NEWS
author haftmann
Sat Sep 15 19:27:42 2007 +0200 (2007-09-15)
changeset 24587 4f2cbf6e563f
parent 24507 ac22a2a67100
child 24606 7acbb982fc77
permissions -rw-r--r--
multi-functional value keyword
     1 Isabelle NEWS -- history user-relevant changes
     2 ==============================================
     3 
     4 New in this Isabelle version
     5 ----------------------------
     6 
     7 *** General ***
     8 
     9 * More uniform information about legacy features, notably a
    10 warning/error of "Legacy feature: ...", depending on the state of the
    11 tolerate_legacy_features flag (default true). FUTURE INCOMPATIBILITY:
    12 legacy features will disappear eventually.
    13 
    14 * Theory syntax: the header format ``theory A = B + C:'' has been
    15 discontinued in favour of ``theory A imports B C begin''.  Use isatool
    16 fixheaders to convert existing theory files.  INCOMPATIBILITY.
    17 
    18 * Theory syntax: the old non-Isar theory file format has been
    19 discontinued altogether.  Note that ML proof scripts may still be used
    20 with Isar theories; migration is usually quite simple with the ML
    21 function use_legacy_bindings.  INCOMPATIBILITY.
    22 
    23 * Theory syntax: some popular names (e.g. 'class', 'declaration',
    24 'fun', 'help', 'if') are now keywords.  INCOMPATIBILITY, use double
    25 quotes.
    26 
    27 * Theory loader: be more serious about observing the static theory
    28 header specifications (including optional directories), but not the
    29 accidental file locations of previously successful loads.  The strict
    30 update policy of former update_thy is now already performed by
    31 use_thy, so the former has been removed; use_thys updates several
    32 theories simultaneously, just as 'imports' within a theory header
    33 specification, but without merging the results.  Potential
    34 INCOMPATIBILITY: may need to refine theory headers and commands
    35 ROOT.ML which depend on load order.
    36 
    37 * Theory loader: optional support for content-based file
    38 identification, instead of the traditional scheme of full physical
    39 path plus date stamp; configured by the ISABELLE_FILE_IDENT setting
    40 (cf. the system manual).  The new scheme allows to work with
    41 non-finished theories in persistent session images, such that source
    42 files may be moved later on without requiring reloads.
    43 
    44 * Theory loader: old-style ML proof scripts being *attached* to a thy
    45 file (with the same base name as the theory) are considered a legacy
    46 feature, which will disappear eventually. Even now, the theory loader no
    47 longer maintains dependencies on such files.
    48 
    49 * Syntax: the scope for resolving ambiguities via type-inference is now
    50 limited to individual terms, instead of whole simultaneous
    51 specifications as before. This greatly reduces the complexity of the
    52 syntax module and improves flexibility by separating parsing and
    53 type-checking. INCOMPATIBILITY: additional type-constraints (explicit
    54 'fixes' etc.) are required in rare situations.
    55 
    56 * Legacy goal package: reduced interface to the bare minimum required
    57 to keep existing proof scripts running.  Most other user-level
    58 functions are now part of the OldGoals structure, which is *not* open
    59 by default (consider isatool expandshort before open OldGoals).
    60 Removed top_sg, prin, printyp, pprint_term/typ altogether, because
    61 these tend to cause confusion about the actual goal (!) context being
    62 used here, which is not necessarily the same as the_context().
    63 
    64 * Command 'find_theorems': supports "*" wild-card in "name:"
    65 criterion; "with_dups" option.  Certain ProofGeneral versions might
    66 support a specific search form (see ProofGeneral/CHANGES).
    67 
    68 * The ``prems limit'' option (cf. ProofContext.prems_limit) is now -1
    69 by default, which means that "prems" (and also "fixed variables") are
    70 suppressed from proof state output.  Note that the ProofGeneral
    71 settings mechanism allows to change and save options persistently, but
    72 older versions of Isabelle will fail to start up if a negative prems
    73 limit is imposed.
    74 
    75 * Local theory targets may be specified by non-nested blocks of
    76 ``context/locale/class ... begin'' followed by ``end''.  The body may
    77 contain definitions, theorems etc., including any derived mechanism
    78 that has been implemented on top of these primitives.  This concept
    79 generalizes the existing ``theorem (in ...)'' towards more versatility
    80 and scalability.
    81 
    82 * Proof General interface: proper undo of final 'end' command;
    83 discontinued Isabelle/classic mode (ML proof scripts).
    84 
    85 
    86 *** Document preparation ***
    87 
    88 * Added antiquotation @{theory name} which prints the given name,
    89 after checking that it refers to a valid ancestor theory in the
    90 current context.
    91 
    92 * Added antiquotations @{ML_type text} and @{ML_struct text} which
    93 check the given source text as ML type/structure, printing verbatim.
    94 
    95 * Added antiquotation @{abbrev "c args"} which prints the abbreviation
    96 "c args == rhs" given in the current context.  (Any number of
    97 arguments may be given on the LHS.)
    98 
    99 
   100 
   101 *** Pure ***
   102 
   103 * code generator: consts in 'consts_code' Isar commands are now referred
   104   to by usual term syntax (including optional type annotations).
   105 
   106 * code generator: 
   107   - Isar 'definition's, 'constdef's and primitive instance definitions are added
   108     explicitly to the table of defining equations
   109   - primitive definitions are not used as defining equations by default any longer
   110   - defining equations are now definitly restricted to meta "==" and object
   111         equality "="
   112   - HOL theories have been adopted accordingly
   113 
   114 * class_package.ML offers a combination of axclasses and locales to
   115 achieve Haskell-like type classes in Isabelle.  See
   116 HOL/ex/Classpackage.thy for examples.
   117 
   118 * Yet another code generator framework allows to generate executable
   119 code for ML and Haskell (including "class"es).  A short usage sketch:
   120 
   121     internal compilation:
   122         code_gen <list of constants (term syntax)> in SML
   123     writing SML code to a file:
   124         code_gen <list of constants (term syntax)> in SML <filename>
   125     writing OCaml code to a file:
   126         code_gen <list of constants (term syntax)> in OCaml <filename>
   127     writing Haskell code to a bunch of files:
   128         code_gen <list of constants (term syntax)> in Haskell <filename>
   129 
   130 Reasonable default setup of framework in HOL/Main.
   131 
   132 Theorem attributs for selecting and transforming function equations theorems:
   133 
   134     [code fun]:        select a theorem as function equation for a specific constant
   135     [code fun del]:    deselect a theorem as function equation for a specific constant
   136     [code inline]:     select an equation theorem for unfolding (inlining) in place
   137     [code inline del]: deselect an equation theorem for unfolding (inlining) in place
   138 
   139 User-defined serializations (target in {SML, OCaml, Haskell}):
   140 
   141     code_const <and-list of constants (term syntax)>
   142       {(target) <and-list of const target syntax>}+
   143 
   144     code_type <and-list of type constructors>
   145       {(target) <and-list of type target syntax>}+
   146 
   147     code_instance <and-list of instances>
   148       {(target)}+
   149         where instance ::= <type constructor> :: <class>
   150 
   151     code_class <and_list of classes>
   152       {(target) <and-list of class target syntax>}+
   153         where class target syntax ::= <class name> {where {<classop> == <target syntax>}+}?
   154 
   155 code_instance and code_class only apply to target Haskell.
   156 
   157 See HOL theories and HOL/ex/Codegenerator*.thy for usage examples.
   158 Doc/Isar/Advanced/Codegen/ provides a tutorial.
   159 
   160 * Command 'no_translations' removes translation rules from theory
   161 syntax.
   162 
   163 * Overloaded definitions are now actually checked for acyclic
   164 dependencies.  The overloading scheme is slightly more general than
   165 that of Haskell98, although Isabelle does not demand an exact
   166 correspondence to type class and instance declarations.
   167 INCOMPATIBILITY, use ``defs (unchecked overloaded)'' to admit more
   168 exotic versions of overloading -- at the discretion of the user!
   169 
   170 Polymorphic constants are represented via type arguments, i.e. the
   171 instantiation that matches an instance against the most general
   172 declaration given in the signature.  For example, with the declaration
   173 c :: 'a => 'a => 'a, an instance c :: nat => nat => nat is represented
   174 as c(nat).  Overloading is essentially simultaneous structural
   175 recursion over such type arguments.  Incomplete specification patterns
   176 impose global constraints on all occurrences, e.g. c('a * 'a) on the
   177 LHS means that more general c('a * 'b) will be disallowed on any RHS.
   178 Command 'print_theory' outputs the normalized system of recursive
   179 equations, see section "definitions".
   180 
   181 * Configuration options are maintained within the theory or proof
   182 context (with name and type bool/int/string), providing a very simple
   183 interface to a poor-man's version of general context data.  Tools may
   184 declare options in ML (e.g. using Attrib.config_int) and then refer to
   185 these values using Config.get etc.  Users may change options via an
   186 associated attribute of the same name.  This form of context
   187 declaration works particularly well with commands 'declare' or
   188 'using', for example ``declare [[foo = 42]]''.  Thus it has become
   189 very easy to avoid global references, which would not observe Isar
   190 toplevel undo/redo and fail to work with multithreading.
   191 
   192 Various global ML references of Pure and HOL have been turned into
   193 configuration options:
   194 
   195   Unify.search_bound		unify_search_bound
   196   Unify.trace_bound		unify_trace_bound
   197   Unify.trace_simp		unify_trace_simp
   198   Unify.trace_types		unify_trace_types
   199   Simplifier.simp_depth_limit	simp_depth_limit
   200   Blast.depth_limit		blast_depth_limit
   201   DatatypeProp.dtK		datatype_distinctness_limit
   202   fast_arith_neq_limit  	fast_arith_neq_limit
   203   fast_arith_split_limit	fast_arith_split_limit
   204 
   205 * Named collections of theorems may be easily installed as context
   206 data using the functor NamedThmsFun (see
   207 src/Pure/Tools/named_thms.ML).  The user may add or delete facts via
   208 attributes; there is also a toplevel print command.  This facility is
   209 just a common case of general context data, which is the preferred way
   210 for anything more complex than just a list of facts in canonical
   211 order.
   212 
   213 * Isar: command 'declaration' augments a local theory by generic
   214 declaration functions written in ML.  This enables arbitrary content
   215 being added to the context, depending on a morphism that tells the
   216 difference of the original declaration context wrt. the application
   217 context encountered later on.
   218 
   219 * Isar: proper interfaces for simplification procedures.  Command
   220 'simproc_setup' declares named simprocs (with match patterns, and body
   221 text in ML).  Attribute "simproc" adds/deletes simprocs in the current
   222 context.  ML antiquotation @{simproc name} retrieves named simprocs.
   223 
   224 * Isar: an extra pair of brackets around attribute declarations
   225 abbreviates a theorem reference involving an internal dummy fact,
   226 which will be ignored later --- only the effect of the attribute on
   227 the background context will persist.  This form of in-place
   228 declarations is particularly useful with commands like 'declare' and
   229 'using', for example ``have A using [[simproc a]] by simp''.
   230 
   231 * Isar: method "assumption" (and implicit closing of subproofs) now
   232 takes simple non-atomic goal assumptions into account: after applying
   233 an assumption as a rule the resulting subgoals are solved by atomic
   234 assumption steps.  This is particularly useful to finish 'obtain'
   235 goals, such as "!!x. (!!x. P x ==> thesis) ==> P x ==> thesis",
   236 without referring to the original premise "!!x. P x ==> thesis" in the
   237 Isar proof context.  POTENTIAL INCOMPATIBILITY: method "assumption" is
   238 more permissive.
   239 
   240 * Isar: implicit use of prems from the Isar proof context is
   241 considered a legacy feature.  Common applications like ``have A .''
   242 may be replaced by ``have A by fact'' or ``note `A`''.  In general,
   243 referencing facts explicitly here improves readability and
   244 maintainability of proof texts.
   245 
   246 * Isar: improper proof element 'guess' is like 'obtain', but derives
   247 the obtained context from the course of reasoning!  For example:
   248 
   249   assume "EX x y. A x & B y"   -- "any previous fact"
   250   then guess x and y by clarify
   251 
   252 This technique is potentially adventurous, depending on the facts and
   253 proof tools being involved here.
   254 
   255 * Isar: known facts from the proof context may be specified as literal
   256 propositions, using ASCII back-quote syntax.  This works wherever
   257 named facts used to be allowed so far, in proof commands, proof
   258 methods, attributes etc.  Literal facts are retrieved from the context
   259 according to unification of type and term parameters.  For example,
   260 provided that "A" and "A ==> B" and "!!x. P x ==> Q x" are known
   261 theorems in the current context, then these are valid literal facts:
   262 `A` and `A ==> B` and `!!x. P x ==> Q x" as well as `P a ==> Q a` etc.
   263 
   264 There is also a proof method "fact" which does the same composition
   265 for explicit goal states, e.g. the following proof texts coincide with
   266 certain special cases of literal facts:
   267 
   268   have "A" by fact                 ==  note `A`
   269   have "A ==> B" by fact           ==  note `A ==> B`
   270   have "!!x. P x ==> Q x" by fact  ==  note `!!x. P x ==> Q x`
   271   have "P a ==> Q a" by fact       ==  note `P a ==> Q a`
   272 
   273 * Isar: ":" (colon) is no longer a symbolic identifier character in
   274 outer syntax.  Thus symbolic identifiers may be used without
   275 additional white space in declarations like this: ``assume *: A''.
   276 
   277 * Isar: 'print_facts' prints all local facts of the current context,
   278 both named and unnamed ones.
   279 
   280 * Isar: 'def' now admits simultaneous definitions, e.g.:
   281 
   282   def x == "t" and y == "u"
   283 
   284 * Isar: added command 'unfolding', which is structurally similar to
   285 'using', but affects both the goal state and facts by unfolding given
   286 rewrite rules.  Thus many occurrences of the 'unfold' method or
   287 'unfolded' attribute may be replaced by first-class proof text.
   288 
   289 * Isar: methods 'unfold' / 'fold', attributes 'unfolded' / 'folded',
   290 and command 'unfolding' now all support object-level equalities
   291 (potentially conditional).  The underlying notion of rewrite rule is
   292 analogous to the 'rule_format' attribute, but *not* that of the
   293 Simplifier (which is usually more generous).
   294 
   295 * Isar: the new attribute [rotated n] (default n = 1) rotates the
   296 premises of a theorem by n. Useful in conjunction with drule.
   297 
   298 * Isar: the goal restriction operator [N] (default N = 1) evaluates a
   299 method expression within a sandbox consisting of the first N
   300 sub-goals, which need to exist.  For example, ``simp_all [3]''
   301 simplifies the first three sub-goals, while (rule foo, simp_all)[]
   302 simplifies all new goals that emerge from applying rule foo to the
   303 originally first one.
   304 
   305 * Isar: schematic goals are no longer restricted to higher-order
   306 patterns; e.g. ``lemma "?P(?x)" by (rule TrueI)'' now works as
   307 expected.
