NEWS
author wenzelm
Wed Nov 22 20:51:00 2006 +0100 (2006-11-22)
changeset 21471 03a5ef1936c5
parent 21462 74ddf3a522f8
child 21537 45b3a85ee548
permissions -rw-r--r--
* settings: ML_IDENTIFIER includes the Isabelle version identifier;
     1 Isabelle NEWS -- history user-relevant changes
     2 ==============================================
     3 
     4 New in this Isabelle version
     5 ----------------------------
     6 
     7 *** General ***
     8 
     9 * Theory syntax: the header format ``theory A = B + C:'' has been
    10 discontinued in favour of ``theory A imports B C begin''.  Use isatool
    11 fixheaders to convert existing theory files.  INCOMPATIBILITY.
    12 
    13 * Theory syntax: the old non-Isar theory file format has been
    14 discontinued altogether.  Note that ML proof scripts may still be used
    15 with Isar theories; migration is usually quite simple with the ML
    16 function use_legacy_bindings.  INCOMPATIBILITY.
    17 
    18 * Theory syntax: some popular names (e.g. "class", "if", "fun") are
    19 now keywords.  INCOMPATIBILITY, use double quotes.
    20 
    21 * Legacy goal package: reduced interface to the bare minimum required
    22 to keep existing proof scripts running.  Most other user-level
    23 functions are now part of the OldGoals structure, which is *not* open
    24 by default (consider isatool expandshort before open OldGoals).
    25 Removed top_sg, prin, printyp, pprint_term/typ altogether, because
    26 these tend to cause confusion about the actual goal (!) context being
    27 used here, which is not necessarily the same as the_context().
    28 
    29 * Command 'find_theorems': support "*" wildcard in "name:" criterion.
    30 
    31 * The ``prems limit'' option (cf. ProofContext.prems_limit) is now -1
    32 by default, which means that "prems" (and also "fixed variables") are
    33 suppressed from proof state output.  Note that the ProofGeneral
    34 settings mechanism allows to change and save options persistently, but
    35 older versions of Isabelle will fail to start up if a negative prems
    36 limit is imposed.
    37 
    38 * Local theory targets may be specified by non-nested blocks of
    39 ``context/locale/class ... begin'' followed by ``end''.  The body may
    40 contain definitions, theorems etc., including any derived mechanism
    41 that has been implemented on top of these primitives.  This concept
    42 generalizes the existing ``theorem (in ...)'' towards more versatility
    43 and scalability.
    44 
    45 
    46 *** Document preparation ***
    47 
    48 * Added antiquotation @{theory name} which prints the name $A$, after
    49 checking that it refers to a valid ancestor theory in the current
    50 context.
    51 
    52 * Added antiquotations @{ML_type text} and @{ML_struct text} which
    53 check the given source text as ML type/structure, printing verbatim.
    54 
    55 
    56 *** Pure ***
    57 
    58 * class_package.ML offers a combination of axclasses and locales to
    59 achieve Haskell-like type classes in Isabelle. See
    60 HOL/ex/Classpackage.thy for examples.
    61 
    62 * Yet another code generator framework allows to generate executable
    63 code for ML and Haskell (including "class"es). A short usage sketch:
    64 
    65     internal compilation:
    66         code_gen <list of constants (term syntax)> (SML *)
    67     writing SML code to a file:
    68         code_gen <list of constants (term syntax)> (SML <filename>)
    69     writing Haskell code to a bunch of files:
    70         code_gen <list of constants (term syntax)> (Haskell <filename>)
    71 
    72 Reasonable default setup of framework in HOL/Main.
    73 
    74 See HOL/ex/Code*.thy for examples.
    75 
    76 Theorem attributs for selecting and transforming function equations theorems:
    77 
    78     [code fun]:       select a theorem as function equation for a specific constant
    79     [code nofun]:     deselect a theorem as function equation for a specific constant
    80     [code inline]:    select an equation theorem for unfolding (inlining) in place
    81     [code noinline]:  deselect an equation theorem for unfolding (inlining) in place
    82 
    83 User-defined serializations (target in {SML, Haskell}):
    84 
    85     code_const <and-list of constants (term syntax)>
    86       {(target) <and-list of const target syntax>}+
    87 
    88     code_type <and-list of type constructors>
    89       {(target) <and-list of type target syntax>}+
    90 
    91     code_instance <and-list of instances>
    92       {(target)}+
    93         where instance ::= <type constructor> :: <class>
    94 
    95     code_class <and_list of classes>
    96       {(target) <and-list of class target syntax>}+
    97         where class target syntax ::= <class name> {where {<classop> == <target syntax>}+}?
    98 
    99 For code_instance and code_class, target SML is silently ignored.
   100 
   101 See HOL theories and HOL/ex/Code*.thy for usage examples.  Doc/Isar/Advanced/Codegen/
   102 provides a tutorial.
   103 
   104 * Command 'no_translations' removes translation rules from theory
   105 syntax.
   106 
   107 * Overloaded definitions are now actually checked for acyclic
   108 dependencies.  The overloading scheme is slightly more general than
   109 that of Haskell98, although Isabelle does not demand an exact
   110 correspondence to type class and instance declarations.
   111 INCOMPATIBILITY, use ``defs (unchecked overloaded)'' to admit more
   112 exotic versions of overloading -- at the discretion of the user!
   113 
   114 Polymorphic constants are represented via type arguments, i.e. the
   115 instantiation that matches an instance against the most general
   116 declaration given in the signature.  For example, with the declaration
   117 c :: 'a => 'a => 'a, an instance c :: nat => nat => nat is represented
   118 as c(nat).  Overloading is essentially simultaneous structural
   119 recursion over such type arguments.  Incomplete specification patterns
   120 impose global constraints on all occurrences, e.g. c('a * 'a) on the
   121 LHS means that more general c('a * 'b) will be disallowed on any RHS.
   122 Command 'print_theory' outputs the normalized system of recursive
   123 equations, see section "definitions".
   124 
   125 * Isar: improper proof element 'guess' is like 'obtain', but derives
   126 the obtained context from the course of reasoning!  For example:
   127 
   128   assume "EX x y. A x & B y"   -- "any previous fact"
   129   then guess x and y by clarify
   130 
   131 This technique is potentially adventurous, depending on the facts and
   132 proof tools being involved here.
   133 
   134 * Isar: known facts from the proof context may be specified as literal
   135 propositions, using ASCII back-quote syntax.  This works wherever
   136 named facts used to be allowed so far, in proof commands, proof
   137 methods, attributes etc.  Literal facts are retrieved from the context
   138 according to unification of type and term parameters.  For example,
   139 provided that "A" and "A ==> B" and "!!x. P x ==> Q x" are known
   140 theorems in the current context, then these are valid literal facts:
   141 `A` and `A ==> B` and `!!x. P x ==> Q x" as well as `P a ==> Q a` etc.
   142 
   143 There is also a proof method "fact" which does the same composition
   144 for explicit goal states, e.g. the following proof texts coincide with
   145 certain special cases of literal facts:
   146 
   147   have "A" by fact                 ==  note `A`
   148   have "A ==> B" by fact           ==  note `A ==> B`
   149   have "!!x. P x ==> Q x" by fact  ==  note `!!x. P x ==> Q x`
   150   have "P a ==> Q a" by fact       ==  note `P a ==> Q a`
   151 
   152 * Isar: ":" (colon) is no longer a symbolic identifier character in
   153 outer syntax.  Thus symbolic identifiers may be used without
   154 additional white space in declarations like this: ``assume *: A''.
   155 
   156 * Isar: 'print_facts' prints all local facts of the current context,
   157 both named and unnamed ones.
   158 
   159 * Isar: 'def' now admits simultaneous definitions, e.g.:
   160 
   161   def x == "t" and y == "u"
   162 
   163 * Isar: added command 'unfolding', which is structurally similar to
   164 'using', but affects both the goal state and facts by unfolding given
   165 rewrite rules.  Thus many occurrences of the 'unfold' method or
   166 'unfolded' attribute may be replaced by first-class proof text.
   167 
   168 * Isar: methods 'unfold' / 'fold', attributes 'unfolded' / 'folded',
   169 and command 'unfolding' now all support object-level equalities
   170 (potentially conditional).  The underlying notion of rewrite rule is
   171 analogous to the 'rule_format' attribute, but *not* that of the
   172 Simplifier (which is usually more generous).
   173 
   174 * Isar: the goal restriction operator [N] (default N = 1) evaluates a
   175 method expression within a sandbox consisting of the first N
   176 sub-goals, which need to exist.  For example, ``simp_all [3]''
   177 simplifies the first three sub-goals, while (rule foo, simp_all)[]
   178 simplifies all new goals that emerge from applying rule foo to the
   179 originally first one.
   180 
   181 * Isar: schematic goals are no longer restricted to higher-order
   182 patterns; e.g. ``lemma "?P(?x)" by (rule TrueI)'' now works as
   183 expected.
   184 
   185 * Isar: the conclusion of a long theorem statement is now either
   186 'shows' (a simultaneous conjunction, as before), or 'obtains'
   187 (essentially a disjunction of cases with local parameters and
   188 assumptions).  The latter allows to express general elimination rules
   189 adequately; in this notation common elimination rules look like this:
   190 
   191   lemma exE:    -- "EX x. P x ==> (!!x. P x ==> thesis) ==> thesis"
   192     assumes "EX x. P x"
   193     obtains x where "P x"
   194 
   195   lemma conjE:  -- "A & B ==> (A ==> B ==> thesis) ==> thesis"
   196     assumes "A & B"
   197     obtains A and B
   198 
   199   lemma disjE:  -- "A | B ==> (A ==> thesis) ==> (B ==> thesis) ==> thesis"
   200     assumes "A | B"
   201     obtains
   202       A
   203     | B
   204 
   205 The subsequent classical rules even refer to the formal "thesis"
   206 explicitly:
   207 
   208   lemma classical:     -- "(~ thesis ==> thesis) ==> thesis"
   209     obtains "~ thesis"
   210 
   211   lemma Peirce's_Law:  -- "((thesis ==> something) ==> thesis) ==> thesis"
   212     obtains "thesis ==> something"
   213 
   214 The actual proof of an 'obtains' statement is analogous to that of the
   215 Isar proof element 'obtain', only that there may be several cases.
   216 Optional case names may be specified in parentheses; these will be
   217 available both in the present proof and as annotations in the
   218 resulting rule, for later use with the 'cases' method (cf. attribute
   219 case_names).
   220 
   221 * Isar: the assumptions of a long theorem statement are available as
   222 "assms" fact in the proof context.  This is more appropriate than the
   223 (historical) "prems", which refers to all assumptions of the current
   224 context, including those from the target locale, proof body etc.
   225 
   226 * Isar: 'print_statement' prints theorems from the current theory or
   227 proof context in long statement form, according to the syntax of a
   228 top-level lemma.
   229 
   230 * Isar: 'obtain' takes an optional case name for the local context
   231 introduction rule (default "that").
   232 
   233 * Isar: removed obsolete 'concl is' patterns.  INCOMPATIBILITY, use
   234 explicit (is "_ ==> ?foo") in the rare cases where this still happens
   235 to occur.
   236 
   237 * Pure: syntax "CONST name" produces a fully internalized constant
   238 according to the current context.  This is particularly useful for
   239 syntax translations that should refer to internal constant
   240 representations independently of name spaces.
   241 
   242 * Isar/locales: 'notation' provides a robust interface to the 'syntax'
   243 primitive that also works in a locale context (both for constants and
   244 fixed variables).  Type declaration and internal syntactic
   245 representation of given constants retrieved from the context.
   246 
   247 * Isar/locales: new derived specification elements 'axiomatization',
   248 'definition', 'abbreviation', which support type-inference, admit
   249 object-level specifications (equality, equivalence).  See also the
   250 isar-ref manual.  Examples:
   251 
   252   axiomatization
   253     eq  (infix "===" 50)
   254     where eq_refl: "x === x" and eq_subst: "x === y ==> P x ==> P y"
   255 
   256   definition
   257     "f x y = x + y + 1"
   258     "g x = f x x"
   259 
   260   abbreviation
   261     neq  (infix "=!=" 50)
   262     "x =!= y == ~ (x === y)"
   263 
   264 These specifications may be also used in a locale context.  Then the
   265 constants being introduced depend on certain fixed parameters, and the
   266 constant name is qualified by the locale base name.  An internal
   267 abbreviation takes care for convenient input and output, making the
   268 parameters implicit and using the original short name.  See also
   269 HOL/ex/Abstract_NAT.thy for an example of deriving polymorphic
   270 entities from a monomorphic theory.
   271 
   272 Presently, abbreviations are only available 'in' a target locale, but
   273 not inherited by general import expressions.  Also note that
   274 'abbreviation' may be used as a type-safe replacement for 'syntax' +
   275 'translations' in common applications.
   276 
   277 Concrete syntax is attached to specified constants in internal form,
   278 independently of name spaces.  The parse tree representation is
   279 slightly different -- use 'notation' instead of raw 'syntax', and
   280 'translations' with explicit "CONST" markup to accommodate this.
   281 
   282 * Isar/locales: improved parameter handling:
   283 - use of locales "var" and "struct" no longer necessary;
   284 - parameter renamings are no longer required to be injective.
   285   This enables, for example, to define a locale for endomorphisms thus:
   286   locale endom = homom mult mult h.
   287 
   288 * Isar/locales: changed the way locales with predicates are defined.
   289 Instead of accumulating the specification, the imported expression is
   290 now an interpretation.
   291 INCOMPATIBILITY: different normal form of locale expressions.
   292 In particular, in interpretations of locales with predicates,
   293 goals repesenting already interpreted fragments are not removed
   294 automatically.  Use methods `intro_locales' and `unfold_locales'; see below.
   295 
   296 * Isar/locales: new methods `intro_locales' and `unfold_locales' provide
   297 backward reasoning on locales predicates.  The methods are aware of
   298 interpretations and discharge corresponding goals.  `intro_locales' is
   299 less aggressive then `unfold_locales' and does not unfold predicates to
   300 assumptions.
   301 
   302 * Isar/locales: the order in which locale fragments are accumulated
   303 has changed.  This enables to override declarations from fragments
   304 due to interpretations -- for example, unwanted simp rules.
   305 
   306 * Provers/induct: improved internal context management to support
   307 local fixes and defines on-the-fly.  Thus explicit meta-level
   308 connectives !! and ==> are rarely required anymore in inductive goals
   309 (using object-logic connectives for this purpose has been long
   310 obsolete anyway).  The subsequent proof patterns illustrate advanced
   311 techniques of natural induction; general datatypes and inductive sets
   312 work analogously (see also src/HOL/Lambda for realistic examples).
   313 
   314 (1) This is how to ``strengthen'' an inductive goal wrt. certain
   315 parameters:
   316 
   317   lemma
   318     fixes n :: nat and x :: 'a
   319     assumes a: "A n x"
   320     shows "P n x"
   321     using a                     -- {* make induct insert fact a *}
   322   proof (induct n arbitrary: x) -- {* generalize goal to "!!x. A n x ==> P n x" *}
   323     case 0
   324     show ?case sorry
   325   next
   326     case (Suc n)
   327     note `!!x. A n x ==> P n x` -- {* induction hypothesis, according to induction rule *}
   328     note `A (Suc n) x`          -- {* induction premise, stemming from fact a *}
   329     show ?case sorry
   330   qed
   331 
   332 (2) This is how to perform induction over ``expressions of a certain
   333 form'', using a locally defined inductive parameter n == "a x"
   334 together with strengthening (the latter is usually required to get
   335 sufficiently flexible induction hypotheses):
   336 
   337   lemma
   338     fixes a :: "'a => nat"
   339     assumes a: "A (a x)"
   340     shows "P (a x)"
   341     using a
   342   proof (induct n == "a x" arbitrary: x)
   343     ...
   344 
   345 See also HOL/Isar_examples/Puzzle.thy for an application of the this
   346 particular technique.
   347 
   348 (3) This is how to perform existential reasoning ('obtains' or
   349 'obtain') by induction, while avoiding explicit object-logic
   350 encodings:
   351 
   352   lemma
   353     fixes n :: nat
   354     obtains x :: 'a where "P n x" and "Q n x"
   355   proof (induct n arbitrary: thesis)
   356     case 0
   357     obtain x where "P 0 x" and "Q 0 x" sorry
   358     then show thesis by (rule 0)
   359   next
   360     case (Suc n)
   361     obtain x where "P n x" and "Q n x" by (rule Suc.hyps)
   362     obtain x where "P (Suc n) x" and "Q (Suc n) x" sorry
   363     then show thesis by (rule Suc.prems)
   364   qed
   365 
   366 Here the 'arbitrary: thesis' specification essentially modifies the
   367 scope of the formal thesis parameter, in order to the get the whole
   368 existence statement through the induction as expected.
   369 
   370 * Provers/induct: mutual induction rules are now specified as a list
   371 of rule sharing the same induction cases.  HOL packages usually
   372 provide foo_bar.inducts for mutually defined items foo and bar
   373 (e.g. inductive sets or datatypes).  INCOMPATIBILITY, users need to
   374 specify mutual induction rules differently, i.e. like this:
   375 
   376   (induct rule: foo_bar.inducts)
   377   (induct set: foo bar)
   378   (induct type: foo bar)
   379 
   380 The ML function ProjectRule.projections turns old-style rules into the
   381 new format.
   382 
   383 * Provers/induct: improved handling of simultaneous goals.  Instead of
   384 introducing object-level conjunction, the statement is now split into
   385 several conclusions, while the corresponding symbolic cases are
   386 nested accordingly.  INCOMPATIBILITY, proofs need to be structured
   387 explicitly.  For example:
   388 
   389   lemma
   390     fixes n :: nat
   391     shows "P n" and "Q n"
   392   proof (induct n)
   393     case 0 case 1
   394     show "P 0" sorry
   395   next
   396     case 0 case 2
   397     show "Q 0" sorry
   398   next
   399     case (Suc n) case 1
   400     note `P n` and `Q n`
   401     show "P (Suc n)" sorry
   402   next
   403     case (Suc n) case 2
   404     note `P n` and `Q n`
   405     show "Q (Suc n)" sorry
   406   qed
   407 
   408 The split into subcases may be deferred as follows -- this is
   409 particularly relevant for goal statements with local premises.
   410 
   411   lemma
   412     fixes n :: nat
   413     shows "A n ==> P n" and "B n ==> Q n"
   414   proof (induct n)
   415     case 0
   416     {
   417       case 1
   418       note `A 0`
   419       show "P 0" sorry
   420     next
   421       case 2
   422       note `B 0`
   423       show "Q 0" sorry
   424     }
   425   next
   426     case (Suc n)
   427     note `A n ==> P n` and `B n ==> Q n`
   428     {
   429       case 1
   430       note `A (Suc n)`
   431       show "P (Suc n)" sorry
   432     next
   433       case 2
   434       note `B (Suc n)`
   435       show "Q (Suc n)" sorry
   436     }
   437   qed
   438 
   439 If simultaneous goals are to be used with mutual rules, the statement
   440 needs to be structured carefully as a two-level conjunction, using
   441 lists of propositions separated by 'and':
   442 
   443   lemma
   444     shows "a : A ==> P1 a"
   445           "a : A ==> P2 a"
   446       and "b : B ==> Q1 b"
   447           "b : B ==> Q2 b"
   448           "b : B ==> Q3 b"
   449   proof (induct set: A B)
   450 
   451 * Provers/induct: support coinduction as well.  See
   452 src/HOL/Library/Coinductive_List.thy for various examples.
   453 
   454 * Attribute "symmetric" produces result with standardized schematic
   455 variables (index 0).  Potential INCOMPATIBILITY.
   456 
   457 * Simplifier: by default the simplifier trace only shows top level rewrites
   458 now. That is, trace_simp_depth_limit is set to 1 by default. Thus there is
   459 less danger of being flooded by the trace. The trace indicates where parts
   460 have been suppressed.
   461   
   462 * Provers/classical: removed obsolete classical version of elim_format
   463 attribute; classical elim/dest rules are now treated uniformly when
   464 manipulating the claset.
   465 
   466 * Provers/classical: stricter checks to ensure that supplied intro,
   467 dest and elim rules are well-formed; dest and elim rules must have at
   468 least one premise.
   469 
   470 * Provers/classical: attributes dest/elim/intro take an optional
   471 weight argument for the rule (just as the Pure versions).  Weights are
   472 ignored by automated tools, but determine the search order of single
   473 rule steps.
   474 
   475 * Syntax: input syntax now supports dummy variable binding "%_. b",
   476 where the body does not mention the bound variable.  Note that dummy
   477 patterns implicitly depend on their context of bounds, which makes
   478 "{_. _}" match any set comprehension as expected.  Potential
   479 INCOMPATIBILITY -- parse translations need to cope with syntactic
   480 constant "_idtdummy" in the binding position.
   481 
   482 * Syntax: removed obsolete syntactic constant "_K" and its associated
   483 parse translation.  INCOMPATIBILITY -- use dummy abstraction instead,
   484 for example "A -> B" => "Pi A (%_. B)".
   485 
   486 * Pure: 'class_deps' command visualizes the subclass relation, using
   487 the graph browser tool.
   488 
   489 * Pure: 'print_theory' now suppresses entities with internal name
   490 (trailing "_") by default; use '!' option for full details.
   491 
   492 
   493 *** HOL ***
   494 
   495 * New syntactic class "size"; overloaded constant "size" now
   496 has type "'a::size ==> bool"
   497 
   498 * Constants "Divides.op div", "Divides.op mod" and "Divides.op dvd" no named
   499   "Divides.div", "Divides.mod" and "Divides.dvd"
   500   INCOMPATIBILITY for ML code directly refering to constant names.
   501 
   502 * Replaced "auto_term" by the conceptually simpler method "relation",
   503 which just applies the instantiated termination rule with no further
   504 simplifications.
   505 INCOMPATIBILITY: 
   506 Replace 
   507         termination by (auto_term "MYREL")
   508 with 
   509         termination by (relation "MYREL") auto
   510 
   511 * Automated termination proofs "by lexicographic_order" are now
   512 included in the abbreviated function command "fun". No explicit
   513 "termination" command is necessary anymore.
   514 INCOMPATIBILITY: If a "fun"-definition cannot be proved terminating by
   515 a lexicographic size order, then the command fails. Use the expanded
   516 version "function" for these cases.
   517 
   518 * New method "lexicographic_order" automatically synthesizes
   519 termination relations as lexicographic combinations of size measures. 
   520 Usage for (function package) termination proofs:
   521 
   522 termination 
   523 by lexicographic_order
   524 
   525 * Records: generalised field-update to take a function on the field 
   526 rather than the new value: r(|A := x|) is translated to A_update (K x) r
   527 The K-combinator that is internally used is called K_record.
   528 INCOMPATIBILITY: Usage of the plain update functions has to be
   529 adapted.
   530  
   531 * axclass "semiring_0" now contains annihilation axioms 
   532 ("x * 0 = 0","0 * x = 0"), which are required for a semiring. Richer 
   533 structures do not inherit from semiring_0 anymore, because this property 
   534 is a theorem there, not an axiom.
   535 INCOMPATIBILITY: In instances of semiring_0, there is more to prove, but 
   536 this is mostly trivial.
   537 
   538 * axclass "recpower" was generalized to arbitrary monoids, not just 
   539 commutative semirings.
   540 INCOMPATIBILITY: If you use recpower and need commutativity or a semiring
   541 property, add the corresponding classes.
   542 
   543 * Locale Lattic_Locales.partial_order changed (to achieve consistency with
   544   axclass order):
   545   - moved to Orderings.partial_order
   546   - additional parameter ``less''
   547   INCOMPATIBILITY.
   548 
   549 * Constant "List.list_all2" in List.thy now uses authentic syntax.
   550 INCOMPATIBILITY: translations containing list_all2 may go wrong. On Isar
   551 level, use abbreviations instead.
   552 
   553 * Constant "List.op mem" in List.thy now has proper name: "List.memberl"
   554 INCOMPATIBILITY: rarely occuring name references (e.g. ``List.op mem.simps'')
   555 require renaming (e.g. ``List.memberl.simps'').
