithms, instantiated to the JVM;
  2806 
  2807 * Bali: Java source language formalization. Type system, operational
  2808 semantics, axiomatic semantics. Supported language features:
  2809 classes, interfaces, objects,virtual methods, static methods,
  2810 static/instance fields, arrays, access modifiers, definite
  2811 assignment, exceptions.
  2812 
  2813 
  2814 *** ZF ***
  2815 
  2816 * ZF/Constructible: consistency proof for AC (Gdel's constructible
  2817 universe, etc.);
  2818 
  2819 * Main ZF: virtually all theories converted to new-style format;
  2820 
  2821 
  2822 *** ML ***
  2823 
  2824 * Pure: Tactic.prove provides sane interface for internal proofs;
  2825 omits the infamous "standard" operation, so this is more appropriate
  2826 than prove_goalw_cterm in many situations (e.g. in simprocs);
  2827 
  2828 * Pure: improved error reporting of simprocs;
  2829 
  2830 * Provers: Simplifier.simproc(_i) provides sane interface for setting
  2831 up simprocs;
  2832 
  2833 
  2834 *** Document preparation ***
  2835 
  2836 * uses \par instead of \\ for line breaks in theory text. This may
  2837 shift some page breaks in large documents. To get the old behaviour
  2838 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
  2839 
  2840 * minimized dependencies of isabelle.sty and isabellesym.sty on
  2841 other packages
  2842 
  2843 * \<euro> now needs package babel/greek instead of marvosym (which
  2844 broke \Rightarrow)
  2845 
  2846 * normal size for \<zero>...\<nine> (uses \mathbf instead of
  2847 textcomp package)
  2848 
  2849 
  2850 
  2851 New in Isabelle2002 (March 2002)
  2852 --------------------------------
  2853 
  2854 *** Document preparation ***
  2855 
  2856 * greatly simplified document preparation setup, including more
  2857 graceful interpretation of isatool usedir -i/-d/-D options, and more
  2858 instructive isatool mkdir; users should basically be able to get
  2859 started with "isatool mkdir HOL Test && isatool make"; alternatively,
  2860 users may run a separate document processing stage manually like this:
  2861 "isatool usedir -D output HOL Test && isatool document Test/output";
  2862 
  2863 * theory dependency graph may now be incorporated into documents;
  2864 isatool usedir -g true will produce session_graph.eps/.pdf for use
  2865 with \includegraphics of LaTeX;
  2866 
  2867 * proper spacing of consecutive markup elements, especially text
  2868 blocks after section headings;
  2869 
  2870 * support bold style (for single symbols only), input syntax is like
  2871 this: "\<^bold>\<alpha>" or "\<^bold>A";
  2872 
  2873 * \<bullet> is now output as bold \cdot by default, which looks much
  2874 better in printed text;
  2875 
  2876 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
  2877 note that these symbols are currently unavailable in Proof General /
  2878 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
  2879 
  2880 * isatool latex no longer depends on changed TEXINPUTS, instead
  2881 isatool document copies the Isabelle style files to the target
  2882 location;
  2883 
  2884 
  2885 *** Isar ***
  2886 
  2887 * Pure/Provers: improved proof by cases and induction;
  2888   - 'case' command admits impromptu naming of parameters (such as
  2889     "case (Suc n)");
  2890   - 'induct' method divinates rule instantiation from the inductive
  2891     claim; no longer requires excessive ?P bindings for proper
  2892     instantiation of cases;
  2893   - 'induct' method properly enumerates all possibilities of set/type
  2894     rules; as a consequence facts may be also passed through *type*
  2895     rules without further ado;
  2896   - 'induct' method now derives symbolic cases from the *rulified*
  2897     rule (before it used to rulify cases stemming from the internal
  2898     atomized version); this means that the context of a non-atomic
  2899     statement becomes is included in the hypothesis, avoiding the
  2900     slightly cumbersome show "PROP ?case" form;
  2901   - 'induct' may now use elim-style induction rules without chaining
  2902     facts, using ``missing'' premises from the goal state; this allows
  2903     rules stemming from inductive sets to be applied in unstructured
  2904     scripts, while still benefitting from proper handling of non-atomic
  2905     statements; NB: major inductive premises need to be put first, all
  2906     the rest of the goal is passed through the induction;
  2907   - 'induct' proper support for mutual induction involving non-atomic
  2908     rule statements (uses the new concept of simultaneous goals, see
  2909     below);
  2910   - append all possible rule selections, but only use the first
  2911     success (no backtracking);
  2912   - removed obsolete "(simplified)" and "(stripped)" options of methods;
  2913   - undeclared rule case names default to numbers 1, 2, 3, ...;
  2914   - added 'print_induct_rules' (covered by help item in recent Proof
  2915     General versions);
  2916   - moved induct/cases attributes to Pure, methods to Provers;
  2917   - generic method setup instantiated for FOL and HOL;
  2918 
  2919 * Pure: support multiple simultaneous goal statements, for example
  2920 "have a: A and b: B" (same for 'theorem' etc.); being a pure
  2921 meta-level mechanism, this acts as if several individual goals had
  2922 been stated separately; in particular common proof methods need to be
  2923 repeated in order to cover all claims; note that a single elimination
  2924 step is *not* sufficient to establish the two conjunctions, so this
  2925 fails:
  2926 
  2927   assume "A & B" then have A and B ..   (*".." fails*)
  2928 
  2929 better use "obtain" in situations as above; alternative refer to
  2930 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
  2931 
  2932 * Pure: proper integration with ``locales''; unlike the original
  2933 version by Florian Kammller, Isar locales package high-level proof
  2934 contexts rather than raw logical ones (e.g. we admit to include
  2935 attributes everywhere); operations on locales include merge and
  2936 rename; support for implicit arguments (``structures''); simultaneous
  2937 type-inference over imports and text; see also HOL/ex/Locales.thy for
  2938 some examples;
  2939 
  2940 * Pure: the following commands have been ``localized'', supporting a
  2941 target locale specification "(in name)": 'lemma', 'theorem',
  2942 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
  2943 stored both within the locale and at the theory level (exported and
  2944 qualified by the locale name);
  2945 
  2946 * Pure: theory goals may now be specified in ``long'' form, with
  2947 ad-hoc contexts consisting of arbitrary locale elements. for example
  2948 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
  2949 definitions may be given, too); the result is a meta-level rule with
  2950 the context elements being discharged in the obvious way;
  2951 
  2952 * Pure: new proof command 'using' allows to augment currently used
  2953 facts after a goal statement ('using' is syntactically analogous to
  2954 'apply', but acts on the goal's facts only); this allows chained facts
  2955 to be separated into parts given before and after a claim, as in
  2956 ``from a and b have C using d and e <proof>'';
  2957 
  2958 * Pure: renamed "antecedent" case to "rule_context";
  2959 
  2960 * Pure: new 'judgment' command records explicit information about the
  2961 object-logic embedding (used by several tools internally); no longer
  2962 use hard-wired "Trueprop";
  2963 
  2964 * Pure: added 'corollary' command;
  2965 
  2966 * Pure: fixed 'token_translation' command;
  2967 
  2968 * Pure: removed obsolete 'exported' attribute;
  2969 
  2970 * Pure: dummy pattern "_" in is/let is now automatically lifted over
  2971 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
  2972 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
  2973 
  2974 * Pure: method 'atomize' presents local goal premises as object-level
  2975 statements (atomic meta-level propositions); setup controlled via
  2976 rewrite rules declarations of 'atomize' attribute; example
  2977 application: 'induct' method with proper rule statements in improper
  2978 proof *scripts*;
  2979 
  2980 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
  2981 now consider the syntactic context of assumptions, giving a better
  2982 chance to get type-inference of the arguments right (this is
  2983 especially important for locales);
  2984 
  2985 * Pure: "sorry" no longer requires quick_and_dirty in interactive
  2986 mode;
  2987 
  2988 * Pure/obtain: the formal conclusion "thesis", being marked as
  2989 ``internal'', may no longer be reference directly in the text;
  2990 potential INCOMPATIBILITY, may need to use "?thesis" in rare
  2991 situations;
  2992 
  2993 * Pure: generic 'sym' attribute which declares a rule both as pure
  2994 'elim?' and for the 'symmetric' operation;
  2995 
  2996 * Pure: marginal comments ``--'' may now occur just anywhere in the
  2997 text; the fixed correlation with particular command syntax has been
  2998 discontinued;
  2999 
  3000 * Pure: new method 'rules' is particularly well-suited for proof
  3001 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
  3002 but often produces more compact proof terms with less detours;
  3003 
  3004 * Pure/Provers/classical: simplified integration with pure rule
  3005 attributes and methods; the classical "intro?/elim?/dest?"