   308 
   309 * Isar: the conclusion of a long theorem statement is now either
   310 'shows' (a simultaneous conjunction, as before), or 'obtains'
   311 (essentially a disjunction of cases with local parameters and
   312 assumptions).  The latter allows to express general elimination rules
   313 adequately; in this notation common elimination rules look like this:
   314 
   315   lemma exE:    -- "EX x. P x ==> (!!x. P x ==> thesis) ==> thesis"
   316     assumes "EX x. P x"
   317     obtains x where "P x"
   318 
   319   lemma conjE:  -- "A & B ==> (A ==> B ==> thesis) ==> thesis"
   320     assumes "A & B"
   321     obtains A and B
   322 
   323   lemma disjE:  -- "A | B ==> (A ==> thesis) ==> (B ==> thesis) ==> thesis"
   324     assumes "A | B"
   325     obtains
   326       A
   327     | B
   328 
   329 The subsequent classical rules even refer to the formal "thesis"
   330 explicitly:
   331 
   332   lemma classical:     -- "(~ thesis ==> thesis) ==> thesis"
   333     obtains "~ thesis"
   334 
   335   lemma Peirce's_Law:  -- "((thesis ==> something) ==> thesis) ==> thesis"
   336     obtains "thesis ==> something"
   337 
   338 The actual proof of an 'obtains' statement is analogous to that of the
   339 Isar proof element 'obtain', only that there may be several cases.
   340 Optional case names may be specified in parentheses; these will be
   341 available both in the present proof and as annotations in the
   342 resulting rule, for later use with the 'cases' method (cf. attribute
   343 case_names).
   344 
   345 * Isar: the assumptions of a long theorem statement are available as
   346 "assms" fact in the proof context.  This is more appropriate than the
   347 (historical) "prems", which refers to all assumptions of the current
   348 context, including those from the target locale, proof body etc.
   349 
   350 * Isar: 'print_statement' prints theorems from the current theory or
   351 proof context in long statement form, according to the syntax of a
   352 top-level lemma.
   353 
   354 * Isar: 'obtain' takes an optional case name for the local context
   355 introduction rule (default "that").
   356 
   357 * Isar: removed obsolete 'concl is' patterns.  INCOMPATIBILITY, use
   358 explicit (is "_ ==> ?foo") in the rare cases where this still happens
   359 to occur.
   360 
   361 * Pure: syntax "CONST name" produces a fully internalized constant
   362 according to the current context.  This is particularly useful for
   363 syntax translations that should refer to internal constant
   364 representations independently of name spaces.
   365 
   366 * Pure: syntax constant for foo (binder "FOO ") is called "foo_binder"
   367 instead of "FOO ". This allows multiple binder declarations to coexist
   368 in the same context.  INCOMPATIBILITY.
   369 
   370 * Isar/locales: 'notation' provides a robust interface to the 'syntax'
   371 primitive that also works in a locale context (both for constants and
   372 fixed variables).  Type declaration and internal syntactic
   373 representation of given constants retrieved from the context.
   374 
   375 * Isar/locales: new derived specification elements 'axiomatization',
   376 'definition', 'abbreviation', which support type-inference, admit
   377 object-level specifications (equality, equivalence).  See also the
   378 isar-ref manual.  Examples:
   379 
   380   axiomatization
   381     eq  (infix "===" 50) where
   382     eq_refl: "x === x" and eq_subst: "x === y ==> P x ==> P y"
   383 
   384   definition "f x y = x + y + 1"
   385   definition g where "g x = f x x"
   386 
   387   abbreviation
   388     neq  (infix "=!=" 50) where
   389     "x =!= y == ~ (x === y)"
   390 
   391 These specifications may be also used in a locale context.  Then the
   392 constants being introduced depend on certain fixed parameters, and the
   393 constant name is qualified by the locale base name.  An internal
   394 abbreviation takes care for convenient input and output, making the
   395 parameters implicit and using the original short name.  See also
   396 HOL/ex/Abstract_NAT.thy for an example of deriving polymorphic
   397 entities from a monomorphic theory.
   398 
   399 Presently, abbreviations are only available 'in' a target locale, but
   400 not inherited by general import expressions.  Also note that
   401 'abbreviation' may be used as a type-safe replacement for 'syntax' +
   402 'translations' in common applications.
   403 
   404 Concrete syntax is attached to specified constants in internal form,
   405 independently of name spaces.  The parse tree representation is
   406 slightly different -- use 'notation' instead of raw 'syntax', and
   407 'translations' with explicit "CONST" markup to accommodate this.
   408 
   409 * Pure: command 'print_abbrevs' prints all constant abbreviations of
   410 the current context.  Print mode "no_abbrevs" prevents inversion of
   411 abbreviations on output.
   412 
   413 * Isar/locales: improved parameter handling:
   414 - use of locales "var" and "struct" no longer necessary;
   415 - parameter renamings are no longer required to be injective.
   416   This enables, for example, to define a locale for endomorphisms thus:
   417   locale endom = homom mult mult h.
   418 
   419 * Isar/locales: changed the way locales with predicates are defined.
   420 Instead of accumulating the specification, the imported expression is
   421 now an interpretation.  INCOMPATIBILITY: different normal form of
   422 locale expressions.  In particular, in interpretations of locales with
   423 predicates, goals repesenting already interpreted fragments are not
   424 removed automatically.  Use methods `intro_locales' and
   425 `unfold_locales'; see below.
   426 
   427 * Isar/locales: new methods `intro_locales' and `unfold_locales'
   428 provide backward reasoning on locales predicates.  The methods are
   429 aware of interpretations and discharge corresponding goals.
   430 `intro_locales' is less aggressive then `unfold_locales' and does not
   431 unfold predicates to assumptions.
   432 
   433 * Isar/locales: the order in which locale fragments are accumulated
   434 has changed.  This enables to override declarations from fragments due
   435 to interpretations -- for example, unwanted simp rules.
   436 
   437 * Isar/locales: interpretation in theories and proof contexts has been
   438 extended.  One may now specify (and prove) equations, which are
   439 unfolded in interpreted theorems.  This is useful for replacing
   440 defined concepts (constants depending on locale parameters) by
   441 concepts already existing in the target context.  Example:
   442 
   443   interpretation partial_order ["op <= :: [int, int] => bool"]
   444     where "partial_order.less (op <=) (x::int) y = (x < y)"
   445 
   446 Typically, the constant `partial_order.less' is created by a definition
   447 specification element in the context of locale partial_order.
   448 
   449 * Provers/induct: improved internal context management to support
   450 local fixes and defines on-the-fly.  Thus explicit meta-level
   451 connectives !! and ==> are rarely required anymore in inductive goals
   452 (using object-logic connectives for this purpose has been long
   453 obsolete anyway).  The subsequent proof patterns illustrate advanced
   454 techniques of natural induction; general datatypes and inductive sets
   455 work analogously (see also src/HOL/Lambda for realistic examples).
   456 
   457 (1) This is how to ``strengthen'' an inductive goal wrt. certain
   458 parameters:
   459 
   460   lemma
   461     fixes n :: nat and x :: 'a
   462     assumes a: "A n x"
   463     shows "P n x"
   464     using a                     -- {* make induct insert fact a *}
   465   proof (induct n arbitrary: x) -- {* generalize goal to "!!x. A n x ==> P n x" *}
   466     case 0
   467     show ?case sorry
   468   next
   469     case (Suc n)
   470     note `!!x. A n x ==> P n x` -- {* induction hypothesis, according to induction rule *}
   471     note `A (Suc n) x`          -- {* induction premise, stemming from fact a *}
   472     show ?case sorry
   473   qed
   474 
   475 (2) This is how to perform induction over ``expressions of a certain
   476 form'', using a locally defined inductive parameter n == "a x"
   477 together with strengthening (the latter is usually required to get
   478 sufficiently flexible induction hypotheses):
   479 
   480   lemma
   481     fixes a :: "'a => nat"
   482     assumes a: "A (a x)"
   483     shows "P (a x)"
   484     using a
   485   proof (induct n == "a x" arbitrary: x)
   486     ...
   487 
   488 See also HOL/Isar_examples/Puzzle.thy for an application of the this
   489 particular technique.
   490 
   491 (3) This is how to perform existential reasoning ('obtains' or
   492 'obtain') by induction, while avoiding explicit object-logic
   493 encodings:
   494 
   495   lemma
   496     fixes n :: nat
   497     obtains x :: 'a where "P n x" and "Q n x"
   498   proof (induct n arbitrary: thesis)
   499     case 0
   500     obtain x where "P 0 x" and "Q 0 x" sorry
   501     then show thesis by (rule 0)
   502   next
   503     case (Suc n)
   504     obtain x where "P n x" and "Q n x" by (rule Suc.hyps)
   505     obtain x where "P (Suc n) x" and "Q (Suc n) x" sorry
   506     then show thesis by (rule Suc.prems)
   507   qed
   508 
   509 Here the 'arbitrary: thesis' specification essentially modifies the
   510 scope of the formal thesis parameter, in order to the get the whole
   511 existence statement through the induction as expected.
   512 
   513 * Provers/induct: mutual induction rules are now specified as a list
   514 of rule sharing the same induction cases.  HOL packages usually
   515 provide foo_bar.inducts for mutually defined items foo and bar
   516 (e.g. inductive sets or datatypes).  INCOMPATIBILITY, users need to
   517 specify mutual induction rules differently, i.e. like this:
   518 
   519   (induct rule: foo_bar.inducts)
   520   (induct set: foo bar)
   521   (induct type: foo bar)
   522 
   523 The ML function ProjectRule.projections turns old-style rules into the
   524 new format.
   525 
   526 * Provers/induct: improved handling of simultaneous goals.  Instead of
   527 introducing object-level conjunction, the statement is now split into
   528 several conclusions, while the corresponding symbolic cases are
   529 nested accordingly.  INCOMPATIBILITY, proofs need to be structured
   530 explicitly.  For example:
   531 
   532   lemma
   533     fixes n :: nat
   534     shows "P n" and "Q n"
   535   proof (induct n)
   536     case 0 case 1
   537     show "P 0" sorry
   538   next
   539     case 0 case 2
   540     show "Q 0" sorry
   541   next
   542     case (Suc n) case 1
   543     note `P n` and `Q n`
   544     show "P (Suc n)" sorry
   545   next
   546     case (Suc n) case 2
   547     note `P n` and `Q n`
   548     show "Q (Suc n)" sorry
   549   qed
   550 
   551 The split into subcases may be deferred as follows -- this is
   552 particularly relevant for goal statements with local premises.
   553 
   554   lemma
   555     fixes n :: nat
   556     shows "A n ==> P n" and "B n ==> Q n"
   557   proof (induct n)
   558     case 0
   559     {
   560       case 1
   561       note `A 0`
   562       show "P 0" sorry
   563     next
   564       case 2
   565       note `B 0`
   566       show "Q 0" sorry
   567     }
   568   next
   569     case (Suc n)
   570     note `A n ==> P n` and `B n ==> Q n`
   571     {
   572       case 1
   573       note `A (Suc n)`
   574       show "P (Suc n)" sorry
   575     next
   576       case 2
   577       note `B (Suc n)`
   578       show "Q (Suc n)" sorry
   579     }
   580   qed
   581 
   582 If simultaneous goals are to be used with mutual rules, the statement
   583 needs to be structured carefully as a two-level conjunction, using
   584 lists of propositions separated by 'and':
   585 
   586   lemma
   587     shows "a : A ==> P1 a"
   588           "a : A ==> P2 a"
   589       and "b : B ==> Q1 b"
   590           "b : B ==> Q2 b"
   591           "b : B ==> Q3 b"
   592   proof (induct set: A B)
   593 
   594 * Provers/induct: support coinduction as well.  See
   595 src/HOL/Library/Coinductive_List.thy for various examples.
   596 
   597 * Attribute "symmetric" produces result with standardized schematic
   598 variables (index 0).  Potential INCOMPATIBILITY.
   599 
   600 * Simplifier: by default the simplifier trace only shows top level
   601 rewrites now. That is, trace_simp_depth_limit is set to 1 by
   602 default. Thus there is less danger of being flooded by the trace. The
   603 trace indicates where parts have been suppressed.
   604   
   605 * Provers/classical: removed obsolete classical version of elim_format
   606 attribute; classical elim/dest rules are now treated uniformly when
   607 manipulating the claset.
   608 
   609 * Provers/classical: stricter checks to ensure that supplied intro,
   610 dest and elim rules are well-formed; dest and elim rules must have at
   611 least one premise.
   612 
   613 * Provers/classical: attributes dest/elim/intro take an optional
   614 weight argument for the rule (just as the Pure versions).  Weights are
   615 ignored by automated tools, but determine the search order of single
   616 rule steps.
   617 
   618 * Syntax: input syntax now supports dummy variable binding "%_. b",
   619 where the body does not mention the bound variable.  Note that dummy
   620 patterns implicitly depend on their context of bounds, which makes
   621 "{_. _}" match any set comprehension as expected.  Potential
   622 INCOMPATIBILITY -- parse translations need to cope with syntactic
   623 constant "_idtdummy" in the binding position.
   624 
   625 * Syntax: removed obsolete syntactic constant "_K" and its associated
   626 parse translation.  INCOMPATIBILITY -- use dummy abstraction instead,
   627 for example "A -> B" => "Pi A (%_. B)".
   628 
   629 * Pure: 'class_deps' command visualizes the subclass relation, using
   630 the graph browser tool.
   631 
   632 * Pure: 'print_theory' now suppresses entities with internal name
   633 (trailing "_") by default; use '!' option for full details.
   634 
   635 
   636 *** HOL ***
   637 
   638 * Formulation of theorem "dense" changed slightly due to integration with new
   639 class dense_linear_order.
   640 
   641 * theory Finite_Set: "name-space" locales Lattice, Distrib_lattice, Linorder etc.
   642 have disappeared; operations defined in terms of fold_set now are named
   643 Inf_fin, Sup_fin.  INCOMPATIBILITY.
   644 
   645 * HOL-Word:
   646   New extensive library and type for generic, fixed size machine
   647   words, with arithemtic, bit-wise, shifting and rotating operations,
   648   reflection into int, nat, and bool lists, automation for linear
   649   arithmetic (by automatic reflection into nat or int), including
   650   lemmas on overflow and monotonicity. Instantiated to all appropriate
   651   arithmetic type classes, supporting automatic simplification of
   652   numerals on all operations. Jointly developed by NICTA, Galois, and
   653   PSU.
   654 
   655 * Library/Boolean_Algebra: locales for abstract boolean algebras.
   656 
   657 * Library/Numeral_Type: numbers as types, e.g. TYPE(32).
   658 
   659 * Code generator library theories:
   660   * Pretty_Int represents HOL integers by big integer literals in target
   661     languages.
   662   * Pretty_Char represents HOL characters by character literals in target
   663     languages.
   664   * Pretty_Char_chr like Pretty_Char, but also offers treatment of character
   665     codes; includes Pretty_Int.
   666   * Executable_Set allows to generate code for finite sets using lists.
   667   * Executable_Rat implements rational numbers as triples (sign, enumerator,
   668     denominator).
   669   * Executable_Real implements a subset of real numbers, namly those
   670     representable by rational numbers.
   671   * Efficient_Nat implements natural numbers by integers, which in general will
   672     result in higher efficency; pattern matching with 0/Suc is eliminated;
   673     includes Pretty_Int.
   674   * ML_String provides an additional datatype ml_string; in the HOL default
   675     setup, strings in HOL are mapped to lists of HOL characters in SML; values
   676     of type ml_string are mapped to strings in SML.
   677   * ML_Int provides an additional datatype ml_int which is mapped to to SML
   678     built-in integers.