   556 
   557 * Renamed constants in HOL.thy:
   558     0      ~> HOL.zero
   559     1      ~> HOL.one
   560 INCOMPATIBILITY: ML code directly refering to constant names may need adaption
   561 This in general only affects hand-written proof tactics, simprocs and so on.
   562 
   563 * New theory Code_Generator providing class 'eq',
   564 allowing for code generation with polymorphic equality.
   565 
   566 * Numeral syntax: type 'bin' which was a mere type copy of 'int' has been
   567 abandoned in favour of plain 'int'. INCOMPATIBILITY -- Significant changes
   568 for setting up numeral syntax for types:
   569 
   570   - new constants Numeral.pred and Numeral.succ instead
   571       of former Numeral.bin_pred and Numeral.bin_succ.
   572   - Use integer operations instead of bin_add, bin_mult and so on.
   573   - Numeral simplification theorems named Numeral.numeral_simps instead of Bin_simps.
   574   - ML structure Bin_Simprocs now named Int_Numeral_Base_Simprocs.
   575 
   576 See HOL/Integ/IntArith.thy for an example setup.
   577 
   578 * New top level command 'normal_form' computes the normal form of a term
   579 that may contain free variables. For example 'normal_form "rev[a,b,c]"'
   580 prints '[b,c,a]'. This command is suitable for heavy-duty computations
   581 because the functions are compiled to ML first.
   582 INCOMPATIBILITY: new keywords 'normal_form' must quoted when used as
   583 an identifier.
   584 
   585 * Alternative iff syntax "A <-> B" for equality on bool (with priority
   586 25 like -->); output depends on the "iff" print_mode, the default is
   587 "A = B" (with priority 50).
   588 
   589 * Renamed constants in HOL.thy and Orderings.thy:
   590     op +   ~> HOL.plus
   591     op -   ~> HOL.minus
   592     uminus ~> HOL.uminus
   593     op *   ~> HOL.times
   594     op <   ~> Orderings.less
   595     op <=  ~> Orderings.less_eq
   596 
   597 Adaptions may be required in the following cases:
   598 
   599 a) User-defined constants using any of the names "plus", "minus", "times",
   600 "less" or "less_eq". The standard syntax translations for "+", "-" and "*"
   601 may go wrong.
   602 INCOMPATIBILITY: use more specific names.
   603 
   604 b) Variables named "plus", "minus", "times", "less", "less_eq"
   605 INCOMPATIBILITY: use more specific names.
   606 
   607 c) Permutative equations (e.g. "a + b = b + a")
   608 Since the change of names also changes the order of terms, permutative
   609 rewrite rules may get applied in a different order. Experience shows that
   610 this is rarely the case (only two adaptions in the whole Isabelle
   611 distribution).
   612 INCOMPATIBILITY: rewrite proofs
   613 
   614 d) ML code directly refering to constant names
   615 This in general only affects hand-written proof tactics, simprocs and so on.
   616 INCOMPATIBILITY: grep your sourcecode and replace names.
   617 
   618 * Relations less (<) and less_eq (<=) are also available on type bool.
   619 Modified syntax to disallow nesting without explicit parentheses,
   620 e.g. "(x < y) < z" or "x < (y < z)", but NOT "x < y < z".
   621 
   622 * "LEAST x:A. P" expands to "LEAST x. x:A & P" (input only).
   623 
   624 * Relation composition operator "op O" now has precedence 75 and binds
   625 stronger than union and intersection. INCOMPATIBILITY.
   626 
   627 * The old set interval syntax "{m..n(}" (and relatives) has been removed.
   628   Use "{m..<n}" (and relatives) instead.
   629 
   630 * In the context of the assumption "~(s = t)" the Simplifier rewrites
   631 "t = s" to False (by simproc "neq_simproc").  For backward
   632 compatibility this can be disabled by ML "reset use_neq_simproc".
   633 
   634 * "m dvd n" where m and n are numbers is evaluated to True/False by simp.
   635 
   636 * Theorem Cons_eq_map_conv no longer has attribute `simp'.
   637 
   638 * Theorem setsum_mult renamed to setsum_right_distrib.
   639 
   640 * Prefer ex1I over ex_ex1I in single-step reasoning, e.g. by the
   641 'rule' method.
   642 
   643 * Tactics 'sat' and 'satx' reimplemented, several improvements: goals
   644 no longer need to be stated as "<prems> ==> False", equivalences (i.e.
   645 "=" on type bool) are handled, variable names of the form "lit_<n>"
   646 are no longer reserved, significant speedup.
   647 
   648 * Tactics 'sat' and 'satx' can now replay MiniSat proof traces.  zChaff is
   649 still supported as well.
   650 
   651 * inductive and datatype: provide projections of mutual rules, bundled
   652 as foo_bar.inducts;
   653 
   654 * Library: moved theories Parity, GCD, Binomial, Infinite_Set to Library.
   655 
   656 * Library: moved theory Accessible_Part to main HOL.
   657 
   658 * Library: added theory Coinductive_List of potentially infinite lists
   659 as greatest fixed-point.
   660 
   661 * Library: added theory AssocList which implements (finite) maps as
   662 association lists.
   663 
   664 * New proof method "evaluation" for efficiently solving a goal (i.e. a
   665 boolean expression) by compiling it to ML. The goal is "proved" (via
   666 the oracle "Evaluation") if it evaluates to True.
   667 
   668 * Linear arithmetic now splits certain operators (e.g. min, max, abs)
   669 also when invoked by the simplifier.  This results in the simplifier
   670 being more powerful on arithmetic goals.  INCOMPATIBILITY.  Set
   671 fast_arith_split_limit to 0 to obtain the old behavior.
   672 
   673 * Support for hex (0x20) and binary (0b1001) numerals. 
   674 
   675 * New method: reify eqs (t), where eqs are equations for an
   676 interpretation I :: 'a list => 'b => 'c and t::'c is an optional
   677 parameter, computes a term s::'b and a list xs::'a list and proves the
   678 theorem I xs s = t. This is also known as reification or quoting. The
   679 resulting theorem is applied to the subgoal to substitute t with I xs
   680 s.  If t is omitted, the subgoal itself is reified.
   681 
   682 * New method: reflection corr_thm eqs (t). The parameters eqs and (t)
   683 are as explained above. corr_thm is a theorem for I vs (f t) = I vs t,
   684 where f is supposed to be a computable function (in the sense of code
   685 generattion). The method uses reify to compute s and xs as above then
   686 applies corr_thm and uses normalization by evaluation to "prove" f s =
   687 r and finally gets the theorem t = r, which is again applied to the
   688 subgoal. An Example is available in HOL/ex/ReflectionEx.thy.
   689 
   690 * Reflection: Automatic refification now handels binding, an example
   691 is available in HOL/ex/ReflectionEx.thy
   692 
   693 
   694 *** HOL-Algebra ***
   695 
   696 * Formalisation of ideals and the quotient construction over rings.
   697 
   698 * Order and lattice theory no longer based on records.
   699 INCOMPATIBILITY.
   700 
   701 * Method algebra is now set up via an attribute.  For examples see
   702 CRing.thy.  INCOMPATIBILITY: the method is now weaker on combinations
   703 of algebraic structures.
   704 
   705 * Renamed `CRing.thy' to `Ring.thy'.  INCOMPATIBILITY.
   706 
   707 
   708 *** HOL-Complex ***
   709 
   710 * Theory Real: new method ferrack implements quantifier elimination
   711 for linear arithmetic over the reals. The quantifier elimination
   712 feature is used only for decision, for compatibility with arith. This
   713 means a goal is either solved or left unchanged, no simplification.
   714 
   715 
   716 *** ML ***
   717 
   718 * Pure/table:
   719 
   720 Function `...tab.foldl` removed.
   721 INCOMPATIBILITY: use `...tabfold` instead
   722 
   723 * Pure/library:
   724 
   725 `gen_rem(s)` abandoned in favour of `remove` / `subtract`.
   726 INCOMPATIBILITY:
   727 rewrite "gen_rem eq (xs, x)" to "remove (eq o swap) x xs"
   728 rewrite "gen_rems eq (xs, ys)" to "subtract (eq o swap) ys xs"
   729 drop "swap" if "eq" is symmetric.
   730 
   731 * Pure/library:
   732 
   733 infixes `ins` `ins_string` `ins_int` have been abandoned in favour of `insert`.
   734 INCOMPATIBILITY: rewrite "x ins(_...) xs" to "insert (op =) x xs"
   735 
   736 * Pure/General/susp.ML:
   737 
   738 New module for delayed evaluations.
   739 
   740 * Pure/library:
   741 
   742 Semantically identical functions "equal_list" and "eq_list" have been
   743 unified to "eq_list".
   744 
   745 * Pure/library:
   746 
   747   val burrow: ('a list -> 'b list) -> 'a list list -> 'b list list
   748   val fold_burrow: ('a list -> 'c -> 'b list * 'd) -> 'a list list -> 'c -> 'b list list * 'd
   749 
   750 The semantics of "burrow" is: "take a function with *simulatanously*
   751 transforms a list of value, and apply it *simulatanously* to a list of
   752 list of values of the appropriate type". Confer this with "map" which
   753 would *not* apply its argument function simulatanously but in
   754 sequence. "fold_burrow" has an additional context.
   755 
   756 Both actually avoid the usage of "unflat" since they hide away
   757 "unflat" from the user.
   758 
   759 * Pure/library: functions map2 and fold2 with curried syntax for
   760 simultanous mapping and folding:
   761 
   762     val map2: ('a -> 'b -> 'c) -> 'a list -> 'b list -> 'c list
   763     val fold2: ('a -> 'b -> 'c -> 'c) -> 'a list -> 'b list -> 'c -> 'c
   764 
   765 * Pure/library: indexed lists - some functions in the Isabelle library
   766 treating lists over 'a as finite mappings from [0...n] to 'a have been
   767 given more convenient names and signatures reminiscent of similar
   768 functions for alists, tables, etc:
   769 
   770   val nth: 'a list -> int -> 'a 
   771   val nth_map: int -> ('a -> 'a) -> 'a list -> 'a list
   772   val fold_index: (int * 'a -> 'b -> 'b) -> 'a list -> 'b -> 'b
   773 
   774 Note that fold_index starts counting at index 0, not 1 like foldln
   775 used to.
   776 
   777 * Pure/library: general ``divide_and_conquer'' combinator on lists.
   778 
   779 * Pure/General/name_mangler.ML provides a functor for generic name
   780 mangling (bijective mapping from expression values to strings).
   781 
   782 * Pure/General/rat.ML implements rational numbers.
   783 
   784 * Pure/General/table.ML: the join operations now works via exceptions
   785 DUP/SAME instead of type option.  This is simpler in simple cases, and
   786 admits slightly more efficient complex applications.
   787 
   788 * Pure: datatype Context.generic joins theory/Proof.context and
   789 provides some facilities for code that works in either kind of
   790 context, notably GenericDataFun for uniform theory and proof data.
   791 
   792 * Pure: 'advanced' translation functions (parse_translation etc.) now
   793 use Context.generic instead of just theory.
   794 
   795 * Pure: simplified internal attribute type, which is now always
   796 Context.generic * thm -> Context.generic * thm.  Global (theory)
   797 vs. local (Proof.context) attributes have been discontinued, while
   798 minimizing code duplication.  Thm.rule_attribute and
   799 Thm.declaration_attribute build canonical attributes; see also
   800 structure Context for further operations on Context.generic, notably
   801 GenericDataFun.  INCOMPATIBILITY, need to adapt attribute type
   802 declarations and definitions.
   803 
   804 * Pure/kernel: consts certification ignores sort constraints given in
   805 signature declarations.  (This information is not relevant to the
   806 logic, but only for type inference.)  IMPORTANT INTERNAL CHANGE.
   807 
   808 * Pure: axiomatic type classes are now purely definitional, with
   809 explicit proofs of class axioms and super class relations performed
   810 internally.  See Pure/axclass.ML for the main internal interfaces --
   811 notably AxClass.define_class supercedes AxClass.add_axclass, and
   812 AxClass.axiomatize_class/classrel/arity supercede
   813 Sign.add_classes/classrel/arities.
   814 
   815 * Pure/Isar: Args/Attrib parsers operate on Context.generic --
   816 global/local versions on theory vs. Proof.context have been
   817 discontinued; Attrib.syntax and Method.syntax have been adapted
   818 accordingly.  INCOMPATIBILITY, need to adapt parser expressions for
   819 attributes, methods, etc.
   820 
   821 * Pure: several functions of signature "... -> theory -> theory * ..."
   822 have been reoriented to "... -> theory -> ... * theory" in order to
   823 allow natural usage in combination with the ||>, ||>>, |-> and
   824 fold_map combinators.
   825 
   826 * Pure: primitive rule lift_rule now takes goal cterm instead of an
   827 actual goal state (thm).  Use Thm.lift_rule (Thm.cprem_of st i) to
   828 achieve the old behaviour.
   829 
   830 * Pure: the "Goal" constant is now called "prop", supporting a
   831 slightly more general idea of ``protecting'' meta-level rule
   832 statements.
   833 
   834 * Pure: Logic.(un)varify only works in a global context, which is now
   835 enforced instead of silently assumed.  INCOMPATIBILITY, may use
   836 Logic.legacy_(un)varify as temporary workaround.
   837 
   838 * Pure: structure Name provides scalable operations for generating
   839 internal variable names, notably Name.variants etc.  This replaces
   840 some popular functions from term.ML:
   841 
   842   Term.variant		->  Name.variant
   843   Term.variantlist	->  Name.variant_list  (*canonical argument order*)
   844   Term.invent_names	->  Name.invent_list
   845 
   846 Note that low-level renaming rarely occurs in new code -- operations
   847 from structure Variable are used instead (see below).
   848 
   849 * Pure: structure Variable provides fundamental operations for proper
   850 treatment of fixed/schematic variables in a context.  For example,
   851 Variable.import introduces fixes for schematics of given facts and
   852 Variable.export reverses the effect (up to renaming) -- this replaces
   853 various freeze_thaw operations.
   854 
   855 * Pure: structure Goal provides simple interfaces for
   856 init/conclude/finish and tactical prove operations (replacing former
   857 Tactic.prove).  Goal.prove is the canonical way to prove results
   858 within a given context; Goal.prove_global is a degraded version for
   859 theory level goals, including a global Drule.standard.  Note that
   860 OldGoals.prove_goalw_cterm has long been obsolete, since it is
   861 ill-behaved in a local proof context (e.g. with local fixes/assumes or
   862 in a locale context).
   863 
   864 * Isar: simplified treatment of user-level errors, using exception
   865 ERROR of string uniformly.  Function error now merely raises ERROR,
   866 without any side effect on output channels.  The Isar toplevel takes
   867 care of proper display of ERROR exceptions.  ML code may use plain
   868 handle/can/try; cat_error may be used to concatenate errors like this:
   869 
   870   ... handle ERROR msg => cat_error msg "..."
   871 
   872 Toplevel ML code (run directly or through the Isar toplevel) may be
   873 embedded into the Isar toplevel with exception display/debug like
   874 this:
   875 
   876   Isar.toplevel (fn () => ...)
   877 
   878 INCOMPATIBILITY, removed special transform_error facilities, removed
   879 obsolete variants of user-level exceptions (ERROR_MESSAGE,
   880 Context.PROOF, ProofContext.CONTEXT, Proof.STATE, ProofHistory.FAIL)
   881 -- use plain ERROR instead.
   882 
   883 * Isar: theory setup now has type (theory -> theory), instead of a
   884 list.  INCOMPATIBILITY, may use #> to compose setup functions.
   885 
   886 * Isar: installed ML toplevel pretty printer for type Proof.context,
   887 subject to ProofContext.debug/verbose flags.
   888 
   889 * Isar: Toplevel.theory_to_proof admits transactions that modify the
   890 theory before entering a proof state.  Transactions now always see a
   891 quasi-functional intermediate checkpoint, both in interactive and
   892 batch mode.
   893 
   894 * Simplifier: the simpset of a running simplification process now
   895 contains a proof context (cf. Simplifier.the_context), which is the
   896 very context that the initial simpset has been retrieved from (by
   897 simpset_of/local_simpset_of).  Consequently, all plug-in components
   898 (solver, looper etc.) may depend on arbitrary proof data.
   899 
   900 * Simplifier.inherit_context inherits the proof context (plus the
   901 local bounds) of the current simplification process; any simproc
   902 etc. that calls the Simplifier recursively should do this!  Removed
   903 former Simplifier.inherit_bounds, which is already included here --
   904 INCOMPATIBILITY.  Tools based on low-level rewriting may even have to
   905 specify an explicit context using Simplifier.context/theory_context.
   906 
   907 * Simplifier/Classical Reasoner: more abstract interfaces
   908 change_simpset/claset for modifying the simpset/claset reference of a
   909 theory; raw versions simpset/claset_ref etc. have been discontinued --
   910 INCOMPATIBILITY.
   911 
   912 * Provers: more generic wrt. syntax of object-logics, avoid hardwired
   913 "Trueprop" etc.
   914 
   915 
   916 *** System ***
   917 
   918 * settings: ML_IDENTIFIER -- which is appended to user specific heap
   919 locations -- now includes the Isabelle version identifier as well.
   920 This simplifies use of multiple Isabelle installations.
   921 
   922 * isabelle-process: option -S (secure mode) disables some critical
   923 operations, notably runtime compilation and evaluation of ML source
   924 code.
   925 
   926 
   927 New in Isabelle2005 (October 2005)
   928 ----------------------------------
   929 
   930 *** General ***
   931 
   932 * Theory headers: the new header syntax for Isar theories is
   933 
   934   theory <name>
   935   imports <theory1> ... <theoryN>
   936   uses <file1> ... <fileM>
   937   begin
   938 
   939 where the 'uses' part is optional.  The previous syntax
   940 
   941   theory <name> = <theory1> + ... + <theoryN>:
   942 
   943 will disappear in the next release.  Use isatool fixheaders to convert
   944 existing theory files.  Note that there is no change in ancient
   945 non-Isar theories now, but these will disappear soon.
   946 
   947 * Theory loader: parent theories can now also be referred to via
   948 relative and absolute paths.
   949 
   950 * Command 'find_theorems' searches for a list of criteria instead of a
   951 list of constants. Known criteria are: intro, elim, dest, name:string,
   952 simp:term, and any term. Criteria can be preceded by '-' to select
   953 theorems that do not match. Intro, elim, dest select theorems that
   954 match the current goal, name:s selects theorems whose fully qualified
   955 name contain s, and simp:term selects all simplification rules whose
   956 lhs match term.  Any other term is interpreted as pattern and selects
   957 all theorems matching the pattern. Available in ProofGeneral under
   958 'ProofGeneral -> Find Theorems' or C-c C-f.  Example:
   959 
   960   C-c C-f (100) "(_::nat) + _ + _" intro -name: "HOL."
   961 
   962 prints the last 100 theorems matching the pattern "(_::nat) + _ + _",
   963 matching the current goal as introduction rule and not having "HOL."
   964 in their name (i.e. not being defined in theory HOL).
   965 
   966 * Command 'thms_containing' has been discontinued in favour of
   967 'find_theorems'; INCOMPATIBILITY.
   968 
   969 * Communication with Proof General is now 8bit clean, which means that
   970 Unicode text in UTF-8 encoding may be used within theory texts (both
   971 formal and informal parts).  Cf. option -U of the Isabelle Proof
   972 General interface.  Here are some simple examples (cf. src/HOL/ex):
   973 
   974   http://isabelle.in.tum.de/library/HOL/ex/Hebrew.html
   975   http://isabelle.in.tum.de/library/HOL/ex/Chinese.html
   976 
   977 * Improved efficiency of the Simplifier and, to a lesser degree, the
   978 Classical Reasoner.  Typical big applications run around 2 times
   979 faster.
   980 
   981 
   982 *** Document preparation ***
   983 
   984 * Commands 'display_drafts' and 'print_drafts' perform simple output
   985 of raw sources.  Only those symbols that do not require additional
   986 LaTeX packages (depending on comments in isabellesym.sty) are
   987 displayed properly, everything else is left verbatim.  isatool display
   988 and isatool print are used as front ends (these are subject to the
   989 DVI/PDF_VIEWER and PRINT_COMMAND settings, respectively).
   990 
   991 * Command tags control specific markup of certain regions of text,
   992 notably folding and hiding.  Predefined tags include "theory" (for
   993 theory begin and end), "proof" for proof commands, and "ML" for
   994 commands involving ML code; the additional tags "visible" and
   995 "invisible" are unused by default.  Users may give explicit tag
   996 specifications in the text, e.g. ''by %invisible (auto)''.  The
   997 interpretation of tags is determined by the LaTeX job during document
   998 preparation: see option -V of isatool usedir, or options -n and -t of
   999 isatool document, or even the LaTeX macros \isakeeptag, \isafoldtag,
  1000 \isadroptag.
  1001 
  1002 Several document versions may be produced at the same time via isatool
  1003 usedir (the generated index.html will link all of them).  Typical
  1004 specifications include ''-V document=theory,proof,ML'' to present
  1005 theory/proof/ML parts faithfully, ''-V outline=/proof,/ML'' to fold
  1006 proof and ML commands, and ''-V mutilated=-theory,-proof,-ML'' to omit
  1007 these parts without any formal replacement text.  The Isabelle site
  1008 default settings produce ''document'' and ''outline'' versions as
  1009 specified above.
  1010 
  1011 * Several new antiquotations:
  1012 
  1013   @{term_type term} prints a term with its type annotated;
  1014 
  1015   @{typeof term} prints the type of a term;
  1016 
  1017   @{const const} is the same as @{term const}, but checks that the
  1018   argument is a known logical constant;
  1019 
  1020   @{term_style style term} and @{thm_style style thm} print a term or
  1021   theorem applying a "style" to it
  1022 
  1023   @{ML text}
  1024 
  1025 Predefined styles are 'lhs' and 'rhs' printing the lhs/rhs of
  1026 definitions, equations, inequations etc., 'concl' printing only the
  1027 conclusion of a meta-logical statement theorem, and 'prem1' .. 'prem19'
  1028 to print the specified premise.  TermStyle.add_style provides an ML
  1029 interface for introducing further styles.  See also the "LaTeX Sugar"
  1030 document practical applications.  The ML antiquotation prints
  1031 type-checked ML expressions verbatim.
  1032 
  1033 * Markup commands 'chapter', 'section', 'subsection', 'subsubsection',
  1034 and 'text' support optional locale specification '(in loc)', which
  1035 specifies the default context for interpreting antiquotations.  For
  1036 example: 'text (in lattice) {* @{thm inf_assoc}*}'.
  1037 
  1038 * Option 'locale=NAME' of antiquotations specifies an alternative
  1039 context interpreting the subsequent argument.  For example: @{thm
  1040 [locale=lattice] inf_assoc}.
  1041 
  1042 * Proper output of proof terms (@{prf ...} and @{full_prf ...}) within
  1043 a proof context.
  1044 
  1045 * Proper output of antiquotations for theory commands involving a
  1046 proof context (such as 'locale' or 'theorem (in loc) ...').
  1047 
  1048 * Delimiters of outer tokens (string etc.) now produce separate LaTeX
  1049 macros (\isachardoublequoteopen, isachardoublequoteclose etc.).
  1050 
  1051 * isatool usedir: new option -C (default true) controls whether option
  1052 -D should include a copy of the original document directory; -C false
  1053 prevents unwanted effects such as copying of administrative CVS data.
  1054 
  1055 
  1056 *** Pure ***
  1057 
  1058 * Considerably improved version of 'constdefs' command.  Now performs
  1059 automatic type-inference of declared constants; additional support for
  1060 local structure declarations (cf. locales and HOL records), see also
  1061 isar-ref manual.  Potential INCOMPATIBILITY: need to observe strictly
  1062 sequential dependencies of definitions within a single 'constdefs'
  1063 section; moreover, the declared name needs to be an identifier.  If
  1064 all fails, consider to fall back on 'consts' and 'defs' separately.
  1065 
  1066 * Improved indexed syntax and implicit structures.  First of all,
  1067 indexed syntax provides a notational device for subscripted
  1068 application, using the new syntax \<^bsub>term\<^esub> for arbitrary
  1069 expressions.  Secondly, in a local context with structure
  1070 declarations, number indexes \<^sub>n or the empty index (default
  1071 number 1) refer to a certain fixed variable implicitly; option
  1072 show_structs controls printing of implicit structures.  Typical
  1073 applications of these concepts involve record types and locales.