  3006 declarations coincide with the pure ones; the "rule" method no longer
  3007 includes classically swapped intros; "intro" and "elim" methods no
  3008 longer pick rules from the context; also got rid of ML declarations
  3009 AddXIs/AddXEs/AddXDs; all of this has some potential for
  3010 INCOMPATIBILITY;
  3011 
  3012 * Provers/classical: attribute 'swapped' produces classical inversions
  3013 of introduction rules;
  3014 
  3015 * Provers/simplifier: 'simplified' attribute may refer to explicit
  3016 rules instead of full simplifier context; 'iff' attribute handles
  3017 conditional rules;
  3018 
  3019 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
  3020 
  3021 * HOL: 'recdef' now fails on unfinished automated proofs, use
  3022 "(permissive)" option to recover old behavior;
  3023 
  3024 * HOL: 'inductive' no longer features separate (collective) attributes
  3025 for 'intros' (was found too confusing);
  3026 
  3027 * HOL: properly declared induction rules less_induct and
  3028 wf_induct_rule;
  3029 
  3030 
  3031 *** HOL ***
  3032 
  3033 * HOL: moved over to sane numeral syntax; the new policy is as
  3034 follows:
  3035 
  3036   - 0 and 1 are polymorphic constants, which are defined on any
  3037   numeric type (nat, int, real etc.);
  3038 
  3039   - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
  3040   binary representation internally;
  3041 
  3042   - type nat has special constructor Suc, and generally prefers Suc 0
  3043   over 1::nat and Suc (Suc 0) over 2::nat;
  3044 
  3045 This change may cause significant problems of INCOMPATIBILITY; here
  3046 are some hints on converting existing sources:
  3047 
  3048   - due to the new "num" token, "-0" and "-1" etc. are now atomic
  3049   entities, so expressions involving "-" (unary or binary minus) need
  3050   to be spaced properly;
  3051 
  3052   - existing occurrences of "1" may need to be constraint "1::nat" or
  3053   even replaced by Suc 0; similar for old "2";
  3054 
  3055   - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
  3056 
  3057   - remove all special provisions on numerals in proofs;
  3058 
  3059 * HOL: simp rules nat_number expand numerals on nat to Suc/0
  3060 representation (depends on bin_arith_simps in the default context);
  3061 
  3062 * HOL: symbolic syntax for x^2 (numeral 2);
  3063 
  3064 * HOL: the class of all HOL types is now called "type" rather than
  3065 "term"; INCOMPATIBILITY, need to adapt references to this type class
  3066 in axclass/classes, instance/arities, and (usually rare) occurrences
  3067 in typings (of consts etc.); internally the class is called
  3068 "HOL.type", ML programs should refer to HOLogic.typeS;
  3069 
  3070 * HOL/record package improvements:
  3071   - new derived operations "fields" to build a partial record section,
  3072     "extend" to promote a fixed record to a record scheme, and
  3073     "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
  3074     declared as simp by default;
  3075   - shared operations ("more", "fields", etc.) now need to be always
  3076     qualified) --- potential INCOMPATIBILITY;
  3077   - removed "make_scheme" operations (use "make" with "extend") --
  3078     INCOMPATIBILITY;
  3079   - removed "more" class (simply use "term") -- INCOMPATIBILITY;
  3080   - provides cases/induct rules for use with corresponding Isar
  3081     methods (for concrete records, record schemes, concrete more
  3082     parts, and schematic more parts -- in that order);
  3083   - internal definitions directly based on a light-weight abstract
  3084     theory of product types over typedef rather than datatype;
  3085 
  3086 * HOL: generic code generator for generating executable ML code from
  3087 specifications; specific support for HOL constructs such as inductive
  3088 datatypes and sets, as well as recursive functions; can be invoked
  3089 via 'generate_code' theory section;
  3090 
  3091 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
  3092 
  3093 * HOL: consolidated and renamed several theories.  In particular:
  3094         Ord.thy has been absorbed into HOL.thy
  3095         String.thy has been absorbed into List.thy
  3096 
  3097 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
  3098 (beware of argument permutation!);
  3099 
  3100 * HOL: linorder_less_split superseded by linorder_cases;
  3101 
  3102 * HOL/List: "nodups" renamed to "distinct";
  3103 
  3104 * HOL: added "The" definite description operator; move Hilbert's "Eps"
  3105 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
  3106   - Ex_def has changed, now need to use some_eq_ex
  3107 
  3108 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
  3109 in this (rare) case use:
  3110 
  3111   delSWrapper "split_all_tac"
  3112   addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
  3113 
  3114 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
  3115 MAY FAIL;
  3116 
  3117 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
  3118 Isabelle's type classes, ^ on functions and relations has too general
  3119 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
  3120 necessary to attach explicit type constraints;
  3121 
  3122 * HOL/Relation: the prefix name of the infix "O" has been changed from
  3123 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
  3124 renamed accordingly (eg "compI" -> "rel_compI").
  3125 
  3126 * HOL: syntax translations now work properly with numerals and records
  3127 expressions;
  3128 
  3129 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
  3130 of "lam" -- INCOMPATIBILITY;
  3131 
  3132 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
  3133 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
  3134 renamed "Product_Type.unit";
  3135 
  3136 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
  3137 
  3138 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
  3139 the "cases" method);
  3140 
  3141 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
  3142 Florian Kammller);
  3143 
  3144 * HOL/IMP: updated and converted to new-style theory format; several
  3145 parts turned into readable document, with proper Isar proof texts and
  3146 some explanations (by Gerwin Klein);
  3147 
  3148 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
  3149 
  3150 * HOL-Hyperreal is now a logic image;
  3151 
  3152 
  3153 *** HOLCF ***
  3154 
  3155 * Isar: consts/constdefs supports mixfix syntax for continuous
  3156 operations;
  3157 
  3158 * Isar: domain package adapted to new-style theory format, e.g. see
  3159 HOLCF/ex/Dnat.thy;
  3160 
  3161 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
  3162 potential INCOMPATIBILITY; now use plain induct_tac instead of former
  3163 lift.induct_tac, always use UU instead of Undef;
  3164 
  3165 * HOLCF/IMP: updated and converted to new-style theory;
  3166 
  3167 
  3168 *** ZF ***
  3169 
  3170 * Isar: proper integration of logic-specific tools and packages,
  3171 including theory commands '(co)inductive', '(co)datatype',
  3172 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
  3173 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
  3174 
  3175 * theory Main no longer includes AC; for the Axiom of Choice, base
  3176 your theory on Main_ZFC;
  3177 
  3178 * the integer library now covers quotients and remainders, with many
  3179 laws relating division to addition, multiplication, etc.;
  3180 
  3181 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
  3182 typeless version of the formalism;
  3183 
  3184 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
  3185 format;
  3186 
  3187 * ZF/Induct: new directory for examples of inductive definitions,
  3188 including theory Multiset for multiset orderings; converted to
  3189 new-style theory format;
  3190 
  3191 * ZF: many new theorems about lists, ordinals, etc.;
  3192 
  3193 
  3194 *** General ***
  3195 
  3196 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
  3197 variable proof controls level of detail: 0 = no proofs (only oracle
  3198 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
  3199 also ref manual for further ML interfaces;
  3200 
  3201 * Pure/axclass: removed obsolete ML interface
  3202 goal_subclass/goal_arity;
  3203 
  3204 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
  3205 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
  3206 separate tokens, so expressions involving minus need to be spaced
  3207 properly;
  3208 
  3209 * Pure/syntax: support non-oriented infixes, using keyword "infix"
  3210 rather than "infixl" or "infixr";
  3211 
  3212 * Pure/syntax: concrete syntax for dummy type variables admits genuine
  3213 sort constraint specifications in type inference; e.g. "x::_::foo"
  3214 ensures that the type of "x" is of sort "foo" (but not necessarily a
  3215 type variable);
  3216 
  3217 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
  3218 control output of nested => (types); the default behavior is
  3219 "type_brackets";
  3220 
  3221 * Pure/syntax: builtin parse translation for "_constify" turns valued
  3222 tokens into AST constants;
  3223 
  3224 * Pure/syntax: prefer later declarations of translations and print
  3225 translation functions; potential INCOMPATIBILITY: need to reverse
  3226 multiple declarations for same syntax element constant;
  3227 
  3228 * Pure/show_hyps reset by default (in accordance to existing Isar
  3229 practice);
  3230 
  3231 * Provers/classical: renamed addaltern to addafter, addSaltern to
  3232 addSafter;
  3233 
  3234 * Provers/clasimp: ``iff'' declarations now handle conditional rules
  3235 as well;
  3236 
  3237 * system: tested support for MacOS X; should be able to get Isabelle +
  3238 Proof General to work in a plain Terminal after installing Poly/ML
  3239 (e.g. from the Isabelle distribution area) and GNU bash alone
  3240 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
  3241 support requires further installations, e.g. from
  3242 http://fink.sourceforge.net/);
  3243 
  3244 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
  3245 
  3246 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
  3247 of 40 MB), cf. ML_OPTIONS;
  3248 
  3249 * system: Proof General keywords specification is now part of the
  3250 Isabelle distribution (see etc/isar-keywords.el);
  3251 
  3252 * system: support for persistent Proof General sessions (refrain from
  3253 outdating all loaded theories on startup); user may create writable
  3254 logic images like this: ``isabelle -q HOL Test'';
  3255 
  3256 * system: smart selection of Isabelle process versus Isabelle
  3257 interface, accommodates case-insensitive file systems (e.g. HFS+); may
  3258 run both "isabelle" and "Isabelle" even if file names are badly
  3259 damaged (executable inspects the case of the first letter of its own
  3260 name); added separate "isabelle-process" and "isabelle-interface";
  3261 
  3262 * system: refrain from any attempt at filtering input streams; no
  3263 longer support ``8bit'' encoding of old isabelle font, instead proper
  3264 iso-latin characters may now be used; the related isatools
  3265 "symbolinput" and "nonascii" have disappeared as well;
  3266 
  3267 * system: removed old "xterm" interface (the print modes "xterm" and
  3268 "xterm_color" are still available for direct use in a suitable
  3269 terminal);
  3270 
  3271 
  3272 
  3273 New in Isabelle99-2 (February 2001)
  3274 -----------------------------------
  3275 
  3276 *** Overview of INCOMPATIBILITIES ***
  3277 
  3278 * HOL: please note that theories in the Library and elsewhere often use the
  3279 new-style (Isar) format; to refer to their theorems in an ML script you must
  3280 bind them to ML identifers by e.g.      val thm_name = thm "thm_name";
  3281 
  3282 * HOL: inductive package no longer splits induction rule aggressively,
  3283 but only as far as specified by the introductions given; the old
  3284 format may be recovered via ML function complete_split_rule or attribute
  3285 'split_rule (complete)';
  3286 
  3287 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
  3288 gfp_Tarski to gfp_unfold;
  3289 
  3290 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
  3291 
  3292 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
  3293 relation); infix "^^" has been renamed "``"; infix "``" has been
  3294 renamed "`"; "univalent" has been renamed "single_valued";
  3295 
  3296 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
  3297 operation;
  3298 
  3299 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
  3300 
  3301 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
  3302 
  3303 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
  3304 consequence, it is no longer monotonic wrt. the local goal context
  3305 (which is now passed through the inductive cases);
  3306 
  3307 * Document preparation: renamed standard symbols \<ll> to \<lless> and
  3308 \<gg> to \<ggreater>;
  3309 
  3310 
  3311 *** Document preparation ***
  3312 
  3313 * \isabellestyle{NAME} selects version of Isabelle output (currently
  3314 available: are "it" for near math-mode best-style output, "sl" for
  3315 slanted text style, and "tt" for plain type-writer; if no
  3316 \isabellestyle command is given, output is according to slanted
  3317 type-writer);
  3318 
  3319 * support sub/super scripts (for single symbols only), input syntax is
  3320 like this: "A\<^sup>*" or "A\<^sup>\<star>";
  3321 
  3322 * some more standard symbols; see Appendix A of the system manual for
  3323 the complete list of symbols defined in isabellesym.sty;
  3324 
  3325 * improved isabelle style files; more abstract symbol implementation
  3326 (should now use \isamath{...} and \isatext{...} in custom symbol
  3327 definitions);
  3328 
  3329 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
  3330 state; Note that presentation of goal states does not conform to
  3331 actual human-readable proof documents.  Please do not include goal
  3332 states into document output unless you really know what you are doing!