   679 
   680 * New package for inductive predicates
   681 
   682   An n-ary predicate p with m parameters z_1, ..., z_m can now be defined via
   683 
   684     inductive
   685       p :: "U_1 => ... => U_m => T_1 => ... => T_n => bool"
   686       for z_1 :: U_1 and ... and z_n :: U_m
   687     where
   688       rule_1: "... ==> p z_1 ... z_m t_1_1 ... t_1_n"
   689     | ...
   690 
   691   rather than
   692 
   693     consts s :: "U_1 => ... => U_m => (T_1 * ... * T_n) set"
   694 
   695     abbreviation p :: "U_1 => ... => U_m => T_1 => ... => T_n => bool"
   696     where "p z_1 ... z_m x_1 ... x_n == (x_1, ..., x_n) : s z_1 ... z_m"
   697 
   698     inductive "s z_1 ... z_m"
   699     intros
   700       rule_1: "... ==> (t_1_1, ..., t_1_n) : s z_1 ... z_m"
   701       ...
   702 
   703   For backward compatibility, there is a wrapper allowing inductive
   704   sets to be defined with the new package via
   705 
   706     inductive_set
   707       s :: "U_1 => ... => U_m => (T_1 * ... * T_n) set"
   708       for z_1 :: U_1 and ... and z_n :: U_m
   709     where
   710       rule_1: "... ==> (t_1_1, ..., t_1_n) : s z_1 ... z_m"
   711     | ...
   712 
   713   or
   714 
   715     inductive_set
   716       s :: "U_1 => ... => U_m => (T_1 * ... * T_n) set"
   717       and p :: "U_1 => ... => U_m => T_1 => ... => T_n => bool"
   718       for z_1 :: U_1 and ... and z_n :: U_m
   719     where
   720       "p z_1 ... z_m x_1 ... x_n == (x_1, ..., x_n) : s z_1 ... z_m"
   721     | rule_1: "... ==> p z_1 ... z_m t_1_1 ... t_1_n"
   722     | ...
   723 
   724   if the additional syntax "p ..." is required.
   725 
   726   Many examples can be found in the subdirectories Auth, Bali, Induct,
   727   or MicroJava.
   728 
   729   INCOMPATIBILITIES:
   730 
   731   - Since declaration and definition of inductive sets or predicates
   732     is no longer separated, abbreviations involving the newly introduced
   733     sets or predicates must be specified together with the introduction
   734     rules after the "where" keyword (see example above), rather than before
   735     the actual inductive definition.
   736 
   737   - The variables in induction and elimination rules are now quantified
   738     in the order of their occurrence in the introduction rules, rather than
   739     in alphabetical order. Since this may break some proofs, these proofs
   740     either have to be repaired, e.g. by reordering the variables
   741     a_i_1 ... a_i_{k_i} in Isar "case" statements of the form
   742 
   743       case (rule_i a_i_1 ... a_i_{k_i})
   744 
   745     or the old order of quantification has to be restored by explicitly adding
   746     meta-level quantifiers in the introduction rules, i.e.
   747 
   748       | rule_i: "!!a_i_1 ... a_i_{k_i}. ... ==> p z_1 ... z_m t_i_1 ... t_i_n"
   749 
   750   - The format of the elimination rules is now
   751 
   752       p z_1 ... z_m x_1 ... x_n ==>
   753         (!!a_1_1 ... a_1_{k_1}. x_1 = t_1_1 ==> ... ==> x_n = t_1_n ==> ... ==> P)
   754         ==> ... ==> P
   755 
   756     for predicates and
   757 
   758       (x_1, ..., x_n) : s z_1 ... z_m ==>
   759         (!!a_1_1 ... a_1_{k_1}. x_1 = t_1_1 ==> ... ==> x_n = t_1_n ==> ... ==> P)
   760         ==> ... ==> P
   761 
   762     for sets rather than
   763 
   764       x : s z_1 ... z_m ==>
   765         (!!a_1_1 ... a_1_{k_1}. x = (t_1_1, ..., t_1_n) ==> ... ==> P)
   766         ==> ... ==> P
   767 
   768     This may require terms in goals to be expanded to n-tuples (e.g. using case_tac
   769     or simplification with the split_paired_all rule) before the above elimination
   770     rule is applicable.
   771 
   772   - The elimination or case analysis rules for (mutually) inductive sets or
   773     predicates are now called "p_1.cases" ... "p_k.cases". The list of rules
   774     "p_1_..._p_k.elims" is no longer available.
   775 
   776 * Method "metis" proves goals by applying the Metis general-purpose
   777 resolution prover.  Examples are in the directory MetisExamples.  See
   778 also http://gilith.com/software/metis/
   779   
   780 * Command 'sledgehammer' invokes external automatic theorem provers as
   781 background processes.  It generates calls to the "metis" method if
   782 successful. These can be pasted into the proof.  Users do not have to
   783 wait for the automatic provers to return.
   784 
   785 * Case-expressions allow arbitrary constructor-patterns (including "_") and
   786   take their order into account, like in functional programming.
   787   Internally, this is translated into nested case-expressions; missing cases
   788   are added and mapped to the predefined constant "undefined". In complicated
   789   cases printing may no longer show the original input but the internal
   790   form. Lambda-abstractions allow the same form of pattern matching:
   791   "% pat1 => e1 | ..." is an abbreviation for
   792   "%x. case x of pat1 => e1 | ..." where x is a new variable.
   793 
   794 * IntDef: The constant "int :: nat => int" has been removed; now "int"
   795   is an abbreviation for "of_nat :: nat => int". The simplification rules
   796   for "of_nat" have been changed to work like "int" did previously.
   797   (potential INCOMPATIBILITY)
   798   - "of_nat (Suc m)" simplifies to "1 + of_nat m" instead of "of_nat m + 1"
   799   - of_nat_diff and of_nat_mult are no longer default simp rules
   800 
   801 * Method "algebra" solves polynomial equations over (semi)rings using
   802   Groebner bases. The (semi)ring structure is defined by locales and
   803   the tool setup depends on that generic context. Installing the
   804   method for a specific type involves instantiating the locale and
   805   possibly adding declarations for computation on the coefficients.
   806   The method is already instantiated for natural numbers and for the
   807   axiomatic class of idoms with numerals.  See also the paper by
   808   Chaieb and Wenzel at CALCULEMUS 2007 for the general principles
   809   underlying this architecture of context-aware proof-tools.
   810 
   811 * constant "List.op @" now named "List.append".  Use ML antiquotations
   812 @{const_name List.append} or @{term " ... @ ... "} to circumvent
   813 possible incompatibilities when working on ML level.
   814 
   815 * Constant renames due to introduction of canonical name prefixing for
   816   class package:
   817 
   818     HOL.abs ~> HOL.minus_class.abs
   819     HOL.divide ~> HOL.divide_class.divide
   820     Nat.power ~> Nat.power_class.power
   821     Nat.size ~> Nat.size_class.size
   822     Numeral.number_of ~> Numeral.number_class.number_of
   823     FixedPoint.Inf ~> FixedPoint.complete_lattice_class.Inf
   824     FixedPoint.Sup ~> FixedPoint.complete_lattice_class.Sup
   825 
   826 * Rudimentary class target mechanism involves constant renames:
   827 
   828     Orderings.min ~> Orderings.ord_class.min
   829     Orderings.max ~> Orderings.ord_class.max
   830 
   831 * primrec: missing cases mapped to "undefined" instead of "arbitrary"
   832 
   833 * new constant "undefined" with axiom "undefined x = undefined"
   834 
   835 * new class "default" with associated constant "default"
   836 
   837 * new function listsum :: 'a list => 'a for arbitrary monoids.
   838   Special syntax: "SUM x <- xs. f x" (and latex variants)
   839 
   840 * new (input only) syntax for Haskell-like list comprehension, eg
   841   [(x,y). x <- xs, y <- ys, x ~= y]
   842   For details see List.thy.
   843 
   844 * The special syntax for function "filter" has changed from [x : xs. P] to
   845   [x <- xs. P] to avoid an ambiguity caused by list comprehension syntax,
   846   and for uniformity. INCOMPATIBILITY
   847 
   848 * Lemma "set_take_whileD" renamed to "set_takeWhileD"
   849 
   850 * function "sgn" is now overloaded and available on int, real, complex
   851   (and other numeric types).
   852   The details: new class "sgn" with function "sgn";
   853   two possible defs of sgn in the classes sgn_if and sgn_div_norm
   854   (as equational assumptions);
   855   ordered_idom now also inherits from sgn_if - INCOMPATIBILITY.
   856 
   857 * New lemma collection field_simps (an extension of ring_simps)
   858   for manipulating (in)equations involving division. Multiplies
   859   with all denominators that can be proved to be non-zero (in equations)
   860   or positive/negative (in inequations).
   861 
   862 * Lemma collections ring_eq_simps, group_eq_simps and ring_distrib
   863   have been improved and renamed to ring_simps, group_simps and ring_distribs.
   864   Removed lemmas field_xyz in Ring_and_Field
   865   because they were subsumed by lemmas xyz.
   866 INCOMPATIBILITY.
   867 
   868 * Library/Pretty_Int.thy: maps HOL numerals on target language integer literals
   869   when generating code.
   870 
   871 * Library/Pretty_Char.thy: maps HOL characters on target language character literals
   872   when generating code.
   873 
   874 * Library/Commutative_Ring.thy: switched from recdef to function package;
   875   constants add, mul, pow now curried.  Infix syntax for algebraic operations.
   876 
   877 * Some steps towards more uniform lattice theory development in HOL.
   878 
   879     constants "meet" and "join" now named "inf" and "sup"
   880     constant "Meet" now named "Inf"
   881 
   882     classes "meet_semilorder" and "join_semilorder" now named
   883       "lower_semilattice" and "upper_semilattice"
   884     class "lorder" now named "lattice"
   885     class "comp_lat" now named "complete_lattice"
   886 
   887     Instantiation of lattice classes allows explicit definitions
   888     for "inf" and "sup" operations (or "Inf" and "Sup" for complete lattices).
   889 
   890   INCOMPATIBILITY.  Theorem renames:
   891 
   892     meet_left_le            ~> inf_le1
   893     meet_right_le           ~> inf_le2
   894     join_left_le            ~> sup_ge1
   895     join_right_le           ~> sup_ge2
   896     meet_join_le            ~> inf_sup_ord
   897     le_meetI                ~> le_infI
   898     join_leI                ~> le_supI
   899     le_meet                 ~> le_inf_iff
   900     le_join                 ~> ge_sup_conv
   901     meet_idempotent         ~> inf_idem
   902     join_idempotent         ~> sup_idem
   903     meet_comm               ~> inf_commute
   904     join_comm               ~> sup_commute
   905     meet_leI1               ~> le_infI1
   906     meet_leI2               ~> le_infI2
   907     le_joinI1               ~> le_supI1
   908     le_joinI2               ~> le_supI2
   909     meet_assoc              ~> inf_assoc
   910     join_assoc              ~> sup_assoc
   911     meet_left_comm          ~> inf_left_commute
   912     meet_left_idempotent    ~> inf_left_idem
   913     join_left_comm          ~> sup_left_commute
   914     join_left_idempotent    ~> sup_left_idem
   915     meet_aci                ~> inf_aci
   916     join_aci                ~> sup_aci
   917     le_def_meet             ~> le_iff_inf
   918     le_def_join             ~> le_iff_sup
   919     join_absorp2            ~> sup_absorb2
   920     join_absorp1            ~> sup_absorb1
   921     meet_absorp1            ~> inf_absorb1
   922     meet_absorp2            ~> inf_absorb2
   923     meet_join_absorp        ~> inf_sup_absorb
   924     join_meet_absorp        ~> sup_inf_absorb
   925     distrib_join_le         ~> distrib_sup_le
   926     distrib_meet_le         ~> distrib_inf_le
   927 
   928     add_meet_distrib_left   ~> add_inf_distrib_left
   929     add_join_distrib_left   ~> add_sup_distrib_left
   930     is_join_neg_meet        ~> is_join_neg_inf
   931     is_meet_neg_join        ~> is_meet_neg_sup
   932     add_meet_distrib_right  ~> add_inf_distrib_right
   933     add_join_distrib_right  ~> add_sup_distrib_right
   934     add_meet_join_distribs  ~> add_sup_inf_distribs
   935     join_eq_neg_meet        ~> sup_eq_neg_inf
   936     meet_eq_neg_join        ~> inf_eq_neg_sup
   937     add_eq_meet_join        ~> add_eq_inf_sup
   938     meet_0_imp_0            ~> inf_0_imp_0
   939     join_0_imp_0            ~> sup_0_imp_0
   940     meet_0_eq_0             ~> inf_0_eq_0
   941     join_0_eq_0             ~> sup_0_eq_0
   942     neg_meet_eq_join        ~> neg_inf_eq_sup
   943     neg_join_eq_meet        ~> neg_sup_eq_inf
   944     join_eq_if              ~> sup_eq_if
   945 
   946     mono_meet               ~> mono_inf
   947     mono_join               ~> mono_sup
   948     meet_bool_eq            ~> inf_bool_eq
   949     join_bool_eq            ~> sup_bool_eq
   950     meet_fun_eq             ~> inf_fun_eq
   951     join_fun_eq             ~> sup_fun_eq
   952     meet_set_eq             ~> inf_set_eq
   953     join_set_eq             ~> sup_set_eq
   954     meet1_iff               ~> inf1_iff
   955     meet2_iff               ~> inf2_iff
   956     meet1I                  ~> inf1I
   957     meet2I                  ~> inf2I
   958     meet1D1                 ~> inf1D1
   959     meet2D1                 ~> inf2D1
   960     meet1D2                 ~> inf1D2
   961     meet2D2                 ~> inf2D2
   962     meet1E                  ~> inf1E
   963     meet2E                  ~> inf2E
   964     join1_iff               ~> sup1_iff
   965     join2_iff               ~> sup2_iff
   966     join1I1                 ~> sup1I1
   967     join2I1                 ~> sup2I1
   968     join1I1                 ~> sup1I1
   969     join2I2                 ~> sup1I2
   970     join1CI                 ~> sup1CI
   971     join2CI                 ~> sup2CI
   972     join1E                  ~> sup1E
   973     join2E                  ~> sup2E
   974 
   975     is_meet_Meet            ~> is_meet_Inf
   976     Meet_bool_def           ~> Inf_bool_def
   977     Meet_fun_def            ~> Inf_fun_def
   978     Meet_greatest           ~> Inf_greatest
   979     Meet_lower              ~> Inf_lower
   980     Meet_set_def            ~> Inf_set_def
   981 
   982     Sup_def                 ~> Sup_Inf
   983     Sup_bool_eq             ~> Sup_bool_def
   984     Sup_fun_eq              ~> Sup_fun_def
   985     Sup_set_eq              ~> Sup_set_def
   986 
   987     listsp_meetI            ~> listsp_infI
   988     listsp_meet_eq          ~> listsp_inf_eq
   989 
   990     meet_min                ~> inf_min
   991     join_max                ~> sup_max
   992 
   993 * Classes "order" and "linorder": facts "refl", "trans" and
   994 "cases" renamed ro "order_refl", "order_trans" and "linorder_cases", to
   995 avoid clashes with HOL "refl" and "trans". INCOMPATIBILITY.