  1074 
  1075 * New command 'no_syntax' removes grammar declarations (and
  1076 translations) resulting from the given syntax specification, which is
  1077 interpreted in the same manner as for the 'syntax' command.
  1078 
  1079 * 'Advanced' translation functions (parse_translation etc.) may depend
  1080 on the signature of the theory context being presently used for
  1081 parsing/printing, see also isar-ref manual.
  1082 
  1083 * Improved 'oracle' command provides a type-safe interface to turn an
  1084 ML expression of type theory -> T -> term into a primitive rule of
  1085 type theory -> T -> thm (i.e. the functionality of Thm.invoke_oracle
  1086 is already included here); see also FOL/ex/IffExample.thy;
  1087 INCOMPATIBILITY.
  1088 
  1089 * axclass: name space prefix for class "c" is now "c_class" (was "c"
  1090 before); "cI" is no longer bound, use "c.intro" instead.
  1091 INCOMPATIBILITY.  This change avoids clashes of fact bindings for
  1092 axclasses vs. locales.
  1093 
  1094 * Improved internal renaming of symbolic identifiers -- attach primes
  1095 instead of base 26 numbers.
  1096 
  1097 * New flag show_question_marks controls printing of leading question
  1098 marks in schematic variable names.
  1099 
  1100 * In schematic variable names, *any* symbol following \<^isub> or
  1101 \<^isup> is now treated as part of the base name.  For example, the
  1102 following works without printing of awkward ".0" indexes:
  1103 
  1104   lemma "x\<^isub>1 = x\<^isub>2 ==> x\<^isub>2 = x\<^isub>1"
  1105     by simp
  1106 
  1107 * Inner syntax includes (*(*nested*) comments*).
  1108 
  1109 * Pretty printer now supports unbreakable blocks, specified in mixfix
  1110 annotations as "(00...)".
  1111 
  1112 * Clear separation of logical types and nonterminals, where the latter
  1113 may only occur in 'syntax' specifications or type abbreviations.
  1114 Before that distinction was only partially implemented via type class
  1115 "logic" vs. "{}".  Potential INCOMPATIBILITY in rare cases of improper
  1116 use of 'types'/'consts' instead of 'nonterminals'/'syntax'.  Some very
  1117 exotic syntax specifications may require further adaption
  1118 (e.g. Cube/Cube.thy).
  1119 
  1120 * Removed obsolete type class "logic", use the top sort {} instead.
  1121 Note that non-logical types should be declared as 'nonterminals'
  1122 rather than 'types'.  INCOMPATIBILITY for new object-logic
  1123 specifications.
  1124 
  1125 * Attributes 'induct' and 'cases': type or set names may now be
  1126 locally fixed variables as well.
  1127 
  1128 * Simplifier: can now control the depth to which conditional rewriting
  1129 is traced via the PG menu Isabelle -> Settings -> Trace Simp Depth
  1130 Limit.
  1131 
  1132 * Simplifier: simplification procedures may now take the current
  1133 simpset into account (cf. Simplifier.simproc(_i) / mk_simproc
  1134 interface), which is very useful for calling the Simplifier
  1135 recursively.  Minor INCOMPATIBILITY: the 'prems' argument of simprocs
  1136 is gone -- use prems_of_ss on the simpset instead.  Moreover, the
  1137 low-level mk_simproc no longer applies Logic.varify internally, to
  1138 allow for use in a context of fixed variables.
  1139 
  1140 * thin_tac now works even if the assumption being deleted contains !!
  1141 or ==>.  More generally, erule now works even if the major premise of
  1142 the elimination rule contains !! or ==>.
  1143 
  1144 * Method 'rules' has been renamed to 'iprover'. INCOMPATIBILITY.
  1145 
  1146 * Reorganized bootstrapping of the Pure theories; CPure is now derived
  1147 from Pure, which contains all common declarations already.  Both
  1148 theories are defined via plain Isabelle/Isar .thy files.
  1149 INCOMPATIBILITY: elements of CPure (such as the CPure.intro /
  1150 CPure.elim / CPure.dest attributes) now appear in the Pure name space;
  1151 use isatool fixcpure to adapt your theory and ML sources.
  1152 
  1153 * New syntax 'name(i-j, i-, i, ...)' for referring to specific
  1154 selections of theorems in named facts via index ranges.
  1155 
  1156 * 'print_theorems': in theory mode, really print the difference
  1157 wrt. the last state (works for interactive theory development only),
  1158 in proof mode print all local facts (cf. 'print_facts');
  1159 
  1160 * 'hide': option '(open)' hides only base names.
  1161 
  1162 * More efficient treatment of intermediate checkpoints in interactive
  1163 theory development.
  1164 
  1165 * Code generator is now invoked via code_module (incremental code
  1166 generation) and code_library (modular code generation, ML structures
  1167 for each theory).  INCOMPATIBILITY: new keywords 'file' and 'contains'
  1168 must be quoted when used as identifiers.
  1169 
  1170 * New 'value' command for reading, evaluating and printing terms using
  1171 the code generator.  INCOMPATIBILITY: command keyword 'value' must be
  1172 quoted when used as identifier.
  1173 
  1174 
  1175 *** Locales ***
  1176 
  1177 * New commands for the interpretation of locale expressions in
  1178 theories (1), locales (2) and proof contexts (3).  These generate
  1179 proof obligations from the expression specification.  After the
  1180 obligations have been discharged, theorems of the expression are added
  1181 to the theory, target locale or proof context.  The synopsis of the
  1182 commands is a follows:
  1183 
  1184   (1) interpretation expr inst
  1185   (2) interpretation target < expr
  1186   (3) interpret expr inst
  1187 
  1188 Interpretation in theories and proof contexts require a parameter
  1189 instantiation of terms from the current context.  This is applied to
  1190 specifications and theorems of the interpreted expression.
  1191 Interpretation in locales only permits parameter renaming through the
  1192 locale expression.  Interpretation is smart in that interpretations
  1193 that are active already do not occur in proof obligations, neither are
  1194 instantiated theorems stored in duplicate.  Use 'print_interps' to
  1195 inspect active interpretations of a particular locale.  For details,
  1196 see the Isar Reference manual.  Examples can be found in
  1197 HOL/Finite_Set.thy and HOL/Algebra/UnivPoly.thy.
  1198 
  1199 INCOMPATIBILITY: former 'instantiate' has been withdrawn, use
  1200 'interpret' instead.
  1201 
  1202 * New context element 'constrains' for adding type constraints to
  1203 parameters.
  1204 
  1205 * Context expressions: renaming of parameters with syntax
  1206 redeclaration.
  1207 
  1208 * Locale declaration: 'includes' disallowed.
  1209 
  1210 * Proper static binding of attribute syntax -- i.e. types / terms /
  1211 facts mentioned as arguments are always those of the locale definition
  1212 context, independently of the context of later invocations.  Moreover,
  1213 locale operations (renaming and type / term instantiation) are applied
  1214 to attribute arguments as expected.
  1215 
  1216 INCOMPATIBILITY of the ML interface: always pass Attrib.src instead of
  1217 actual attributes; rare situations may require Attrib.attribute to
  1218 embed those attributes into Attrib.src that lack concrete syntax.
  1219 Attribute implementations need to cooperate properly with the static
  1220 binding mechanism.  Basic parsers Args.XXX_typ/term/prop and
  1221 Attrib.XXX_thm etc. already do the right thing without further
  1222 intervention.  Only unusual applications -- such as "where" or "of"
  1223 (cf. src/Pure/Isar/attrib.ML), which process arguments depending both
  1224 on the context and the facts involved -- may have to assign parsed
  1225 values to argument tokens explicitly.
  1226 
  1227 * Changed parameter management in theorem generation for long goal
  1228 statements with 'includes'.  INCOMPATIBILITY: produces a different
  1229 theorem statement in rare situations.
  1230 
  1231 * Locale inspection command 'print_locale' omits notes elements.  Use
  1232 'print_locale!' to have them included in the output.
  1233 
  1234 
  1235 *** Provers ***
  1236 
  1237 * Provers/hypsubst.ML: improved version of the subst method, for
  1238 single-step rewriting: it now works in bound variable contexts. New is
  1239 'subst (asm)', for rewriting an assumption.  INCOMPATIBILITY: may
  1240 rewrite a different subterm than the original subst method, which is
  1241 still available as 'simplesubst'.
  1242 
  1243 * Provers/quasi.ML: new transitivity reasoners for transitivity only
  1244 and quasi orders.
  1245 
  1246 * Provers/trancl.ML: new transitivity reasoner for transitive and
  1247 reflexive-transitive closure of relations.
  1248 
  1249 * Provers/blast.ML: new reference depth_limit to make blast's depth
  1250 limit (previously hard-coded with a value of 20) user-definable.
  1251 
  1252 * Provers/simplifier.ML has been moved to Pure, where Simplifier.setup
  1253 is peformed already.  Object-logics merely need to finish their
  1254 initial simpset configuration as before.  INCOMPATIBILITY.
  1255 
  1256 
  1257 *** HOL ***
  1258 
  1259 * Symbolic syntax of Hilbert Choice Operator is now as follows:
  1260 
  1261   syntax (epsilon)
  1262     "_Eps" :: "[pttrn, bool] => 'a"    ("(3\<some>_./ _)" [0, 10] 10)
  1263 
  1264 The symbol \<some> is displayed as the alternative epsilon of LaTeX
  1265 and x-symbol; use option '-m epsilon' to get it actually printed.
  1266 Moreover, the mathematically important symbolic identifier \<epsilon>
  1267 becomes available as variable, constant etc.  INCOMPATIBILITY,
  1268 
  1269 * "x > y" abbreviates "y < x" and "x >= y" abbreviates "y <= x".
  1270 Similarly for all quantifiers: "ALL x > y" etc.  The x-symbol for >=
  1271 is \<ge>. New transitivity rules have been added to HOL/Orderings.thy to
  1272 support corresponding Isar calculations.
  1273 
  1274 * "{x:A. P}" abbreviates "{x. x:A & P}", and similarly for "\<in>"
  1275 instead of ":".
  1276 
  1277 * theory SetInterval: changed the syntax for open intervals:
  1278 
  1279   Old       New
  1280   {..n(}    {..<n}
  1281   {)n..}    {n<..}
  1282   {m..n(}   {m..<n}
  1283   {)m..n}   {m<..n}
  1284   {)m..n(}  {m<..<n}
  1285 
  1286 The old syntax is still supported but will disappear in the next
  1287 release.  For conversion use the following Emacs search and replace
  1288 patterns (these are not perfect but work quite well):
  1289 
  1290   {)\([^\.]*\)\.\.  ->  {\1<\.\.}
  1291   \.\.\([^(}]*\)(}  ->  \.\.<\1}
  1292 
  1293 * Theory Commutative_Ring (in Library): method comm_ring for proving
  1294 equalities in commutative rings; method 'algebra' provides a generic
  1295 interface.
  1296 
  1297 * Theory Finite_Set: changed the syntax for 'setsum', summation over
  1298 finite sets: "setsum (%x. e) A", which used to be "\<Sum>x:A. e", is
  1299 now either "SUM x:A. e" or "\<Sum>x \<in> A. e". The bound variable can
  1300 be a tuple pattern.
  1301 
  1302 Some new syntax forms are available:
  1303 
  1304   "\<Sum>x | P. e"      for     "setsum (%x. e) {x. P}"
  1305   "\<Sum>x = a..b. e"   for     "setsum (%x. e) {a..b}"
  1306   "\<Sum>x = a..<b. e"  for     "setsum (%x. e) {a..<b}"
  1307   "\<Sum>x < k. e"      for     "setsum (%x. e) {..<k}"
  1308 
  1309 The latter form "\<Sum>x < k. e" used to be based on a separate
  1310 function "Summation", which has been discontinued.
  1311 
  1312 * theory Finite_Set: in structured induction proofs, the insert case
  1313 is now 'case (insert x F)' instead of the old counterintuitive 'case
  1314 (insert F x)'.
  1315 
  1316 * The 'refute' command has been extended to support a much larger
  1317 fragment of HOL, including axiomatic type classes, constdefs and
  1318 typedefs, inductive datatypes and recursion.
  1319 
  1320 * New tactics 'sat' and 'satx' to prove propositional tautologies.
  1321 Requires zChaff with proof generation to be installed.  See
  1322 HOL/ex/SAT_Examples.thy for examples.
  1323 
  1324 * Datatype induction via method 'induct' now preserves the name of the
  1325 induction variable. For example, when proving P(xs::'a list) by
  1326 induction on xs, the induction step is now P(xs) ==> P(a#xs) rather
  1327 than P(list) ==> P(a#list) as previously.  Potential INCOMPATIBILITY
  1328 in unstructured proof scripts.
  1329 
  1330 * Reworked implementation of records.  Improved scalability for
  1331 records with many fields, avoiding performance problems for type
  1332 inference. Records are no longer composed of nested field types, but
  1333 of nested extension types. Therefore the record type only grows linear
  1334 in the number of extensions and not in the number of fields.  The
  1335 top-level (users) view on records is preserved.  Potential
  1336 INCOMPATIBILITY only in strange cases, where the theory depends on the
  1337 old record representation. The type generated for a record is called
  1338 <record_name>_ext_type.
  1339 
  1340 Flag record_quick_and_dirty_sensitive can be enabled to skip the
  1341 proofs triggered by a record definition or a simproc (if
  1342 quick_and_dirty is enabled).  Definitions of large records can take
  1343 quite long.
  1344 
  1345 New simproc record_upd_simproc for simplification of multiple record
  1346 updates enabled by default.  Moreover, trivial updates are also
  1347 removed: r(|x := x r|) = r.  INCOMPATIBILITY: old proofs break
  1348 occasionally, since simplification is more powerful by default.
  1349 
  1350 * typedef: proper support for polymorphic sets, which contain extra
  1351 type-variables in the term.
  1352 
  1353 * Simplifier: automatically reasons about transitivity chains
  1354 involving "trancl" (r^+) and "rtrancl" (r^*) by setting up tactics
  1355 provided by Provers/trancl.ML as additional solvers.  INCOMPATIBILITY:
  1356 old proofs break occasionally as simplification may now solve more
  1357 goals than previously.
  1358 
  1359 * Simplifier: converts x <= y into x = y if assumption y <= x is
  1360 present.  Works for all partial orders (class "order"), in particular
  1361 numbers and sets.  For linear orders (e.g. numbers) it treats ~ x < y
  1362 just like y <= x.
  1363 
  1364 * Simplifier: new simproc for "let x = a in f x".  If a is a free or
  1365 bound variable or a constant then the let is unfolded.  Otherwise
  1366 first a is simplified to b, and then f b is simplified to g. If
  1367 possible we abstract b from g arriving at "let x = b in h x",
  1368 otherwise we unfold the let and arrive at g.  The simproc can be
  1369 enabled/disabled by the reference use_let_simproc.  Potential
  1370 INCOMPATIBILITY since simplification is more powerful by default.
  1371 
  1372 * Classical reasoning: the meson method now accepts theorems as arguments.
  1373 
  1374 * Prover support: pre-release of the Isabelle-ATP linkup, which runs background
  1375 jobs to provide advice on the provability of subgoals.
  1376 
  1377 * Theory OrderedGroup and Ring_and_Field: various additions and
  1378 improvements to faciliate calculations involving equalities and
  1379 inequalities.
  1380 
  1381 The following theorems have been eliminated or modified
  1382 (INCOMPATIBILITY):
  1383 
  1384   abs_eq             now named abs_of_nonneg
  1385   abs_of_ge_0        now named abs_of_nonneg
  1386   abs_minus_eq       now named abs_of_nonpos
  1387   imp_abs_id         now named abs_of_nonneg
  1388   imp_abs_neg_id     now named abs_of_nonpos
  1389   mult_pos           now named mult_pos_pos
  1390   mult_pos_le        now named mult_nonneg_nonneg
  1391   mult_pos_neg_le    now named mult_nonneg_nonpos
  1392   mult_pos_neg2_le   now named mult_nonneg_nonpos2
  1393   mult_neg           now named mult_neg_neg
  1394   mult_neg_le        now named mult_nonpos_nonpos
  1395 
  1396 * Theory Parity: added rules for simplifying exponents.
  1397 
  1398 * Theory List:
  1399 
  1400 The following theorems have been eliminated or modified
  1401 (INCOMPATIBILITY):
  1402 
  1403   list_all_Nil       now named list_all.simps(1)
  1404   list_all_Cons      now named list_all.simps(2)
  1405   list_all_conv      now named list_all_iff
  1406   set_mem_eq         now named mem_iff
  1407 
  1408 * Theories SetsAndFunctions and BigO (see HOL/Library) support
  1409 asymptotic "big O" calculations.  See the notes in BigO.thy.
  1410 
  1411 
  1412 *** HOL-Complex ***
  1413 
  1414 * Theory RealDef: better support for embedding natural numbers and
  1415 integers in the reals.
  1416 
  1417 The following theorems have been eliminated or modified
  1418 (INCOMPATIBILITY):
  1419 
  1420   exp_ge_add_one_self  now requires no hypotheses
  1421   real_of_int_add      reversed direction of equality (use [symmetric])
  1422   real_of_int_minus    reversed direction of equality (use [symmetric])
  1423   real_of_int_diff     reversed direction of equality (use [symmetric])
  1424   real_of_int_mult     reversed direction of equality (use [symmetric])
  1425 
  1426 * Theory RComplete: expanded support for floor and ceiling functions.
  1427 
  1428 * Theory Ln is new, with properties of the natural logarithm
  1429 
  1430 * Hyperreal: There is a new type constructor "star" for making
  1431 nonstandard types.  The old type names are now type synonyms:
  1432 
  1433   hypreal = real star
  1434   hypnat = nat star
  1435   hcomplex = complex star
  1436 
  1437 * Hyperreal: Many groups of similarly-defined constants have been
  1438 replaced by polymorphic versions (INCOMPATIBILITY):
  1439 
  1440   star_of <-- hypreal_of_real, hypnat_of_nat, hcomplex_of_complex
  1441 
  1442   starset      <-- starsetNat, starsetC
  1443   *s*          <-- *sNat*, *sc*
  1444   starset_n    <-- starsetNat_n, starsetC_n
  1445   *sn*         <-- *sNatn*, *scn*
  1446   InternalSets <-- InternalNatSets, InternalCSets
  1447 
  1448   starfun      <-- starfun{Nat,Nat2,C,RC,CR}
  1449   *f*          <-- *fNat*, *fNat2*, *fc*, *fRc*, *fcR*
  1450   starfun_n    <-- starfun{Nat,Nat2,C,RC,CR}_n
  1451   *fn*         <-- *fNatn*, *fNat2n*, *fcn*, *fRcn*, *fcRn*
  1452   InternalFuns <-- InternalNatFuns, InternalNatFuns2, Internal{C,RC,CR}Funs
  1453 
  1454 * Hyperreal: Many type-specific theorems have been removed in favor of
  1455 theorems specific to various axiomatic type classes (INCOMPATIBILITY):
  1456 
  1457   add_commute <-- {hypreal,hypnat,hcomplex}_add_commute
  1458   add_assoc   <-- {hypreal,hypnat,hcomplex}_add_assocs
  1459   OrderedGroup.add_0 <-- {hypreal,hypnat,hcomplex}_add_zero_left
  1460   OrderedGroup.add_0_right <-- {hypreal,hcomplex}_add_zero_right
  1461   right_minus <-- hypreal_add_minus
  1462   left_minus <-- {hypreal,hcomplex}_add_minus_left
  1463   mult_commute <-- {hypreal,hypnat,hcomplex}_mult_commute
  1464   mult_assoc <-- {hypreal,hypnat,hcomplex}_mult_assoc
  1465   mult_1_left <-- {hypreal,hypnat}_mult_1, hcomplex_mult_one_left
  1466   mult_1_right <-- hcomplex_mult_one_right
  1467   mult_zero_left <-- hcomplex_mult_zero_left
  1468   left_distrib <-- {hypreal,hypnat,hcomplex}_add_mult_distrib
  1469   right_distrib <-- hypnat_add_mult_distrib2
  1470   zero_neq_one <-- {hypreal,hypnat,hcomplex}_zero_not_eq_one
  1471   right_inverse <-- hypreal_mult_inverse
  1472   left_inverse <-- hypreal_mult_inverse_left, hcomplex_mult_inv_left
  1473   order_refl <-- {hypreal,hypnat}_le_refl
  1474   order_trans <-- {hypreal,hypnat}_le_trans
  1475   order_antisym <-- {hypreal,hypnat}_le_anti_sym
  1476   order_less_le <-- {hypreal,hypnat}_less_le
  1477   linorder_linear <-- {hypreal,hypnat}_le_linear
  1478   add_left_mono <-- {hypreal,hypnat}_add_left_mono
  1479   mult_strict_left_mono <-- {hypreal,hypnat}_mult_less_mono2
  1480   add_nonneg_nonneg <-- hypreal_le_add_order
  1481 
  1482 * Hyperreal: Separate theorems having to do with type-specific
  1483 versions of constants have been merged into theorems that apply to the
  1484 new polymorphic constants (INCOMPATIBILITY):
  1485 
  1486   STAR_UNIV_set <-- {STAR_real,NatStar_real,STARC_complex}_set
  1487   STAR_empty_set <-- {STAR,NatStar,STARC}_empty_set
  1488   STAR_Un <-- {STAR,NatStar,STARC}_Un
  1489   STAR_Int <-- {STAR,NatStar,STARC}_Int
  1490   STAR_Compl <-- {STAR,NatStar,STARC}_Compl
  1491   STAR_subset <-- {STAR,NatStar,STARC}_subset
  1492   STAR_mem <-- {STAR,NatStar,STARC}_mem
  1493   STAR_mem_Compl <-- {STAR,STARC}_mem_Compl
  1494   STAR_diff <-- {STAR,STARC}_diff
  1495   STAR_star_of_image_subset <-- {STAR_hypreal_of_real, NatStar_hypreal_of_real,
  1496     STARC_hcomplex_of_complex}_image_subset
  1497   starset_n_Un <-- starset{Nat,C}_n_Un
  1498   starset_n_Int <-- starset{Nat,C}_n_Int
  1499   starset_n_Compl <-- starset{Nat,C}_n_Compl
  1500   starset_n_diff <-- starset{Nat,C}_n_diff
  1501   InternalSets_Un <-- Internal{Nat,C}Sets_Un
  1502   InternalSets_Int <-- Internal{Nat,C}Sets_Int
  1503   InternalSets_Compl <-- Internal{Nat,C}Sets_Compl
  1504   InternalSets_diff <-- Internal{Nat,C}Sets_diff
  1505   InternalSets_UNIV_diff <-- Internal{Nat,C}Sets_UNIV_diff
  1506   InternalSets_starset_n <-- Internal{Nat,C}Sets_starset{Nat,C}_n
  1507   starset_starset_n_eq <-- starset{Nat,C}_starset{Nat,C}_n_eq
  1508   starset_n_starset <-- starset{Nat,C}_n_starset{Nat,C}
  1509   starfun_n_starfun <-- starfun{Nat,Nat2,C,RC,CR}_n_starfun{Nat,Nat2,C,RC,CR}
  1510   starfun <-- starfun{Nat,Nat2,C,RC,CR}
  1511   starfun_mult <-- starfun{Nat,Nat2,C,RC,CR}_mult
  1512   starfun_add <-- starfun{Nat,Nat2,C,RC,CR}_add
  1513   starfun_minus <-- starfun{Nat,Nat2,C,RC,CR}_minus
  1514   starfun_diff <-- starfun{C,RC,CR}_diff
  1515   starfun_o <-- starfun{NatNat2,Nat2,_stafunNat,C,C_starfunRC,_starfunCR}_o
  1516   starfun_o2 <-- starfun{NatNat2,_stafunNat,C,C_starfunRC,_starfunCR}_o2
  1517   starfun_const_fun <-- starfun{Nat,Nat2,C,RC,CR}_const_fun
  1518   starfun_inverse <-- starfun{Nat,C,RC,CR}_inverse
  1519   starfun_eq <-- starfun{Nat,Nat2,C,RC,CR}_eq
  1520   starfun_eq_iff <-- starfun{C,RC,CR}_eq_iff
  1521   starfun_Id <-- starfunC_Id
  1522   starfun_approx <-- starfun{Nat,CR}_approx
  1523   starfun_capprox <-- starfun{C,RC}_capprox
  1524   starfun_abs <-- starfunNat_rabs
  1525   starfun_lambda_cancel <-- starfun{C,CR,RC}_lambda_cancel
  1526   starfun_lambda_cancel2 <-- starfun{C,CR,RC}_lambda_cancel2
  1527   starfun_mult_HFinite_approx <-- starfunCR_mult_HFinite_capprox
  1528   starfun_mult_CFinite_capprox <-- starfun{C,RC}_mult_CFinite_capprox
  1529   starfun_add_capprox <-- starfun{C,RC}_add_capprox
  1530   starfun_add_approx <-- starfunCR_add_approx
  1531   starfun_inverse_inverse <-- starfunC_inverse_inverse
  1532   starfun_divide <-- starfun{C,CR,RC}_divide
  1533   starfun_n <-- starfun{Nat,C}_n
  1534   starfun_n_mult <-- starfun{Nat,C}_n_mult
  1535   starfun_n_add <-- starfun{Nat,C}_n_add
  1536   starfun_n_add_minus <-- starfunNat_n_add_minus
  1537   starfun_n_const_fun <-- starfun{Nat,C}_n_const_fun
  1538   starfun_n_minus <-- starfun{Nat,C}_n_minus
  1539   starfun_n_eq <-- starfun{Nat,C}_n_eq
  1540 
  1541   star_n_add <-- {hypreal,hypnat,hcomplex}_add
  1542   star_n_minus <-- {hypreal,hcomplex}_minus
  1543   star_n_diff <-- {hypreal,hcomplex}_diff
  1544   star_n_mult <-- {hypreal,hcomplex}_mult
  1545   star_n_inverse <-- {hypreal,hcomplex}_inverse
  1546   star_n_le <-- {hypreal,hypnat}_le
  1547   star_n_less <-- {hypreal,hypnat}_less
  1548   star_n_zero_num <-- {hypreal,hypnat,hcomplex}_zero_num
  1549   star_n_one_num <-- {hypreal,hypnat,hcomplex}_one_num
  1550   star_n_abs <-- hypreal_hrabs
  1551   star_n_divide <-- hcomplex_divide
  1552 
  1553   star_of_add <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_add
  1554   star_of_minus <-- {hypreal_of_real,hcomplex_of_complex}_minus
  1555   star_of_diff <-- hypreal_of_real_diff
  1556   star_of_mult <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_mult
  1557   star_of_one <-- {hypreal_of_real,hcomplex_of_complex}_one
  1558   star_of_zero <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_zero
  1559   star_of_le <-- {hypreal_of_real,hypnat_of_nat}_le_iff
  1560   star_of_less <-- {hypreal_of_real,hypnat_of_nat}_less_iff
  1561   star_of_eq <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_eq_iff
  1562   star_of_inverse <-- {hypreal_of_real,hcomplex_of_complex}_inverse
  1563   star_of_divide <-- {hypreal_of_real,hcomplex_of_complex}_divide
  1564   star_of_of_nat <-- {hypreal_of_real,hcomplex_of_complex}_of_nat
  1565   star_of_of_int <-- {hypreal_of_real,hcomplex_of_complex}_of_int
  1566   star_of_number_of <-- {hypreal,hcomplex}_number_of
  1567   star_of_number_less <-- number_of_less_hypreal_of_real_iff
  1568   star_of_number_le <-- number_of_le_hypreal_of_real_iff
  1569   star_of_eq_number <-- hypreal_of_real_eq_number_of_iff
  1570   star_of_less_number <-- hypreal_of_real_less_number_of_iff
  1571   star_of_le_number <-- hypreal_of_real_le_number_of_iff
  1572   star_of_power <-- hypreal_of_real_power
  1573   star_of_eq_0 <-- hcomplex_of_complex_zero_iff
  1574 
  1575 * Hyperreal: new method "transfer" that implements the transfer
  1576 principle of nonstandard analysis. With a subgoal that mentions
  1577 nonstandard types like "'a star", the command "apply transfer"
  1578 replaces it with an equivalent one that mentions only standard types.