  3333 
  3334 * proper indentation of antiquoted output with proportional LaTeX
  3335 fonts;
  3336 
  3337 * no_document ML operator temporarily disables LaTeX document
  3338 generation;
  3339 
  3340 * isatool unsymbolize tunes sources for plain ASCII communication;
  3341 
  3342 
  3343 *** Isar ***
  3344 
  3345 * Pure: Isar now suffers initial goal statements to contain unbound
  3346 schematic variables (this does not conform to actual readable proof
  3347 documents, due to unpredictable outcome and non-compositional proof
  3348 checking); users who know what they are doing may use schematic goals
  3349 for Prolog-style synthesis of proven results;
  3350 
  3351 * Pure: assumption method (an implicit finishing) now handles actual
  3352 rules as well;
  3353 
  3354 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
  3355 initial goal, declare "that" only as Pure intro (only for single
  3356 steps); the "that" rule assumption may now be involved in implicit
  3357 finishing, thus ".." becomes a feasible for trivial obtains;
  3358 
  3359 * Pure: default proof step now includes 'intro_classes'; thus trivial
  3360 instance proofs may be performed by "..";
  3361 
  3362 * Pure: ?thesis / ?this / "..." now work for pure meta-level
  3363 statements as well;
  3364 
  3365 * Pure: more robust selection of calculational rules;
  3366 
  3367 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
  3368 rule (as well as the assumption rule);
  3369 
  3370 * Pure: 'thm_deps' command visualizes dependencies of theorems and
  3371 lemmas, using the graph browser tool;
  3372 
  3373 * Pure: predict failure of "show" in interactive mode;
  3374 
  3375 * Pure: 'thms_containing' now takes actual terms as arguments;
  3376 
  3377 * HOL: improved method 'induct' --- now handles non-atomic goals
  3378 (potential INCOMPATIBILITY); tuned error handling;
  3379 
  3380 * HOL: cases and induct rules now provide explicit hints about the
  3381 number of facts to be consumed (0 for "type" and 1 for "set" rules);
  3382 any remaining facts are inserted into the goal verbatim;
  3383 
  3384 * HOL: local contexts (aka cases) may now contain term bindings as
  3385 well; the 'cases' and 'induct' methods new provide a ?case binding for
  3386 the result to be shown in each case;
  3387 
  3388 * HOL: added 'recdef_tc' command;
  3389 
  3390 * isatool convert assists in eliminating legacy ML scripts;
  3391 
  3392 
  3393 *** HOL ***
  3394 
  3395 * HOL/Library: a collection of generic theories to be used together
  3396 with main HOL; the theory loader path already includes this directory
  3397 by default; the following existing theories have been moved here:
  3398 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
  3399 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
  3400 
  3401 * HOL/Unix: "Some aspects of Unix file-system security", a typical
  3402 modelling and verification task performed in Isabelle/HOL +
  3403 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
  3404 
  3405 * HOL/Algebra: special summation operator SUM no longer exists, it has
  3406 been replaced by setsum; infix 'assoc' now has priority 50 (like
  3407 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
  3408 'domain', this makes the theory consistent with mathematical
  3409 literature;
  3410 
  3411 * HOL basics: added overloaded operations "inverse" and "divide"
  3412 (infix "/"), syntax for generic "abs" operation, generic summation
  3413 operator \<Sum>;
  3414 
  3415 * HOL/typedef: simplified package, provide more useful rules (see also
  3416 HOL/subset.thy);
  3417 
  3418 * HOL/datatype: induction rule for arbitrarily branching datatypes is
  3419 now expressed as a proper nested rule (old-style tactic scripts may
  3420 require atomize_strip_tac to cope with non-atomic premises);
  3421 
  3422 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
  3423 to "split_conv" (old name still available for compatibility);
  3424 
  3425 * HOL: improved concrete syntax for strings (e.g. allows translation
  3426 rules with string literals);
  3427 
  3428 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
  3429  and Fleuriot's mechanization of analysis, including the transcendental
  3430  functions for the reals;
  3431 
  3432 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
  3433 
  3434 
  3435 *** CTT ***
  3436 
  3437 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
  3438 "lam" is displayed as TWO lambda-symbols
  3439 
  3440 * CTT: theory Main now available, containing everything (that is, Bool
  3441 and Arith);
  3442 
  3443 
  3444 *** General ***
  3445 
  3446 * Pure: the Simplifier has been implemented properly as a derived rule
  3447 outside of the actual kernel (at last!); the overall performance
  3448 penalty in practical applications is about 50%, while reliability of
  3449 the Isabelle inference kernel has been greatly improved;
  3450 
  3451 * print modes "brackets" and "no_brackets" control output of nested =>
  3452 (types) and ==> (props); the default behaviour is "brackets";
  3453 
  3454 * Provers: fast_tac (and friends) now handle actual object-logic rules
  3455 as assumptions as well;
  3456 
  3457 * system: support Poly/ML 4.0;
  3458 
  3459 * system: isatool install handles KDE version 1 or 2;
  3460 
  3461 
  3462 
  3463 New in Isabelle99-1 (October 2000)
  3464 ----------------------------------
  3465 
  3466 *** Overview of INCOMPATIBILITIES ***
  3467 
  3468 * HOL: simplification of natural numbers is much changed; to partly
  3469 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
  3470 issue the following ML commands:
  3471 
  3472   Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
  3473   Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
  3474 
  3475 * HOL: simplification no longer dives into case-expressions; this is
  3476 controlled by "t.weak_case_cong" for each datatype t;
  3477 
  3478 * HOL: nat_less_induct renamed to less_induct;
  3479 
  3480 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
  3481 fixsome to patch .thy and .ML sources automatically;
  3482 
  3483   select_equality  -> some_equality
  3484   select_eq_Ex     -> some_eq_ex
  3485   selectI2EX       -> someI2_ex
  3486   selectI2         -> someI2
  3487   selectI          -> someI
  3488   select1_equality -> some1_equality
  3489   Eps_sym_eq       -> some_sym_eq_trivial
  3490   Eps_eq           -> some_eq_trivial
  3491 
  3492 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
  3493 
  3494 * HOL: removed obsolete theorem binding expand_if (refer to split_if
  3495 instead);
  3496 
  3497 * HOL: the recursion equations generated by 'recdef' are now called
  3498 f.simps instead of f.rules;
  3499 
  3500 * HOL: qed_spec_mp now also handles bounded ALL as well;
  3501 
  3502 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
  3503 sometimes be needed;
  3504 
  3505 * HOL: the constant for "f``x" is now "image" rather than "op ``";
  3506 
  3507 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
  3508 
  3509 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
  3510 product is now "<*>" instead of "Times"; the lexicographic product is
  3511 now "<*lex*>" instead of "**";
  3512 
  3513 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
  3514 of main HOL, but was unused); better use HOL's datatype package;
  3515 
  3516 * HOL: removed "symbols" syntax for constant "override" of theory Map;
  3517 the old syntax may be recovered as follows:
  3518 
  3519   syntax (symbols)
  3520     override  :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
  3521       (infixl "\\<oplus>" 100)
  3522 
  3523 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3524 
  3525 * HOL/ML: even fewer consts are declared as global (see theories Ord,
  3526 Lfp, Gfp, WF); this only affects ML packages that refer to const names
  3527 internally;
  3528 
  3529 * HOL and ZF: syntax for quotienting wrt an equivalence relation
  3530 changed from A/r to A//r;
  3531 
  3532 * ZF: new treatment of arithmetic (nat & int) may break some old
  3533 proofs;
  3534 
  3535 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
  3536 rulify -> rule_format, elimify -> elim_format, ...);
  3537 
  3538 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3539 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
  3540 should have to change intro!! to intro? only); replaced "delrule" by
  3541 "rule del";
  3542 
  3543 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
  3544 
  3545 * Provers: strengthened force_tac by using new first_best_tac;
  3546 
  3547 * LaTeX document preparation: several changes of isabelle.sty (see
  3548 lib/texinputs);
  3549 
  3550 
  3551 *** Document preparation ***
  3552 
  3553 * formal comments (text blocks etc.) in new-style theories may now
  3554 contain antiquotations of thm/prop/term/typ/text to be presented
  3555 according to latex print mode; concrete syntax is like this:
  3556 @{term[show_types] "f(x) = a + x"};
  3557 
  3558 * isatool mkdir provides easy setup of Isabelle session directories,
  3559 including proper document sources;
  3560 
  3561 * generated LaTeX sources are now deleted after successful run
  3562 (isatool document -c); may retain a copy somewhere else via -D option
  3563 of isatool usedir;
  3564 
  3565 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
  3566 style files, achieving self-contained LaTeX sources and simplifying
  3567 LaTeX debugging;
  3568 
  3569 * old-style theories now produce (crude) LaTeX output as well;
  3570 
  3571 * browser info session directories are now self-contained (may be put
  3572 on WWW server seperately); improved graphs of nested sessions; removed
  3573 graph for 'all sessions';
  3574 
  3575 * several improvements in isabelle style files; \isabellestyle{it}
  3576 produces fake math mode output; \isamarkupheader is now \section by
  3577 default; see lib/texinputs/isabelle.sty etc.;
  3578 
  3579 
  3580 *** Isar ***
  3581 
  3582 * Isar/Pure: local results and corresponding term bindings are now
  3583 subject to Hindley-Milner polymorphism (similar to ML); this
  3584 accommodates incremental type-inference very nicely;
  3585 
  3586 * Isar/Pure: new derived language element 'obtain' supports
  3587 generalized existence reasoning;
  3588 
  3589 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
  3590 support accumulation of results, without applying any rules yet;
  3591 useful to collect intermediate results without explicit name
  3592 references, and for use with transitivity rules with more than 2
  3593 premises;
  3594 
  3595 * Isar/Pure: scalable support for case-analysis type proofs: new
  3596 'case' language element refers to local contexts symbolically, as
  3597 produced by certain proof methods; internally, case names are attached
  3598 to theorems as "tags";
  3599 
  3600 * Isar/Pure: theory command 'hide' removes declarations from
  3601 class/type/const name spaces;
  3602 
  3603 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
  3604 indicate potential overloading;
  3605 
  3606 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
  3607 
  3608 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
  3609 "{a,b,c}" instead of {a,b,c};
  3610 
  3611 * Isar/Pure now provides its own version of intro/elim/dest
  3612 attributes; useful for building new logics, but beware of confusion
  3613 with the version in Provers/classical;
  3614 
  3615 * Isar/Pure: the local context of (non-atomic) goals is provided via
  3616 case name 'antecedent';
  3617 
  3618 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
  3619 to the current context is now done automatically);
  3620 
  3621 * Isar/Pure: theory command 'method_setup' provides a simple interface
  3622 for definining proof methods in ML;
  3623 
  3624 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3625 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
  3626 most cases, one should have to change intro!! to intro? only);
  3627 replaced "delrule" by "rule del";
  3628 
  3629 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
  3630 'symmetric' attribute (the latter supercedes [RS sym]);
  3631 
  3632 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
  3633 method modifier); 'simp' method: 'only:' modifier removes loopers as
  3634 well (including splits);
  3635 
  3636 * Isar/Provers: Simplifier and Classical methods now support all kind
  3637 of modifiers used in the past, including 'cong', 'iff', etc.