   996 
   997 * Classes "order" and "linorder": 
   998 potential INCOMPATIBILITY: order of proof goals in order/linorder instance
   999 proofs changed.
  1000 
  1001 * Dropped lemma duplicate def_imp_eq in favor of meta_eq_to_obj_eq.
  1002 INCOMPATIBILITY.
  1003 
  1004 * Dropped lemma duplicate if_def2 in favor of if_bool_eq_conj.
  1005 INCOMPATIBILITY.
  1006 
  1007 * Added syntactic class "size"; overloaded constant "size" now has
  1008 type "'a::size ==> bool"
  1009 
  1010 * Renamed constants "Divides.op div", "Divides.op mod" and "Divides.op
  1011 dvd" to "Divides.div_class.div", "Divides.div_class.mod" and "Divides.dvd". INCOMPATIBILITY.
  1012 
  1013 * Added method "lexicographic_order" automatically synthesizes
  1014 termination relations as lexicographic combinations of size measures
  1015 -- 'function' package.
  1016 
  1017 * HOL/records: generalised field-update to take a function on the
  1018 field rather than the new value: r(|A := x|) is translated to A_update
  1019 (K x) r The K-combinator that is internally used is called K_record.
  1020 INCOMPATIBILITY: Usage of the plain update functions has to be
  1021 adapted.
  1022  
  1023 * axclass "semiring_0" now contains annihilation axioms x * 0 = 0 and
  1024 0 * x = 0, which are required for a semiring.  Richer structures do
  1025 not inherit from semiring_0 anymore, because this property is a
  1026 theorem there, not an axiom.  INCOMPATIBILITY: In instances of
  1027 semiring_0, there is more to prove, but this is mostly trivial.
  1028 
  1029 * axclass "recpower" was generalized to arbitrary monoids, not just
  1030 commutative semirings.  INCOMPATIBILITY: If you use recpower and need
  1031 commutativity or a semiring property, add the corresponding classes.
  1032 
  1033 * Unified locale partial_order with class definition (cf. theory
  1034 Orderings), added parameter ``less''.  INCOMPATIBILITY.
  1035 
  1036 * Constant "List.list_all2" in List.thy now uses authentic syntax.
  1037 INCOMPATIBILITY: translations containing list_all2 may go wrong.  On
  1038 Isar level, use abbreviations instead.
  1039 
  1040 * Renamed constant "List.op mem" to "List.memberl" INCOMPATIBILITY:
  1041 rarely occuring name references (e.g. ``List.op mem.simps'') require
  1042 renaming (e.g. ``List.memberl.simps'').
  1043 
  1044 * Renamed constants "0" to "HOL.zero_class.zero" and "1" to "HOL.one_class.one".
  1045 INCOMPATIBILITY.
  1046 
  1047 * Added class "HOL.eq", allowing for code generation with polymorphic equality.
  1048 
  1049 * Numeral syntax: type 'bin' which was a mere type copy of 'int' has
  1050 been abandoned in favour of plain 'int'. INCOMPATIBILITY --
  1051 significant changes for setting up numeral syntax for types:
  1052 
  1053   - new constants Numeral.pred and Numeral.succ instead
  1054       of former Numeral.bin_pred and Numeral.bin_succ.
  1055   - Use integer operations instead of bin_add, bin_mult and so on.
  1056   - Numeral simplification theorems named Numeral.numeral_simps instead of Bin_simps.
  1057   - ML structure Bin_Simprocs now named Int_Numeral_Base_Simprocs.
  1058 
  1059 See HOL/Integ/IntArith.thy for an example setup.
  1060 
  1061 * New top level command 'normal_form' computes the normal form of a
  1062 term that may contain free variables. For example ``normal_form
  1063 "rev[a,b,c]"'' produces ``[b,c,a]'' (without proof).  This command is
  1064 suitable for heavy-duty computations because the functions are
  1065 compiled to ML first.
  1066 
  1067 * Alternative iff syntax "A <-> B" for equality on bool (with priority
  1068 25 like -->); output depends on the "iff" print_mode, the default is
  1069 "A = B" (with priority 50).
  1070 
  1071 * Renamed constants in HOL.thy and Orderings.thy:
  1072     op +   ~> HOL.plus_class.plus
  1073     op -   ~> HOL.minus_class.minus
  1074     uminus ~> HOL.minus_class.uminus
  1075     abs    ~> HOL.abs_class.abs
  1076     op *   ~> HOL.times_class.times
  1077     op <   ~> HOL.ord_class.less
  1078     op <=  ~> HOL.ord_class.less_eq
  1079 
  1080 Adaptions may be required in the following cases:
  1081 
  1082 a) User-defined constants using any of the names "plus", "minus", "times",
  1083 "less" or "less_eq". The standard syntax translations for "+", "-" and "*"
  1084 may go wrong.
  1085 INCOMPATIBILITY: use more specific names.
  1086 
  1087 b) Variables named "plus", "minus", "times", "less", "less_eq"
  1088 INCOMPATIBILITY: use more specific names.
  1089 
  1090 c) Permutative equations (e.g. "a + b = b + a")
  1091 Since the change of names also changes the order of terms, permutative
  1092 rewrite rules may get applied in a different order. Experience shows that
  1093 this is rarely the case (only two adaptions in the whole Isabelle
  1094 distribution).
  1095 INCOMPATIBILITY: rewrite proofs
  1096 
  1097 d) ML code directly refering to constant names
  1098 This in general only affects hand-written proof tactics, simprocs and so on.
  1099 INCOMPATIBILITY: grep your sourcecode and replace names.  Consider use
  1100 of const_name ML antiquotations.
  1101 
  1102 * Relations less (<) and less_eq (<=) are also available on type bool.
  1103 Modified syntax to disallow nesting without explicit parentheses,
  1104 e.g. "(x < y) < z" or "x < (y < z)", but NOT "x < y < z".
  1105 
  1106 * "LEAST x:A. P" expands to "LEAST x. x:A & P" (input only).
  1107 
  1108 * Relation composition operator "op O" now has precedence 75 and binds
  1109 stronger than union and intersection. INCOMPATIBILITY.
  1110 
  1111 * The old set interval syntax "{m..n(}" (and relatives) has been
  1112 removed.  Use "{m..<n}" (and relatives) instead.
  1113 
  1114 * In the context of the assumption "~(s = t)" the Simplifier rewrites
  1115 "t = s" to False (by simproc "neq_simproc").  For backward
  1116 compatibility this can be disabled by ML "reset use_neq_simproc".
  1117 
  1118 * "m dvd n" where m and n are numbers is evaluated to True/False by
  1119 simp.
  1120 
  1121 * Theorem Cons_eq_map_conv no longer declared as ``simp''.
  1122 
  1123 * Theorem setsum_mult renamed to setsum_right_distrib.
  1124 
  1125 * Prefer ex1I over ex_ex1I in single-step reasoning, e.g. by the
  1126 ``rule'' method.
  1127 
  1128 * Reimplemented methods ``sat'' and ``satx'', with several
  1129 improvements: goals no longer need to be stated as "<prems> ==>
  1130 False", equivalences (i.e. "=" on type bool) are handled, variable
  1131 names of the form "lit_<n>" are no longer reserved, significant
  1132 speedup.
  1133 
  1134 * Methods ``sat'' and ``satx'' can now replay MiniSat proof traces.
  1135 zChaff is still supported as well.
  1136 
  1137 * 'inductive' and 'datatype': provide projections of mutual rules,
  1138 bundled as foo_bar.inducts;
  1139 
  1140 * Library: moved theories Parity, GCD, Binomial, Infinite_Set to
  1141 Library.
  1142 
  1143 * Library: moved theory Accessible_Part to main HOL.
  1144 
  1145 * Library: added theory Coinductive_List of potentially infinite lists
  1146 as greatest fixed-point.
  1147 
  1148 * Library: added theory AssocList which implements (finite) maps as
  1149 association lists.
  1150 
  1151 * Added proof method ``evaluation'' for efficiently solving a goal
  1152 (i.e. a boolean expression) by compiling it to ML. The goal is
  1153 "proved" (via an oracle) if it evaluates to True.
  1154 
  1155 * Linear arithmetic now splits certain operators (e.g. min, max, abs)
  1156 also when invoked by the simplifier.  This results in the simplifier
  1157 being more powerful on arithmetic goals.  INCOMPATIBILITY.  Set
  1158 fast_arith_split_limit to 0 to obtain the old behavior.
  1159 
  1160 * Support for hex (0x20) and binary (0b1001) numerals.
  1161 
  1162 * New method: reify eqs (t), where eqs are equations for an
  1163 interpretation I :: 'a list => 'b => 'c and t::'c is an optional
  1164 parameter, computes a term s::'b and a list xs::'a list and proves the
  1165 theorem I xs s = t. This is also known as reification or quoting. The
  1166 resulting theorem is applied to the subgoal to substitute t with I xs
  1167 s.  If t is omitted, the subgoal itself is reified.
  1168 
  1169 * New method: reflection corr_thm eqs (t). The parameters eqs and (t)
  1170 are as explained above. corr_thm is a theorem for I vs (f t) = I vs t,
  1171 where f is supposed to be a computable function (in the sense of code
  1172 generattion). The method uses reify to compute s and xs as above then
  1173 applies corr_thm and uses normalization by evaluation to "prove" f s =
  1174 r and finally gets the theorem t = r, which is again applied to the
  1175 subgoal. An Example is available in HOL/ex/ReflectionEx.thy.
  1176 
  1177 * Reflection: Automatic reification now handels binding, an example
  1178 is available in HOL/ex/ReflectionEx.thy
  1179 
  1180 
  1181 *** HOL-Algebra ***
  1182 
  1183 * Formalisation of ideals and the quotient construction over rings.
  1184 
  1185 * Order and lattice theory no longer based on records.
  1186 INCOMPATIBILITY.
  1187 
  1188 * Renamed lemmas least_carrier -> least_closed and greatest_carrier ->
  1189 greatest_closed.  INCOMPATIBILITY.
  1190 
  1191 * Method algebra is now set up via an attribute.  For examples see
  1192 Ring.thy.  INCOMPATIBILITY: the method is now weaker on combinations
  1193 of algebraic structures.
  1194 
  1195 * Renamed theory CRing to Ring.
  1196 
  1197 
  1198 *** HOL-Complex ***
  1199 
  1200 * Theory Real: new method ferrack implements quantifier elimination
  1201 for linear arithmetic over the reals. The quantifier elimination
  1202 feature is used only for decision, for compatibility with arith. This
  1203 means a goal is either solved or left unchanged, no simplification.
  1204 
  1205 * Hyperreal: Functions root and sqrt are now defined on negative real
  1206 inputs so that root n (- x) = - root n x and sqrt (- x) = - sqrt x.
  1207 Nonnegativity side conditions have been removed from many lemmas, so
  1208 that more subgoals may now be solved by simplification; potential
  1209 INCOMPATIBILITY.
  1210 
  1211 * Real: New axiomatic classes formalize real normed vector spaces and
  1212 algebras, using new overloaded constants scaleR :: real => 'a => 'a
  1213 and norm :: 'a => real.
  1214 
  1215 * Real: New constant of_real :: real => 'a::real_algebra_1 injects
  1216 from reals into other types. The overloaded constant Reals :: 'a set
  1217 is now defined as range of_real; potential INCOMPATIBILITY.
  1218 
  1219 * Real: ML code generation is supported now and hence also quickcheck.
  1220 Reals are implemented as arbitrary precision rationals.
  1221 
  1222 * Hyperreal: Several constants that previously worked only for the
  1223 reals have been generalized, so they now work over arbitrary vector
  1224 spaces. Type annotations may need to be added in some cases; potential
  1225 INCOMPATIBILITY.
  1226 
  1227   Infinitesimal  :: ('a::real_normed_vector) star set
  1228   HFinite        :: ('a::real_normed_vector) star set
  1229   HInfinite      :: ('a::real_normed_vector) star set
  1230   approx         :: ('a::real_normed_vector) star => 'a star => bool
  1231   monad          :: ('a::real_normed_vector) star => 'a star set
  1232   galaxy         :: ('a::real_normed_vector) star => 'a star set
  1233   (NS)LIMSEQ     :: [nat => 'a::real_normed_vector, 'a] => bool
  1234   (NS)convergent :: (nat => 'a::real_normed_vector) => bool
  1235   (NS)Bseq       :: (nat => 'a::real_normed_vector) => bool
  1236   (NS)Cauchy     :: (nat => 'a::real_normed_vector) => bool
  1237   (NS)LIM        :: ['a::real_normed_vector => 'b::real_normed_vector, 'a, 'b] => bool
  1238   is(NS)Cont     :: ['a::real_normed_vector => 'b::real_normed_vector, 'a] => bool
  1239   deriv          :: ['a::real_normed_field => 'a, 'a, 'a] => bool
  1240   sgn            :: 'a::real_normed_vector => 'a
  1241   exp            :: 'a::{recpower,real_normed_field,banach} => 'a
  1242 
  1243 * Complex: Some complex-specific constants are now abbreviations for
  1244 overloaded ones: complex_of_real = of_real, cmod = norm, hcmod =
  1245 hnorm.  Other constants have been entirely removed in favor of the
  1246 polymorphic versions (INCOMPATIBILITY):
  1247 
  1248   approx        <-- capprox
  1249   HFinite       <-- CFinite
  1250   HInfinite     <-- CInfinite
  1251   Infinitesimal <-- CInfinitesimal
  1252   monad         <-- cmonad
  1253   galaxy        <-- cgalaxy
  1254   (NS)LIM       <-- (NS)CLIM, (NS)CRLIM
  1255   is(NS)Cont    <-- is(NS)Contc, is(NS)contCR
  1256   (ns)deriv     <-- (ns)cderiv
  1257 
  1258 
  1259 *** ML ***
  1260 
  1261 * Generic arithmetic modules: Tools/integer.ML, Tools/rat.ML, Tools/float.ML
  1262 
  1263 * Context data interfaces (Theory/Proof/GenericDataFun): removed
  1264 name/print, uninitialized data defaults to ad-hoc copy of empty value,
  1265 init only required for impure data.  INCOMPATIBILITY: empty really
  1266 need to be empty (no dependencies on theory content!)
  1267 
  1268 * ML within Isar: antiquotations allow to embed statically-checked
  1269 formal entities in the source, referring to the context available at
  1270 compile-time.  For example:
  1271 
  1272 ML {* @{typ "'a => 'b"} *}
  1273 ML {* @{term "%x. x"} *}
  1274 ML {* @{prop "x == y"} *}
  1275 ML {* @{ctyp "'a => 'b"} *}
  1276 ML {* @{cterm "%x. x"} *}
  1277 ML {* @{cprop "x == y"} *}
  1278 ML {* @{thm asm_rl} *}
  1279 ML {* @{thms asm_rl} *}
  1280 ML {* @{const_name c} *}
  1281 ML {* @{const_syntax c} *}
  1282 ML {* @{context} *}
  1283 ML {* @{theory} *}
  1284 ML {* @{theory Pure} *}
  1285 ML {* @{simpset} *}
  1286 ML {* @{claset} *}
  1287 ML {* @{clasimpset} *}
  1288 
  1289 The same works for sources being ``used'' within an Isar context.
  1290 
  1291 * ML in Isar: improved error reporting; extra verbosity with
  1292 Toplevel.debug enabled.
  1293 
  1294 * Pure/library:
  1295 
  1296   val burrow: ('a list -> 'b list) -> 'a list list -> 'b list list
  1297   val fold_burrow: ('a list -> 'c -> 'b list * 'd) -> 'a list list -> 'c -> 'b list list * 'd
  1298 
  1299 The semantics of "burrow" is: "take a function with *simulatanously*
  1300 transforms a list of value, and apply it *simulatanously* to a list of
  1301 list of values of the appropriate type". Compare this with "map" which
  1302 would *not* apply its argument function simulatanously but in
  1303 sequence; "fold_burrow" has an additional context.