  1579 To be successful, all free variables must have standard types; non-
  1580 standard variables must have explicit universal quantifiers.
  1581 
  1582 * Hyperreal: A theory of Taylor series.
  1583 
  1584 
  1585 *** HOLCF ***
  1586 
  1587 * Discontinued special version of 'constdefs' (which used to support
  1588 continuous functions) in favor of the general Pure one with full
  1589 type-inference.
  1590 
  1591 * New simplification procedure for solving continuity conditions; it
  1592 is much faster on terms with many nested lambda abstractions (cubic
  1593 instead of exponential time).
  1594 
  1595 * New syntax for domain package: selector names are now optional.
  1596 Parentheses should be omitted unless argument is lazy, for example:
  1597 
  1598   domain 'a stream = cons "'a" (lazy "'a stream")
  1599 
  1600 * New command 'fixrec' for defining recursive functions with pattern
  1601 matching; defining multiple functions with mutual recursion is also
  1602 supported.  Patterns may include the constants cpair, spair, up, sinl,
  1603 sinr, or any data constructor defined by the domain package. The given
  1604 equations are proven as rewrite rules. See HOLCF/ex/Fixrec_ex.thy for
  1605 syntax and examples.
  1606 
  1607 * New commands 'cpodef' and 'pcpodef' for defining predicate subtypes
  1608 of cpo and pcpo types. Syntax is exactly like the 'typedef' command,
  1609 but the proof obligation additionally includes an admissibility
  1610 requirement. The packages generate instances of class cpo or pcpo,
  1611 with continuity and strictness theorems for Rep and Abs.
  1612 
  1613 * HOLCF: Many theorems have been renamed according to a more standard naming
  1614 scheme (INCOMPATIBILITY):
  1615 
  1616   foo_inject:  "foo$x = foo$y ==> x = y"
  1617   foo_eq:      "(foo$x = foo$y) = (x = y)"
  1618   foo_less:    "(foo$x << foo$y) = (x << y)"
  1619   foo_strict:  "foo$UU = UU"
  1620   foo_defined: "... ==> foo$x ~= UU"
  1621   foo_defined_iff: "(foo$x = UU) = (x = UU)"
  1622 
  1623 
  1624 *** ZF ***
  1625 
  1626 * ZF/ex: theories Group and Ring provide examples in abstract algebra,
  1627 including the First Isomorphism Theorem (on quotienting by the kernel
  1628 of a homomorphism).
  1629 
  1630 * ZF/Simplifier: install second copy of type solver that actually
  1631 makes use of TC rules declared to Isar proof contexts (or locales);
  1632 the old version is still required for ML proof scripts.
  1633 
  1634 
  1635 *** Cube ***
  1636 
  1637 * Converted to Isar theory format; use locales instead of axiomatic
  1638 theories.
  1639 
  1640 
  1641 *** ML ***
  1642 
  1643 * Pure/library.ML: added ##>, ##>>, #>> -- higher-order counterparts
  1644 for ||>, ||>>, |>>,
  1645 
  1646 * Pure/library.ML no longer defines its own option datatype, but uses
  1647 that of the SML basis, which has constructors NONE and SOME instead of
  1648 None and Some, as well as exception Option.Option instead of OPTION.
  1649 The functions the, if_none, is_some, is_none have been adapted
  1650 accordingly, while Option.map replaces apsome.
  1651 
  1652 * Pure/library.ML: the exception LIST has been given up in favour of
  1653 the standard exceptions Empty and Subscript, as well as
  1654 Library.UnequalLengths.  Function like Library.hd and Library.tl are
  1655 superceded by the standard hd and tl functions etc.
  1656 
  1657 A number of basic list functions are no longer exported to the ML
  1658 toplevel, as they are variants of predefined functions.  The following
  1659 suggests how one can translate existing code:
  1660 
  1661     rev_append xs ys = List.revAppend (xs, ys)
  1662     nth_elem (i, xs) = List.nth (xs, i)
  1663     last_elem xs = List.last xs
  1664     flat xss = List.concat xss
  1665     seq fs = List.app fs
  1666     partition P xs = List.partition P xs
  1667     mapfilter f xs = List.mapPartial f xs
  1668 
  1669 * Pure/library.ML: several combinators for linear functional
  1670 transformations, notably reverse application and composition:
  1671 
  1672   x |> f                f #> g
  1673   (x, y) |-> f          f #-> g
  1674 
  1675 * Pure/library.ML: introduced/changed precedence of infix operators:
  1676 
  1677   infix 1 |> |-> ||> ||>> |>> |>>> #> #->;
  1678   infix 2 ?;
  1679   infix 3 o oo ooo oooo;
  1680   infix 4 ~~ upto downto;
  1681 
  1682 Maybe INCOMPATIBILITY when any of those is used in conjunction with other
  1683 infix operators.
  1684 
  1685 * Pure/library.ML: natural list combinators fold, fold_rev, and
  1686 fold_map support linear functional transformations and nesting.  For
  1687 example:
  1688 
  1689   fold f [x1, ..., xN] y =
  1690     y |> f x1 |> ... |> f xN
  1691 
  1692   (fold o fold) f [xs1, ..., xsN] y =
  1693     y |> fold f xs1 |> ... |> fold f xsN
  1694 
  1695   fold f [x1, ..., xN] =
  1696     f x1 #> ... #> f xN
  1697 
  1698   (fold o fold) f [xs1, ..., xsN] =
  1699     fold f xs1 #> ... #> fold f xsN
  1700 
  1701 * Pure/library.ML: the following selectors on type 'a option are
  1702 available:
  1703 
  1704   the:               'a option -> 'a  (*partial*)
  1705   these:             'a option -> 'a  where 'a = 'b list
  1706   the_default: 'a -> 'a option -> 'a
  1707   the_list:          'a option -> 'a list
  1708 
  1709 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1710 basic operations for association lists, following natural argument
  1711 order; moreover the explicit equality predicate passed here avoids
  1712 potentially expensive polymorphic runtime equality checks.
  1713 The old functions may be expressed as follows:
  1714 
  1715   assoc = uncurry (AList.lookup (op =))
  1716   assocs = these oo AList.lookup (op =)
  1717   overwrite = uncurry (AList.update (op =)) o swap
  1718 
  1719 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  1720 
  1721   val make: ('a -> 'b) -> 'a list -> ('a * 'b) list
  1722   val find: ('a * 'b -> bool) -> ('c * 'b) list -> 'a -> 'c list
  1723 
  1724 replacing make_keylist and keyfilter (occassionally used)
  1725 Naive rewrites:
  1726 
  1727   make_keylist = AList.make
  1728   keyfilter = AList.find (op =)
  1729 
  1730 * eq_fst and eq_snd now take explicit equality parameter, thus
  1731   avoiding eqtypes. Naive rewrites:
  1732 
  1733     eq_fst = eq_fst (op =)
  1734     eq_snd = eq_snd (op =)
  1735 
  1736 * Removed deprecated apl and apr (rarely used).
  1737   Naive rewrites:
  1738 
  1739     apl (n, op) =>>= curry op n
  1740     apr (op, m) =>>= fn n => op (n, m)
  1741 
  1742 * Pure/General: structure OrdList (cf. Pure/General/ord_list.ML)
  1743 provides a reasonably efficient light-weight implementation of sets as
  1744 lists.
  1745 
  1746 * Pure/General: generic tables (cf. Pure/General/table.ML) provide a
  1747 few new operations; existing lookup and update are now curried to
  1748 follow natural argument order (for use with fold etc.);
  1749 INCOMPATIBILITY, use (uncurry Symtab.lookup) etc. as last resort.
  1750 
  1751 * Pure/General: output via the Isabelle channels of
  1752 writeln/warning/error etc. is now passed through Output.output, with a
  1753 hook for arbitrary transformations depending on the print_mode
  1754 (cf. Output.add_mode -- the first active mode that provides a output
  1755 function wins).  Already formatted output may be embedded into further
  1756 text via Output.raw; the result of Pretty.string_of/str_of and derived
  1757 functions (string_of_term/cterm/thm etc.) is already marked raw to
  1758 accommodate easy composition of diagnostic messages etc.  Programmers
  1759 rarely need to care about Output.output or Output.raw at all, with
  1760 some notable exceptions: Output.output is required when bypassing the
  1761 standard channels (writeln etc.), or in token translations to produce
  1762 properly formatted results; Output.raw is required when capturing
  1763 already output material that will eventually be presented to the user
  1764 a second time.  For the default print mode, both Output.output and
  1765 Output.raw have no effect.
  1766 
  1767 * Pure/General: Output.time_accumulator NAME creates an operator ('a
  1768 -> 'b) -> 'a -> 'b to measure runtime and count invocations; the
  1769 cumulative results are displayed at the end of a batch session.
  1770 
  1771 * Pure/General: File.sysify_path and File.quote_sysify path have been
  1772 replaced by File.platform_path and File.shell_path (with appropriate
  1773 hooks).  This provides a clean interface for unusual systems where the
  1774 internal and external process view of file names are different.
  1775 
  1776 * Pure: more efficient orders for basic syntactic entities: added
  1777 fast_string_ord, fast_indexname_ord, fast_term_ord; changed sort_ord
  1778 and typ_ord to use fast_string_ord and fast_indexname_ord (term_ord is
  1779 NOT affected); structures Symtab, Vartab, Typtab, Termtab use the fast
  1780 orders now -- potential INCOMPATIBILITY for code that depends on a
  1781 particular order for Symtab.keys, Symtab.dest, etc. (consider using
  1782 Library.sort_strings on result).
  1783 
  1784 * Pure/term.ML: combinators fold_atyps, fold_aterms, fold_term_types,
  1785 fold_types traverse types/terms from left to right, observing natural
  1786 argument order.  Supercedes previous foldl_XXX versions, add_frees,
  1787 add_vars etc. have been adapted as well: INCOMPATIBILITY.
  1788 
  1789 * Pure: name spaces have been refined, with significant changes of the
  1790 internal interfaces -- INCOMPATIBILITY.  Renamed cond_extern(_table)
  1791 to extern(_table).  The plain name entry path is superceded by a
  1792 general 'naming' context, which also includes the 'policy' to produce
  1793 a fully qualified name and external accesses of a fully qualified
  1794 name; NameSpace.extend is superceded by context dependent
  1795 Sign.declare_name.  Several theory and proof context operations modify
  1796 the naming context.  Especially note Theory.restore_naming and
  1797 ProofContext.restore_naming to get back to a sane state; note that
  1798 Theory.add_path is no longer sufficient to recover from
  1799 Theory.absolute_path in particular.
  1800 
  1801 * Pure: new flags short_names (default false) and unique_names
  1802 (default true) for controlling output of qualified names.  If
  1803 short_names is set, names are printed unqualified.  If unique_names is
  1804 reset, the name prefix is reduced to the minimum required to achieve
  1805 the original result when interning again, even if there is an overlap
  1806 with earlier declarations.
  1807 
  1808 * Pure/TheoryDataFun: change of the argument structure; 'prep_ext' is
  1809 now 'extend', and 'merge' gets an additional Pretty.pp argument
  1810 (useful for printing error messages).  INCOMPATIBILITY.
  1811 
  1812 * Pure: major reorganization of the theory context.  Type Sign.sg and
  1813 Theory.theory are now identified, referring to the universal
  1814 Context.theory (see Pure/context.ML).  Actual signature and theory
  1815 content is managed as theory data.  The old code and interfaces were
  1816 spread over many files and structures; the new arrangement introduces
  1817 considerable INCOMPATIBILITY to gain more clarity:
  1818 
  1819   Context -- theory management operations (name, identity, inclusion,
  1820     parents, ancestors, merge, etc.), plus generic theory data;
  1821 
  1822   Sign -- logical signature and syntax operations (declaring consts,
  1823     types, etc.), plus certify/read for common entities;
  1824 
  1825   Theory -- logical theory operations (stating axioms, definitions,
  1826     oracles), plus a copy of logical signature operations (consts,
  1827     types, etc.); also a few basic management operations (Theory.copy,
  1828     Theory.merge, etc.)
  1829 
  1830 The most basic sign_of operations (Theory.sign_of, Thm.sign_of_thm
  1831 etc.) as well as the sign field in Thm.rep_thm etc. have been retained
  1832 for convenience -- they merely return the theory.
  1833 
  1834 * Pure: type Type.tsig is superceded by theory in most interfaces.
  1835 
  1836 * Pure: the Isar proof context type is already defined early in Pure
  1837 as Context.proof (note that ProofContext.context and Proof.context are
  1838 aliases, where the latter is the preferred name).  This enables other
  1839 Isabelle components to refer to that type even before Isar is present.
  1840 
  1841 * Pure/sign/theory: discontinued named name spaces (i.e. classK,
  1842 typeK, constK, axiomK, oracleK), but provide explicit operations for
  1843 any of these kinds.  For example, Sign.intern typeK is now
  1844 Sign.intern_type, Theory.hide_space Sign.typeK is now
  1845 Theory.hide_types.  Also note that former
  1846 Theory.hide_classes/types/consts are now
  1847 Theory.hide_classes_i/types_i/consts_i, while the non '_i' versions
  1848 internalize their arguments!  INCOMPATIBILITY.
  1849 
  1850 * Pure: get_thm interface (of PureThy and ProofContext) expects
  1851 datatype thmref (with constructors Name and NameSelection) instead of
  1852 plain string -- INCOMPATIBILITY;
  1853 
  1854 * Pure: cases produced by proof methods specify options, where NONE
  1855 means to remove case bindings -- INCOMPATIBILITY in
  1856 (RAW_)METHOD_CASES.
  1857 
  1858 * Pure: the following operations retrieve axioms or theorems from a
  1859 theory node or theory hierarchy, respectively:
  1860 
  1861   Theory.axioms_of: theory -> (string * term) list
  1862   Theory.all_axioms_of: theory -> (string * term) list
  1863   PureThy.thms_of: theory -> (string * thm) list
  1864   PureThy.all_thms_of: theory -> (string * thm) list
  1865 
  1866 * Pure: print_tac now outputs the goal through the trace channel.
  1867 
  1868 * Isar toplevel: improved diagnostics, mostly for Poly/ML only.
  1869 Reference Toplevel.debug (default false) controls detailed printing
  1870 and tracing of low-level exceptions; Toplevel.profiling (default 0)
  1871 controls execution profiling -- set to 1 for time and 2 for space
  1872 (both increase the runtime).
  1873 
  1874 * Isar session: The initial use of ROOT.ML is now always timed,
  1875 i.e. the log will show the actual process times, in contrast to the
  1876 elapsed wall-clock time that the outer shell wrapper produces.
  1877 
  1878 * Simplifier: improved handling of bound variables (nameless
  1879 representation, avoid allocating new strings).  Simprocs that invoke
  1880 the Simplifier recursively should use Simplifier.inherit_bounds to
  1881 avoid local name clashes.  Failure to do so produces warnings
  1882 "Simplifier: renamed bound variable ..."; set Simplifier.debug_bounds
  1883 for further details.
  1884 
  1885 * ML functions legacy_bindings and use_legacy_bindings produce ML fact
  1886 bindings for all theorems stored within a given theory; this may help
  1887 in porting non-Isar theories to Isar ones, while keeping ML proof
  1888 scripts for the time being.
  1889 
  1890 * ML operator HTML.with_charset specifies the charset begin used for
  1891 generated HTML files.  For example:
  1892 
  1893   HTML.with_charset "utf-8" use_thy "Hebrew";
  1894   HTML.with_charset "utf-8" use_thy "Chinese";
  1895 
  1896 
  1897 *** System ***
  1898 
  1899 * Allow symlinks to all proper Isabelle executables (Isabelle,
  1900 isabelle, isatool etc.).
  1901 
  1902 * ISABELLE_DOC_FORMAT setting specifies preferred document format (for
  1903 isatool doc, isatool mkdir, display_drafts etc.).
  1904 
  1905 * isatool usedir: option -f allows specification of the ML file to be
  1906 used by Isabelle; default is ROOT.ML.
  1907 
  1908 * New isatool version outputs the version identifier of the Isabelle
  1909 distribution being used.
  1910 
  1911 * HOL: new isatool dimacs2hol converts files in DIMACS CNF format
  1912 (containing Boolean satisfiability problems) into Isabelle/HOL
  1913 theories.
  1914 
  1915 
  1916 
  1917 New in Isabelle2004 (April 2004)
  1918 --------------------------------
  1919 
  1920 *** General ***
  1921 
  1922 * Provers/order.ML:  new efficient reasoner for partial and linear orders.
  1923   Replaces linorder.ML.
  1924 
  1925 * Pure: Greek letters (except small lambda, \<lambda>), as well as Gothic
  1926   (\<aa>...\<zz>\<AA>...\<ZZ>), calligraphic (\<A>...\<Z>), and Euler
  1927   (\<a>...\<z>), are now considered normal letters, and can therefore
  1928   be used anywhere where an ASCII letter (a...zA...Z) has until
  1929   now. COMPATIBILITY: This obviously changes the parsing of some
  1930   terms, especially where a symbol has been used as a binder, say
  1931   '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed
  1932   as an identifier.  Fix it by inserting a space around former
  1933   symbols.  Call 'isatool fixgreek' to try to fix parsing errors in
  1934   existing theory and ML files.
  1935 
  1936 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.
  1937 
  1938 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now
  1939   allowed in identifiers. Similar to Greek letters \<^isub> is now considered
  1940   a normal (but invisible) letter. For multiple letter subscripts repeat
  1941   \<^isub> like this: x\<^isub>1\<^isub>2.
  1942 
  1943 * Pure: There are now sub-/superscripts that can span more than one
  1944   character. Text between \<^bsub> and \<^esub> is set in subscript in
  1945   ProofGeneral and LaTeX, text between \<^bsup> and \<^esup> in
  1946   superscript. The new control characters are not identifier parts.
  1947 
  1948 * Pure: Control-symbols of the form \<^raw:...> will literally print the
  1949   content of "..." to the latex file instead of \isacntrl... . The "..."
  1950   may consist of any printable characters excluding the end bracket >.
  1951 
  1952 * Pure: Using new Isar command "finalconsts" (or the ML functions
  1953   Theory.add_finals or Theory.add_finals_i) it is now possible to
  1954   declare constants "final", which prevents their being given a definition
  1955   later.  It is useful for constants whose behaviour is fixed axiomatically
  1956   rather than definitionally, such as the meta-logic connectives.
  1957 
  1958 * Pure: 'instance' now handles general arities with general sorts
  1959   (i.e. intersections of classes),
  1960 
  1961 * Presentation: generated HTML now uses a CSS style sheet to make layout
  1962   (somewhat) independent of content. It is copied from lib/html/isabelle.css.
  1963   It can be changed to alter the colors/layout of generated pages.
  1964 
  1965 
  1966 *** Isar ***
  1967 
  1968 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,
  1969   cut_tac, subgoal_tac and thin_tac:
  1970   - Now understand static (Isar) contexts.  As a consequence, users of Isar
  1971     locales are no longer forced to write Isar proof scripts.
  1972     For details see Isar Reference Manual, paragraph 4.3.2: Further tactic
  1973     emulations.
  1974   - INCOMPATIBILITY: names of variables to be instantiated may no
  1975     longer be enclosed in quotes.  Instead, precede variable name with `?'.
  1976     This is consistent with the instantiation attribute "where".
  1977 
  1978 * Attributes "where" and "of":
  1979   - Now take type variables of instantiated theorem into account when reading
  1980     the instantiation string.  This fixes a bug that caused instantiated
  1981     theorems to have too special types in some circumstances.
  1982   - "where" permits explicit instantiations of type variables.
  1983 
  1984 * Calculation commands "moreover" and "also" no longer interfere with
  1985   current facts ("this"), admitting arbitrary combinations with "then"
  1986   and derived forms.
  1987 
  1988 * Locales:
  1989   - Goal statements involving the context element "includes" no longer
  1990     generate theorems with internal delta predicates (those ending on
  1991     "_axioms") in the premise.
  1992     Resolve particular premise with <locale>.intro to obtain old form.
  1993   - Fixed bug in type inference ("unify_frozen") that prevented mix of target
  1994     specification and "includes" elements in goal statement.