  3638 
  3639 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
  3640 of Simplifier and Classical reasoner);
  3641 
  3642 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
  3643 now support named cases; major packages (inductive, datatype, primrec,
  3644 recdef) support case names and properly name parameters;
  3645 
  3646 * Isar/HOL: new transitivity rules for substitution in inequalities --
  3647 monotonicity conditions are extracted to be proven at end of
  3648 calculations;
  3649 
  3650 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
  3651 method anyway;
  3652 
  3653 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
  3654 split_if split_if_asm; datatype package provides theorems foo.splits =
  3655 foo.split foo.split_asm for each datatype;
  3656 
  3657 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
  3658 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
  3659 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
  3660 use "(cases (simplified))" method in proper proof texts);
  3661 
  3662 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
  3663 
  3664 * Isar: names of theorems etc. may be natural numbers as well;
  3665 
  3666 * Isar: 'pr' command: optional arguments for goals_limit and
  3667 ProofContext.prems_limit; no longer prints theory contexts, but only
  3668 proof states;
  3669 
  3670 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
  3671 additional print modes to be specified; e.g. "pr(latex)" will print
  3672 proof state according to the Isabelle LaTeX style;
  3673 
  3674 * Isar: improved support for emulating tactic scripts, including proof
  3675 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
  3676 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
  3677 (for HOL datatypes);
  3678 
  3679 * Isar: simplified (more robust) goal selection of proof methods: 1st
  3680 goal, all goals, or explicit goal specifier (tactic emulation); thus
  3681 'proof method scripts' have to be in depth-first order;
  3682 
  3683 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
  3684 
  3685 * Isar: removed 'help' command, which hasn't been too helpful anyway;
  3686 should instead use individual commands for printing items
  3687 (print_commands, print_methods etc.);
  3688 
  3689 * Isar: added 'nothing' --- the empty list of theorems;
  3690 
  3691 
  3692 *** HOL ***
  3693 
  3694 * HOL/MicroJava: formalization of a fragment of Java, together with a
  3695 corresponding virtual machine and a specification of its bytecode
  3696 verifier and a lightweight bytecode verifier, including proofs of
  3697 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
  3698 Cornelia Pusch (see also the homepage of project Bali at
  3699 http://isabelle.in.tum.de/Bali/);
  3700 
  3701 * HOL/Algebra: new theory of rings and univariate polynomials, by
  3702 Clemens Ballarin;
  3703 
  3704 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
  3705 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
  3706 Rasmussen;
  3707 
  3708 * HOL/Lattice: fundamental concepts of lattice theory and order
  3709 structures, including duals, properties of bounds versus algebraic
  3710 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
  3711 Theorem for complete lattices etc.; may also serve as a demonstration
  3712 for abstract algebraic reasoning using axiomatic type classes, and
  3713 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
  3714 
  3715 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
  3716 von Oheimb;
  3717 
  3718 * HOL/IMPP: extension of IMP with local variables and mutually
  3719 recursive procedures, by David von Oheimb;
  3720 
  3721 * HOL/Lambda: converted into new-style theory and document;
  3722 
  3723 * HOL/ex/Multiquote: example of multiple nested quotations and
  3724 anti-quotations -- basically a generalized version of de-Bruijn
  3725 representation; very useful in avoiding lifting of operations;
  3726 
  3727 * HOL/record: added general record equality rule to simpset; fixed
  3728 select-update simplification procedure to handle extended records as
  3729 well; admit "r" as field name;
  3730 
  3731 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
  3732 other numeric types and also as the identity of groups, rings, etc.;
  3733 
  3734 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
  3735 Types nat and int belong to this axclass;
  3736 
  3737 * HOL: greatly improved simplification involving numerals of type nat, int, real:
  3738    (i + #8 + j) = Suc k simplifies to  #7 + (i + j) = k
  3739    i*j + k + j*#3*i     simplifies to  #4*(i*j) + k
  3740   two terms #m*u and #n*u are replaced by #(m+n)*u
  3741     (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
  3742   and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
  3743     or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
  3744 
  3745 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
  3746 powerful prover for predicate logic but knows nothing of clasets; see
  3747 ex/mesontest.ML and ex/mesontest2.ML for example applications;
  3748 
  3749 * HOL: new version of "case_tac" subsumes both boolean case split and
  3750 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
  3751 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
  3752 
  3753 * HOL: simplification no longer dives into case-expressions: only the
  3754 selector expression is simplified, but not the remaining arms; to
  3755 enable full simplification of case-expressions for datatype t, you may
  3756 remove t.weak_case_cong from the simpset, either globally (Delcongs
  3757 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
  3758 
  3759 * HOL/recdef: the recursion equations generated by 'recdef' for
  3760 function 'f' are now called f.simps instead of f.rules; if all
  3761 termination conditions are proved automatically, these simplification
  3762 rules are added to the simpset, as in primrec; rules may be named
  3763 individually as well, resulting in a separate list of theorems for
  3764 each equation;
  3765 
  3766 * HOL/While is a new theory that provides a while-combinator. It
  3767 permits the definition of tail-recursive functions without the
  3768 provision of a termination measure. The latter is necessary once the
  3769 invariant proof rule for while is applied.
  3770 
  3771 * HOL: new (overloaded) notation for the set of elements below/above
  3772 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
  3773 
  3774 * HOL: theorems impI, allI, ballI bound as "strip";
  3775 
  3776 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
  3777 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
  3778 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
  3779 
  3780 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3781 
  3782 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
  3783 main HOL, but was unused);
  3784 
  3785 * HOL: fewer consts declared as global (e.g. have to refer to
  3786 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
  3787 
  3788 * HOL: tuned AST representation of nested pairs, avoiding bogus output
  3789 in case of overlap with user translations (e.g. judgements over
  3790 tuples); (note that the underlying logical represenation is still
  3791 bogus);
  3792 
  3793 
  3794 *** ZF ***
  3795 
  3796 * ZF: simplification automatically cancels common terms in arithmetic
  3797 expressions over nat and int;
  3798 
  3799 * ZF: new treatment of nat to minimize type-checking: all operators
  3800 coerce their operands to a natural number using the function natify,
  3801 making the algebraic laws unconditional;
  3802 
  3803 * ZF: as above, for int: operators coerce their operands to an integer
  3804 using the function intify;
  3805 
  3806 * ZF: the integer library now contains many of the usual laws for the
  3807 orderings, including $<=, and monotonicity laws for $+ and $*;
  3808 
  3809 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
  3810 simplification;
  3811 
  3812 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
  3813 to the simplifier and classical reasoner simultaneously;
  3814 
  3815 
  3816 *** General ***
  3817 
  3818 * Provers: blast_tac now handles actual object-logic rules as
  3819 assumptions; note that auto_tac uses blast_tac internally as well;
  3820 
  3821 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
  3822 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
  3823 
  3824 * Provers: delrules now handles destruct rules as well (no longer need
  3825 explicit make_elim);
  3826 
  3827 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
  3828   [| inj ?f;          ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3829 use instead the strong form,
  3830   [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3831 in HOL, FOL and ZF the function cla_make_elim will create such rules
  3832 from destruct-rules;
  3833 
  3834 * Provers: Simplifier.easy_setup provides a fast path to basic
  3835 Simplifier setup for new object-logics;
  3836 
  3837 * Pure: AST translation rules no longer require constant head on LHS;
  3838 
  3839 * Pure: improved name spaces: ambiguous output is qualified; support
  3840 for hiding of names;
  3841 
  3842 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
  3843 XSYMBOL_HOME; no longer need to do manual configuration in most
  3844 situations;
  3845 
  3846 * system: compression of ML heaps images may now be controlled via -c
  3847 option of isabelle and isatool usedir (currently only observed by
  3848 Poly/ML);
  3849 
  3850 * system: isatool installfonts may handle X-Symbol fonts as well (very
  3851 useful for remote X11);
  3852 
  3853 * system: provide TAGS file for Isabelle sources;
  3854 
  3855 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
  3856 order;
  3857 
  3858 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
  3859 timing flag supersedes proof_timing and Toplevel.trace;
  3860 
  3861 * ML: new combinators |>> and |>>> for incremental transformations
  3862 with secondary results (e.g. certain theory extensions):
  3863 
  3864 * ML: PureThy.add_defs gets additional argument to indicate potential
  3865 overloading (usually false);
  3866 
  3867 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
  3868 results;
  3869 
  3870 
  3871 
  3872 New in Isabelle99 (October 1999)
  3873 --------------------------------
  3874 
  3875 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3876 
  3877 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
  3878 are no longer simplified.  (This allows the simplifier to unfold recursive
  3879 functional programs.)  To restore the old behaviour, declare
  3880 
  3881     Delcongs [if_weak_cong];
  3882 
  3883 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
  3884 complement;
  3885 
  3886 * HOL: the predicate "inj" is now defined by translation to "inj_on";
  3887 
  3888 * HOL/datatype: mutual_induct_tac no longer exists --
  3889   use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
  3890 
  3891 * HOL/typedef: fixed type inference for representing set; type
  3892 arguments now have to occur explicitly on the rhs as type constraints;
  3893 
  3894 * ZF: The con_defs part of an inductive definition may no longer refer
  3895 to constants declared in the same theory;
  3896 
  3897 * HOL, ZF: the function mk_cases, generated by the inductive
  3898 definition package, has lost an argument.  To simplify its result, it
  3899 uses the default simpset instead of a supplied list of theorems.
  3900 
  3901 * HOL/List: the constructors of type list are now Nil and Cons;
  3902 
  3903 * Simplifier: the type of the infix ML functions
  3904         setSSolver addSSolver setSolver addSolver
  3905 is now  simpset * solver -> simpset  where `solver' is a new abstract type
  3906 for packaging solvers. A solver is created via
  3907         mk_solver: string -> (thm list -> int -> tactic) -> solver
  3908 where the string argument is only a comment.
  3909 
  3910 
  3911 *** Proof tools ***
  3912 
  3913 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
  3914 decision procedure for linear arithmetic. Currently it is used for
  3915 types `nat', `int', and `real' in HOL (see below); it can, should and
  3916 will be instantiated for other types and logics as well.
  3917 
  3918 * The simplifier now accepts rewrite rules with flexible heads, eg
  3919      hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
  3920   They are applied like any rule with a non-pattern lhs, i.e. by first-order
  3921   matching.