  1304 
  1305 * Pure/library: functions map2 and fold2 with curried syntax for
  1306 simultanous mapping and folding:
  1307 
  1308     val map2: ('a -> 'b -> 'c) -> 'a list -> 'b list -> 'c list
  1309     val fold2: ('a -> 'b -> 'c -> 'c) -> 'a list -> 'b list -> 'c -> 'c
  1310 
  1311 * Pure/library: indexed lists - some functions in the Isabelle library
  1312 treating lists over 'a as finite mappings from [0...n] to 'a have been
  1313 given more convenient names and signatures reminiscent of similar
  1314 functions for alists, tables, etc:
  1315 
  1316   val nth: 'a list -> int -> 'a 
  1317   val nth_map: int -> ('a -> 'a) -> 'a list -> 'a list
  1318   val fold_index: (int * 'a -> 'b -> 'b) -> 'a list -> 'b -> 'b
  1319 
  1320 Note that fold_index starts counting at index 0, not 1 like foldln
  1321 used to.
  1322 
  1323 * Pure/library: added general ``divide_and_conquer'' combinator on
  1324 lists.
  1325 
  1326 * Pure/General/table.ML: the join operations now works via exceptions
  1327 DUP/SAME instead of type option.  This is simpler in simple cases, and
  1328 admits slightly more efficient complex applications.
  1329 
  1330 * Pure: datatype Context.generic joins theory/Proof.context and
  1331 provides some facilities for code that works in either kind of
  1332 context, notably GenericDataFun for uniform theory and proof data.
  1333 
  1334 * Pure: 'advanced' translation functions (parse_translation etc.) now
  1335 use Context.generic instead of just theory.
  1336 
  1337 * Pure: simplified internal attribute type, which is now always
  1338 Context.generic * thm -> Context.generic * thm.  Global (theory)
  1339 vs. local (Proof.context) attributes have been discontinued, while
  1340 minimizing code duplication.  Thm.rule_attribute and
  1341 Thm.declaration_attribute build canonical attributes; see also
  1342 structure Context for further operations on Context.generic, notably
  1343 GenericDataFun.  INCOMPATIBILITY, need to adapt attribute type
  1344 declarations and definitions.
  1345 
  1346 * Pure/kernel: consts certification ignores sort constraints given in
  1347 signature declarations.  (This information is not relevant to the
  1348 logic, but only for type inference.)  IMPORTANT INTERNAL CHANGE,
  1349 potential INCOMPATIBILITY.
  1350 
  1351 * Pure: axiomatic type classes are now purely definitional, with
  1352 explicit proofs of class axioms and super class relations performed
  1353 internally.  See Pure/axclass.ML for the main internal interfaces --
  1354 notably AxClass.define_class supercedes AxClass.add_axclass, and
  1355 AxClass.axiomatize_class/classrel/arity supercede
  1356 Sign.add_classes/classrel/arities.
  1357 
  1358 * Pure/Isar: Args/Attrib parsers operate on Context.generic --
  1359 global/local versions on theory vs. Proof.context have been
  1360 discontinued; Attrib.syntax and Method.syntax have been adapted
  1361 accordingly.  INCOMPATIBILITY, need to adapt parser expressions for
  1362 attributes, methods, etc.
  1363 
  1364 * Pure: several functions of signature "... -> theory -> theory * ..."
  1365 have been reoriented to "... -> theory -> ... * theory" in order to
  1366 allow natural usage in combination with the ||>, ||>>, |-> and
  1367 fold_map combinators.
  1368 
  1369 * Pure: official theorem names (closed derivations) and additional
  1370 comments (tags) are now strictly separate.  Name hints -- which are
  1371 maintained as tags -- may be attached any time without affecting the
  1372 derivation.
  1373 
  1374 * Pure: primitive rule lift_rule now takes goal cterm instead of an
  1375 actual goal state (thm).  Use Thm.lift_rule (Thm.cprem_of st i) to
  1376 achieve the old behaviour.
  1377 
  1378 * Pure: the "Goal" constant is now called "prop", supporting a
  1379 slightly more general idea of ``protecting'' meta-level rule
  1380 statements.
  1381 
  1382 * Pure: Logic.(un)varify only works in a global context, which is now
  1383 enforced instead of silently assumed.  INCOMPATIBILITY, may use
  1384 Logic.legacy_(un)varify as temporary workaround.
  1385 
  1386 * Pure: structure Name provides scalable operations for generating
  1387 internal variable names, notably Name.variants etc.  This replaces
  1388 some popular functions from term.ML:
  1389 
  1390   Term.variant		->  Name.variant
  1391   Term.variantlist	->  Name.variant_list  (*canonical argument order*)
  1392   Term.invent_names	->  Name.invent_list
  1393 
  1394 Note that low-level renaming rarely occurs in new code -- operations
  1395 from structure Variable are used instead (see below).
  1396 
  1397 * Pure: structure Variable provides fundamental operations for proper
  1398 treatment of fixed/schematic variables in a context.  For example,
  1399 Variable.import introduces fixes for schematics of given facts and
  1400 Variable.export reverses the effect (up to renaming) -- this replaces
  1401 various freeze_thaw operations.
  1402 
  1403 * Pure: structure Goal provides simple interfaces for
  1404 init/conclude/finish and tactical prove operations (replacing former
  1405 Tactic.prove).  Goal.prove is the canonical way to prove results
  1406 within a given context; Goal.prove_global is a degraded version for
  1407 theory level goals, including a global Drule.standard.  Note that
  1408 OldGoals.prove_goalw_cterm has long been obsolete, since it is
  1409 ill-behaved in a local proof context (e.g. with local fixes/assumes or
  1410 in a locale context).
  1411 
  1412 * Isar: simplified treatment of user-level errors, using exception
  1413 ERROR of string uniformly.  Function error now merely raises ERROR,
  1414 without any side effect on output channels.  The Isar toplevel takes
  1415 care of proper display of ERROR exceptions.  ML code may use plain
  1416 handle/can/try; cat_error may be used to concatenate errors like this:
  1417 
  1418   ... handle ERROR msg => cat_error msg "..."
  1419 
  1420 Toplevel ML code (run directly or through the Isar toplevel) may be
  1421 embedded into the Isar toplevel with exception display/debug like
  1422 this:
  1423 
  1424   Isar.toplevel (fn () => ...)
  1425 
  1426 INCOMPATIBILITY, removed special transform_error facilities, removed
  1427 obsolete variants of user-level exceptions (ERROR_MESSAGE,
  1428 Context.PROOF, ProofContext.CONTEXT, Proof.STATE, ProofHistory.FAIL)
  1429 -- use plain ERROR instead.
  1430 
  1431 * Isar: theory setup now has type (theory -> theory), instead of a
  1432 list.  INCOMPATIBILITY, may use #> to compose setup functions.
  1433 
  1434 * Isar: installed ML toplevel pretty printer for type Proof.context,
  1435 subject to ProofContext.debug/verbose flags.
  1436 
  1437 * Isar: Toplevel.theory_to_proof admits transactions that modify the
  1438 theory before entering a proof state.  Transactions now always see a
  1439 quasi-functional intermediate checkpoint, both in interactive and
  1440 batch mode.
  1441 
  1442 * Simplifier: the simpset of a running simplification process now
  1443 contains a proof context (cf. Simplifier.the_context), which is the
  1444 very context that the initial simpset has been retrieved from (by
  1445 simpset_of/local_simpset_of).  Consequently, all plug-in components
  1446 (solver, looper etc.) may depend on arbitrary proof data.
  1447 
  1448 * Simplifier.inherit_context inherits the proof context (plus the
  1449 local bounds) of the current simplification process; any simproc
  1450 etc. that calls the Simplifier recursively should do this!  Removed
  1451 former Simplifier.inherit_bounds, which is already included here --
  1452 INCOMPATIBILITY.  Tools based on low-level rewriting may even have to
  1453 specify an explicit context using Simplifier.context/theory_context.
  1454 
  1455 * Simplifier/Classical Reasoner: more abstract interfaces
  1456 change_simpset/claset for modifying the simpset/claset reference of a
  1457 theory; raw versions simpset/claset_ref etc. have been discontinued --
  1458 INCOMPATIBILITY.
  1459 
  1460 * Provers: more generic wrt. syntax of object-logics, avoid hardwired
  1461 "Trueprop" etc.
  1462 
  1463 
  1464 *** System ***
  1465 
  1466 * settings: ML_IDENTIFIER -- which is appended to user specific heap
  1467 locations -- now includes the Isabelle version identifier as well.
  1468 This simplifies use of multiple Isabelle installations.
  1469 
  1470 * isabelle-process: option -S (secure mode) disables some critical
  1471 operations, notably runtime compilation and evaluation of ML source
  1472 code.
  1473 
  1474 * Support for multithreading, using Poly/ML 5.1 (internal version from
  1475 CVS). The theory loader exploits parallelism when processing independent
  1476 theories, following the header specifications. The maximum number of
  1477 worker threads is specified via usedir option -M or the "max-threads"
  1478 setting in Proof General. A speedup factor of 1.5--3.5 can be expected
  1479 on a 4-core machine, and up to 6 on a 8-core machine. User-code needs to
  1480 observe certain guidelines for thread-safe programming, see appendix A
  1481 in the Isar Implementation manual.
  1482 
  1483 
  1484 New in Isabelle2005 (October 2005)
  1485 ----------------------------------
  1486 
  1487 *** General ***
  1488 
  1489 * Theory headers: the new header syntax for Isar theories is
  1490 
  1491   theory <name>
  1492   imports <theory1> ... <theoryN>
  1493   uses <file1> ... <fileM>
  1494   begin
  1495 
  1496 where the 'uses' part is optional.  The previous syntax
  1497 
  1498   theory <name> = <theory1> + ... + <theoryN>:
  1499 
  1500 will disappear in the next release.  Use isatool fixheaders to convert
  1501 existing theory files.  Note that there is no change in ancient
  1502 non-Isar theories now, but these will disappear soon.
  1503 
  1504 * Theory loader: parent theories can now also be referred to via
  1505 relative and absolute paths.
  1506 
  1507 * Command 'find_theorems' searches for a list of criteria instead of a
  1508 list of constants. Known criteria are: intro, elim, dest, name:string,
  1509 simp:term, and any term. Criteria can be preceded by '-' to select
  1510 theorems that do not match. Intro, elim, dest select theorems that
  1511 match the current goal, name:s selects theorems whose fully qualified
  1512 name contain s, and simp:term selects all simplification rules whose
  1513 lhs match term.  Any other term is interpreted as pattern and selects
  1514 all theorems matching the pattern. Available in ProofGeneral under
  1515 'ProofGeneral -> Find Theorems' or C-c C-f.  Example:
  1516 
  1517   C-c C-f (100) "(_::nat) + _ + _" intro -name: "HOL."
  1518 
  1519 prints the last 100 theorems matching the pattern "(_::nat) + _ + _",
  1520 matching the current goal as introduction rule and not having "HOL."
  1521 in their name (i.e. not being defined in theory HOL).
  1522 
  1523 * Command 'thms_containing' has been discontinued in favour of
  1524 'find_theorems'; INCOMPATIBILITY.
  1525 
  1526 * Communication with Proof General is now 8bit clean, which means that
  1527 Unicode text in UTF-8 encoding may be used within theory texts (both
  1528 formal and informal parts).  Cf. option -U of the Isabelle Proof
  1529 General interface.  Here are some simple examples (cf. src/HOL/ex):
  1530 
  1531   http://isabelle.in.tum.de/library/HOL/ex/Hebrew.html
  1532   http://isabelle.in.tum.de/library/HOL/ex/Chinese.html
  1533 
  1534 * Improved efficiency of the Simplifier and, to a lesser degree, the
  1535 Classical Reasoner.  Typical big applications run around 2 times
  1536 faster.
  1537 
  1538 
  1539 *** Document preparation ***
  1540 
  1541 * Commands 'display_drafts' and 'print_drafts' perform simple output
  1542 of raw sources.  Only those symbols that do not require additional
  1543 LaTeX packages (depending on comments in isabellesym.sty) are
  1544 displayed properly, everything else is left verbatim.  isatool display
  1545 and isatool print are used as front ends (these are subject to the
  1546 DVI/PDF_VIEWER and PRINT_COMMAND settings, respectively).
  1547 
  1548 * Command tags control specific markup of certain regions of text,
  1549 notably folding and hiding.  Predefined tags include "theory" (for
  1550 theory begin and end), "proof" for proof commands, and "ML" for
  1551 commands involving ML code; the additional tags "visible" and
  1552 "invisible" are unused by default.  Users may give explicit tag
  1553 specifications in the text, e.g. ''by %invisible (auto)''.  The
  1554 interpretation of tags is determined by the LaTeX job during document
  1555 preparation: see option -V of isatool usedir, or options -n and -t of
  1556 isatool document, or even the LaTeX macros \isakeeptag, \isafoldtag,
  1557 \isadroptag.
  1558 
  1559 Several document versions may be produced at the same time via isatool
  1560 usedir (the generated index.html will link all of them).  Typical
  1561 specifications include ''-V document=theory,proof,ML'' to present
  1562 theory/proof/ML parts faithfully, ''-V outline=/proof,/ML'' to fold
  1563 proof and ML commands, and ''-V mutilated=-theory,-proof,-ML'' to omit
  1564 these parts without any formal replacement text.  The Isabelle site
  1565 default settings produce ''document'' and ''outline'' versions as
  1566 specified above.
  1567 
  1568 * Several new antiquotations:
  1569 
  1570   @{term_type term} prints a term with its type annotated;
  1571 
  1572   @{typeof term} prints the type of a term;
  1573 
  1574   @{const const} is the same as @{term const}, but checks that the
  1575   argument is a known logical constant;
  1576 
  1577   @{term_style style term} and @{thm_style style thm} print a term or
  1578   theorem applying a "style" to it
  1579 
  1580   @{ML text}
  1581 
  1582 Predefined styles are 'lhs' and 'rhs' printing the lhs/rhs of
  1583 definitions, equations, inequations etc., 'concl' printing only the
  1584 conclusion of a meta-logical statement theorem, and 'prem1' .. 'prem19'
  1585 to print the specified premise.  TermStyle.add_style provides an ML
  1586 interface for introducing further styles.  See also the "LaTeX Sugar"
  1587 document practical applications.  The ML antiquotation prints
  1588 type-checked ML expressions verbatim.
  1589 
  1590 * Markup commands 'chapter', 'section', 'subsection', 'subsubsection',
  1591 and 'text' support optional locale specification '(in loc)', which
  1592 specifies the default context for interpreting antiquotations.  For
  1593 example: 'text (in lattice) {* @{thm inf_assoc}*}'.
  1594 
  1595 * Option 'locale=NAME' of antiquotations specifies an alternative
  1596 context interpreting the subsequent argument.  For example: @{thm
  1597 [locale=lattice] inf_assoc}.