  1995   - Rule sets <locale>.intro and <locale>.axioms no longer declared as
  1996     [intro?] and [elim?] (respectively) by default.
  1997   - Experimental command for instantiation of locales in proof contexts:
  1998         instantiate <label>[<attrs>]: <loc>
  1999     Instantiates locale <loc> and adds all its theorems to the current context
  2000     taking into account their attributes.  Label and attrs are optional
  2001     modifiers, like in theorem declarations.  If present, names of
  2002     instantiated theorems are qualified with <label>, and the attributes
  2003     <attrs> are applied after any attributes these theorems might have already.
  2004       If the locale has assumptions, a chained fact of the form
  2005     "<loc> t1 ... tn" is expected from which instantiations of the parameters
  2006     are derived.  The command does not support old-style locales declared
  2007     with "locale (open)".
  2008       A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.
  2009 
  2010 * HOL: Tactic emulation methods induct_tac and case_tac understand static
  2011   (Isar) contexts.
  2012 
  2013 
  2014 *** HOL ***
  2015 
  2016 * Proof import: new image HOL4 contains the imported library from
  2017   the HOL4 system with about 2500 theorems. It is imported by
  2018   replaying proof terms produced by HOL4 in Isabelle. The HOL4 image
  2019   can be used like any other Isabelle image.  See
  2020   HOL/Import/HOL/README for more information.
  2021 
  2022 * Simplifier:
  2023   - Much improved handling of linear and partial orders.
  2024     Reasoners for linear and partial orders are set up for type classes
  2025     "linorder" and "order" respectively, and are added to the default simpset
  2026     as solvers.  This means that the simplifier can build transitivity chains
  2027     to solve goals from the assumptions.
  2028   - INCOMPATIBILITY: old proofs break occasionally.  Typically, applications
  2029     of blast or auto after simplification become unnecessary because the goal
  2030     is solved by simplification already.
  2031 
  2032 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,
  2033     all proved in axiomatic type classes for semirings, rings and fields.
  2034 
  2035 * Numerics:
  2036   - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are
  2037     now formalized using the Ring_and_Field theory mentioned above.
  2038   - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently
  2039     than before, because now they are set up once in a generic manner.
  2040   - INCOMPATIBILITY: many type-specific arithmetic laws have gone.
  2041     Look for the general versions in Ring_and_Field (and Power if they concern
  2042     exponentiation).
  2043 
  2044 * Type "rat" of the rational numbers is now available in HOL-Complex.
  2045 
  2046 * Records:
  2047   - Record types are now by default printed with their type abbreviation
  2048     instead of the list of all field types. This can be configured via
  2049     the reference "print_record_type_abbr".
  2050   - Simproc "record_upd_simproc" for simplification of multiple updates added
  2051     (not enabled by default).
  2052   - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.
  2053     EX x. x = sel r to True (not enabled by default).
  2054   - Tactic "record_split_simp_tac" to split and simplify records added.
  2055 
  2056 * 'specification' command added, allowing for definition by
  2057   specification.  There is also an 'ax_specification' command that
  2058   introduces the new constants axiomatically.
  2059 
  2060 * arith(_tac) is now able to generate counterexamples for reals as well.
  2061 
  2062 * HOL-Algebra: new locale "ring" for non-commutative rings.
  2063 
  2064 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function
  2065   definitions, thanks to Sava Krsti\'{c} and John Matthews.
  2066 
  2067 * HOL-Matrix: a first theory for matrices in HOL with an application of
  2068   matrix theory to linear programming.
  2069 
  2070 * Unions and Intersections:
  2071   The latex output syntax of UN and INT has been changed
  2072   from "\Union x \in A. B" to "\Union_{x \in A} B"
  2073   i.e. the index formulae has become a subscript.
  2074   Similarly for "\Union x. B", and for \Inter instead of \Union.
  2075 
  2076 * Unions and Intersections over Intervals:
  2077   There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is
  2078   also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"
  2079   like in normal math, and corresponding versions for < and for intersection.
  2080 
  2081 * HOL/List: Ordering "lexico" is renamed "lenlex" and the standard
  2082   lexicographic dictonary ordering has been added as "lexord".
  2083 
  2084 * ML: the legacy theory structures Int and List have been removed. They had
  2085   conflicted with ML Basis Library structures having the same names.
  2086 
  2087 * 'refute' command added to search for (finite) countermodels.  Only works
  2088   for a fragment of HOL.  The installation of an external SAT solver is
  2089   highly recommended.  See "HOL/Refute.thy" for details.
  2090 
  2091 * 'quickcheck' command: Allows to find counterexamples by evaluating
  2092   formulae under an assignment of free variables to random values.
  2093   In contrast to 'refute', it can deal with inductive datatypes,
  2094   but cannot handle quantifiers. See "HOL/ex/Quickcheck_Examples.thy"
  2095   for examples.
  2096 
  2097 
  2098 *** HOLCF ***
  2099 
  2100 * Streams now come with concatenation and are part of the HOLCF image
  2101 
  2102 
  2103 
  2104 New in Isabelle2003 (May 2003)
  2105 ------------------------------
  2106 
  2107 *** General ***
  2108 
  2109 * Provers/simplifier:
  2110 
  2111   - Completely reimplemented method simp (ML: Asm_full_simp_tac):
  2112     Assumptions are now subject to complete mutual simplification,
  2113     not just from left to right. The simplifier now preserves
  2114     the order of assumptions.
  2115 
  2116     Potential INCOMPATIBILITY:
  2117 
  2118     -- simp sometimes diverges where the old version did
  2119        not, e.g. invoking simp on the goal
  2120 
  2121         [| P (f x); y = x; f x = f y |] ==> Q
  2122 
  2123        now gives rise to the infinite reduction sequence
  2124 
  2125         P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
  2126 
  2127        Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
  2128        kind of problem.
  2129 
  2130     -- Tactics combining classical reasoner and simplification (such as auto)
  2131        are also affected by this change, because many of them rely on
  2132        simp. They may sometimes diverge as well or yield a different numbers
  2133        of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
  2134        in case of problems. Sometimes subsequent calls to the classical
  2135        reasoner will fail because a preceeding call to the simplifier too
  2136        eagerly simplified the goal, e.g. deleted redundant premises.
  2137 
  2138   - The simplifier trace now shows the names of the applied rewrite rules
  2139 
  2140   - You can limit the number of recursive invocations of the simplifier
  2141     during conditional rewriting (where the simplifie tries to solve the
  2142     conditions before applying the rewrite rule):
  2143     ML "simp_depth_limit := n"
  2144     where n is an integer. Thus you can force termination where previously
  2145     the simplifier would diverge.
  2146 
  2147   - Accepts free variables as head terms in congruence rules.  Useful in Isar.
  2148 
  2149   - No longer aborts on failed congruence proof.  Instead, the
  2150     congruence is ignored.
  2151 
  2152 * Pure: New generic framework for extracting programs from constructive
  2153   proofs. See HOL/Extraction.thy for an example instantiation, as well
  2154   as HOL/Extraction for some case studies.
  2155 
  2156 * Pure: The main goal of the proof state is no longer shown by default, only
  2157 the subgoals. This behaviour is controlled by a new flag.
  2158    PG menu: Isabelle/Isar -> Settings -> Show Main Goal
  2159 (ML: Proof.show_main_goal).
  2160 
  2161 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
  2162 rules whose conclusion matches subgoal 1:
  2163       PG menu: Isabelle/Isar -> Show me -> matching rules
  2164 The rules are ordered by how closely they match the subgoal.
  2165 In particular, rules that solve a subgoal outright are displayed first
  2166 (or rather last, the way they are printed).
  2167 (ML: ProofGeneral.print_intros())
  2168 
  2169 * Pure: New flag trace_unify_fail causes unification to print
  2170 diagnostic information (PG: in trace buffer) when it fails. This is
  2171 useful for figuring out why single step proofs like rule, erule or
  2172 assumption failed.
  2173 
  2174 * Pure: Locale specifications now produce predicate definitions
  2175 according to the body of text (covering assumptions modulo local
  2176 definitions); predicate "loc_axioms" covers newly introduced text,
  2177 while "loc" is cumulative wrt. all included locale expressions; the
  2178 latter view is presented only on export into the global theory
  2179 context; potential INCOMPATIBILITY, use "(open)" option to fall back
  2180 on the old view without predicates;
  2181 
  2182 * Pure: predefined locales "var" and "struct" are useful for sharing
  2183 parameters (as in CASL, for example); just specify something like
  2184 ``var x + var y + struct M'' as import;
  2185 
  2186 * Pure: improved thms_containing: proper indexing of facts instead of
  2187 raw theorems; check validity of results wrt. current name space;
  2188 include local facts of proof configuration (also covers active
  2189 locales), cover fixed variables in index; may use "_" in term
  2190 specification; an optional limit for the number of printed facts may
  2191 be given (the default is 40);
  2192 
  2193 * Pure: disallow duplicate fact bindings within new-style theory files
  2194 (batch-mode only);
  2195 
  2196 * Provers: improved induct method: assumptions introduced by case
  2197 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
  2198 the goal statement); "foo" still refers to all facts collectively;
  2199 
  2200 * Provers: the function blast.overloaded has been removed: all constants
  2201 are regarded as potentially overloaded, which improves robustness in exchange
  2202 for slight decrease in efficiency;
  2203 
  2204 * Provers/linorder: New generic prover for transitivity reasoning over
  2205 linear orders.  Note: this prover is not efficient!
  2206 
  2207 * Isar: preview of problems to finish 'show' now produce an error
  2208 rather than just a warning (in interactive mode);
  2209 
  2210 
  2211 *** HOL ***
  2212 
  2213 * arith(_tac)
  2214 
  2215  - Produces a counter example if it cannot prove a goal.
  2216    Note that the counter example may be spurious if the goal is not a formula
  2217    of quantifier-free linear arithmetic.
  2218    In ProofGeneral the counter example appears in the trace buffer.
  2219 
  2220  - Knows about div k and mod k where k is a numeral of type nat or int.
  2221 
  2222  - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
  2223    linear arithmetic fails. This takes account of quantifiers and divisibility.
  2224    Presburger arithmetic can also be called explicitly via presburger(_tac).
  2225 
  2226 * simp's arithmetic capabilities have been enhanced a bit: it now
  2227 takes ~= in premises into account (by performing a case split);
  2228 
  2229 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
  2230 are distributed over a sum of terms;
  2231 
  2232 * New tactic "trans_tac" and method "trans" instantiate
  2233 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
  2234 "<=", "<" and "=").
  2235 
  2236 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main
  2237 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
  2238 
  2239 * 'typedef' command has new option "open" to suppress the set
  2240 definition;
  2241 
  2242 * functions Min and Max on finite sets have been introduced (theory
  2243 Finite_Set);
  2244 
  2245 * attribute [symmetric] now works for relations as well; it turns
  2246 (x,y) : R^-1 into (y,x) : R, and vice versa;
  2247 
  2248 * induct over a !!-quantified statement (say !!x1..xn):
  2249   each "case" automatically performs "fix x1 .. xn" with exactly those names.
  2250 
  2251 * Map: `empty' is no longer a constant but a syntactic abbreviation for
  2252 %x. None. Warning: empty_def now refers to the previously hidden definition
  2253 of the empty set.
  2254 
  2255 * Algebra: formalization of classical algebra.  Intended as base for
  2256 any algebraic development in Isabelle.  Currently covers group theory
  2257 (up to Sylow's theorem) and ring theory (Universal Property of
  2258 Univariate Polynomials).  Contributions welcome;
  2259 
  2260 * GroupTheory: deleted, since its material has been moved to Algebra;
  2261 
  2262 * Complex: new directory of the complex numbers with numeric constants,
  2263 nonstandard complex numbers, and some complex analysis, standard and
  2264 nonstandard (Jacques Fleuriot);
  2265 
  2266 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
  2267 
  2268 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques
  2269 Fleuriot);
  2270 
  2271 * Real/HahnBanach: updated and adapted to locales;
  2272 
  2273 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
  2274 Gray and Kramer);
  2275 
  2276 * UNITY: added the Meier-Sanders theory of progress sets;
  2277 
  2278 * MicroJava: bytecode verifier and lightweight bytecode verifier
  2279 as abstract algorithms, instantiated to the JVM;
  2280 
  2281 * Bali: Java source language formalization. Type system, operational
  2282 semantics, axiomatic semantics. Supported language features:
  2283 classes, interfaces, objects,virtual methods, static methods,
  2284 static/instance fields, arrays, access modifiers, definite
  2285 assignment, exceptions.
  2286 
  2287 
  2288 *** ZF ***
  2289 
  2290 * ZF/Constructible: consistency proof for AC (Gdel's constructible
  2291 universe, etc.);
  2292 
  2293 * Main ZF: virtually all theories converted to new-style format;
  2294 
  2295 
  2296 *** ML ***
  2297 
  2298 * Pure: Tactic.prove provides sane interface for internal proofs;
  2299 omits the infamous "standard" operation, so this is more appropriate
  2300 than prove_goalw_cterm in many situations (e.g. in simprocs);
  2301 
  2302 * Pure: improved error reporting of simprocs;
  2303 
  2304 * Provers: Simplifier.simproc(_i) provides sane interface for setting
  2305 up simprocs;
  2306 
  2307 
  2308 *** Document preparation ***
  2309 
  2310 * uses \par instead of \\ for line breaks in theory text. This may
  2311 shift some page breaks in large documents. To get the old behaviour
  2312 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
  2313 
  2314 * minimized dependencies of isabelle.sty and isabellesym.sty on
  2315 other packages
  2316 
  2317 * \<euro> now needs package babel/greek instead of marvosym (which
  2318 broke \Rightarrow)
  2319 
  2320 * normal size for \<zero>...\<nine> (uses \mathbf instead of
  2321 textcomp package)
  2322 
  2323 
  2324 
  2325 New in Isabelle2002 (March 2002)
  2326 --------------------------------
  2327 
  2328 *** Document preparation ***
  2329 
  2330 * greatly simplified document preparation setup, including more
  2331 graceful interpretation of isatool usedir -i/-d/-D options, and more
  2332 instructive isatool mkdir; users should basically be able to get
  2333 started with "isatool mkdir HOL Test && isatool make"; alternatively,
  2334 users may run a separate document processing stage manually like this:
  2335 "isatool usedir -D output HOL Test && isatool document Test/output";
  2336 
  2337 * theory dependency graph may now be incorporated into documents;
  2338 isatool usedir -g true will produce session_graph.eps/.pdf for use
  2339 with \includegraphics of LaTeX;
  2340 
  2341 * proper spacing of consecutive markup elements, especially text
  2342 blocks after section headings;
  2343 
  2344 * support bold style (for single symbols only), input syntax is like
  2345 this: "\<^bold>\<alpha>" or "\<^bold>A";
  2346 
  2347 * \<bullet> is now output as bold \cdot by default, which looks much
  2348 better in printed text;
  2349 
  2350 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
  2351 note that these symbols are currently unavailable in Proof General /
  2352 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
  2353 
  2354 * isatool latex no longer depends on changed TEXINPUTS, instead
  2355 isatool document copies the Isabelle style files to the target
  2356 location;
  2357 
  2358 
  2359 *** Isar ***
  2360 
  2361 * Pure/Provers: improved proof by cases and induction;
  2362   - 'case' command admits impromptu naming of parameters (such as
  2363     "case (Suc n)");
  2364   - 'induct' method divinates rule instantiation from the inductive
  2365     claim; no longer requires excessive ?P bindings for proper
  2366     instantiation of cases;
  2367   - 'induct' method properly enumerates all possibilities of set/type
  2368     rules; as a consequence facts may be also passed through *type*
  2369     rules without further ado;
  2370   - 'induct' method now derives symbolic cases from the *rulified*
  2371     rule (before it used to rulify cases stemming from the internal
  2372     atomized version); this means that the context of a non-atomic
  2373     statement becomes is included in the hypothesis, avoiding the
  2374     slightly cumbersome show "PROP ?case" form;
  2375   - 'induct' may now use elim-style induction rules without chaining
  2376     facts, using ``missing'' premises from the goal state; this allows
  2377     rules stemming from inductive sets to be applied in unstructured
  2378     scripts, while still benefitting from proper handling of non-atomic
  2379     statements; NB: major inductive premises need to be put first, all
  2380     the rest of the goal is passed through the induction;
  2381   - 'induct' proper support for mutual induction involving non-atomic
  2382     rule statements (uses the new concept of simultaneous goals, see
  2383     below);
  2384   - append all possible rule selections, but only use the first
  2385     success (no backtracking);
  2386   - removed obsolete "(simplified)" and "(stripped)" options of methods;
  2387   - undeclared rule case names default to numbers 1, 2, 3, ...;
  2388   - added 'print_induct_rules' (covered by help item in recent Proof
  2389     General versions);
  2390   - moved induct/cases attributes to Pure, methods to Provers;
  2391   - generic method setup instantiated for FOL and HOL;
  2392 
  2393 * Pure: support multiple simultaneous goal statements, for example
  2394 "have a: A and b: B" (same for 'theorem' etc.); being a pure
  2395 meta-level mechanism, this acts as if several individual goals had
  2396 been stated separately; in particular common proof methods need to be
  2397 repeated in order to cover all claims; note that a single elimination
  2398 step is *not* sufficient to establish the two conjunctions, so this
  2399 fails:
  2400 
  2401   assume "A & B" then have A and B ..   (*".." fails*)
  2402 
  2403 better use "obtain" in situations as above; alternative refer to
  2404 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
  2405 
  2406 * Pure: proper integration with ``locales''; unlike the original
  2407 version by Florian Kammller, Isar locales package high-level proof
  2408 contexts rather than raw logical ones (e.g. we admit to include
  2409 attributes everywhere); operations on locales include merge and
  2410 rename; support for implicit arguments (``structures''); simultaneous
  2411 type-inference over imports and text; see also HOL/ex/Locales.thy for
  2412 some examples;
  2413 
  2414 * Pure: the following commands have been ``localized'', supporting a
  2415 target locale specification "(in name)": 'lemma', 'theorem',
  2416 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
  2417 stored both within the locale and at the theory level (exported and
  2418 qualified by the locale name);
  2419 
  2420 * Pure: theory goals may now be specified in ``long'' form, with
  2421 ad-hoc contexts consisting of arbitrary locale elements. for example
  2422 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
  2423 definitions may be given, too); the result is a meta-level rule with
  2424 the context elements being discharged in the obvious way;
  2425 
  2426 * Pure: new proof command 'using' allows to augment currently used
  2427 facts after a goal statement ('using' is syntactically analogous to
  2428 'apply', but acts on the goal's facts only); this allows chained facts
  2429 to be separated into parts given before and after a claim, as in
  2430 ``from a and b have C using d and e <proof>'';
  2431 
  2432 * Pure: renamed "antecedent" case to "rule_context";
  2433 
  2434 * Pure: new 'judgment' command records explicit information about the
  2435 object-logic embedding (used by several tools internally); no longer
  2436 use hard-wired "Trueprop";
  2437 
  2438 * Pure: added 'corollary' command;
  2439 
  2440 * Pure: fixed 'token_translation' command;
  2441 
  2442 * Pure: removed obsolete 'exported' attribute;
  2443 
  2444 * Pure: dummy pattern "_" in is/let is now automatically lifted over
  2445 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
  2446 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
  2447 
  2448 * Pure: method 'atomize' presents local goal premises as object-level
  2449 statements (atomic meta-level propositions); setup controlled via
  2450 rewrite rules declarations of 'atomize' attribute; example
  2451 application: 'induct' method with proper rule statements in improper
  2452 proof *scripts*;
  2453 
  2454 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
  2455 now consider the syntactic context of assumptions, giving a better
  2456 chance to get type-inference of the arguments right (this is
  2457 especially important for locales);
  2458 
  2459 * Pure: "sorry" no longer requires quick_and_dirty in interactive
  2460 mode;
  2461 
  2462 * Pure/obtain: the formal conclusion "thesis", being marked as
  2463 ``internal'', may no longer be reference directly in the text;
  2464 potential INCOMPATIBILITY, may need to use "?thesis" in rare
  2465 situations;
  2466 
  2467 * Pure: generic 'sym' attribute which declares a rule both as pure
  2468 'elim?' and for the 'symmetric' operation;
  2469 
  2470 * Pure: marginal comments ``--'' may now occur just anywhere in the
  2471 text; the fixed correlation with particular command syntax has been
  2472 discontinued;
  2473 
  2474 * Pure: new method 'rules' is particularly well-suited for proof
  2475 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
  2476 but often produces more compact proof terms with less detours;
  2477 
  2478 * Pure/Provers/classical: simplified integration with pure rule
  2479 attributes and methods; the classical "intro?/elim?/dest?"
  2480 declarations coincide with the pure ones; the "rule" method no longer
  2481 includes classically swapped intros; "intro" and "elim" methods no
  2482 longer pick rules from the context; also got rid of ML declarations
  2483 AddXIs/AddXEs/AddXDs; all of this has some potential for
  2484 INCOMPATIBILITY;
  2485 
  2486 * Provers/classical: attribute 'swapped' produces classical inversions
  2487 of introduction rules;
  2488 
  2489 * Provers/simplifier: 'simplified' attribute may refer to explicit
  2490 rules instead of full simplifier context; 'iff' attribute handles
  2491 conditional rules;
  2492 
  2493 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
  2494 
  2495 * HOL: 'recdef' now fails on unfinished automated proofs, use
  2496 "(permissive)" option to recover old behavior;
  2497 
  2498 * HOL: 'inductive' no longer features separate (collective) attributes
  2499 for 'intros' (was found too confusing);
  2500 
  2501 * HOL: properly declared induction rules less_induct and
  2502 wf_induct_rule;
  2503 
  2504 
  2505 *** HOL ***
  2506 
  2507 * HOL: moved over to sane numeral syntax; the new policy is as
  2508 follows:
  2509 
  2510   - 0 and 1 are polymorphic constants, which are defined on any
  2511   numeric type (nat, int, real etc.);
  2512 
  2513   - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
  2514   binary representation internally;
  2515 
  2516   - type nat has special constructor Suc, and generally prefers Suc 0
  2517   over 1::nat and Suc (Suc 0) over 2::nat;
  2518 
  2519 This change may cause significant problems of INCOMPATIBILITY; here
  2520 are some hints on converting existing sources:
  2521 
  2522   - due to the new "num" token, "-0" and "-1" etc. are now atomic
  2523   entities, so expressions involving "-" (unary or binary minus) need
  2524   to be spaced properly;
  2525 
  2526   - existing occurrences of "1" may need to be constraint "1::nat" or
  2527   even replaced by Suc 0; similar for old "2";
  2528 
  2529   - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
  2530 
  2531   - remove all special provisions on numerals in proofs;
  2532 
  2533 * HOL: simp rules nat_number expand numerals on nat to Suc/0
  2534 representation (depends on bin_arith_simps in the default context);
  2535 
  2536 * HOL: symbolic syntax for x^2 (numeral 2);
  2537 
  2538 * HOL: the class of all HOL types is now called "type" rather than
  2539 "term"; INCOMPATIBILITY, need to adapt references to this type class
  2540 in axclass/classes, instance/arities, and (usually rare) occurrences
  2541 in typings (of consts etc.); internally the class is called
  2542 "HOL.type", ML programs should refer to HOLogic.typeS;
  2543 
  2544 * HOL/record package improvements:
  2545   - new derived operations "fields" to build a partial record section,
  2546     "extend" to promote a fixed record to a record scheme, and
  2547     "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
  2548     declared as simp by default;
  2549   - shared operations ("more", "fields", etc.) now need to be always
  2550     qualified) --- potential INCOMPATIBILITY;
  2551   - removed "make_scheme" operations (use "make" with "extend") --
  2552     INCOMPATIBILITY;
  2553   - removed "more" class (simply use "term") -- INCOMPATIBILITY;
  2554   - provides cases/induct rules for use with corresponding Isar
  2555     methods (for concrete records, record schemes, concrete more
  2556     parts, and schematic more parts -- in that order);
  2557   - internal definitions directly based on a light-weight abstract
  2558     theory of product types over typedef rather than datatype;
  2559 
  2560 * HOL: generic code generator for generating executable ML code from
  2561 specifications; specific support for HOL constructs such as inductive
  2562 datatypes and sets, as well as recursive functions; can be invoked
  2563 via 'generate_code' theory section;
  2564 
  2565 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
  2566 
  2567 * HOL: consolidated and renamed several theories.  In particular:
  2568         Ord.thy has been absorbed into HOL.thy
  2569         String.thy has been absorbed into List.thy
  2570 
  2571 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
  2572 (beware of argument permutation!);
  2573 
  2574 * HOL: linorder_less_split superseded by linorder_cases;
  2575 
  2576 * HOL/List: "nodups" renamed to "distinct";
  2577 
  2578 * HOL: added "The" definite description operator; move Hilbert's "Eps"
  2579 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
  2580   - Ex_def has changed, now need to use some_eq_ex
  2581 
  2582 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
  2583 in this (rare) case use:
  2584 
  2585   delSWrapper "split_all_tac"
  2586   addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
  2587 
  2588 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
  2589 MAY FAIL;
  2590 
  2591 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
  2592 Isabelle's type classes, ^ on functions and relations has too general
  2593 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
  2594 necessary to attach explicit type constraints;
  2595 
  2596 * HOL/Relation: the prefix name of the infix "O" has been changed from
  2597 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
  2598 renamed accordingly (eg "compI" -> "rel_compI").