  3922 
  3923 
  3924 *** General ***
  3925 
  3926 * New Isabelle/Isar subsystem provides an alternative to traditional
  3927 tactical theorem proving; together with the ProofGeneral/isar user
  3928 interface it offers an interactive environment for developing human
  3929 readable proof documents (Isar == Intelligible semi-automated
  3930 reasoning); for further information see isatool doc isar-ref,
  3931 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
  3932 
  3933 * improved and simplified presentation of theories: better HTML markup
  3934 (including colors), graph views in several sizes; isatool usedir now
  3935 provides a proper interface for user theories (via -P option); actual
  3936 document preparation based on (PDF)LaTeX is available as well (for
  3937 new-style theories only); see isatool doc system for more information;
  3938 
  3939 * native support for Proof General, both for classic Isabelle and
  3940 Isabelle/Isar;
  3941 
  3942 * ML function thm_deps visualizes dependencies of theorems and lemmas,
  3943 using the graph browser tool;
  3944 
  3945 * Isabelle manuals now also available as PDF;
  3946 
  3947 * theory loader rewritten from scratch (may not be fully
  3948 bug-compatible); old loadpath variable has been replaced by show_path,
  3949 add_path, del_path, reset_path functions; new operations such as
  3950 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
  3951 isatool doc ref);
  3952 
  3953 * improved isatool install: option -k creates KDE application icon,
  3954 option -p DIR installs standalone binaries;
  3955 
  3956 * added ML_PLATFORM setting (useful for cross-platform installations);
  3957 more robust handling of platform specific ML images for SML/NJ;
  3958 
  3959 * the settings environment is now statically scoped, i.e. it is never
  3960 created again in sub-processes invoked from isabelle, isatool, or
  3961 Isabelle;
  3962 
  3963 * path element specification '~~' refers to '$ISABELLE_HOME';
  3964 
  3965 * in locales, the "assumes" and "defines" parts may be omitted if
  3966 empty;
  3967 
  3968 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
  3969 long arrows);
  3970 
  3971 * new print_mode "HTML";
  3972 
  3973 * new flag show_tags controls display of tags of theorems (which are
  3974 basically just comments that may be attached by some tools);
  3975 
  3976 * Isamode 2.6 requires patch to accomodate change of Isabelle font
  3977 mode and goal output format:
  3978 
  3979 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
  3980 244c244
  3981 <       (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
  3982 ---
  3983 >       (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
  3984 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
  3985 181c181
  3986 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
  3987 ---
  3988 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
  3989 
  3990 * function bind_thms stores lists of theorems (cf. bind_thm);
  3991 
  3992 * new shorthand tactics ftac, eatac, datac, fatac;
  3993 
  3994 * qed (and friends) now accept "" as result name; in that case the
  3995 theorem is not stored, but proper checks and presentation of the
  3996 result still apply;
  3997 
  3998 * theorem database now also indexes constants "Trueprop", "all",
  3999 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
  4000 
  4001 
  4002 *** HOL ***
  4003 
  4004 ** HOL arithmetic **
  4005 
  4006 * There are now decision procedures for linear arithmetic over nat and
  4007 int:
  4008 
  4009 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
  4010 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
  4011 are treated as atomic; subformulae not involving type `nat' or `int'
  4012 are ignored; quantified subformulae are ignored unless they are
  4013 positive universal or negative existential. The tactic has to be
  4014 invoked by hand and can be a little bit slow. In particular, the
  4015 running time is exponential in the number of occurrences of `min' and
  4016 `max', and `-' on `nat'.
  4017 
  4018 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
  4019 (negated) (in)equalities among the premises and the conclusion into
  4020 account (i.e. no compound formulae) and does not know about `min' and
  4021 `max', and `-' on `nat'. It is fast and is used automatically by the
  4022 simplifier.
  4023 
  4024 NB: At the moment, these decision procedures do not cope with mixed
  4025 nat/int formulae where the two parts interact, such as `m < n ==>
  4026 int(m) < int(n)'.
  4027 
  4028 * HOL/Numeral provides a generic theory of numerals (encoded
  4029 efficiently as bit strings); setup for types nat/int/real is in place;
  4030 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
  4031 int, existing theories and proof scripts may require a few additional
  4032 type constraints;
  4033 
  4034 * integer division and remainder can now be performed on constant
  4035 arguments;
  4036 
  4037 * many properties of integer multiplication, division and remainder
  4038 are now available;
  4039 
  4040 * An interface to the Stanford Validity Checker (SVC) is available through the
  4041 tactic svc_tac.  Propositional tautologies and theorems of linear arithmetic
  4042 are proved automatically.  SVC must be installed separately, and its results
  4043 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
  4044 invocation of the underlying oracle).  For SVC see
  4045   http://verify.stanford.edu/SVC
  4046 
  4047 * IsaMakefile: the HOL-Real target now builds an actual image;
  4048 
  4049 
  4050 ** HOL misc **
  4051 
  4052 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
  4053 (in Isabelle/Isar) -- by Gertrud Bauer;
  4054 
  4055 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
  4056 analysis for assembly languages with subtypes;
  4057 
  4058 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
  4059 -- avoids syntactic ambiguities and treats state, transition, and
  4060 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
  4061 changed syntax and (many) tactics;
  4062 
  4063 * HOL/inductive: Now also handles more general introduction rules such
  4064   as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
  4065   theorems are now maintained within the theory (maintained via the
  4066   "mono" attribute);
  4067 
  4068 * HOL/datatype: Now also handles arbitrarily branching datatypes
  4069   (using function types) such as
  4070 
  4071   datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
  4072 
  4073 * HOL/record: record_simproc (part of the default simpset) takes care
  4074 of selectors applied to updated records; record_split_tac is no longer
  4075 part of the default claset; update_defs may now be removed from the
  4076 simpset in many cases; COMPATIBILITY: old behavior achieved by
  4077 
  4078   claset_ref () := claset() addSWrapper record_split_wrapper;
  4079   Delsimprocs [record_simproc]
  4080 
  4081 * HOL/typedef: fixed type inference for representing set; type
  4082 arguments now have to occur explicitly on the rhs as type constraints;
  4083 
  4084 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
  4085 names rather than an ML expression;
  4086 
  4087 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
  4088 supplied later.  Program schemes can be defined, such as
  4089     "While B C s = (if B s then While B C (C s) else s)"
  4090 where the well-founded relation can be chosen after B and C have been given.
  4091 
  4092 * HOL/List: the constructors of type list are now Nil and Cons;
  4093 INCOMPATIBILITY: while [] and infix # syntax is still there, of
  4094 course, ML tools referring to List.list.op # etc. have to be adapted;
  4095 
  4096 * HOL_quantifiers flag superseded by "HOL" print mode, which is
  4097 disabled by default; run isabelle with option -m HOL to get back to
  4098 the original Gordon/HOL-style output;
  4099 
  4100 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
  4101 ALL x<=y. P, EX x<y. P, EX x<=y. P;
  4102 
  4103 * HOL basic syntax simplified (more orthogonal): all variants of
  4104 All/Ex now support plain / symbolic / HOL notation; plain syntax for
  4105 Eps operator is provided as well: "SOME x. P[x]";
  4106 
  4107 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
  4108 
  4109 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
  4110 thus available for user theories;
  4111 
  4112 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
  4113 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
  4114 time;
  4115 
  4116 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
  4117 several times and then mp;
  4118 
  4119 
  4120 *** LK ***
  4121 
  4122 * the notation <<...>> is now available as a notation for sequences of
  4123 formulas;
  4124 
  4125 * the simplifier is now installed
  4126 
  4127 * the axiom system has been generalized (thanks to Soren Heilmann)
  4128 
  4129 * the classical reasoner now has a default rule database
  4130 
  4131 
  4132 *** ZF ***
  4133 
  4134 * new primrec section allows primitive recursive functions to be given
  4135 directly (as in HOL) over datatypes and the natural numbers;
  4136 
  4137 * new tactics induct_tac and exhaust_tac for induction (or case
  4138 analysis) over datatypes and the natural numbers;
  4139 
  4140 * the datatype declaration of type T now defines the recursor T_rec;
  4141 
  4142 * simplification automatically does freeness reasoning for datatype
  4143 constructors;
  4144 
  4145 * automatic type-inference, with AddTCs command to insert new
  4146 type-checking rules;
  4147 
  4148 * datatype introduction rules are now added as Safe Introduction rules
  4149 to the claset;
  4150 
  4151 * the syntax "if P then x else y" is now available in addition to
  4152 if(P,x,y);
  4153 
  4154 
  4155 *** Internal programming interfaces ***
  4156 
  4157 * tuned simplifier trace output; new flag debug_simp;
  4158 
  4159 * structures Vartab / Termtab (instances of TableFun) offer efficient
  4160 tables indexed by indexname_ord / term_ord (compatible with aconv);
  4161 
  4162 * AxClass.axclass_tac lost the theory argument;
  4163 
  4164 * tuned current_goals_markers semantics: begin / end goal avoids
  4165 printing empty lines;
  4166 
  4167 * removed prs and prs_fn hook, which was broken because it did not
  4168 include \n in its semantics, forcing writeln to add one
  4169 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
  4170 string -> unit if you really want to output text without newline;
  4171 
  4172 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
  4173 plain output, interface builders may have to enable 'isabelle_font'
  4174 mode to get Isabelle font glyphs as before;
  4175 
  4176 * refined token_translation interface; INCOMPATIBILITY: output length
  4177 now of type real instead of int;
  4178 
  4179 * theory loader actions may be traced via new ThyInfo.add_hook
  4180 interface (see src/Pure/Thy/thy_info.ML); example application: keep
  4181 your own database of information attached to *whole* theories -- as
  4182 opposed to intra-theory data slots offered via TheoryDataFun;
  4183 
  4184 * proper handling of dangling sort hypotheses (at last!);
  4185 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
  4186 extra sort hypotheses that can be witnessed from the type signature;
  4187 the force_strip_shyps flag is gone, any remaining shyps are simply
  4188 left in the theorem (with a warning issued by strip_shyps_warning);
  4189 
  4190 
  4191 
  4192 New in Isabelle98-1 (October 1998)
  4193 ----------------------------------
  4194 
  4195 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  4196 
  4197 * several changes of automated proof tools;
  4198 
  4199 * HOL: major changes to the inductive and datatype packages, including
  4200 some minor incompatibilities of theory syntax;
  4201 
  4202 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
  4203 called `inj_on';
  4204 
  4205 * HOL: removed duplicate thms in Arith:
  4206   less_imp_add_less  should be replaced by  trans_less_add1
  4207   le_imp_add_le      should be replaced by  trans_le_add1
  4208 
  4209 * HOL: unary minus is now overloaded (new type constraints may be
  4210 required);
  4211 
  4212 * HOL and ZF: unary minus for integers is now #- instead of #~.  In
  4213 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
  4214 now taken as an integer constant.