  1598 
  1599 * Proper output of proof terms (@{prf ...} and @{full_prf ...}) within
  1600 a proof context.
  1601 
  1602 * Proper output of antiquotations for theory commands involving a
  1603 proof context (such as 'locale' or 'theorem (in loc) ...').
  1604 
  1605 * Delimiters of outer tokens (string etc.) now produce separate LaTeX
  1606 macros (\isachardoublequoteopen, isachardoublequoteclose etc.).
  1607 
  1608 * isatool usedir: new option -C (default true) controls whether option
  1609 -D should include a copy of the original document directory; -C false
  1610 prevents unwanted effects such as copying of administrative CVS data.
  1611 
  1612 
  1613 *** Pure ***
  1614 
  1615 * Considerably improved version of 'constdefs' command.  Now performs
  1616 automatic type-inference of declared constants; additional support for
  1617 local structure declarations (cf. locales and HOL records), see also
  1618 isar-ref manual.  Potential INCOMPATIBILITY: need to observe strictly
  1619 sequential dependencies of definitions within a single 'constdefs'
  1620 section; moreover, the declared name needs to be an identifier.  If
  1621 all fails, consider to fall back on 'consts' and 'defs' separately.
  1622 
  1623 * Improved indexed syntax and implicit structures.  First of all,
  1624 indexed syntax provides a notational device for subscripted
  1625 application, using the new syntax \<^bsub>term\<^esub> for arbitrary
  1626 expressions.  Secondly, in a local context with structure
  1627 declarations, number indexes \<^sub>n or the empty index (default
  1628 number 1) refer to a certain fixed variable implicitly; option
  1629 show_structs controls printing of implicit structures.  Typical
  1630 applications of these concepts involve record types and locales.
  1631 
  1632 * New command 'no_syntax' removes grammar declarations (and
  1633 translations) resulting from the given syntax specification, which is
  1634 interpreted in the same manner as for the 'syntax' command.
  1635 
  1636 * 'Advanced' translation functions (parse_translation etc.) may depend
  1637 on the signature of the theory context being presently used for
  1638 parsing/printing, see also isar-ref manual.
  1639 
  1640 * Improved 'oracle' command provides a type-safe interface to turn an
  1641 ML expression of type theory -> T -> term into a primitive rule of
  1642 type theory -> T -> thm (i.e. the functionality of Thm.invoke_oracle
  1643 is already included here); see also FOL/ex/IffExample.thy;
  1644 INCOMPATIBILITY.
  1645 
  1646 * axclass: name space prefix for class "c" is now "c_class" (was "c"
  1647 before); "cI" is no longer bound, use "c.intro" instead.
  1648 INCOMPATIBILITY.  This change avoids clashes of fact bindings for
  1649 axclasses vs. locales.
  1650 
  1651 * Improved internal renaming of symbolic identifiers -- attach primes
  1652 instead of base 26 numbers.
  1653 
  1654 * New flag show_question_marks controls printing of leading question
  1655 marks in schematic variable names.
  1656 
  1657 * In schematic variable names, *any* symbol following \<^isub> or
  1658 \<^isup> is now treated as part of the base name.  For example, the
  1659 following works without printing of awkward ".0" indexes:
  1660 
  1661   lemma "x\<^isub>1 = x\<^isub>2 ==> x\<^isub>2 = x\<^isub>1"
  1662     by simp
  1663 
  1664 * Inner syntax includes (*(*nested*) comments*).
  1665 
  1666 * Pretty printer now supports unbreakable blocks, specified in mixfix
  1667 annotations as "(00...)".
  1668 
  1669 * Clear separation of logical types and nonterminals, where the latter
  1670 may only occur in 'syntax' specifications or type abbreviations.
  1671 Before that distinction was only partially implemented via type class
  1672 "logic" vs. "{}".  Potential INCOMPATIBILITY in rare cases of improper
  1673 use of 'types'/'consts' instead of 'nonterminals'/'syntax'.  Some very
  1674 exotic syntax specifications may require further adaption
  1675 (e.g. Cube/Cube.thy).
  1676 
  1677 * Removed obsolete type class "logic", use the top sort {} instead.
  1678 Note that non-logical types should be declared as 'nonterminals'
  1679 rather than 'types'.  INCOMPATIBILITY for new object-logic
  1680 specifications.
  1681 
  1682 * Attributes 'induct' and 'cases': type or set names may now be
  1683 locally fixed variables as well.
  1684 
  1685 * Simplifier: can now control the depth to which conditional rewriting
  1686 is traced via the PG menu Isabelle -> Settings -> Trace Simp Depth
  1687 Limit.
  1688 
  1689 * Simplifier: simplification procedures may now take the current
  1690 simpset into account (cf. Simplifier.simproc(_i) / mk_simproc
  1691 interface), which is very useful for calling the Simplifier
  1692 recursively.  Minor INCOMPATIBILITY: the 'prems' argument of simprocs
  1693 is gone -- use prems_of_ss on the simpset instead.  Moreover, the
  1694 low-level mk_simproc no longer applies Logic.varify internally, to
  1695 allow for use in a context of fixed variables.
  1696 
  1697 * thin_tac now works even if the assumption being deleted contains !!
  1698 or ==>.  More generally, erule now works even if the major premise of
  1699 the elimination rule contains !! or ==>.
  1700 
  1701 * Method 'rules' has been renamed to 'iprover'. INCOMPATIBILITY.
  1702 
  1703 * Reorganized bootstrapping of the Pure theories; CPure is now derived
  1704 from Pure, which contains all common declarations already.  Both
  1705 theories are defined via plain Isabelle/Isar .thy files.
  1706 INCOMPATIBILITY: elements of CPure (such as the CPure.intro /
  1707 CPure.elim / CPure.dest attributes) now appear in the Pure name space;
  1708 use isatool fixcpure to adapt your theory and ML sources.
  1709 
  1710 * New syntax 'name(i-j, i-, i, ...)' for referring to specific
  1711 selections of theorems in named facts via index ranges.
  1712 
  1713 * 'print_theorems': in theory mode, really print the difference
  1714 wrt. the last state (works for interactive theory development only),
  1715 in proof mode print all local facts (cf. 'print_facts');
  1716 
  1717 * 'hide': option '(open)' hides only base names.
  1718 
  1719 * More efficient treatment of intermediate checkpoints in interactive
  1720 theory development.
  1721 
  1722 * Code generator is now invoked via code_module (incremental code
  1723 generation) and code_library (modular code generation, ML structures
  1724 for each theory).  INCOMPATIBILITY: new keywords 'file' and 'contains'
  1725 must be quoted when used as identifiers.
  1726 
  1727 * New 'value' command for reading, evaluating and printing terms using
  1728 the code generator.  INCOMPATIBILITY: command keyword 'value' must be
  1729 quoted when used as identifier.
  1730 
  1731 
  1732 *** Locales ***
  1733 
  1734 * New commands for the interpretation of locale expressions in
  1735 theories (1), locales (2) and proof contexts (3).  These generate
  1736 proof obligations from the expression specification.  After the
  1737 obligations have been discharged, theorems of the expression are added
  1738 to the theory, target locale or proof context.  The synopsis of the
  1739 commands is a follows:
  1740 
  1741   (1) interpretation expr inst
  1742   (2) interpretation target < expr
  1743   (3) interpret expr inst
  1744 
  1745 Interpretation in theories and proof contexts require a parameter
  1746 instantiation of terms from the current context.  This is applied to
  1747 specifications and theorems of the interpreted expression.
  1748 Interpretation in locales only permits parameter renaming through the
  1749 locale expression.  Interpretation is smart in that interpretations
  1750 that are active already do not occur in proof obligations, neither are
  1751 instantiated theorems stored in duplicate.  Use 'print_interps' to
  1752 inspect active interpretations of a particular locale.  For details,
  1753 see the Isar Reference manual.  Examples can be found in
  1754 HOL/Finite_Set.thy and HOL/Algebra/UnivPoly.thy.
  1755 
  1756 INCOMPATIBILITY: former 'instantiate' has been withdrawn, use
  1757 'interpret' instead.
  1758 
  1759 * New context element 'constrains' for adding type constraints to
  1760 parameters.
  1761 
  1762 * Context expressions: renaming of parameters with syntax
  1763 redeclaration.
  1764 
  1765 * Locale declaration: 'includes' disallowed.
  1766 
  1767 * Proper static binding of attribute syntax -- i.e. types / terms /
  1768 facts mentioned as arguments are always those of the locale definition
  1769 context, independently of the context of later invocations.  Moreover,
  1770 locale operations (renaming and type / term instantiation) are applied
  1771 to attribute arguments as expected.
  1772 
  1773 INCOMPATIBILITY of the ML interface: always pass Attrib.src instead of
  1774 actual attributes; rare situations may require Attrib.attribute to
  1775 embed those attributes into Attrib.src that lack concrete syntax.
  1776 Attribute implementations need to cooperate properly with the static
  1777 binding mechanism.  Basic parsers Args.XXX_typ/term/prop and
  1778 Attrib.XXX_thm etc. already do the right thing without further
  1779 intervention.  Only unusual applications -- such as "where" or "of"
  1780 (cf. src/Pure/Isar/attrib.ML), which process arguments depending both
  1781 on the context and the facts involved -- may have to assign parsed
  1782 values to argument tokens explicitly.
  1783 
  1784 * Changed parameter management in theorem generation for long goal
  1785 statements with 'includes'.  INCOMPATIBILITY: produces a different
  1786 theorem statement in rare situations.
  1787 
  1788 * Locale inspection command 'print_locale' omits notes elements.  Use
  1789 'print_locale!' to have them included in the output.
  1790 
  1791 
  1792 *** Provers ***
  1793 
  1794 * Provers/hypsubst.ML: improved version of the subst method, for
  1795 single-step rewriting: it now works in bound variable contexts. New is
  1796 'subst (asm)', for rewriting an assumption.  INCOMPATIBILITY: may
  1797 rewrite a different subterm than the original subst method, which is
  1798 still available as 'simplesubst'.
  1799 
  1800 * Provers/quasi.ML: new transitivity reasoners for transitivity only
  1801 and quasi orders.
  1802 
  1803 * Provers/trancl.ML: new transitivity reasoner for transitive and
  1804 reflexive-transitive closure of relations.
  1805 
  1806 * Provers/blast.ML: new reference depth_limit to make blast's depth
  1807 limit (previously hard-coded with a value of 20) user-definable.
  1808 
  1809 * Provers/simplifier.ML has been moved to Pure, where Simplifier.setup
  1810 is peformed already.  Object-logics merely need to finish their
  1811 initial simpset configuration as before.  INCOMPATIBILITY.
  1812 
  1813 
  1814 *** HOL ***
  1815 
  1816 * Symbolic syntax of Hilbert Choice Operator is now as follows:
  1817 
  1818   syntax (epsilon)
  1819     "_Eps" :: "[pttrn, bool] => 'a"    ("(3\<some>_./ _)" [0, 10] 10)
  1820 
  1821 The symbol \<some> is displayed as the alternative epsilon of LaTeX
  1822 and x-symbol; use option '-m epsilon' to get it actually printed.
  1823 Moreover, the mathematically important symbolic identifier \<epsilon>
  1824 becomes available as variable, constant etc.  INCOMPATIBILITY,
  1825 
  1826 * "x > y" abbreviates "y < x" and "x >= y" abbreviates "y <= x".
  1827 Similarly for all quantifiers: "ALL x > y" etc.  The x-symbol for >=
  1828 is \<ge>. New transitivity rules have been added to HOL/Orderings.thy to
  1829 support corresponding Isar calculations.
  1830 
  1831 * "{x:A. P}" abbreviates "{x. x:A & P}", and similarly for "\<in>"
  1832 instead of ":".
  1833 
  1834 * theory SetInterval: changed the syntax for open intervals:
  1835 
  1836   Old       New
  1837   {..n(}    {..<n}
  1838   {)n..}    {n<..}
  1839   {m..n(}   {m..<n}
  1840   {)m..n}   {m<..n}
  1841   {)m..n(}  {m<..<n}
  1842 
  1843 The old syntax is still supported but will disappear in the next
  1844 release.  For conversion use the following Emacs search and replace
  1845 patterns (these are not perfect but work quite well):
  1846 
  1847   {)\([^\.]*\)\.\.  ->  {\1<\.\.}
  1848   \.\.\([^(}]*\)(}  ->  \.\.<\1}
  1849 
  1850 * Theory Commutative_Ring (in Library): method comm_ring for proving
  1851 equalities in commutative rings; method 'algebra' provides a generic
  1852 interface.
  1853 
  1854 * Theory Finite_Set: changed the syntax for 'setsum', summation over
  1855 finite sets: "setsum (%x. e) A", which used to be "\<Sum>x:A. e", is
  1856 now either "SUM x:A. e" or "\<Sum>x \<in> A. e". The bound variable can
  1857 be a tuple pattern.
  1858 
  1859 Some new syntax forms are available:
  1860 
  1861   "\<Sum>x | P. e"      for     "setsum (%x. e) {x. P}"
  1862   "\<Sum>x = a..b. e"   for     "setsum (%x. e) {a..b}"
  1863   "\<Sum>x = a..<b. e"  for     "setsum (%x. e) {a..<b}"
  1864   "\<Sum>x < k. e"      for     "setsum (%x. e) {..<k}"
  1865 
  1866 The latter form "\<Sum>x < k. e" used to be based on a separate
  1867 function "Summation", which has been discontinued.
  1868 
  1869 * theory Finite_Set: in structured induction proofs, the insert case
  1870 is now 'case (insert x F)' instead of the old counterintuitive 'case
  1871 (insert F x)'.
  1872 
  1873 * The 'refute' command has been extended to support a much larger
  1874 fragment of HOL, including axiomatic type classes, constdefs and
  1875 typedefs, inductive datatypes and recursion.
  1876 
  1877 * New tactics 'sat' and 'satx' to prove propositional tautologies.
  1878 Requires zChaff with proof generation to be installed.  See
  1879 HOL/ex/SAT_Examples.thy for examples.
  1880 
  1881 * Datatype induction via method 'induct' now preserves the name of the
  1882 induction variable. For example, when proving P(xs::'a list) by
  1883 induction on xs, the induction step is now P(xs) ==> P(a#xs) rather
  1884 than P(list) ==> P(a#list) as previously.  Potential INCOMPATIBILITY
  1885 in unstructured proof scripts.
  1886 
  1887 * Reworked implementation of records.  Improved scalability for
  1888 records with many fields, avoiding performance problems for type
  1889 inference. Records are no longer composed of nested field types, but
  1890 of nested extension types. Therefore the record type only grows linear
  1891 in the number of extensions and not in the number of fields.  The
  1892 top-level (users) view on records is preserved.  Potential
  1893 INCOMPATIBILITY only in strange cases, where the theory depends on the
  1894 old record representation. The type generated for a record is called
  1895 <record_name>_ext_type.
  1896 
  1897 Flag record_quick_and_dirty_sensitive can be enabled to skip the
  1898 proofs triggered by a record definition or a simproc (if
  1899 quick_and_dirty is enabled).  Definitions of large records can take
  1900 quite long.
  1901 
  1902 New simproc record_upd_simproc for simplification of multiple record
  1903 updates enabled by default.  Moreover, trivial updates are also
  1904 removed: r(|x := x r|) = r.  INCOMPATIBILITY: old proofs break
  1905 occasionally, since simplification is more powerful by default.