  2599 
  2600 * HOL: syntax translations now work properly with numerals and records
  2601 expressions;
  2602 
  2603 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
  2604 of "lam" -- INCOMPATIBILITY;
  2605 
  2606 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
  2607 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
  2608 renamed "Product_Type.unit";
  2609 
  2610 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
  2611 
  2612 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
  2613 the "cases" method);
  2614 
  2615 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
  2616 Florian Kammller);
  2617 
  2618 * HOL/IMP: updated and converted to new-style theory format; several
  2619 parts turned into readable document, with proper Isar proof texts and
  2620 some explanations (by Gerwin Klein);
  2621 
  2622 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
  2623 
  2624 * HOL-Hyperreal is now a logic image;
  2625 
  2626 
  2627 *** HOLCF ***
  2628 
  2629 * Isar: consts/constdefs supports mixfix syntax for continuous
  2630 operations;
  2631 
  2632 * Isar: domain package adapted to new-style theory format, e.g. see
  2633 HOLCF/ex/Dnat.thy;
  2634 
  2635 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
  2636 potential INCOMPATIBILITY; now use plain induct_tac instead of former
  2637 lift.induct_tac, always use UU instead of Undef;
  2638 
  2639 * HOLCF/IMP: updated and converted to new-style theory;
  2640 
  2641 
  2642 *** ZF ***
  2643 
  2644 * Isar: proper integration of logic-specific tools and packages,
  2645 including theory commands '(co)inductive', '(co)datatype',
  2646 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
  2647 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
  2648 
  2649 * theory Main no longer includes AC; for the Axiom of Choice, base
  2650 your theory on Main_ZFC;
  2651 
  2652 * the integer library now covers quotients and remainders, with many
  2653 laws relating division to addition, multiplication, etc.;
  2654 
  2655 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
  2656 typeless version of the formalism;
  2657 
  2658 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
  2659 format;
  2660 
  2661 * ZF/Induct: new directory for examples of inductive definitions,
  2662 including theory Multiset for multiset orderings; converted to
  2663 new-style theory format;
  2664 
  2665 * ZF: many new theorems about lists, ordinals, etc.;
  2666 
  2667 
  2668 *** General ***
  2669 
  2670 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
  2671 variable proof controls level of detail: 0 = no proofs (only oracle
  2672 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
  2673 also ref manual for further ML interfaces;
  2674 
  2675 * Pure/axclass: removed obsolete ML interface
  2676 goal_subclass/goal_arity;
  2677 
  2678 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
  2679 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
  2680 separate tokens, so expressions involving minus need to be spaced
  2681 properly;
  2682 
  2683 * Pure/syntax: support non-oriented infixes, using keyword "infix"
  2684 rather than "infixl" or "infixr";
  2685 
  2686 * Pure/syntax: concrete syntax for dummy type variables admits genuine
  2687 sort constraint specifications in type inference; e.g. "x::_::foo"
  2688 ensures that the type of "x" is of sort "foo" (but not necessarily a
  2689 type variable);
  2690 
  2691 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
  2692 control output of nested => (types); the default behavior is
  2693 "type_brackets";
  2694 
  2695 * Pure/syntax: builtin parse translation for "_constify" turns valued
  2696 tokens into AST constants;
  2697 
  2698 * Pure/syntax: prefer later declarations of translations and print
  2699 translation functions; potential INCOMPATIBILITY: need to reverse
  2700 multiple declarations for same syntax element constant;
  2701 
  2702 * Pure/show_hyps reset by default (in accordance to existing Isar
  2703 practice);
  2704 
  2705 * Provers/classical: renamed addaltern to addafter, addSaltern to
  2706 addSafter;
  2707 
  2708 * Provers/clasimp: ``iff'' declarations now handle conditional rules
  2709 as well;
  2710 
  2711 * system: tested support for MacOS X; should be able to get Isabelle +
  2712 Proof General to work in a plain Terminal after installing Poly/ML
  2713 (e.g. from the Isabelle distribution area) and GNU bash alone
  2714 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
  2715 support requires further installations, e.g. from
  2716 http://fink.sourceforge.net/);
  2717 
  2718 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
  2719 
  2720 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
  2721 of 40 MB), cf. ML_OPTIONS;
  2722 
  2723 * system: Proof General keywords specification is now part of the
  2724 Isabelle distribution (see etc/isar-keywords.el);
  2725 
  2726 * system: support for persistent Proof General sessions (refrain from
  2727 outdating all loaded theories on startup); user may create writable
  2728 logic images like this: ``isabelle -q HOL Test'';
  2729 
  2730 * system: smart selection of Isabelle process versus Isabelle
  2731 interface, accommodates case-insensitive file systems (e.g. HFS+); may
  2732 run both "isabelle" and "Isabelle" even if file names are badly
  2733 damaged (executable inspects the case of the first letter of its own
  2734 name); added separate "isabelle-process" and "isabelle-interface";
  2735 
  2736 * system: refrain from any attempt at filtering input streams; no
  2737 longer support ``8bit'' encoding of old isabelle font, instead proper
  2738 iso-latin characters may now be used; the related isatools
  2739 "symbolinput" and "nonascii" have disappeared as well;
  2740 
  2741 * system: removed old "xterm" interface (the print modes "xterm" and
  2742 "xterm_color" are still available for direct use in a suitable
  2743 terminal);
  2744 
  2745 
  2746 
  2747 New in Isabelle99-2 (February 2001)
  2748 -----------------------------------
  2749 
  2750 *** Overview of INCOMPATIBILITIES ***
  2751 
  2752 * HOL: please note that theories in the Library and elsewhere often use the
  2753 new-style (Isar) format; to refer to their theorems in an ML script you must
  2754 bind them to ML identifers by e.g.      val thm_name = thm "thm_name";
  2755 
  2756 * HOL: inductive package no longer splits induction rule aggressively,
  2757 but only as far as specified by the introductions given; the old
  2758 format may be recovered via ML function complete_split_rule or attribute
  2759 'split_rule (complete)';
  2760 
  2761 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
  2762 gfp_Tarski to gfp_unfold;
  2763 
  2764 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
  2765 
  2766 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
  2767 relation); infix "^^" has been renamed "``"; infix "``" has been
  2768 renamed "`"; "univalent" has been renamed "single_valued";
  2769 
  2770 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
  2771 operation;
  2772 
  2773 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
  2774 
  2775 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
  2776 
  2777 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
  2778 consequence, it is no longer monotonic wrt. the local goal context
  2779 (which is now passed through the inductive cases);
  2780 
  2781 * Document preparation: renamed standard symbols \<ll> to \<lless> and
  2782 \<gg> to \<ggreater>;
  2783 
  2784 
  2785 *** Document preparation ***
  2786 
  2787 * \isabellestyle{NAME} selects version of Isabelle output (currently
  2788 available: are "it" for near math-mode best-style output, "sl" for
  2789 slanted text style, and "tt" for plain type-writer; if no
  2790 \isabellestyle command is given, output is according to slanted
  2791 type-writer);
  2792 
  2793 * support sub/super scripts (for single symbols only), input syntax is
  2794 like this: "A\<^sup>*" or "A\<^sup>\<star>";
  2795 
  2796 * some more standard symbols; see Appendix A of the system manual for
  2797 the complete list of symbols defined in isabellesym.sty;
  2798 
  2799 * improved isabelle style files; more abstract symbol implementation
  2800 (should now use \isamath{...} and \isatext{...} in custom symbol
  2801 definitions);
  2802 
  2803 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
  2804 state; Note that presentation of goal states does not conform to
  2805 actual human-readable proof documents.  Please do not include goal
  2806 states into document output unless you really know what you are doing!
  2807 
  2808 * proper indentation of antiquoted output with proportional LaTeX
  2809 fonts;
  2810 
  2811 * no_document ML operator temporarily disables LaTeX document
  2812 generation;
  2813 
  2814 * isatool unsymbolize tunes sources for plain ASCII communication;
  2815 
  2816 
  2817 *** Isar ***
  2818 
  2819 * Pure: Isar now suffers initial goal statements to contain unbound
  2820 schematic variables (this does not conform to actual readable proof
  2821 documents, due to unpredictable outcome and non-compositional proof
  2822 checking); users who know what they are doing may use schematic goals
  2823 for Prolog-style synthesis of proven results;
  2824 
  2825 * Pure: assumption method (an implicit finishing) now handles actual
  2826 rules as well;
  2827 
  2828 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
  2829 initial goal, declare "that" only as Pure intro (only for single
  2830 steps); the "that" rule assumption may now be involved in implicit
  2831 finishing, thus ".." becomes a feasible for trivial obtains;
  2832 
  2833 * Pure: default proof step now includes 'intro_classes'; thus trivial
  2834 instance proofs may be performed by "..";
  2835 
  2836 * Pure: ?thesis / ?this / "..." now work for pure meta-level
  2837 statements as well;
  2838 
  2839 * Pure: more robust selection of calculational rules;
  2840 
  2841 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
  2842 rule (as well as the assumption rule);
  2843 
  2844 * Pure: 'thm_deps' command visualizes dependencies of theorems and
  2845 lemmas, using the graph browser tool;
  2846 
  2847 * Pure: predict failure of "show" in interactive mode;
  2848 
  2849 * Pure: 'thms_containing' now takes actual terms as arguments;
  2850 
  2851 * HOL: improved method 'induct' --- now handles non-atomic goals
  2852 (potential INCOMPATIBILITY); tuned error handling;
  2853 
  2854 * HOL: cases and induct rules now provide explicit hints about the
  2855 number of facts to be consumed (0 for "type" and 1 for "set" rules);
  2856 any remaining facts are inserted into the goal verbatim;
  2857 
  2858 * HOL: local contexts (aka cases) may now contain term bindings as
  2859 well; the 'cases' and 'induct' methods new provide a ?case binding for
  2860 the result to be shown in each case;
  2861 
  2862 * HOL: added 'recdef_tc' command;
  2863 
  2864 * isatool convert assists in eliminating legacy ML scripts;
  2865 
  2866 
  2867 *** HOL ***
  2868 
  2869 * HOL/Library: a collection of generic theories to be used together
  2870 with main HOL; the theory loader path already includes this directory
  2871 by default; the following existing theories have been moved here:
  2872 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
  2873 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
  2874 
  2875 * HOL/Unix: "Some aspects of Unix file-system security", a typical
  2876 modelling and verification task performed in Isabelle/HOL +
  2877 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
  2878 
  2879 * HOL/Algebra: special summation operator SUM no longer exists, it has
  2880 been replaced by setsum; infix 'assoc' now has priority 50 (like
  2881 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
  2882 'domain', this makes the theory consistent with mathematical
  2883 literature;
  2884 
  2885 * HOL basics: added overloaded operations "inverse" and "divide"
  2886 (infix "/"), syntax for generic "abs" operation, generic summation
  2887 operator \<Sum>;
  2888 
  2889 * HOL/typedef: simplified package, provide more useful rules (see also
  2890 HOL/subset.thy);
  2891 
  2892 * HOL/datatype: induction rule for arbitrarily branching datatypes is
  2893 now expressed as a proper nested rule (old-style tactic scripts may
  2894 require atomize_strip_tac to cope with non-atomic premises);
  2895 
  2896 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
  2897 to "split_conv" (old name still available for compatibility);
  2898 
  2899 * HOL: improved concrete syntax for strings (e.g. allows translation
  2900 rules with string literals);
  2901 
  2902 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
  2903  and Fleuriot's mechanization of analysis, including the transcendental
  2904  functions for the reals;
  2905 
  2906 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
  2907 
  2908 
  2909 *** CTT ***
  2910 
  2911 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
  2912 "lam" is displayed as TWO lambda-symbols
  2913 
  2914 * CTT: theory Main now available, containing everything (that is, Bool
  2915 and Arith);
  2916 
  2917 
  2918 *** General ***
  2919 
  2920 * Pure: the Simplifier has been implemented properly as a derived rule
  2921 outside of the actual kernel (at last!); the overall performance
  2922 penalty in practical applications is about 50%, while reliability of
  2923 the Isabelle inference kernel has been greatly improved;
  2924 
  2925 * print modes "brackets" and "no_brackets" control output of nested =>
  2926 (types) and ==> (props); the default behaviour is "brackets";
  2927 
  2928 * Provers: fast_tac (and friends) now handle actual object-logic rules
  2929 as assumptions as well;
  2930 
  2931 * system: support Poly/ML 4.0;
  2932 
  2933 * system: isatool install handles KDE version 1 or 2;
  2934 
  2935 
  2936 
  2937 New in Isabelle99-1 (October 2000)
  2938 ----------------------------------
  2939 
  2940 *** Overview of INCOMPATIBILITIES ***
  2941 
  2942 * HOL: simplification of natural numbers is much changed; to partly
  2943 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
  2944 issue the following ML commands:
  2945 
  2946   Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
  2947   Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
  2948 
  2949 * HOL: simplification no longer dives into case-expressions; this is
  2950 controlled by "t.weak_case_cong" for each datatype t;
  2951 
  2952 * HOL: nat_less_induct renamed to less_induct;
  2953 
  2954 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
  2955 fixsome to patch .thy and .ML sources automatically;
  2956 
  2957   select_equality  -> some_equality
  2958   select_eq_Ex     -> some_eq_ex
  2959   selectI2EX       -> someI2_ex
  2960   selectI2         -> someI2
  2961   selectI          -> someI
  2962   select1_equality -> some1_equality
  2963   Eps_sym_eq       -> some_sym_eq_trivial
  2964   Eps_eq           -> some_eq_trivial
  2965 
  2966 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
  2967 
  2968 * HOL: removed obsolete theorem binding expand_if (refer to split_if
  2969 instead);
  2970 
  2971 * HOL: the recursion equations generated by 'recdef' are now called
  2972 f.simps instead of f.rules;
  2973 
  2974 * HOL: qed_spec_mp now also handles bounded ALL as well;
  2975 
  2976 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
  2977 sometimes be needed;
  2978 
  2979 * HOL: the constant for "f``x" is now "image" rather than "op ``";
  2980 
  2981 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
  2982 
  2983 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
  2984 product is now "<*>" instead of "Times"; the lexicographic product is
  2985 now "<*lex*>" instead of "**";
  2986 
  2987 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
  2988 of main HOL, but was unused); better use HOL's datatype package;
  2989 
  2990 * HOL: removed "symbols" syntax for constant "override" of theory Map;
  2991 the old syntax may be recovered as follows:
  2992 
  2993   syntax (symbols)
  2994     override  :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
  2995       (infixl "\\<oplus>" 100)
  2996 
  2997 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  2998 
  2999 * HOL/ML: even fewer consts are declared as global (see theories Ord,
  3000 Lfp, Gfp, WF); this only affects ML packages that refer to const names
  3001 internally;
  3002 
  3003 * HOL and ZF: syntax for quotienting wrt an equivalence relation
  3004 changed from A/r to A//r;
  3005 
  3006 * ZF: new treatment of arithmetic (nat & int) may break some old
  3007 proofs;
  3008 
  3009 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
  3010 rulify -> rule_format, elimify -> elim_format, ...);
  3011 
  3012 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3013 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
  3014 should have to change intro!! to intro? only); replaced "delrule" by
  3015 "rule del";
  3016 
  3017 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
  3018 
  3019 * Provers: strengthened force_tac by using new first_best_tac;
  3020 
  3021 * LaTeX document preparation: several changes of isabelle.sty (see
  3022 lib/texinputs);
  3023 
  3024 
  3025 *** Document preparation ***
  3026 
  3027 * formal comments (text blocks etc.) in new-style theories may now
  3028 contain antiquotations of thm/prop/term/typ/text to be presented
  3029 according to latex print mode; concrete syntax is like this:
  3030 @{term[show_types] "f(x) = a + x"};
  3031 
  3032 * isatool mkdir provides easy setup of Isabelle session directories,
  3033 including proper document sources;
  3034 
  3035 * generated LaTeX sources are now deleted after successful run
  3036 (isatool document -c); may retain a copy somewhere else via -D option
  3037 of isatool usedir;
  3038 
  3039 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
  3040 style files, achieving self-contained LaTeX sources and simplifying
  3041 LaTeX debugging;
  3042 
  3043 * old-style theories now produce (crude) LaTeX output as well;
  3044 
  3045 * browser info session directories are now self-contained (may be put
  3046 on WWW server seperately); improved graphs of nested sessions; removed
  3047 graph for 'all sessions';
  3048 
  3049 * several improvements in isabelle style files; \isabellestyle{it}
  3050 produces fake math mode output; \isamarkupheader is now \section by
  3051 default; see lib/texinputs/isabelle.sty etc.;
  3052 
  3053 
  3054 *** Isar ***
  3055 
  3056 * Isar/Pure: local results and corresponding term bindings are now
  3057 subject to Hindley-Milner polymorphism (similar to ML); this
  3058 accommodates incremental type-inference very nicely;
  3059 
  3060 * Isar/Pure: new derived language element 'obtain' supports
  3061 generalized existence reasoning;
  3062 
  3063 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
  3064 support accumulation of results, without applying any rules yet;
  3065 useful to collect intermediate results without explicit name
  3066 references, and for use with transitivity rules with more than 2
  3067 premises;
  3068 
  3069 * Isar/Pure: scalable support for case-analysis type proofs: new
  3070 'case' language element refers to local contexts symbolically, as
  3071 produced by certain proof methods; internally, case names are attached
  3072 to theorems as "tags";
  3073 
  3074 * Isar/Pure: theory command 'hide' removes declarations from
  3075 class/type/const name spaces;
  3076 
  3077 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
  3078 indicate potential overloading;
  3079 
  3080 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
  3081 
  3082 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
  3083 "{a,b,c}" instead of {a,b,c};
  3084 
  3085 * Isar/Pure now provides its own version of intro/elim/dest
  3086 attributes; useful for building new logics, but beware of confusion
  3087 with the version in Provers/classical;
  3088 
  3089 * Isar/Pure: the local context of (non-atomic) goals is provided via
  3090 case name 'antecedent';
  3091 
  3092 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
  3093 to the current context is now done automatically);
  3094 
  3095 * Isar/Pure: theory command 'method_setup' provides a simple interface
  3096 for definining proof methods in ML;
  3097 
  3098 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3099 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
  3100 most cases, one should have to change intro!! to intro? only);
  3101 replaced "delrule" by "rule del";
  3102 
  3103 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
  3104 'symmetric' attribute (the latter supercedes [RS sym]);
  3105 
  3106 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
  3107 method modifier); 'simp' method: 'only:' modifier removes loopers as
  3108 well (including splits);
  3109 
  3110 * Isar/Provers: Simplifier and Classical methods now support all kind
  3111 of modifiers used in the past, including 'cong', 'iff', etc.
  3112 
  3113 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
  3114 of Simplifier and Classical reasoner);
  3115 
  3116 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
  3117 now support named cases; major packages (inductive, datatype, primrec,
  3118 recdef) support case names and properly name parameters;
  3119 
  3120 * Isar/HOL: new transitivity rules for substitution in inequalities --
  3121 monotonicity conditions are extracted to be proven at end of
  3122 calculations;
  3123 
  3124 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
  3125 method anyway;
  3126 
  3127 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
  3128 split_if split_if_asm; datatype package provides theorems foo.splits =
  3129 foo.split foo.split_asm for each datatype;
  3130 
  3131 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
  3132 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
  3133 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
  3134 use "(cases (simplified))" method in proper proof texts);
  3135 
  3136 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
  3137 
  3138 * Isar: names of theorems etc. may be natural numbers as well;
  3139 
  3140 * Isar: 'pr' command: optional arguments for goals_limit and
  3141 ProofContext.prems_limit; no longer prints theory contexts, but only
  3142 proof states;
  3143 
  3144 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
  3145 additional print modes to be specified; e.g. "pr(latex)" will print
  3146 proof state according to the Isabelle LaTeX style;
  3147 
  3148 * Isar: improved support for emulating tactic scripts, including proof
  3149 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
  3150 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
  3151 (for HOL datatypes);
  3152 
  3153 * Isar: simplified (more robust) goal selection of proof methods: 1st
  3154 goal, all goals, or explicit goal specifier (tactic emulation); thus
  3155 'proof method scripts' have to be in depth-first order;
  3156 
  3157 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
  3158 
  3159 * Isar: removed 'help' command, which hasn't been too helpful anyway;
  3160 should instead use individual commands for printing items
  3161 (print_commands, print_methods etc.);
  3162 
  3163 * Isar: added 'nothing' --- the empty list of theorems;
  3164 
  3165 
  3166 *** HOL ***
  3167 
  3168 * HOL/MicroJava: formalization of a fragment of Java, together with a
  3169 corresponding virtual machine and a specification of its bytecode
  3170 verifier and a lightweight bytecode verifier, including proofs of
  3171 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
  3172 Cornelia Pusch (see also the homepage of project Bali at
  3173 http://isabelle.in.tum.de/Bali/);
  3174 
  3175 * HOL/Algebra: new theory of rings and univariate polynomials, by
  3176 Clemens Ballarin;
  3177 
  3178 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
  3179 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
  3180 Rasmussen;
  3181 
  3182 * HOL/Lattice: fundamental concepts of lattice theory and order
  3183 structures, including duals, properties of bounds versus algebraic
  3184 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
  3185 Theorem for complete lattices etc.; may also serve as a demonstration
  3186 for abstract algebraic reasoning using axiomatic type classes, and
  3187 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
  3188 
  3189 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
  3190 von Oheimb;
  3191 
  3192 * HOL/IMPP: extension of IMP with local variables and mutually
  3193 recursive procedures, by David von Oheimb;
  3194 
  3195 * HOL/Lambda: converted into new-style theory and document;
  3196 
  3197 * HOL/ex/Multiquote: example of multiple nested quotations and
  3198 anti-quotations -- basically a generalized version of de-Bruijn
  3199 representation; very useful in avoiding lifting of operations;
  3200 
  3201 * HOL/record: added general record equality rule to simpset; fixed
  3202 select-update simplification procedure to handle extended records as
  3203 well; admit "r" as field name;
  3204 
  3205 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
  3206 other numeric types and also as the identity of groups, rings, etc.;
  3207 
  3208 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
  3209 Types nat and int belong to this axclass;
  3210 
  3211 * HOL: greatly improved simplification involving numerals of type nat, int, real:
  3212    (i + #8 + j) = Suc k simplifies to  #7 + (i + j) = k
  3213    i*j + k + j*#3*i     simplifies to  #4*(i*j) + k
  3214   two terms #m*u and #n*u are replaced by #(m+n)*u
  3215     (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
  3216   and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
  3217     or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
  3218 
  3219 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
  3220 powerful prover for predicate logic but knows nothing of clasets; see
  3221 ex/mesontest.ML and ex/mesontest2.ML for example applications;
  3222 
  3223 * HOL: new version of "case_tac" subsumes both boolean case split and
  3224 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
  3225 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
  3226 
  3227 * HOL: simplification no longer dives into case-expressions: only the
  3228 selector expression is simplified, but not the remaining arms; to
  3229 enable full simplification of case-expressions for datatype t, you may
  3230 remove t.weak_case_cong from the simpset, either globally (Delcongs
  3231 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
  3232 
  3233 * HOL/recdef: the recursion equations generated by 'recdef' for
  3234 function 'f' are now called f.simps instead of f.rules; if all
  3235 termination conditions are proved automatically, these simplification
  3236 rules are added to the simpset, as in primrec; rules may be named
  3237 individually as well, resulting in a separate list of theorems for
  3238 each equation;
  3239 
  3240 * HOL/While is a new theory that provides a while-combinator. It
  3241 permits the definition of tail-recursive functions without the
  3242 provision of a termination measure. The latter is necessary once the
  3243 invariant proof rule for while is applied.