  4215 
  4216 * Pure: ML function 'theory_of' renamed to 'theory';
  4217 
  4218 
  4219 *** Proof tools ***
  4220 
  4221 * Simplifier:
  4222   1. Asm_full_simp_tac is now more aggressive.
  4223      1. It will sometimes reorient premises if that increases their power to
  4224         simplify.
  4225      2. It does no longer proceed strictly from left to right but may also
  4226         rotate premises to achieve further simplification.
  4227      For compatibility reasons there is now Asm_lr_simp_tac which is like the
  4228      old Asm_full_simp_tac in that it does not rotate premises.
  4229   2. The simplifier now knows a little bit about nat-arithmetic.
  4230 
  4231 * Classical reasoner: wrapper mechanism for the classical reasoner now
  4232 allows for selected deletion of wrappers, by introduction of names for
  4233 wrapper functionals.  This implies that addbefore, addSbefore,
  4234 addaltern, and addSaltern now take a pair (name, tactic) as argument,
  4235 and that adding two tactics with the same name overwrites the first
  4236 one (emitting a warning).
  4237   type wrapper = (int -> tactic) -> (int -> tactic)
  4238   setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
  4239   addWrapper, addSWrapper: claset * (string * wrapper) -> claset
  4240   delWrapper, delSWrapper: claset *  string            -> claset
  4241   getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
  4242 
  4243 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
  4244 semantics; addbefore now affects only the unsafe part of step_tac
  4245 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
  4246 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
  4247 by Force_tac;
  4248 
  4249 * Classical reasoner: setwrapper to setWrapper and compwrapper to
  4250 compWrapper; added safe wrapper (and access functions for it);
  4251 
  4252 * HOL/split_all_tac is now much faster and fails if there is nothing
  4253 to split.  Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
  4254 and the names of the automatically generated variables have changed.
  4255 split_all_tac has moved within claset() from unsafe wrappers to safe
  4256 wrappers, which means that !!-bound variables are split much more
  4257 aggressively, and safe_tac and clarify_tac now split such variables.
  4258 If this splitting is not appropriate, use delSWrapper "split_all_tac".
  4259 Note: the same holds for record_split_tac, which does the job of
  4260 split_all_tac for record fields.
  4261 
  4262 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
  4263 permanently to the default simpset using Addsplits just like
  4264 Addsimps. They can be removed via Delsplits just like
  4265 Delsimps. Lower-case versions are also available.
  4266 
  4267 * HOL/Simplifier: The rule split_if is now part of the default
  4268 simpset. This means that the simplifier will eliminate all occurrences
  4269 of if-then-else in the conclusion of a goal. To prevent this, you can
  4270 either remove split_if completely from the default simpset by
  4271 `Delsplits [split_if]' or remove it in a specific call of the
  4272 simplifier using `... delsplits [split_if]'.  You can also add/delete
  4273 other case splitting rules to/from the default simpset: every datatype
  4274 generates suitable rules `split_t_case' and `split_t_case_asm' (where
  4275 t is the name of the datatype).
  4276 
  4277 * Classical reasoner / Simplifier combination: new force_tac (and
  4278 derivatives Force_tac, force) combines rewriting and classical
  4279 reasoning (and whatever other tools) similarly to auto_tac, but is
  4280 aimed to solve the given subgoal completely.
  4281 
  4282 
  4283 *** General ***
  4284 
  4285 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
  4286 and `goalw': the theory is no longer needed as an explicit argument -
  4287 the current theory context is used; assumptions are no longer returned
  4288 at the ML-level unless one of them starts with ==> or !!; it is
  4289 recommended to convert to these new commands using isatool fixgoal
  4290 (backup your sources first!);
  4291 
  4292 * new top-level commands 'thm' and 'thms' for retrieving theorems from
  4293 the current theory context, and 'theory' to lookup stored theories;
  4294 
  4295 * new theory section 'locale' for declaring constants, assumptions and
  4296 definitions that have local scope;
  4297 
  4298 * new theory section 'nonterminals' for purely syntactic types;
  4299 
  4300 * new theory section 'setup' for generic ML setup functions
  4301 (e.g. package initialization);
  4302 
  4303 * the distribution now includes Isabelle icons: see
  4304 lib/logo/isabelle-{small,tiny}.xpm;
  4305 
  4306 * isatool install - install binaries with absolute references to
  4307 ISABELLE_HOME/bin;
  4308 
  4309 * isatool logo -- create instances of the Isabelle logo (as EPS);
  4310 
  4311 * print mode 'emacs' reserved for Isamode;
  4312 
  4313 * support multiple print (ast) translations per constant name;
  4314 
  4315 * theorems involving oracles are now printed with a suffixed [!];
  4316 
  4317 
  4318 *** HOL ***
  4319 
  4320 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
  4321 
  4322 * HOL/inductive package reorganized and improved: now supports mutual
  4323 definitions such as
  4324 
  4325   inductive EVEN ODD
  4326     intrs
  4327       null "0 : EVEN"
  4328       oddI "n : EVEN ==> Suc n : ODD"
  4329       evenI "n : ODD ==> Suc n : EVEN"
  4330 
  4331 new theorem list "elims" contains an elimination rule for each of the
  4332 recursive sets; inductive definitions now handle disjunctive premises
  4333 correctly (also ZF);
  4334 
  4335 INCOMPATIBILITIES: requires Inductive as an ancestor; component
  4336 "mutual_induct" no longer exists - the induction rule is always
  4337 contained in "induct";
  4338 
  4339 
  4340 * HOL/datatype package re-implemented and greatly improved: now
  4341 supports mutually recursive datatypes such as
  4342 
  4343   datatype
  4344     'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
  4345             | SUM ('a aexp) ('a aexp)
  4346             | DIFF ('a aexp) ('a aexp)
  4347             | NUM 'a
  4348   and
  4349     'a bexp = LESS ('a aexp) ('a aexp)
  4350             | AND ('a bexp) ('a bexp)
  4351             | OR ('a bexp) ('a bexp)
  4352 
  4353 as well as indirectly recursive datatypes such as
  4354 
  4355   datatype
  4356     ('a, 'b) term = Var 'a
  4357                   | App 'b ((('a, 'b) term) list)
  4358 
  4359 The new tactic  mutual_induct_tac [<var_1>, ..., <var_n>] i  performs
  4360 induction on mutually / indirectly recursive datatypes.
  4361 
  4362 Primrec equations are now stored in theory and can be accessed via
  4363 <function_name>.simps.
  4364 
  4365 INCOMPATIBILITIES:
  4366 
  4367   - Theories using datatypes must now have theory Datatype as an
  4368     ancestor.
  4369   - The specific <typename>.induct_tac no longer exists - use the
  4370     generic induct_tac instead.
  4371   - natE has been renamed to nat.exhaust - use exhaust_tac
  4372     instead of res_inst_tac ... natE. Note that the variable
  4373     names in nat.exhaust differ from the names in natE, this
  4374     may cause some "fragile" proofs to fail.
  4375   - The theorems split_<typename>_case and split_<typename>_case_asm
  4376     have been renamed to <typename>.split and <typename>.split_asm.
  4377   - Since default sorts of type variables are now handled correctly,
  4378     some datatype definitions may have to be annotated with explicit
  4379     sort constraints.
  4380   - Primrec definitions no longer require function name and type
  4381     of recursive argument.
  4382 
  4383 Consider using isatool fixdatatype to adapt your theories and proof
  4384 scripts to the new package (backup your sources first!).
  4385 
  4386 
  4387 * HOL/record package: considerably improved implementation; now
  4388 includes concrete syntax for record types, terms, updates; theorems
  4389 for surjective pairing and splitting !!-bound record variables; proof
  4390 support is as follows:
  4391 
  4392   1) standard conversions (selectors or updates applied to record
  4393 constructor terms) are part of the standard simpset;
  4394 
  4395   2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
  4396 made part of standard simpset and claset via addIffs;
  4397 
  4398   3) a tactic for record field splitting (record_split_tac) is part of
  4399 the standard claset (addSWrapper);
  4400 
  4401 To get a better idea about these rules you may retrieve them via
  4402 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
  4403 the name of your record type.
  4404 
  4405 The split tactic 3) conceptually simplifies by the following rule:
  4406 
  4407   "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
  4408 
  4409 Thus any record variable that is bound by meta-all will automatically
  4410 blow up into some record constructor term, consequently the
  4411 simplifications of 1), 2) apply.  Thus force_tac, auto_tac etc. shall
  4412 solve record problems automatically.
  4413 
  4414 
  4415 * reorganized the main HOL image: HOL/Integ and String loaded by
  4416 default; theory Main includes everything;
  4417 
  4418 * automatic simplification of integer sums and comparisons, using cancellation;
  4419 
  4420 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
  4421 
  4422 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
  4423 
  4424 * many new identities for unions, intersections, set difference, etc.;
  4425 
  4426 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
  4427 called split_if, split_split, split_sum_case and split_nat_case (to go
  4428 with add/delsplits);
  4429 
  4430 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
  4431 (?x::unit) = (); this is made part of the default simpset, which COULD
  4432 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
  4433 'Delsimprocs [unit_eq_proc];' as last resort); also note that
  4434 unit_abs_eta_conv is added in order to counter the effect of
  4435 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
  4436 %u.f();
  4437 
  4438 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
  4439 makes more sense);
  4440 
  4441 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  4442   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  4443   disjointness reasoning but breaking a few old proofs.
  4444 
  4445 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
  4446 to 'converse' from 'inverse' (for compatibility with ZF and some
  4447 literature);
  4448 
  4449 * HOL/recdef can now declare non-recursive functions, with {} supplied as
  4450 the well-founded relation;
  4451 
  4452 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
  4453     Compl A.  The "Compl" syntax remains available as input syntax for this
  4454     release ONLY.
  4455 
  4456 * HOL/Update: new theory of function updates:
  4457     f(a:=b) == %x. if x=a then b else f x
  4458 may also be iterated as in f(a:=b,c:=d,...);
  4459 
  4460 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
  4461 
  4462 * HOL/List:
  4463   - new function list_update written xs[i:=v] that updates the i-th
  4464     list position. May also be iterated as in xs[i:=a,j:=b,...].
  4465   - new function `upt' written [i..j(] which generates the list
  4466     [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
  4467     bound write [i..j], which is a shorthand for [i..j+1(].
  4468   - new lexicographic orderings and corresponding wellfoundedness theorems.
  4469 
  4470 * HOL/Arith:
  4471   - removed 'pred' (predecessor) function;
  4472   - generalized some theorems about n-1;
  4473   - many new laws about "div" and "mod";
  4474   - new laws about greatest common divisors (see theory ex/Primes);
  4475 
  4476 * HOL/Relation: renamed the relational operator r^-1 "converse"
  4477 instead of "inverse";
  4478 
  4479 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
  4480   of the multiset ordering;
  4481 
  4482 * directory HOL/Real: a construction of the reals using Dedekind cuts
  4483   (not included by default);
  4484 
  4485 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
  4486 
  4487 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
  4488   programs, i.e. different program variables may have different types.