  1906 
  1907 * typedef: proper support for polymorphic sets, which contain extra
  1908 type-variables in the term.
  1909 
  1910 * Simplifier: automatically reasons about transitivity chains
  1911 involving "trancl" (r^+) and "rtrancl" (r^*) by setting up tactics
  1912 provided by Provers/trancl.ML as additional solvers.  INCOMPATIBILITY:
  1913 old proofs break occasionally as simplification may now solve more
  1914 goals than previously.
  1915 
  1916 * Simplifier: converts x <= y into x = y if assumption y <= x is
  1917 present.  Works for all partial orders (class "order"), in particular
  1918 numbers and sets.  For linear orders (e.g. numbers) it treats ~ x < y
  1919 just like y <= x.
  1920 
  1921 * Simplifier: new simproc for "let x = a in f x".  If a is a free or
  1922 bound variable or a constant then the let is unfolded.  Otherwise
  1923 first a is simplified to b, and then f b is simplified to g. If
  1924 possible we abstract b from g arriving at "let x = b in h x",
  1925 otherwise we unfold the let and arrive at g.  The simproc can be
  1926 enabled/disabled by the reference use_let_simproc.  Potential
  1927 INCOMPATIBILITY since simplification is more powerful by default.
  1928 
  1929 * Classical reasoning: the meson method now accepts theorems as arguments.
  1930 
  1931 * Prover support: pre-release of the Isabelle-ATP linkup, which runs background
  1932 jobs to provide advice on the provability of subgoals.
  1933 
  1934 * Theory OrderedGroup and Ring_and_Field: various additions and
  1935 improvements to faciliate calculations involving equalities and
  1936 inequalities.
  1937 
  1938 The following theorems have been eliminated or modified
  1939 (INCOMPATIBILITY):
  1940 
  1941   abs_eq             now named abs_of_nonneg
  1942   abs_of_ge_0        now named abs_of_nonneg
  1943   abs_minus_eq       now named abs_of_nonpos
  1944   imp_abs_id         now named abs_of_nonneg
  1945   imp_abs_neg_id     now named abs_of_nonpos
  1946   mult_pos           now named mult_pos_pos
  1947   mult_pos_le        now named mult_nonneg_nonneg
  1948   mult_pos_neg_le    now named mult_nonneg_nonpos
  1949   mult_pos_neg2_le   now named mult_nonneg_nonpos2
  1950   mult_neg           now named mult_neg_neg
  1951   mult_neg_le        now named mult_nonpos_nonpos
  1952 
  1953 * The following lemmas in Ring_and_Field have been added to the simplifier:
  1954      
  1955      zero_le_square
  1956      not_square_less_zero 
  1957 
  1958   The following lemmas have been deleted from Real/RealPow:
  1959   
  1960      realpow_zero_zero
  1961      realpow_two
  1962      realpow_less
  1963      zero_le_power
  1964      realpow_two_le
  1965      abs_realpow_two
  1966      realpow_two_abs     
  1967 
  1968 * Theory Parity: added rules for simplifying exponents.
  1969 
  1970 * Theory List:
  1971 
  1972 The following theorems have been eliminated or modified
  1973 (INCOMPATIBILITY):
  1974 
  1975   list_all_Nil       now named list_all.simps(1)
  1976   list_all_Cons      now named list_all.simps(2)
  1977   list_all_conv      now named list_all_iff
  1978   set_mem_eq         now named mem_iff
  1979 
  1980 * Theories SetsAndFunctions and BigO (see HOL/Library) support
  1981 asymptotic "big O" calculations.  See the notes in BigO.thy.
  1982 
  1983 
  1984 *** HOL-Complex ***
  1985 
  1986 * Theory RealDef: better support for embedding natural numbers and
  1987 integers in the reals.
  1988 
  1989 The following theorems have been eliminated or modified
  1990 (INCOMPATIBILITY):
  1991 
  1992   exp_ge_add_one_self  now requires no hypotheses
  1993   real_of_int_add      reversed direction of equality (use [symmetric])
  1994   real_of_int_minus    reversed direction of equality (use [symmetric])
  1995   real_of_int_diff     reversed direction of equality (use [symmetric])
  1996   real_of_int_mult     reversed direction of equality (use [symmetric])
  1997 
  1998 * Theory RComplete: expanded support for floor and ceiling functions.
  1999 
  2000 * Theory Ln is new, with properties of the natural logarithm
  2001 
  2002 * Hyperreal: There is a new type constructor "star" for making
  2003 nonstandard types.  The old type names are now type synonyms:
  2004 
  2005   hypreal = real star
  2006   hypnat = nat star
  2007   hcomplex = complex star
  2008 
  2009 * Hyperreal: Many groups of similarly-defined constants have been
  2010 replaced by polymorphic versions (INCOMPATIBILITY):
  2011 
  2012   star_of <-- hypreal_of_real, hypnat_of_nat, hcomplex_of_complex
  2013 
  2014   starset      <-- starsetNat, starsetC
  2015   *s*          <-- *sNat*, *sc*
  2016   starset_n    <-- starsetNat_n, starsetC_n
  2017   *sn*         <-- *sNatn*, *scn*
  2018   InternalSets <-- InternalNatSets, InternalCSets
  2019 
  2020   starfun      <-- starfun{Nat,Nat2,C,RC,CR}
  2021   *f*          <-- *fNat*, *fNat2*, *fc*, *fRc*, *fcR*
  2022   starfun_n    <-- starfun{Nat,Nat2,C,RC,CR}_n
  2023   *fn*         <-- *fNatn*, *fNat2n*, *fcn*, *fRcn*, *fcRn*
  2024   InternalFuns <-- InternalNatFuns, InternalNatFuns2, Internal{C,RC,CR}Funs
  2025 
  2026 * Hyperreal: Many type-specific theorems have been removed in favor of
  2027 theorems specific to various axiomatic type classes (INCOMPATIBILITY):
  2028 
  2029   add_commute <-- {hypreal,hypnat,hcomplex}_add_commute
  2030   add_assoc   <-- {hypreal,hypnat,hcomplex}_add_assocs
  2031   OrderedGroup.add_0 <-- {hypreal,hypnat,hcomplex}_add_zero_left
  2032   OrderedGroup.add_0_right <-- {hypreal,hcomplex}_add_zero_right
  2033   right_minus <-- hypreal_add_minus
  2034   left_minus <-- {hypreal,hcomplex}_add_minus_left
  2035   mult_commute <-- {hypreal,hypnat,hcomplex}_mult_commute
  2036   mult_assoc <-- {hypreal,hypnat,hcomplex}_mult_assoc
  2037   mult_1_left <-- {hypreal,hypnat}_mult_1, hcomplex_mult_one_left
  2038   mult_1_right <-- hcomplex_mult_one_right
  2039   mult_zero_left <-- hcomplex_mult_zero_left
  2040   left_distrib <-- {hypreal,hypnat,hcomplex}_add_mult_distrib
  2041   right_distrib <-- hypnat_add_mult_distrib2
  2042   zero_neq_one <-- {hypreal,hypnat,hcomplex}_zero_not_eq_one
  2043   right_inverse <-- hypreal_mult_inverse
  2044   left_inverse <-- hypreal_mult_inverse_left, hcomplex_mult_inv_left
  2045   order_refl <-- {hypreal,hypnat}_le_refl
  2046   order_trans <-- {hypreal,hypnat}_le_trans
  2047   order_antisym <-- {hypreal,hypnat}_le_anti_sym
  2048   order_less_le <-- {hypreal,hypnat}_less_le
  2049   linorder_linear <-- {hypreal,hypnat}_le_linear
  2050   add_left_mono <-- {hypreal,hypnat}_add_left_mono
  2051   mult_strict_left_mono <-- {hypreal,hypnat}_mult_less_mono2
  2052   add_nonneg_nonneg <-- hypreal_le_add_order
  2053 
  2054 * Hyperreal: Separate theorems having to do with type-specific
  2055 versions of constants have been merged into theorems that apply to the
  2056 new polymorphic constants (INCOMPATIBILITY):
  2057 
  2058   STAR_UNIV_set <-- {STAR_real,NatStar_real,STARC_complex}_set
  2059   STAR_empty_set <-- {STAR,NatStar,STARC}_empty_set
  2060   STAR_Un <-- {STAR,NatStar,STARC}_Un
  2061   STAR_Int <-- {STAR,NatStar,STARC}_Int
  2062   STAR_Compl <-- {STAR,NatStar,STARC}_Compl
  2063   STAR_subset <-- {STAR,NatStar,STARC}_subset
  2064   STAR_mem <-- {STAR,NatStar,STARC}_mem
  2065   STAR_mem_Compl <-- {STAR,STARC}_mem_Compl
  2066   STAR_diff <-- {STAR,STARC}_diff
  2067   STAR_star_of_image_subset <-- {STAR_hypreal_of_real, NatStar_hypreal_of_real,
  2068     STARC_hcomplex_of_complex}_image_subset
  2069   starset_n_Un <-- starset{Nat,C}_n_Un
  2070   starset_n_Int <-- starset{Nat,C}_n_Int
  2071   starset_n_Compl <-- starset{Nat,C}_n_Compl
  2072   starset_n_diff <-- starset{Nat,C}_n_diff
  2073   InternalSets_Un <-- Internal{Nat,C}Sets_Un
  2074   InternalSets_Int <-- Internal{Nat,C}Sets_Int
  2075   InternalSets_Compl <-- Internal{Nat,C}Sets_Compl
  2076   InternalSets_diff <-- Internal{Nat,C}Sets_diff
  2077   InternalSets_UNIV_diff <-- Internal{Nat,C}Sets_UNIV_diff
  2078   InternalSets_starset_n <-- Internal{Nat,C}Sets_starset{Nat,C}_n
  2079   starset_starset_n_eq <-- starset{Nat,C}_starset{Nat,C}_n_eq
  2080   starset_n_starset <-- starset{Nat,C}_n_starset{Nat,C}
  2081   starfun_n_starfun <-- starfun{Nat,Nat2,C,RC,CR}_n_starfun{Nat,Nat2,C,RC,CR}
  2082   starfun <-- starfun{Nat,Nat2,C,RC,CR}
  2083   starfun_mult <-- starfun{Nat,Nat2,C,RC,CR}_mult
  2084   starfun_add <-- starfun{Nat,Nat2,C,RC,CR}_add
  2085   starfun_minus <-- starfun{Nat,Nat2,C,RC,CR}_minus
  2086   starfun_diff <-- starfun{C,RC,CR}_diff
  2087   starfun_o <-- starfun{NatNat2,Nat2,_stafunNat,C,C_starfunRC,_starfunCR}_o
  2088   starfun_o2 <-- starfun{NatNat2,_stafunNat,C,C_starfunRC,_starfunCR}_o2
  2089   starfun_const_fun <-- starfun{Nat,Nat2,C,RC,CR}_const_fun
  2090   starfun_inverse <-- starfun{Nat,C,RC,CR}_inverse
  2091   starfun_eq <-- starfun{Nat,Nat2,C,RC,CR}_eq
  2092   starfun_eq_iff <-- starfun{C,RC,CR}_eq_iff
  2093   starfun_Id <-- starfunC_Id
  2094   starfun_approx <-- starfun{Nat,CR}_approx
  2095   starfun_capprox <-- starfun{C,RC}_capprox
  2096   starfun_abs <-- starfunNat_rabs
  2097   starfun_lambda_cancel <-- starfun{C,CR,RC}_lambda_cancel
  2098   starfun_lambda_cancel2 <-- starfun{C,CR,RC}_lambda_cancel2
  2099   starfun_mult_HFinite_approx <-- starfunCR_mult_HFinite_capprox
  2100   starfun_mult_CFinite_capprox <-- starfun{C,RC}_mult_CFinite_capprox
  2101   starfun_add_capprox <-- starfun{C,RC}_add_capprox
  2102   starfun_add_approx <-- starfunCR_add_approx
  2103   starfun_inverse_inverse <-- starfunC_inverse_inverse
  2104   starfun_divide <-- starfun{C,CR,RC}_divide
  2105   starfun_n <-- starfun{Nat,C}_n
  2106   starfun_n_mult <-- starfun{Nat,C}_n_mult
  2107   starfun_n_add <-- starfun{Nat,C}_n_add
  2108   starfun_n_add_minus <-- starfunNat_n_add_minus
  2109   starfun_n_const_fun <-- starfun{Nat,C}_n_const_fun
  2110   starfun_n_minus <-- starfun{Nat,C}_n_minus
  2111   starfun_n_eq <-- starfun{Nat,C}_n_eq
  2112 
  2113   star_n_add <-- {hypreal,hypnat,hcomplex}_add
  2114   star_n_minus <-- {hypreal,hcomplex}_minus
  2115   star_n_diff <-- {hypreal,hcomplex}_diff
  2116   star_n_mult <-- {hypreal,hcomplex}_mult
  2117   star_n_inverse <-- {hypreal,hcomplex}_inverse
  2118   star_n_le <-- {hypreal,hypnat}_le
  2119   star_n_less <-- {hypreal,hypnat}_less
  2120   star_n_zero_num <-- {hypreal,hypnat,hcomplex}_zero_num
  2121   star_n_one_num <-- {hypreal,hypnat,hcomplex}_one_num
  2122   star_n_abs <-- hypreal_hrabs
  2123   star_n_divide <-- hcomplex_divide
  2124 
  2125   star_of_add <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_add
  2126   star_of_minus <-- {hypreal_of_real,hcomplex_of_complex}_minus
  2127   star_of_diff <-- hypreal_of_real_diff
  2128   star_of_mult <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_mult
  2129   star_of_one <-- {hypreal_of_real,hcomplex_of_complex}_one
  2130   star_of_zero <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_zero
  2131   star_of_le <-- {hypreal_of_real,hypnat_of_nat}_le_iff
  2132   star_of_less <-- {hypreal_of_real,hypnat_of_nat}_less_iff
  2133   star_of_eq <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_eq_iff
  2134   star_of_inverse <-- {hypreal_of_real,hcomplex_of_complex}_inverse
  2135   star_of_divide <-- {hypreal_of_real,hcomplex_of_complex}_divide
  2136   star_of_of_nat <-- {hypreal_of_real,hcomplex_of_complex}_of_nat
  2137   star_of_of_int <-- {hypreal_of_real,hcomplex_of_complex}_of_int
  2138   star_of_number_of <-- {hypreal,hcomplex}_number_of
  2139   star_of_number_less <-- number_of_less_hypreal_of_real_iff
  2140   star_of_number_le <-- number_of_le_hypreal_of_real_iff
  2141   star_of_eq_number <-- hypreal_of_real_eq_number_of_iff
  2142   star_of_less_number <-- hypreal_of_real_less_number_of_iff
  2143   star_of_le_number <-- hypreal_of_real_le_number_of_iff
  2144   star_of_power <-- hypreal_of_real_power
  2145   star_of_eq_0 <-- hcomplex_of_complex_zero_iff
  2146 
  2147 * Hyperreal: new method "transfer" that implements the transfer
  2148 principle of nonstandard analysis. With a subgoal that mentions
  2149 nonstandard types like "'a star", the command "apply transfer"
  2150 replaces it with an equivalent one that mentions only standard types.