  3244 
  3245 * HOL: new (overloaded) notation for the set of elements below/above
  3246 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
  3247 
  3248 * HOL: theorems impI, allI, ballI bound as "strip";
  3249 
  3250 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
  3251 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
  3252 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
  3253 
  3254 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3255 
  3256 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
  3257 main HOL, but was unused);
  3258 
  3259 * HOL: fewer consts declared as global (e.g. have to refer to
  3260 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
  3261 
  3262 * HOL: tuned AST representation of nested pairs, avoiding bogus output
  3263 in case of overlap with user translations (e.g. judgements over
  3264 tuples); (note that the underlying logical represenation is still
  3265 bogus);
  3266 
  3267 
  3268 *** ZF ***
  3269 
  3270 * ZF: simplification automatically cancels common terms in arithmetic
  3271 expressions over nat and int;
  3272 
  3273 * ZF: new treatment of nat to minimize type-checking: all operators
  3274 coerce their operands to a natural number using the function natify,
  3275 making the algebraic laws unconditional;
  3276 
  3277 * ZF: as above, for int: operators coerce their operands to an integer
  3278 using the function intify;
  3279 
  3280 * ZF: the integer library now contains many of the usual laws for the
  3281 orderings, including $<=, and monotonicity laws for $+ and $*;
  3282 
  3283 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
  3284 simplification;
  3285 
  3286 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
  3287 to the simplifier and classical reasoner simultaneously;
  3288 
  3289 
  3290 *** General ***
  3291 
  3292 * Provers: blast_tac now handles actual object-logic rules as
  3293 assumptions; note that auto_tac uses blast_tac internally as well;
  3294 
  3295 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
  3296 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
  3297 
  3298 * Provers: delrules now handles destruct rules as well (no longer need
  3299 explicit make_elim);
  3300 
  3301 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
  3302   [| inj ?f;          ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3303 use instead the strong form,
  3304   [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3305 in HOL, FOL and ZF the function cla_make_elim will create such rules
  3306 from destruct-rules;
  3307 
  3308 * Provers: Simplifier.easy_setup provides a fast path to basic
  3309 Simplifier setup for new object-logics;
  3310 
  3311 * Pure: AST translation rules no longer require constant head on LHS;
  3312 
  3313 * Pure: improved name spaces: ambiguous output is qualified; support
  3314 for hiding of names;
  3315 
  3316 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
  3317 XSYMBOL_HOME; no longer need to do manual configuration in most
  3318 situations;
  3319 
  3320 * system: compression of ML heaps images may now be controlled via -c
  3321 option of isabelle and isatool usedir (currently only observed by
  3322 Poly/ML);
  3323 
  3324 * system: isatool installfonts may handle X-Symbol fonts as well (very
  3325 useful for remote X11);
  3326 
  3327 * system: provide TAGS file for Isabelle sources;
  3328 
  3329 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
  3330 order;
  3331 
  3332 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
  3333 timing flag supersedes proof_timing and Toplevel.trace;
  3334 
  3335 * ML: new combinators |>> and |>>> for incremental transformations
  3336 with secondary results (e.g. certain theory extensions):
  3337 
  3338 * ML: PureThy.add_defs gets additional argument to indicate potential
  3339 overloading (usually false);
  3340 
  3341 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
  3342 results;
  3343 
  3344 
  3345 
  3346 New in Isabelle99 (October 1999)
  3347 --------------------------------
  3348 
  3349 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3350 
  3351 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
  3352 are no longer simplified.  (This allows the simplifier to unfold recursive
  3353 functional programs.)  To restore the old behaviour, declare
  3354 
  3355     Delcongs [if_weak_cong];
  3356 
  3357 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
  3358 complement;
  3359 
  3360 * HOL: the predicate "inj" is now defined by translation to "inj_on";
  3361 
  3362 * HOL/datatype: mutual_induct_tac no longer exists --
  3363   use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
  3364 
  3365 * HOL/typedef: fixed type inference for representing set; type
  3366 arguments now have to occur explicitly on the rhs as type constraints;
  3367 
  3368 * ZF: The con_defs part of an inductive definition may no longer refer
  3369 to constants declared in the same theory;
  3370 
  3371 * HOL, ZF: the function mk_cases, generated by the inductive
  3372 definition package, has lost an argument.  To simplify its result, it
  3373 uses the default simpset instead of a supplied list of theorems.
  3374 
  3375 * HOL/List: the constructors of type list are now Nil and Cons;
  3376 
  3377 * Simplifier: the type of the infix ML functions
  3378         setSSolver addSSolver setSolver addSolver
  3379 is now  simpset * solver -> simpset  where `solver' is a new abstract type
  3380 for packaging solvers. A solver is created via
  3381         mk_solver: string -> (thm list -> int -> tactic) -> solver
  3382 where the string argument is only a comment.
  3383 
  3384 
  3385 *** Proof tools ***
  3386 
  3387 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
  3388 decision procedure for linear arithmetic. Currently it is used for
  3389 types `nat', `int', and `real' in HOL (see below); it can, should and
  3390 will be instantiated for other types and logics as well.
  3391 
  3392 * The simplifier now accepts rewrite rules with flexible heads, eg
  3393      hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
  3394   They are applied like any rule with a non-pattern lhs, i.e. by first-order
  3395   matching.
  3396 
  3397 
  3398 *** General ***
  3399 
  3400 * New Isabelle/Isar subsystem provides an alternative to traditional
  3401 tactical theorem proving; together with the ProofGeneral/isar user
  3402 interface it offers an interactive environment for developing human
  3403 readable proof documents (Isar == Intelligible semi-automated
  3404 reasoning); for further information see isatool doc isar-ref,
  3405 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
  3406 
  3407 * improved and simplified presentation of theories: better HTML markup
  3408 (including colors), graph views in several sizes; isatool usedir now
  3409 provides a proper interface for user theories (via -P option); actual
  3410 document preparation based on (PDF)LaTeX is available as well (for
  3411 new-style theories only); see isatool doc system for more information;
  3412 
  3413 * native support for Proof General, both for classic Isabelle and
  3414 Isabelle/Isar;
  3415 
  3416 * ML function thm_deps visualizes dependencies of theorems and lemmas,
  3417 using the graph browser tool;
  3418 
  3419 * Isabelle manuals now also available as PDF;
  3420 
  3421 * theory loader rewritten from scratch (may not be fully
  3422 bug-compatible); old loadpath variable has been replaced by show_path,
  3423 add_path, del_path, reset_path functions; new operations such as
  3424 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
  3425 isatool doc ref);
  3426 
  3427 * improved isatool install: option -k creates KDE application icon,
  3428 option -p DIR installs standalone binaries;
  3429 
  3430 * added ML_PLATFORM setting (useful for cross-platform installations);
  3431 more robust handling of platform specific ML images for SML/NJ;
  3432 
  3433 * the settings environment is now statically scoped, i.e. it is never
  3434 created again in sub-processes invoked from isabelle, isatool, or
  3435 Isabelle;
  3436 
  3437 * path element specification '~~' refers to '$ISABELLE_HOME';
  3438 
  3439 * in locales, the "assumes" and "defines" parts may be omitted if
  3440 empty;
  3441 
  3442 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
  3443 long arrows);
  3444 
  3445 * new print_mode "HTML";
  3446 
  3447 * new flag show_tags controls display of tags of theorems (which are
  3448 basically just comments that may be attached by some tools);
  3449 
  3450 * Isamode 2.6 requires patch to accomodate change of Isabelle font
  3451 mode and goal output format:
  3452 
  3453 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
  3454 244c244
  3455 <       (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
  3456 ---
  3457 >       (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
  3458 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
  3459 181c181
  3460 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
  3461 ---
  3462 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
  3463 
  3464 * function bind_thms stores lists of theorems (cf. bind_thm);
  3465 
  3466 * new shorthand tactics ftac, eatac, datac, fatac;
  3467 
  3468 * qed (and friends) now accept "" as result name; in that case the
  3469 theorem is not stored, but proper checks and presentation of the
  3470 result still apply;
  3471 
  3472 * theorem database now also indexes constants "Trueprop", "all",
  3473 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
  3474 
  3475 
  3476 *** HOL ***
  3477 
  3478 ** HOL arithmetic **
  3479 
  3480 * There are now decision procedures for linear arithmetic over nat and
  3481 int:
  3482 
  3483 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
  3484 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
  3485 are treated as atomic; subformulae not involving type `nat' or `int'
  3486 are ignored; quantified subformulae are ignored unless they are
  3487 positive universal or negative existential. The tactic has to be
  3488 invoked by hand and can be a little bit slow. In particular, the
  3489 running time is exponential in the number of occurrences of `min' and
  3490 `max', and `-' on `nat'.
  3491 
  3492 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
  3493 (negated) (in)equalities among the premises and the conclusion into
  3494 account (i.e. no compound formulae) and does not know about `min' and
  3495 `max', and `-' on `nat'. It is fast and is used automatically by the
  3496 simplifier.
  3497 
  3498 NB: At the moment, these decision procedures do not cope with mixed
  3499 nat/int formulae where the two parts interact, such as `m < n ==>
  3500 int(m) < int(n)'.
  3501 
  3502 * HOL/Numeral provides a generic theory of numerals (encoded
  3503 efficiently as bit strings); setup for types nat/int/real is in place;
  3504 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
  3505 int, existing theories and proof scripts may require a few additional
  3506 type constraints;
  3507 
  3508 * integer division and remainder can now be performed on constant
  3509 arguments;
  3510 
  3511 * many properties of integer multiplication, division and remainder
  3512 are now available;
  3513 
  3514 * An interface to the Stanford Validity Checker (SVC) is available through the
  3515 tactic svc_tac.  Propositional tautologies and theorems of linear arithmetic
  3516 are proved automatically.  SVC must be installed separately, and its results
  3517 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
  3518 invocation of the underlying oracle).  For SVC see
  3519   http://verify.stanford.edu/SVC
  3520 
  3521 * IsaMakefile: the HOL-Real target now builds an actual image;
  3522 
  3523 
  3524 ** HOL misc **
  3525 
  3526 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
  3527 (in Isabelle/Isar) -- by Gertrud Bauer;
  3528 
  3529 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
  3530 analysis for assembly languages with subtypes;
  3531 
  3532 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
  3533 -- avoids syntactic ambiguities and treats state, transition, and
  3534 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
  3535 changed syntax and (many) tactics;
  3536 
  3537 * HOL/inductive: Now also handles more general introduction rules such
  3538   as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
  3539   theorems are now maintained within the theory (maintained via the
  3540   "mono" attribute);
  3541 
  3542 * HOL/datatype: Now also handles arbitrarily branching datatypes
  3543   (using function types) such as
  3544 
  3545   datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
  3546 
  3547 * HOL/record: record_simproc (part of the default simpset) takes care
  3548 of selectors applied to updated records; record_split_tac is no longer
  3549 part of the default claset; update_defs may now be removed from the
  3550 simpset in many cases; COMPATIBILITY: old behavior achieved by
  3551 
  3552   claset_ref () := claset() addSWrapper record_split_wrapper;
  3553   Delsimprocs [record_simproc]
  3554 
  3555 * HOL/typedef: fixed type inference for representing set; type
  3556 arguments now have to occur explicitly on the rhs as type constraints;
  3557 
  3558 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
  3559 names rather than an ML expression;
  3560 
  3561 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
  3562 supplied later.  Program schemes can be defined, such as
  3563     "While B C s = (if B s then While B C (C s) else s)"
  3564 where the well-founded relation can be chosen after B and C have been given.
  3565 
  3566 * HOL/List: the constructors of type list are now Nil and Cons;
  3567 INCOMPATIBILITY: while [] and infix # syntax is still there, of
  3568 course, ML tools referring to List.list.op # etc. have to be adapted;
  3569 
  3570 * HOL_quantifiers flag superseded by "HOL" print mode, which is
  3571 disabled by default; run isabelle with option -m HOL to get back to
  3572 the original Gordon/HOL-style output;
  3573 
  3574 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
  3575 ALL x<=y. P, EX x<y. P, EX x<=y. P;
  3576 
  3577 * HOL basic syntax simplified (more orthogonal): all variants of
  3578 All/Ex now support plain / symbolic / HOL notation; plain syntax for
  3579 Eps operator is provided as well: "SOME x. P[x]";
  3580 
  3581 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
  3582 
  3583 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
  3584 thus available for user theories;
  3585 
  3586 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
  3587 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
  3588 time;
  3589 
  3590 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
  3591 several times and then mp;
  3592 
  3593 
  3594 *** LK ***
  3595 
  3596 * the notation <<...>> is now available as a notation for sequences of
  3597 formulas;
  3598 
  3599 * the simplifier is now installed
  3600 
  3601 * the axiom system has been generalized (thanks to Soren Heilmann)
  3602 
  3603 * the classical reasoner now has a default rule database
  3604 
  3605 
  3606 *** ZF ***
  3607 
  3608 * new primrec section allows primitive recursive functions to be given
  3609 directly (as in HOL) over datatypes and the natural numbers;
  3610 
  3611 * new tactics induct_tac and exhaust_tac for induction (or case
  3612 analysis) over datatypes and the natural numbers;
  3613 
  3614 * the datatype declaration of type T now defines the recursor T_rec;
  3615 
  3616 * simplification automatically does freeness reasoning for datatype
  3617 constructors;
  3618 
  3619 * automatic type-inference, with AddTCs command to insert new
  3620 type-checking rules;
  3621 
  3622 * datatype introduction rules are now added as Safe Introduction rules
  3623 to the claset;
  3624 
  3625 * the syntax "if P then x else y" is now available in addition to
  3626 if(P,x,y);
  3627 
  3628 
  3629 *** Internal programming interfaces ***
  3630 
  3631 * tuned simplifier trace output; new flag debug_simp;
  3632 
  3633 * structures Vartab / Termtab (instances of TableFun) offer efficient
  3634 tables indexed by indexname_ord / term_ord (compatible with aconv);
  3635 
  3636 * AxClass.axclass_tac lost the theory argument;
  3637 
  3638 * tuned current_goals_markers semantics: begin / end goal avoids
  3639 printing empty lines;
  3640 
  3641 * removed prs and prs_fn hook, which was broken because it did not
  3642 include \n in its semantics, forcing writeln to add one
  3643 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
  3644 string -> unit if you really want to output text without newline;
  3645 
  3646 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
  3647 plain output, interface builders may have to enable 'isabelle_font'
  3648 mode to get Isabelle font glyphs as before;
  3649 
  3650 * refined token_translation interface; INCOMPATIBILITY: output length
  3651 now of type real instead of int;
  3652 
  3653 * theory loader actions may be traced via new ThyInfo.add_hook
  3654 interface (see src/Pure/Thy/thy_info.ML); example application: keep
  3655 your own database of information attached to *whole* theories -- as
  3656 opposed to intra-theory data slots offered via TheoryDataFun;
  3657 
  3658 * proper handling of dangling sort hypotheses (at last!);
  3659 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
  3660 extra sort hypotheses that can be witnessed from the type signature;
  3661 the force_strip_shyps flag is gone, any remaining shyps are simply
  3662 left in the theorem (with a warning issued by strip_shyps_warning);
  3663 
  3664 
  3665 
  3666 New in Isabelle98-1 (October 1998)
  3667 ----------------------------------
  3668 
  3669 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3670 
  3671 * several changes of automated proof tools;
  3672 
  3673 * HOL: major changes to the inductive and datatype packages, including
  3674 some minor incompatibilities of theory syntax;
  3675 
  3676 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
  3677 called `inj_on';
  3678 
  3679 * HOL: removed duplicate thms in Arith:
  3680   less_imp_add_less  should be replaced by  trans_less_add1
  3681   le_imp_add_le      should be replaced by  trans_le_add1
  3682 
  3683 * HOL: unary minus is now overloaded (new type constraints may be
  3684 required);
  3685 
  3686 * HOL and ZF: unary minus for integers is now #- instead of #~.  In
  3687 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
  3688 now taken as an integer constant.
  3689 
  3690 * Pure: ML function 'theory_of' renamed to 'theory';
  3691 
  3692 
  3693 *** Proof tools ***
  3694 
  3695 * Simplifier:
  3696   1. Asm_full_simp_tac is now more aggressive.
  3697      1. It will sometimes reorient premises if that increases their power to
  3698         simplify.
  3699      2. It does no longer proceed strictly from left to right but may also
  3700         rotate premises to achieve further simplification.
  3701      For compatibility reasons there is now Asm_lr_simp_tac which is like the
  3702      old Asm_full_simp_tac in that it does not rotate premises.
  3703   2. The simplifier now knows a little bit about nat-arithmetic.
  3704 
  3705 * Classical reasoner: wrapper mechanism for the classical reasoner now
  3706 allows for selected deletion of wrappers, by introduction of names for
  3707 wrapper functionals.  This implies that addbefore, addSbefore,
  3708 addaltern, and addSaltern now take a pair (name, tactic) as argument,
  3709 and that adding two tactics with the same name overwrites the first
  3710 one (emitting a warning).
  3711   type wrapper = (int -> tactic) -> (int -> tactic)
  3712   setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
  3713   addWrapper, addSWrapper: claset * (string * wrapper) -> claset
  3714   delWrapper, delSWrapper: claset *  string            -> claset
  3715   getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
  3716 
  3717 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
  3718 semantics; addbefore now affects only the unsafe part of step_tac
  3719 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
  3720 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
  3721 by Force_tac;
  3722 
  3723 * Classical reasoner: setwrapper to setWrapper and compwrapper to
  3724 compWrapper; added safe wrapper (and access functions for it);
  3725 
  3726 * HOL/split_all_tac is now much faster and fails if there is nothing
  3727 to split.  Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
  3728 and the names of the automatically generated variables have changed.
  3729 split_all_tac has moved within claset() from unsafe wrappers to safe
  3730 wrappers, which means that !!-bound variables are split much more
  3731 aggressively, and safe_tac and clarify_tac now split such variables.
  3732 If this splitting is not appropriate, use delSWrapper "split_all_tac".
  3733 Note: the same holds for record_split_tac, which does the job of
  3734 split_all_tac for record fields.
  3735 
  3736 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
  3737 permanently to the default simpset using Addsplits just like
  3738 Addsimps. They can be removed via Delsplits just like
  3739 Delsimps. Lower-case versions are also available.
  3740 
  3741 * HOL/Simplifier: The rule split_if is now part of the default
  3742 simpset. This means that the simplifier will eliminate all occurrences
  3743 of if-then-else in the conclusion of a goal. To prevent this, you can
  3744 either remove split_if completely from the default simpset by
  3745 `Delsplits [split_if]' or remove it in a specific call of the
  3746 simplifier using `... delsplits [split_if]'.  You can also add/delete
  3747 other case splitting rules to/from the default simpset: every datatype
  3748 generates suitable rules `split_t_case' and `split_t_case_asm' (where
  3749 t is the name of the datatype).
  3750 
  3751 * Classical reasoner / Simplifier combination: new force_tac (and
  3752 derivatives Force_tac, force) combines rewriting and classical
  3753 reasoning (and whatever other tools) similarly to auto_tac, but is
  3754 aimed to solve the given subgoal completely.
  3755 
  3756 
  3757 *** General ***
  3758 
  3759 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
  3760 and `goalw': the theory is no longer needed as an explicit argument -
  3761 the current theory context is used; assumptions are no longer returned
  3762 at the ML-level unless one of them starts with ==> or !!; it is
  3763 recommended to convert to these new commands using isatool fixgoal
  3764 (backup your sources first!);
  3765 
  3766 * new top-level commands 'thm' and 'thms' for retrieving theorems from
  3767 the current theory context, and 'theory' to lookup stored theories;
  3768 
  3769 * new theory section 'locale' for declaring constants, assumptions and
  3770 definitions that have local scope;
  3771 
  3772 * new theory section 'nonterminals' for purely syntactic types;
  3773 
  3774 * new theory section 'setup' for generic ML setup functions
  3775 (e.g. package initialization);
  3776 
  3777 * the distribution now includes Isabelle icons: see
  3778 lib/logo/isabelle-{small,tiny}.xpm;
  3779 
  3780 * isatool install - install binaries with absolute references to
  3781 ISABELLE_HOME/bin;
  3782 
  3783 * isatool logo -- create instances of the Isabelle logo (as EPS);
  3784 
  3785 * print mode 'emacs' reserved for Isamode;
  3786 
  3787 * support multiple print (ast) translations per constant name;
  3788 
  3789 * theorems involving oracles are now printed with a suffixed [!];
  3790 
  3791 
  3792 *** HOL ***
  3793 
  3794 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
  3795 
  3796 * HOL/inductive package reorganized and improved: now supports mutual
  3797 definitions such as
  3798 
  3799   inductive EVEN ODD
  3800     intrs
  3801       null "0 : EVEN"
  3802       oddI "n : EVEN ==> Suc n : ODD"
  3803       evenI "n : ODD ==> Suc n : EVEN"
  3804 
  3805 new theorem list "elims" contains an elimination rule for each of the
  3806 recursive sets; inductive definitions now handle disjunctive premises
  3807 correctly (also ZF);
  3808 
  3809 INCOMPATIBILITIES: requires Inductive as an ancestor; component
  3810 "mutual_induct" no longer exists - the induction rule is always
  3811 contained in "induct";
  3812 
  3813 
  3814 * HOL/datatype package re-implemented and greatly improved: now
  3815 supports mutually recursive datatypes such as
  3816 
  3817   datatype
  3818     'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
  3819             | SUM ('a aexp) ('a aexp)
  3820             | DIFF ('a aexp) ('a aexp)
  3821             | NUM 'a
  3822   and
  3823     'a bexp = LESS ('a aexp) ('a aexp)
  3824             | AND ('a bexp) ('a bexp)
  3825             | OR ('a bexp) ('a bexp)
  3826 
  3827 as well as indirectly recursive datatypes such as
  3828 
  3829   datatype
  3830     ('a, 'b) term = Var 'a
  3831                   | App 'b ((('a, 'b) term) list)
  3832 
  3833 The new tactic  mutual_induct_tac [<var_1>, ..., <var_n>] i  performs
  3834 induction on mutually / indirectly recursive datatypes.
  3835 
  3836 Primrec equations are now stored in theory and can be accessed via
  3837 <function_name>.simps.
  3838 
  3839 INCOMPATIBILITIES:
  3840 
  3841   - Theories using datatypes must now have theory Datatype as an
  3842     ancestor.
  3843   - The specific <typename>.induct_tac no longer exists - use the
  3844     generic induct_tac instead.
  3845   - natE has been renamed to nat.exhaust - use exhaust_tac
  3846     instead of res_inst_tac ... natE. Note that the variable
  3847     names in nat.exhaust differ from the names in natE, this
  3848     may cause some "fragile" proofs to fail.
  3849   - The theorems split_<typename>_case and split_<typename>_case_asm
  3850     have been renamed to <typename>.split and <typename>.split_asm.
  3851   - Since default sorts of type variables are now handled correctly,
  3852     some datatype definitions may have to be annotated with explicit
  3853     sort constraints.
  3854   - Primrec definitions no longer require function name and type
  3855     of recursive argument.
  3856 
  3857 Consider using isatool fixdatatype to adapt your theories and proof
  3858 scripts to the new package (backup your sources first!).