  4489 
  4490 * calling (stac rew i) now fails if "rew" has no effect on the goal
  4491   [previously, this check worked only if the rewrite rule was unconditional]
  4492   Now rew can involve either definitions or equalities (either == or =).
  4493 
  4494 
  4495 *** ZF ***
  4496 
  4497 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
  4498   only the theorems proved on ZF.ML;
  4499 
  4500 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  4501   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  4502   disjointness reasoning but breaking a few old proofs.
  4503 
  4504 * ZF/Update: new theory of function updates
  4505     with default rewrite rule  f(x:=y) ` z = if(z=x, y, f`z)
  4506   may also be iterated as in f(a:=b,c:=d,...);
  4507 
  4508 * in  let x=t in u(x), neither t nor u(x) has to be an FOL term.
  4509 
  4510 * calling (stac rew i) now fails if "rew" has no effect on the goal
  4511   [previously, this check worked only if the rewrite rule was unconditional]
  4512   Now rew can involve either definitions or equalities (either == or =).
  4513 
  4514 * case_tac provided for compatibility with HOL
  4515     (like the old excluded_middle_tac, but with subgoals swapped)
  4516 
  4517 
  4518 *** Internal programming interfaces ***
  4519 
  4520 * Pure: several new basic modules made available for general use, see
  4521 also src/Pure/README;
  4522 
  4523 * improved the theory data mechanism to support encapsulation (data
  4524 kind name replaced by private Object.kind, acting as authorization
  4525 key); new type-safe user interface via functor TheoryDataFun; generic
  4526 print_data function becomes basically useless;
  4527 
  4528 * removed global_names compatibility flag -- all theory declarations
  4529 are qualified by default;
  4530 
  4531 * module Pure/Syntax now offers quote / antiquote translation
  4532 functions (useful for Hoare logic etc. with implicit dependencies);
  4533 see HOL/ex/Antiquote for an example use;
  4534 
  4535 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
  4536 cterm -> thm;
  4537 
  4538 * new tactical CHANGED_GOAL for checking that a tactic modifies a
  4539 subgoal;
  4540 
  4541 * Display.print_goals function moved to Locale.print_goals;
  4542 
  4543 * standard print function for goals supports current_goals_markers
  4544 variable for marking begin of proof, end of proof, start of goal; the
  4545 default is ("", "", ""); setting current_goals_markers := ("<proof>",
  4546 "</proof>", "<goal>") causes SGML like tagged proof state printing,
  4547 for example;
  4548 
  4549 
  4550 
  4551 New in Isabelle98 (January 1998)
  4552 --------------------------------
  4553 
  4554 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  4555 
  4556 * changed lexical syntax of terms / types: dots made part of long
  4557 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
  4558 
  4559 * simpset (and claset) reference variable replaced by functions
  4560 simpset / simpset_ref;
  4561 
  4562 * no longer supports theory aliases (via merge) and non-trivial
  4563 implicit merge of thms' signatures;
  4564 
  4565 * most internal names of constants changed due to qualified names;
  4566 
  4567 * changed Pure/Sequence interface (see Pure/seq.ML);
  4568 
  4569 
  4570 *** General Changes ***
  4571 
  4572 * hierachically structured name spaces (for consts, types, axms, thms
  4573 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
  4574 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
  4575 isatool fixdots ensures space after dots (e.g. "%x. x"); set
  4576 long_names for fully qualified output names; NOTE: ML programs
  4577 (special tactics, packages etc.) referring to internal names may have
  4578 to be adapted to cope with fully qualified names; in case of severe
  4579 backward campatibility problems try setting 'global_names' at compile
  4580 time to have enrything declared within a flat name space; one may also
  4581 fine tune name declarations in theories via the 'global' and 'local'
  4582 section;
  4583 
  4584 * reimplemented the implicit simpset and claset using the new anytype
  4585 data filed in signatures; references simpset:simpset ref etc. are
  4586 replaced by functions simpset:unit->simpset and
  4587 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
  4588 to patch your ML files accordingly;
  4589 
  4590 * HTML output now includes theory graph data for display with Java
  4591 applet or isatool browser; data generated automatically via isatool
  4592 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
  4593 
  4594 * defs may now be conditional; improved rewrite_goals_tac to handle
  4595 conditional equations;
  4596 
  4597 * defs now admits additional type arguments, using TYPE('a) syntax;
  4598 
  4599 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
  4600 creates a new theory node; implicit merge of thms' signatures is
  4601 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
  4602 transfer:theory->thm->thm in (rare) cases;
  4603 
  4604 * improved handling of draft signatures / theories; draft thms (and
  4605 ctyps, cterms) are automatically promoted to real ones;
  4606 
  4607 * slightly changed interfaces for oracles: admit many per theory, named
  4608 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
  4609 
  4610 * print_goals: optional output of const types (set show_consts and
  4611 show_types);
  4612 
  4613 * improved output of warnings (###) and errors (***);
  4614 
  4615 * subgoal_tac displays a warning if the new subgoal has type variables;
  4616 
  4617 * removed old README and Makefiles;
  4618 
  4619 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
  4620 
  4621 * removed obsolete init_pps and init_database;
  4622 
  4623 * deleted the obsolete tactical STATE, which was declared by
  4624     fun STATE tacfun st = tacfun st st;
  4625 
  4626 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
  4627 (which abbreviates $HOME);
  4628 
  4629 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
  4630 use isatool fixseq to adapt your ML programs (this works for fully
  4631 qualified references to the Sequence structure only!);
  4632 
  4633 * use_thy no longer requires writable current directory; it always
  4634 reloads .ML *and* .thy file, if either one is out of date;
  4635 
  4636 
  4637 *** Classical Reasoner ***
  4638 
  4639 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
  4640 tactics that use classical reasoning to simplify a subgoal without
  4641 splitting it into several subgoals;
  4642 
  4643 * Safe_tac: like safe_tac but uses the default claset;
  4644 
  4645 
  4646 *** Simplifier ***
  4647 
  4648 * added simplification meta rules:
  4649     (asm_)(full_)simplify: simpset -> thm -> thm;
  4650 
  4651 * simplifier.ML no longer part of Pure -- has to be loaded by object
  4652 logics (again);
  4653 
  4654 * added prems argument to simplification procedures;
  4655 
  4656 * HOL, FOL, ZF: added infix function `addsplits':
  4657   instead of `<simpset> setloop (split_tac <thms>)'
  4658   you can simply write `<simpset> addsplits <thms>'
  4659 
  4660 
  4661 *** Syntax ***
  4662 
  4663 * TYPE('a) syntax for type reflection terms;
  4664 
  4665 * no longer handles consts with name "" -- declare as 'syntax' instead;
  4666 
  4667 * pretty printer: changed order of mixfix annotation preference (again!);
  4668 
  4669 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
  4670 
  4671 
  4672 *** HOL ***
  4673 
  4674 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4675   with `addloop' of the simplifier to faciliate case splitting in premises.
  4676 
  4677 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
  4678 
  4679 * HOL/Auth: new protocol proofs including some for the Internet
  4680   protocol TLS;
  4681 
  4682 * HOL/Map: new theory of `maps' a la VDM;
  4683 
  4684 * HOL/simplifier: simplification procedures nat_cancel_sums for
  4685 cancelling out common nat summands from =, <, <= (in)equalities, or
  4686 differences; simplification procedures nat_cancel_factor for
  4687 cancelling common factor from =, <, <= (in)equalities over natural
  4688 sums; nat_cancel contains both kinds of procedures, it is installed by
  4689 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
  4690 
  4691 * HOL/simplifier: terms of the form
  4692   `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)'  (or t=x)
  4693   are rewritten to
  4694   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
  4695   and those of the form
  4696   `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)'  (or t=x)
  4697   are rewritten to
  4698   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
  4699 
  4700 * HOL/datatype
  4701   Each datatype `t' now comes with a theorem `split_t_case' of the form
  4702 
  4703   P(t_case f1 ... fn x) =
  4704      ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
  4705         ...
  4706        (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
  4707      )
  4708 
  4709   and a theorem `split_t_case_asm' of the form
  4710 
  4711   P(t_case f1 ... fn x) =
  4712     ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
  4713         ...
  4714        (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
  4715      )
  4716   which can be added to a simpset via `addsplits'. The existing theorems
  4717   expand_list_case and expand_option_case have been renamed to
  4718   split_list_case and split_option_case.
  4719 
  4720 * HOL/Arithmetic:
  4721   - `pred n' is automatically converted to `n-1'.
  4722     Users are strongly encouraged not to use `pred' any longer,
  4723     because it will disappear altogether at some point.
  4724   - Users are strongly encouraged to write "0 < n" rather than
  4725     "n ~= 0". Theorems and proof tools have been modified towards this
  4726     `standard'.
  4727 
  4728 * HOL/Lists:
  4729   the function "set_of_list" has been renamed "set" (and its theorems too);
  4730   the function "nth" now takes its arguments in the reverse order and
  4731   has acquired the infix notation "!" as in "xs!n".
  4732 
  4733 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
  4734 
  4735 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
  4736   specialist theorems (like UN1_I) are gone.  Similarly for (INT x.B x);
  4737 
  4738 * HOL/record: extensible records with schematic structural subtyping
  4739 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
  4740 still lacks various theorems and concrete record syntax;
  4741 
  4742 
  4743 *** HOLCF ***
  4744 
  4745 * removed "axioms" and "generated by" sections;
  4746 
  4747 * replaced "ops" section by extended "consts" section, which is capable of
  4748   handling the continuous function space "->" directly;
  4749 
  4750 * domain package:
  4751   . proves theorems immediately and stores them in the theory,
  4752   . creates hierachical name space,
  4753   . now uses normal mixfix annotations (instead of cinfix...),
  4754   . minor changes to some names and values (for consistency),
  4755   . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
  4756   . separator between mutual domain defs: changed "," to "and",
  4757   . improved handling of sort constraints;  now they have to
  4758     appear on the left-hand side of the equations only;
  4759 
  4760 * fixed LAM <x,y,zs>.b syntax;
  4761 
  4762 * added extended adm_tac to simplifier in HOLCF -- can now discharge
  4763 adm (%x. P (t x)), where P is chainfinite and t continuous;
  4764 
  4765 
  4766 *** FOL and ZF ***
  4767 
  4768 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4769   with `addloop' of the simplifier to faciliate case splitting in premises.
  4770 
  4771 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
  4772 in HOL, they strip ALL and --> from proved theorems;
  4773 
  4774 
  4775 
  4776 New in Isabelle94-8 (May 1997)
  4777 ------------------------------
  4778 
  4779 *** General Changes ***
  4780 
  4781 * new utilities to build / run / maintain Isabelle etc. (in parts
  4782 still somewhat experimental); old Makefiles etc. still functional;
  4783 
  4784 * new 'Isabelle System Manual';
  4785 
  4786 * INSTALL text, together with ./configure and ./build scripts;
  4787 
  4788 * reimplemented type inference for greater efficiency, better error
  4789 messages and clean internal interface;
  4790 
  4791 * prlim command for dealing with lots of subgoals (an easier way of
  4792 setting goals_limit);
  4793 
  4794 
  4795 *** Syntax ***
  4796 
  4797 * supports alternative (named) syntax tables (parser and pretty
  4798 printer); internal interface is provided by add_modesyntax(_i);
  4799 
  4800 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
  4801 be used in conjunction with the Isabelle symbol font; uses the
  4802 "symbols" syntax table;
  4803 
  4804 * added token_translation interface (may translate name tokens in
  4805 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
  4806 the current print_mode); IMPORTANT: user print translation functions
  4807 are responsible for marking newly introduced bounds
  4808 (Syntax.mark_boundT);
  4809 
  4810 * token translations for modes "xterm" and "xterm_color" that display
  4811 names in bold, underline etc. or colors (which requires a color
  4812 version of xterm);
  4813 
  4814 * infixes may now be declared with names independent of their syntax;
  4815 
  4816 * added typed_print_translation (like print_translation, but may
  4817 access type of constant);
  4818 
  4819 
  4820 *** Classical Reasoner ***
  4821 
  4822 Blast_tac: a new tactic!  It is often more powerful than fast_tac, but has
  4823 some limitations.  Blast_tac...
  4824   + ignores addss, addbefore, addafter; this restriction is intrinsic
  4825   + ignores elimination rules that don't have the correct format
  4826         (the conclusion MUST be a formula variable)
  4827   + ignores types, which can make HOL proofs fail
  4828   + rules must not require higher-order unification, e.g. apply_type in ZF
  4829     [message "Function Var's argument not a bound variable" relates to this]
  4830   + its proof strategy is more general but can actually be slower
  4831 
  4832 * substitution with equality assumptions no longer permutes other
  4833 assumptions;
  4834 
  4835 * minor changes in semantics of addafter (now called addaltern); renamed
  4836 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
  4837 (and access functions for it);
  4838 
  4839 * improved combination of classical reasoner and simplifier:
  4840   + functions for handling clasimpsets
  4841   + improvement of addss: now the simplifier is called _after_ the
  4842     safe steps.
  4843   + safe variant of addss called addSss: uses safe simplifications
  4844     _during_ the safe steps. It is more complete as it allows multiple
  4845     instantiations of unknowns (e.g. with slow_tac).
  4846 
  4847 *** Simplifier ***
  4848 
  4849 * added interface for simplification procedures (functions that
  4850 produce *proven* rewrite rules on the fly, depending on current
  4851 redex);
  4852 
  4853 * ordering on terms as parameter (used for ordered rewriting);
  4854 
  4855 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
  4856 
  4857 * the solver is now split into a safe and an unsafe part.
  4858 This should be invisible for the normal user, except that the
  4859 functions setsolver and addsolver have been renamed to setSolver and
  4860 addSolver; added safe_asm_full_simp_tac;
  4861 
  4862 
  4863 *** HOL ***
  4864 
  4865 * a generic induction tactic `induct_tac' which works for all datatypes and
  4866 also for type `nat';
  4867 
  4868 * a generic case distinction tactic `exhaust_tac' which works for all
  4869 datatypes and also for type `nat';
  4870 
  4871 * each datatype comes with a function `size';
  4872 
  4873 * patterns in case expressions allow tuple patterns as arguments to
  4874 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
  4875 
  4876 * primrec now also works with type nat;
  4877 
  4878 * recdef: a new declaration form, allows general recursive functions to be
  4879 defined in theory files.  See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
  4880 
  4881 * the constant for negation has been renamed from "not" to "Not" to
  4882 harmonize with FOL, ZF, LK, etc.;
  4883 
  4884 * HOL/ex/LFilter theory of a corecursive "filter" functional for
  4885 infinite lists;
  4886 
  4887 * HOL/Modelcheck demonstrates invocation of model checker oracle;
  4888 
  4889 * HOL/ex/Ring.thy declares cring_simp, which solves equational
  4890 problems in commutative rings, using axiomatic type classes for + and *;
  4891 
  4892 * more examples in HOL/MiniML and HOL/Auth;
  4893 
  4894 * more default rewrite rules for quantifiers, union/intersection;
  4895 
  4896 * a new constant `arbitrary == @x.False';
  4897 
  4898 * HOLCF/IOA replaces old HOL/IOA;
  4899 
  4900 * HOLCF changes: derived all rules and arities
  4901   + axiomatic type classes instead of classes
  4902   + typedef instead of faking type definitions
  4903   + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
  4904   + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
  4905   + eliminated the types void, one, tr
  4906   + use unit lift and bool lift (with translations) instead of one and tr
  4907   + eliminated blift from Lift3.thy (use Def instead of blift)
  4908   all eliminated rules are derived as theorems --> no visible changes ;
  4909 
  4910 
  4911 *** ZF ***
  4912 
  4913 * ZF now has Fast_tac, Simp_tac and Auto_tac.  Union_iff is a now a default
  4914 rewrite rule; this may affect some proofs.  eq_cs is gone but can be put back
  4915 as ZF_cs addSIs [equalityI];
  4916 
  4917 
  4918 
  4919 New in Isabelle94-7 (November 96)
  4920 ---------------------------------
  4921 
  4922 * allowing negative levels (as offsets) in prlev and choplev;
  4923 
  4924 * super-linear speedup for large simplifications;
  4925 
  4926 * FOL, ZF and HOL now use miniscoping: rewriting pushes
  4927 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
  4928 FAIL); can suppress it using the command Delsimps (ex_simps @
  4929 all_simps); De Morgan laws are also now included, by default;
  4930 
  4931 * improved printing of ==>  :  ~:
  4932 
  4933 * new object-logic "Sequents" adds linear logic, while replacing LK
  4934 and Modal (thanks to Sara Kalvala);
  4935 
  4936 * HOL/Auth: correctness proofs for authentication protocols;
  4937 
  4938 * HOL: new auto_tac combines rewriting and classical reasoning (many
  4939 examples on HOL/Auth);
  4940 
  4941 * HOL: new command AddIffs for declaring theorems of the form P=Q to
  4942 the rewriter and classical reasoner simultaneously;
  4943 
  4944 * function uresult no longer returns theorems in "standard" format;
  4945 regain previous version by: val uresult = standard o uresult;
  4946 
  4947 
  4948 
  4949 New in Isabelle94-6
  4950 -------------------
  4951 
  4952 * oracles -- these establish an interface between Isabelle and trusted
  4953 external reasoners, which may deliver results as theorems;
  4954 
  4955 * proof objects (in particular record all uses of oracles);
  4956 
  4957 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
  4958 
  4959 * "constdefs" section in theory files;
  4960 
  4961 * "primrec" section (HOL) no longer requires names;
  4962 
  4963 * internal type "tactic" now simply "thm -> thm Sequence.seq";
  4964 
  4965 
  4966 
  4967 New in Isabelle94-5
  4968 -------------------
  4969 
  4970 * reduced space requirements;
  4971 
  4972 * automatic HTML generation from theories;
  4973 
  4974 * theory files no longer require "..." (quotes) around most types;
  4975 
  4976 * new examples, including two proofs of the Church-Rosser theorem;
  4977 
  4978 * non-curried (1994) version of HOL is no longer distributed;
  4979 
  4980 
  4981 
  4982 New in Isabelle94-4
  4983 -------------------
  4984 
  4985 * greatly reduced space requirements;
  4986 
  4987 * theory files (.thy) no longer require \...\ escapes at line breaks;
  4988 
  4989 * searchable theorem database (see the section "Retrieving theorems" on
  4990 page 8 of the Reference Manual);
  4991 
  4992 * new examples, including Grabczewski's monumental case study of the
  4993 Axiom of Choice;
  4994 
  4995 * The previous version of HOL renamed to Old_HOL;
  4996 
  4997 * The new version of HOL (previously called CHOL) uses a curried syntax
  4998 for functions.  Application looks like f a b instead of f(a,b);
  4999 
  5000 * Mutually recursive inductive definitions finally work in HOL;
  5001 
  5002 * In ZF, pattern-matching on tuples is now available in all abstractions and
  5003 translates to the operator "split";
  5004 
  5005 
  5006 
  5007 New in Isabelle94-3
  5008 -------------------
  5009 
  5010 * new infix operator, addss, allowing the classical reasoner to
  5011 perform simplification at each step of its search.  Example:
  5012         fast_tac (cs addss ss)
  5013 
  5014 * a new logic, CHOL, the same as HOL, but with a curried syntax
  5015 for functions.  Application looks like f a b instead of f(a,b).  Also pairs
  5016 look like (a,b) instead of <a,b>;
  5017 
  5018 * PLEASE NOTE: CHOL will eventually replace HOL!
  5019 
  5020 * In CHOL, pattern-matching on tuples is now available in all abstractions.
  5021 It translates to the operator "split".  A new theory of integers is available;
  5022 
  5023 * In ZF, integer numerals now denote two's-complement binary integers.
  5024 Arithmetic operations can be performed by rewriting.  See ZF/ex/Bin.ML;
  5025 
  5026 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
  5027 of the Axiom of Choice;
  5028 
  5029 
  5030 
  5031 New in Isabelle94-2
  5032 -------------------
  5033 
  5034 * Significantly faster resolution;
  5035 
  5036 * the different sections in a .thy file can now be mixed and repeated
  5037 freely;
  5038 
  5039 * Database of theorems for FOL, HOL and ZF.  New
  5040 commands including qed, qed_goal and bind_thm store theorems in the database.
  5041 
  5042 * Simple database queries: return a named theorem (get_thm) or all theorems of
  5043 a given theory (thms_of), or find out what theory a theorem was proved in
  5044 (theory_of_thm);
  5045 
  5046 * Bugs fixed in the inductive definition and datatype packages;
  5047 
  5048 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
  5049 and HOL_dup_cs obsolete;
  5050 
  5051 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
  5052 have been removed;
  5053 
  5054 * Simpler definition of function space in ZF;
  5055 
  5056 * new results about cardinal and ordinal arithmetic in ZF;
  5057 
  5058 * 'subtype' facility in HOL for introducing new types as subsets of existing
  5059 types;
  5060 
  5061 
  5062 $Id$