  3859 
  3860 
  3861 * HOL/record package: considerably improved implementation; now
  3862 includes concrete syntax for record types, terms, updates; theorems
  3863 for surjective pairing and splitting !!-bound record variables; proof
  3864 support is as follows:
  3865 
  3866   1) standard conversions (selectors or updates applied to record
  3867 constructor terms) are part of the standard simpset;
  3868 
  3869   2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
  3870 made part of standard simpset and claset via addIffs;
  3871 
  3872   3) a tactic for record field splitting (record_split_tac) is part of
  3873 the standard claset (addSWrapper);
  3874 
  3875 To get a better idea about these rules you may retrieve them via
  3876 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
  3877 the name of your record type.
  3878 
  3879 The split tactic 3) conceptually simplifies by the following rule:
  3880 
  3881   "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
  3882 
  3883 Thus any record variable that is bound by meta-all will automatically
  3884 blow up into some record constructor term, consequently the
  3885 simplifications of 1), 2) apply.  Thus force_tac, auto_tac etc. shall
  3886 solve record problems automatically.
  3887 
  3888 
  3889 * reorganized the main HOL image: HOL/Integ and String loaded by
  3890 default; theory Main includes everything;
  3891 
  3892 * automatic simplification of integer sums and comparisons, using cancellation;
  3893 
  3894 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
  3895 
  3896 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
  3897 
  3898 * many new identities for unions, intersections, set difference, etc.;
  3899 
  3900 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
  3901 called split_if, split_split, split_sum_case and split_nat_case (to go
  3902 with add/delsplits);
  3903 
  3904 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
  3905 (?x::unit) = (); this is made part of the default simpset, which COULD
  3906 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
  3907 'Delsimprocs [unit_eq_proc];' as last resort); also note that
  3908 unit_abs_eta_conv is added in order to counter the effect of
  3909 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
  3910 %u.f();
  3911 
  3912 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
  3913 makes more sense);
  3914 
  3915 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  3916   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  3917   disjointness reasoning but breaking a few old proofs.
  3918 
  3919 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
  3920 to 'converse' from 'inverse' (for compatibility with ZF and some
  3921 literature);
  3922 
  3923 * HOL/recdef can now declare non-recursive functions, with {} supplied as
  3924 the well-founded relation;
  3925 
  3926 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
  3927     Compl A.  The "Compl" syntax remains available as input syntax for this
  3928     release ONLY.
  3929 
  3930 * HOL/Update: new theory of function updates:
  3931     f(a:=b) == %x. if x=a then b else f x
  3932 may also be iterated as in f(a:=b,c:=d,...);
  3933 
  3934 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
  3935 
  3936 * HOL/List:
  3937   - new function list_update written xs[i:=v] that updates the i-th
  3938     list position. May also be iterated as in xs[i:=a,j:=b,...].
  3939   - new function `upt' written [i..j(] which generates the list
  3940     [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
  3941     bound write [i..j], which is a shorthand for [i..j+1(].
  3942   - new lexicographic orderings and corresponding wellfoundedness theorems.
  3943 
  3944 * HOL/Arith:
  3945   - removed 'pred' (predecessor) function;
  3946   - generalized some theorems about n-1;
  3947   - many new laws about "div" and "mod";
  3948   - new laws about greatest common divisors (see theory ex/Primes);
  3949 
  3950 * HOL/Relation: renamed the relational operator r^-1 "converse"
  3951 instead of "inverse";
  3952 
  3953 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
  3954   of the multiset ordering;
  3955 
  3956 * directory HOL/Real: a construction of the reals using Dedekind cuts
  3957   (not included by default);
  3958 
  3959 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
  3960 
  3961 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
  3962   programs, i.e. different program variables may have different types.
  3963 
  3964 * calling (stac rew i) now fails if "rew" has no effect on the goal
  3965   [previously, this check worked only if the rewrite rule was unconditional]
  3966   Now rew can involve either definitions or equalities (either == or =).
  3967 
  3968 
  3969 *** ZF ***
  3970 
  3971 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
  3972   only the theorems proved on ZF.ML;
  3973 
  3974 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  3975   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  3976   disjointness reasoning but breaking a few old proofs.
  3977 
  3978 * ZF/Update: new theory of function updates
  3979     with default rewrite rule  f(x:=y) ` z = if(z=x, y, f`z)
  3980   may also be iterated as in f(a:=b,c:=d,...);
  3981 
  3982 * in  let x=t in u(x), neither t nor u(x) has to be an FOL term.
  3983 
  3984 * calling (stac rew i) now fails if "rew" has no effect on the goal
  3985   [previously, this check worked only if the rewrite rule was unconditional]
  3986   Now rew can involve either definitions or equalities (either == or =).
  3987 
  3988 * case_tac provided for compatibility with HOL
  3989     (like the old excluded_middle_tac, but with subgoals swapped)
  3990 
  3991 
  3992 *** Internal programming interfaces ***
  3993 
  3994 * Pure: several new basic modules made available for general use, see
  3995 also src/Pure/README;
  3996 
  3997 * improved the theory data mechanism to support encapsulation (data
  3998 kind name replaced by private Object.kind, acting as authorization
  3999 key); new type-safe user interface via functor TheoryDataFun; generic
  4000 print_data function becomes basically useless;
  4001 
  4002 * removed global_names compatibility flag -- all theory declarations
  4003 are qualified by default;
  4004 
  4005 * module Pure/Syntax now offers quote / antiquote translation
  4006 functions (useful for Hoare logic etc. with implicit dependencies);
  4007 see HOL/ex/Antiquote for an example use;
  4008 
  4009 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
  4010 cterm -> thm;
  4011 
  4012 * new tactical CHANGED_GOAL for checking that a tactic modifies a
  4013 subgoal;
  4014 
  4015 * Display.print_goals function moved to Locale.print_goals;
  4016 
  4017 * standard print function for goals supports current_goals_markers
  4018 variable for marking begin of proof, end of proof, start of goal; the
  4019 default is ("", "", ""); setting current_goals_markers := ("<proof>",
  4020 "</proof>", "<goal>") causes SGML like tagged proof state printing,
  4021 for example;
  4022 
  4023 
  4024 
  4025 New in Isabelle98 (January 1998)
  4026 --------------------------------
  4027 
  4028 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  4029 
  4030 * changed lexical syntax of terms / types: dots made part of long
  4031 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
  4032 
  4033 * simpset (and claset) reference variable replaced by functions
  4034 simpset / simpset_ref;
  4035 
  4036 * no longer supports theory aliases (via merge) and non-trivial
  4037 implicit merge of thms' signatures;
  4038 
  4039 * most internal names of constants changed due to qualified names;
  4040 
  4041 * changed Pure/Sequence interface (see Pure/seq.ML);
  4042 
  4043 
  4044 *** General Changes ***
  4045 
  4046 * hierachically structured name spaces (for consts, types, axms, thms
  4047 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
  4048 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
  4049 isatool fixdots ensures space after dots (e.g. "%x. x"); set
  4050 long_names for fully qualified output names; NOTE: ML programs
  4051 (special tactics, packages etc.) referring to internal names may have
  4052 to be adapted to cope with fully qualified names; in case of severe
  4053 backward campatibility problems try setting 'global_names' at compile
  4054 time to have enrything declared within a flat name space; one may also
  4055 fine tune name declarations in theories via the 'global' and 'local'
  4056 section;
  4057 
  4058 * reimplemented the implicit simpset and claset using the new anytype
  4059 data filed in signatures; references simpset:simpset ref etc. are
  4060 replaced by functions simpset:unit->simpset and
  4061 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
  4062 to patch your ML files accordingly;
  4063 
  4064 * HTML output now includes theory graph data for display with Java
  4065 applet or isatool browser; data generated automatically via isatool
  4066 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
  4067 
  4068 * defs may now be conditional; improved rewrite_goals_tac to handle
  4069 conditional equations;
  4070 
  4071 * defs now admits additional type arguments, using TYPE('a) syntax;
  4072 
  4073 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
  4074 creates a new theory node; implicit merge of thms' signatures is
  4075 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
  4076 transfer:theory->thm->thm in (rare) cases;
  4077 
  4078 * improved handling of draft signatures / theories; draft thms (and
  4079 ctyps, cterms) are automatically promoted to real ones;
  4080 
  4081 * slightly changed interfaces for oracles: admit many per theory, named
  4082 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
  4083 
  4084 * print_goals: optional output of const types (set show_consts and
  4085 show_types);
  4086 
  4087 * improved output of warnings (###) and errors (***);
  4088 
  4089 * subgoal_tac displays a warning if the new subgoal has type variables;
  4090 
  4091 * removed old README and Makefiles;
  4092 
  4093 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
  4094 
  4095 * removed obsolete init_pps and init_database;
  4096 
  4097 * deleted the obsolete tactical STATE, which was declared by
  4098     fun STATE tacfun st = tacfun st st;
  4099 
  4100 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
  4101 (which abbreviates $HOME);
  4102 
  4103 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
  4104 use isatool fixseq to adapt your ML programs (this works for fully
  4105 qualified references to the Sequence structure only!);
  4106 
  4107 * use_thy no longer requires writable current directory; it always
  4108 reloads .ML *and* .thy file, if either one is out of date;
  4109 
  4110 
  4111 *** Classical Reasoner ***
  4112 
  4113 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
  4114 tactics that use classical reasoning to simplify a subgoal without
  4115 splitting it into several subgoals;
  4116 
  4117 * Safe_tac: like safe_tac but uses the default claset;
  4118 
  4119 
  4120 *** Simplifier ***
  4121 
  4122 * added simplification meta rules:
  4123     (asm_)(full_)simplify: simpset -> thm -> thm;
  4124 
  4125 * simplifier.ML no longer part of Pure -- has to be loaded by object
  4126 logics (again);
  4127 
  4128 * added prems argument to simplification procedures;
  4129 
  4130 * HOL, FOL, ZF: added infix function `addsplits':
  4131   instead of `<simpset> setloop (split_tac <thms>)'
  4132   you can simply write `<simpset> addsplits <thms>'
  4133 
  4134 
  4135 *** Syntax ***
  4136 
  4137 * TYPE('a) syntax for type reflection terms;
  4138 
  4139 * no longer handles consts with name "" -- declare as 'syntax' instead;
  4140 
  4141 * pretty printer: changed order of mixfix annotation preference (again!);
  4142 
  4143 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
  4144 
  4145 
  4146 *** HOL ***
  4147 
  4148 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4149   with `addloop' of the simplifier to faciliate case splitting in premises.
  4150 
  4151 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
  4152 
  4153 * HOL/Auth: new protocol proofs including some for the Internet
  4154   protocol TLS;
  4155 
  4156 * HOL/Map: new theory of `maps' a la VDM;
  4157 
  4158 * HOL/simplifier: simplification procedures nat_cancel_sums for
  4159 cancelling out common nat summands from =, <, <= (in)equalities, or
  4160 differences; simplification procedures nat_cancel_factor for
  4161 cancelling common factor from =, <, <= (in)equalities over natural
  4162 sums; nat_cancel contains both kinds of procedures, it is installed by
  4163 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
  4164 
  4165 * HOL/simplifier: terms of the form
  4166   `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)'  (or t=x)
  4167   are rewritten to
  4168   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
  4169   and those of the form
  4170   `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)'  (or t=x)
  4171   are rewritten to
  4172   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
  4173 
  4174 * HOL/datatype
  4175   Each datatype `t' now comes with a theorem `split_t_case' of the form
  4176 
  4177   P(t_case f1 ... fn x) =
  4178      ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
  4179         ...
  4180        (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
  4181      )
  4182 
  4183   and a theorem `split_t_case_asm' of the form
  4184 
  4185   P(t_case f1 ... fn x) =
  4186     ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
  4187         ...
  4188        (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
  4189      )
  4190   which can be added to a simpset via `addsplits'. The existing theorems
  4191   expand_list_case and expand_option_case have been renamed to
  4192   split_list_case and split_option_case.
  4193 
  4194 * HOL/Arithmetic:
  4195   - `pred n' is automatically converted to `n-1'.
  4196     Users are strongly encouraged not to use `pred' any longer,
  4197     because it will disappear altogether at some point.
  4198   - Users are strongly encouraged to write "0 < n" rather than
  4199     "n ~= 0". Theorems and proof tools have been modified towards this
  4200     `standard'.
  4201 
  4202 * HOL/Lists:
  4203   the function "set_of_list" has been renamed "set" (and its theorems too);
  4204   the function "nth" now takes its arguments in the reverse order and
  4205   has acquired the infix notation "!" as in "xs!n".
  4206 
  4207 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
  4208 
  4209 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
  4210   specialist theorems (like UN1_I) are gone.  Similarly for (INT x.B x);
  4211 
  4212 * HOL/record: extensible records with schematic structural subtyping
  4213 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
  4214 still lacks various theorems and concrete record syntax;
  4215 
  4216 
  4217 *** HOLCF ***
  4218 
  4219 * removed "axioms" and "generated by" sections;
  4220 
  4221 * replaced "ops" section by extended "consts" section, which is capable of
  4222   handling the continuous function space "->" directly;
  4223 
  4224 * domain package:
  4225   . proves theorems immediately and stores them in the theory,
  4226   . creates hierachical name space,
  4227   . now uses normal mixfix annotations (instead of cinfix...),
  4228   . minor changes to some names and values (for consistency),
  4229   . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
  4230   . separator between mutual domain defs: changed "," to "and",
  4231   . improved handling of sort constraints;  now they have to
  4232     appear on the left-hand side of the equations only;
  4233 
  4234 * fixed LAM <x,y,zs>.b syntax;
  4235 
  4236 * added extended adm_tac to simplifier in HOLCF -- can now discharge
  4237 adm (%x. P (t x)), where P is chainfinite and t continuous;
  4238 
  4239 
  4240 *** FOL and ZF ***
  4241 
  4242 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4243   with `addloop' of the simplifier to faciliate case splitting in premises.
  4244 
  4245 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
  4246 in HOL, they strip ALL and --> from proved theorems;
  4247 
  4248 
  4249 
  4250 New in Isabelle94-8 (May 1997)
  4251 ------------------------------
  4252 
  4253 *** General Changes ***
  4254 
  4255 * new utilities to build / run / maintain Isabelle etc. (in parts
  4256 still somewhat experimental); old Makefiles etc. still functional;
  4257 
  4258 * new 'Isabelle System Manual';
  4259 
  4260 * INSTALL text, together with ./configure and ./build scripts;
  4261 
  4262 * reimplemented type inference for greater efficiency, better error
  4263 messages and clean internal interface;
  4264 
  4265 * prlim command for dealing with lots of subgoals (an easier way of
  4266 setting goals_limit);
  4267 
  4268 
  4269 *** Syntax ***
  4270 
  4271 * supports alternative (named) syntax tables (parser and pretty
  4272 printer); internal interface is provided by add_modesyntax(_i);
  4273 
  4274 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
  4275 be used in conjunction with the Isabelle symbol font; uses the
  4276 "symbols" syntax table;
  4277 
  4278 * added token_translation interface (may translate name tokens in
  4279 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
  4280 the current print_mode); IMPORTANT: user print translation functions
  4281 are responsible for marking newly introduced bounds
  4282 (Syntax.mark_boundT);
  4283 
  4284 * token translations for modes "xterm" and "xterm_color" that display
  4285 names in bold, underline etc. or colors (which requires a color
  4286 version of xterm);
  4287 
  4288 * infixes may now be declared with names independent of their syntax;
  4289 
  4290 * added typed_print_translation (like print_translation, but may
  4291 access type of constant);
  4292 
  4293 
  4294 *** Classical Reasoner ***
  4295 
  4296 Blast_tac: a new tactic!  It is often more powerful than fast_tac, but has
  4297 some limitations.  Blast_tac...
  4298   + ignores addss, addbefore, addafter; this restriction is intrinsic
  4299   + ignores elimination rules that don't have the correct format
  4300         (the conclusion MUST be a formula variable)
  4301   + ignores types, which can make HOL proofs fail
  4302   + rules must not require higher-order unification, e.g. apply_type in ZF
  4303     [message "Function Var's argument not a bound variable" relates to this]
  4304   + its proof strategy is more general but can actually be slower
  4305 
  4306 * substitution with equality assumptions no longer permutes other
  4307 assumptions;
  4308 
  4309 * minor changes in semantics of addafter (now called addaltern); renamed
  4310 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
  4311 (and access functions for it);
  4312 
  4313 * improved combination of classical reasoner and simplifier:
  4314   + functions for handling clasimpsets
  4315   + improvement of addss: now the simplifier is called _after_ the
  4316     safe steps.
  4317   + safe variant of addss called addSss: uses safe simplifications
  4318     _during_ the safe steps. It is more complete as it allows multiple
  4319     instantiations of unknowns (e.g. with slow_tac).
  4320 
  4321 *** Simplifier ***
  4322 
  4323 * added interface for simplification procedures (functions that
  4324 produce *proven* rewrite rules on the fly, depending on current
  4325 redex);
  4326 
  4327 * ordering on terms as parameter (used for ordered rewriting);
  4328 
  4329 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
  4330 
  4331 * the solver is now split into a safe and an unsafe part.
  4332 This should be invisible for the normal user, except that the
  4333 functions setsolver and addsolver have been renamed to setSolver and
  4334 addSolver; added safe_asm_full_simp_tac;
  4335 
  4336 
  4337 *** HOL ***
  4338 
  4339 * a generic induction tactic `induct_tac' which works for all datatypes and
  4340 also for type `nat';
  4341 
  4342 * a generic case distinction tactic `exhaust_tac' which works for all
  4343 datatypes and also for type `nat';
  4344 
  4345 * each datatype comes with a function `size';
  4346 
  4347 * patterns in case expressions allow tuple patterns as arguments to
  4348 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
  4349 
  4350 * primrec now also works with type nat;
  4351 
  4352 * recdef: a new declaration form, allows general recursive functions to be
  4353 defined in theory files.  See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
  4354 
  4355 * the constant for negation has been renamed from "not" to "Not" to
  4356 harmonize with FOL, ZF, LK, etc.;
  4357 
  4358 * HOL/ex/LFilter theory of a corecursive "filter" functional for
  4359 infinite lists;
  4360 
  4361 * HOL/Modelcheck demonstrates invocation of model checker oracle;
  4362 
  4363 * HOL/ex/Ring.thy declares cring_simp, which solves equational
  4364 problems in commutative rings, using axiomatic type classes for + and *;
  4365 
  4366 * more examples in HOL/MiniML and HOL/Auth;
  4367 
  4368 * more default rewrite rules for quantifiers, union/intersection;
  4369 
  4370 * a new constant `arbitrary == @x.False';
  4371 
  4372 * HOLCF/IOA replaces old HOL/IOA;
  4373 
  4374 * HOLCF changes: derived all rules and arities
  4375   + axiomatic type classes instead of classes
  4376   + typedef instead of faking type definitions
  4377   + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
  4378   + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
  4379   + eliminated the types void, one, tr
  4380   + use unit lift and bool lift (with translations) instead of one and tr
  4381   + eliminated blift from Lift3.thy (use Def instead of blift)
  4382   all eliminated rules are derived as theorems --> no visible changes ;
  4383 
  4384 
  4385 *** ZF ***
  4386 
  4387 * ZF now has Fast_tac, Simp_tac and Auto_tac.  Union_iff is a now a default
  4388 rewrite rule; this may affect some proofs.  eq_cs is gone but can be put back
  4389 as ZF_cs addSIs [equalityI];
  4390 
  4391 
  4392 
  4393 New in Isabelle94-7 (November 96)
  4394 ---------------------------------
  4395 
  4396 * allowing negative levels (as offsets) in prlev and choplev;
  4397 
  4398 * super-linear speedup for large simplifications;
  4399 
  4400 * FOL, ZF and HOL now use miniscoping: rewriting pushes
  4401 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
  4402 FAIL); can suppress it using the command Delsimps (ex_simps @
  4403 all_simps); De Morgan laws are also now included, by default;
  4404 
  4405 * improved printing of ==>  :  ~:
  4406 
  4407 * new object-logic "Sequents" adds linear logic, while replacing LK
  4408 and Modal (thanks to Sara Kalvala);
  4409 
  4410 * HOL/Auth: correctness proofs for authentication protocols;
  4411 
  4412 * HOL: new auto_tac combines rewriting and classical reasoning (many
  4413 examples on HOL/Auth);
  4414 
  4415 * HOL: new command AddIffs for declaring theorems of the form P=Q to
  4416 the rewriter and classical reasoner simultaneously;
  4417 
  4418 * function uresult no longer returns theorems in "standard" format;
  4419 regain previous version by: val uresult = standard o uresult;
  4420 
  4421 
  4422 
  4423 New in Isabelle94-6
  4424 -------------------
  4425 
  4426 * oracles -- these establish an interface between Isabelle and trusted
  4427 external reasoners, which may deliver results as theorems;
  4428 
  4429 * proof objects (in particular record all uses of oracles);
  4430 
  4431 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
  4432 
  4433 * "constdefs" section in theory files;
  4434 
  4435 * "primrec" section (HOL) no longer requires names;
  4436 
  4437 * internal type "tactic" now simply "thm -> thm Sequence.seq";
  4438 
  4439 
  4440 
  4441 New in Isabelle94-5
  4442 -------------------
  4443 
  4444 * reduced space requirements;
  4445 
  4446 * automatic HTML generation from theories;
  4447 
  4448 * theory files no longer require "..." (quotes) around most types;
  4449 
  4450 * new examples, including two proofs of the Church-Rosser theorem;
  4451 
  4452 * non-curried (1994) version of HOL is no longer distributed;
  4453 
  4454 
  4455 
  4456 New in Isabelle94-4
  4457 -------------------
  4458 
  4459 * greatly reduced space requirements;
  4460 
  4461 * theory files (.thy) no longer require \...\ escapes at line breaks;
  4462 
  4463 * searchable theorem database (see the section "Retrieving theorems" on
  4464 page 8 of the Reference Manual);
  4465 
  4466 * new examples, including Grabczewski's monumental case study of the
  4467 Axiom of Choice;
  4468 
  4469 * The previous version of HOL renamed to Old_HOL;
  4470 
  4471 * The new version of HOL (previously called CHOL) uses a curried syntax
  4472 for functions.  Application looks like f a b instead of f(a,b);
  4473 
  4474 * Mutually recursive inductive definitions finally work in HOL;
  4475 
  4476 * In ZF, pattern-matching on tuples is now available in all abstractions and
  4477 translates to the operator "split";
  4478 
  4479 
  4480 
  4481 New in Isabelle94-3
  4482 -------------------
  4483 
  4484 * new infix operator, addss, allowing the classical reasoner to
  4485 perform simplification at each step of its search.  Example:
  4486         fast_tac (cs addss ss)
  4487 
  4488 * a new logic, CHOL, the same as HOL, but with a curried syntax
  4489 for functions.  Application looks like f a b instead of f(a,b).  Also pairs
  4490 look like (a,b) instead of <a,b>;
  4491 
  4492 * PLEASE NOTE: CHOL will eventually replace HOL!
  4493 
  4494 * In CHOL, pattern-matching on tuples is now available in all abstractions.
  4495 It translates to the operator "split".  A new theory of integers is available;
  4496 
  4497 * In ZF, integer numerals now denote two's-complement binary integers.
  4498 Arithmetic operations can be performed by rewriting.  See ZF/ex/Bin.ML;
  4499 
  4500 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
  4501 of the Axiom of Choice;
  4502 
  4503 
  4504 
  4505 New in Isabelle94-2
  4506 -------------------
  4507 
  4508 * Significantly faster resolution;
  4509 
  4510 * the different sections in a .thy file can now be mixed and repeated
  4511 freely;
  4512 
  4513 * Database of theorems for FOL, HOL and ZF.  New
  4514 commands including qed, qed_goal and bind_thm store theorems in the database.
  4515 
  4516 * Simple database queries: return a named theorem (get_thm) or all theorems of
  4517 a given theory (thms_of), or find out what theory a theorem was proved in
  4518 (theory_of_thm);
  4519 
  4520 * Bugs fixed in the inductive definition and datatype packages;
  4521 
  4522 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
  4523 and HOL_dup_cs obsolete;
  4524 
  4525 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
  4526 have been removed;
  4527 
  4528 * Simpler definition of function space in ZF;
  4529 
  4530 * new results about cardinal and ordinal arithmetic in ZF;
  4531 
  4532 * 'subtype' facility in HOL for introducing new types as subsets of existing
  4533 types;
  4534 
  4535 
  4536 $Id$