  2151 To be successful, all free variables must have standard types; non-
  2152 standard variables must have explicit universal quantifiers.
  2153 
  2154 * Hyperreal: A theory of Taylor series.
  2155 
  2156 
  2157 *** HOLCF ***
  2158 
  2159 * Discontinued special version of 'constdefs' (which used to support
  2160 continuous functions) in favor of the general Pure one with full
  2161 type-inference.
  2162 
  2163 * New simplification procedure for solving continuity conditions; it
  2164 is much faster on terms with many nested lambda abstractions (cubic
  2165 instead of exponential time).
  2166 
  2167 * New syntax for domain package: selector names are now optional.
  2168 Parentheses should be omitted unless argument is lazy, for example:
  2169 
  2170   domain 'a stream = cons "'a" (lazy "'a stream")
  2171 
  2172 * New command 'fixrec' for defining recursive functions with pattern
  2173 matching; defining multiple functions with mutual recursion is also
  2174 supported.  Patterns may include the constants cpair, spair, up, sinl,
  2175 sinr, or any data constructor defined by the domain package. The given
  2176 equations are proven as rewrite rules. See HOLCF/ex/Fixrec_ex.thy for
  2177 syntax and examples.
  2178 
  2179 * New commands 'cpodef' and 'pcpodef' for defining predicate subtypes
  2180 of cpo and pcpo types. Syntax is exactly like the 'typedef' command,
  2181 but the proof obligation additionally includes an admissibility
  2182 requirement. The packages generate instances of class cpo or pcpo,
  2183 with continuity and strictness theorems for Rep and Abs.
  2184 
  2185 * HOLCF: Many theorems have been renamed according to a more standard naming
  2186 scheme (INCOMPATIBILITY):
  2187 
  2188   foo_inject:  "foo$x = foo$y ==> x = y"
  2189   foo_eq:      "(foo$x = foo$y) = (x = y)"
  2190   foo_less:    "(foo$x << foo$y) = (x << y)"
  2191   foo_strict:  "foo$UU = UU"
  2192   foo_defined: "... ==> foo$x ~= UU"
  2193   foo_defined_iff: "(foo$x = UU) = (x = UU)"
  2194 
  2195 
  2196 *** ZF ***
  2197 
  2198 * ZF/ex: theories Group and Ring provide examples in abstract algebra,
  2199 including the First Isomorphism Theorem (on quotienting by the kernel
  2200 of a homomorphism).
  2201 
  2202 * ZF/Simplifier: install second copy of type solver that actually
  2203 makes use of TC rules declared to Isar proof contexts (or locales);
  2204 the old version is still required for ML proof scripts.
  2205 
  2206 
  2207 *** Cube ***
  2208 
  2209 * Converted to Isar theory format; use locales instead of axiomatic
  2210 theories.
  2211 
  2212 
  2213 *** ML ***
  2214 
  2215 * Pure/library.ML: added ##>, ##>>, #>> -- higher-order counterparts
  2216 for ||>, ||>>, |>>,
  2217 
  2218 * Pure/library.ML no longer defines its own option datatype, but uses
  2219 that of the SML basis, which has constructors NONE and SOME instead of
  2220 None and Some, as well as exception Option.Option instead of OPTION.
  2221 The functions the, if_none, is_some, is_none have been adapted
  2222 accordingly, while Option.map replaces apsome.
  2223 
  2224 * Pure/library.ML: the exception LIST has been given up in favour of
  2225 the standard exceptions Empty and Subscript, as well as
  2226 Library.UnequalLengths.  Function like Library.hd and Library.tl are
  2227 superceded by the standard hd and tl functions etc.
  2228 
  2229 A number of basic list functions are no longer exported to the ML
  2230 toplevel, as they are variants of predefined functions.  The following
  2231 suggests how one can translate existing code:
  2232 
  2233     rev_append xs ys = List.revAppend (xs, ys)
  2234     nth_elem (i, xs) = List.nth (xs, i)
  2235     last_elem xs = List.last xs
  2236     flat xss = List.concat xss
  2237     seq fs = List.app fs
  2238     partition P xs = List.partition P xs
  2239     mapfilter f xs = List.mapPartial f xs
  2240 
  2241 * Pure/library.ML: several combinators for linear functional
  2242 transformations, notably reverse application and composition:
  2243 
  2244   x |> f                f #> g
  2245   (x, y) |-> f          f #-> g
  2246 
  2247 * Pure/library.ML: introduced/changed precedence of infix operators:
  2248 
  2249   infix 1 |> |-> ||> ||>> |>> |>>> #> #->;
  2250   infix 2 ?;
  2251   infix 3 o oo ooo oooo;
  2252   infix 4 ~~ upto downto;
  2253 
  2254 Maybe INCOMPATIBILITY when any of those is used in conjunction with other
  2255 infix operators.
  2256 
  2257 * Pure/library.ML: natural list combinators fold, fold_rev, and
  2258 fold_map support linear functional transformations and nesting.  For
  2259 example:
  2260 
  2261   fold f [x1, ..., xN] y =
  2262     y |> f x1 |> ... |> f xN
  2263 
  2264   (fold o fold) f [xs1, ..., xsN] y =
  2265     y |> fold f xs1 |> ... |> fold f xsN
  2266 
  2267   fold f [x1, ..., xN] =
  2268     f x1 #> ... #> f xN
  2269 
  2270   (fold o fold) f [xs1, ..., xsN] =
  2271     fold f xs1 #> ... #> fold f xsN
  2272 
  2273 * Pure/library.ML: the following selectors on type 'a option are
  2274 available:
  2275 
  2276   the:               'a option -> 'a  (*partial*)
  2277   these:             'a option -> 'a  where 'a = 'b list
  2278   the_default: 'a -> 'a option -> 'a
  2279   the_list:          'a option -> 'a list
  2280 
  2281 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  2282 basic operations for association lists, following natural argument
  2283 order; moreover the explicit equality predicate passed here avoids
  2284 potentially expensive polymorphic runtime equality checks.
  2285 The old functions may be expressed as follows:
  2286 
  2287   assoc = uncurry (AList.lookup (op =))
  2288   assocs = these oo AList.lookup (op =)
  2289   overwrite = uncurry (AList.update (op =)) o swap
  2290 
  2291 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides
  2292 
  2293   val make: ('a -> 'b) -> 'a list -> ('a * 'b) list
  2294   val find: ('a * 'b -> bool) -> ('c * 'b) list -> 'a -> 'c list
  2295 
  2296 replacing make_keylist and keyfilter (occassionally used)
  2297 Naive rewrites:
  2298 
  2299   make_keylist = AList.make
  2300   keyfilter = AList.find (op =)
  2301 
  2302 * eq_fst and eq_snd now take explicit equality parameter, thus
  2303   avoiding eqtypes. Naive rewrites:
  2304 
  2305     eq_fst = eq_fst (op =)
  2306     eq_snd = eq_snd (op =)
  2307 
  2308 * Removed deprecated apl and apr (rarely used).
  2309   Naive rewrites:
  2310 
  2311     apl (n, op) =>>= curry op n
  2312     apr (op, m) =>>= fn n => op (n, m)
  2313 
  2314 * Pure/General: structure OrdList (cf. Pure/General/ord_list.ML)
  2315 provides a reasonably efficient light-weight implementation of sets as
  2316 lists.
  2317 
  2318 * Pure/General: generic tables (cf. Pure/General/table.ML) provide a
  2319 few new operations; existing lookup and update are now curried to
  2320 follow natural argument order (for use with fold etc.);
  2321 INCOMPATIBILITY, use (uncurry Symtab.lookup) etc. as last resort.
  2322 
  2323 * Pure/General: output via the Isabelle channels of
  2324 writeln/warning/error etc. is now passed through Output.output, with a
  2325 hook for arbitrary transformations depending on the print_mode
  2326 (cf. Output.add_mode -- the first active mode that provides a output
  2327 function wins).  Already formatted output may be embedded into further
  2328 text via Output.raw; the result of Pretty.string_of/str_of and derived
  2329 functions (string_of_term/cterm/thm etc.) is already marked raw to
  2330 accommodate easy composition of diagnostic messages etc.  Programmers
  2331 rarely need to care about Output.output or Output.raw at all, with
  2332 some notable exceptions: Output.output is required when bypassing the
  2333 standard channels (writeln etc.), or in token translations to produce
  2334 properly formatted results; Output.raw is required when capturing
  2335 already output material that will eventually be presented to the user
  2336 a second time.  For the default print mode, both Output.output and
  2337 Output.raw have no effect.
  2338 
  2339 * Pure/General: Output.time_accumulator NAME creates an operator ('a
  2340 -> 'b) -> 'a -> 'b to measure runtime and count invocations; the
  2341 cumulative results are displayed at the end of a batch session.
  2342 
  2343 * Pure/General: File.sysify_path and File.quote_sysify path have been
  2344 replaced by File.platform_path and File.shell_path (with appropriate
  2345 hooks).  This provides a clean interface for unusual systems where the
  2346 internal and external process view of file names are different.
  2347 
  2348 * Pure: more efficient orders for basic syntactic entities: added
  2349 fast_string_ord, fast_indexname_ord, fast_term_ord; changed sort_ord
  2350 and typ_ord to use fast_string_ord and fast_indexname_ord (term_ord is
  2351 NOT affected); structures Symtab, Vartab, Typtab, Termtab use the fast
  2352 orders now -- potential INCOMPATIBILITY for code that depends on a
  2353 particular order for Symtab.keys, Symtab.dest, etc. (consider using
  2354 Library.sort_strings on result).
  2355 
  2356 * Pure/term.ML: combinators fold_atyps, fold_aterms, fold_term_types,
  2357 fold_types traverse types/terms from left to right, observing natural
  2358 argument order.  Supercedes previous foldl_XXX versions, add_frees,
  2359 add_vars etc. have been adapted as well: INCOMPATIBILITY.
  2360 
  2361 * Pure: name spaces have been refined, with significant changes of the
  2362 internal interfaces -- INCOMPATIBILITY.  Renamed cond_extern(_table)
  2363 to extern(_table).  The plain name entry path is superceded by a
  2364 general 'naming' context, which also includes the 'policy' to produce
  2365 a fully qualified name and external accesses of a fully qualified
  2366 name; NameSpace.extend is superceded by context dependent
  2367 Sign.declare_name.  Several theory and proof context operations modify
  2368 the naming context.  Especially note Theory.restore_naming and
  2369 ProofContext.restore_naming to get back to a sane state; note that
  2370 Theory.add_path is no longer sufficient to recover from
  2371 Theory.absolute_path in particular.
  2372 
  2373 * Pure: new flags short_names (default false) and unique_names
  2374 (default true) for controlling output of qualified names.  If
  2375 short_names is set, names are printed unqualified.  If unique_names is
  2376 reset, the name prefix is reduced to the minimum required to achieve
  2377 the original result when interning again, even if there is an overlap
  2378 with earlier declarations.
  2379 
  2380 * Pure/TheoryDataFun: change of the argument structure; 'prep_ext' is
  2381 now 'extend', and 'merge' gets an additional Pretty.pp argument
  2382 (useful for printing error messages).  INCOMPATIBILITY.
  2383 
  2384 * Pure: major reorganization of the theory context.  Type Sign.sg and
  2385 Theory.theory are now identified, referring to the universal
  2386 Context.theory (see Pure/context.ML).  Actual signature and theory
  2387 content is managed as theory data.  The old code and interfaces were
  2388 spread over many files and structures; the new arrangement introduces
  2389 considerable INCOMPATIBILITY to gain more clarity:
  2390 
  2391   Context -- theory management operations (name, identity, inclusion,
  2392     parents, ancestors, merge, etc.), plus generic theory data;
  2393 
  2394   Sign -- logical signature and syntax operations (declaring consts,
  2395     types, etc.), plus certify/read for common entities;
  2396 
  2397   Theory -- logical theory operations (stating axioms, definitions,
  2398     oracles), plus a copy of logical signature operations (consts,
  2399     types, etc.); also a few basic management operations (Theory.copy,
  2400     Theory.merge, etc.)
  2401 
  2402 The most basic sign_of operations (Theory.sign_of, Thm.sign_of_thm
  2403 etc.) as well as the sign field in Thm.rep_thm etc. have been retained
  2404 for convenience -- they merely return the theory.
  2405 
  2406 * Pure: type Type.tsig is superceded by theory in most interfaces.
  2407 
  2408 * Pure: the Isar proof context type is already defined early in Pure
  2409 as Context.proof (note that ProofContext.context and Proof.context are
  2410 aliases, where the latter is the preferred name).  This enables other
  2411 Isabelle components to refer to that type even before Isar is present.
  2412 
  2413 * Pure/sign/theory: discontinued named name spaces (i.e. classK,
  2414 typeK, constK, axiomK, oracleK), but provide explicit operations for
  2415 any of these kinds.  For example, Sign.intern typeK is now
  2416 Sign.intern_type, Theory.hide_space Sign.typeK is now
  2417 Theory.hide_types.  Also note that former
  2418 Theory.hide_classes/types/consts are now
  2419 Theory.hide_classes_i/types_i/consts_i, while the non '_i' versions
  2420 internalize their arguments!  INCOMPATIBILITY.
  2421 
  2422 * Pure: get_thm interface (of PureThy and ProofContext) expects
  2423 datatype thmref (with constructors Name and NameSelection) instead of
  2424 plain string -- INCOMPATIBILITY;
  2425 
  2426 * Pure: cases produced by proof methods specify options, where NONE
  2427 means to remove case bindings -- INCOMPATIBILITY in
  2428 (RAW_)METHOD_CASES.
  2429 
  2430 * Pure: the following operations retrieve axioms or theorems from a
  2431 theory node or theory hierarchy, respectively:
  2432 
  2433   Theory.axioms_of: theory -> (string * term) list
  2434   Theory.all_axioms_of: theory -> (string * term) list
  2435   PureThy.thms_of: theory -> (string * thm) list
  2436   PureThy.all_thms_of: theory -> (string * thm) list
  2437 
  2438 * Pure: print_tac now outputs the goal through the trace channel.
  2439 
  2440 * Isar toplevel: improved diagnostics, mostly for Poly/ML only.
  2441 Reference Toplevel.debug (default false) controls detailed printing
  2442 and tracing of low-level exceptions; Toplevel.profiling (default 0)
  2443 controls execution profiling -- set to 1 for time and 2 for space
  2444 (both increase the runtime).
  2445 
  2446 * Isar session: The initial use of ROOT.ML is now always timed,
  2447 i.e. the log will show the actual process times, in contrast to the
  2448 elapsed wall-clock time that the outer shell wrapper produces.
  2449 
  2450 * Simplifier: improved handling of bound variables (nameless
  2451 representation, avoid allocating new strings).  Simprocs that invoke
  2452 the Simplifier recursively should use Simplifier.inherit_bounds to
  2453 avoid local name clashes.  Failure to do so produces warnings
  2454 "Simplifier: renamed bound variable ..."; set Simplifier.debug_bounds
  2455 for further details.
  2456 
  2457 * ML functions legacy_bindings and use_legacy_bindings produce ML fact
  2458 bindings for all theorems stored within a given theory; this may help
  2459 in porting non-Isar theories to Isar ones, while keeping ML proof
  2460 scripts for the time being.
  2461 
  2462 * ML operator HTML.with_charset specifies the charset begin used for
  2463 generated HTML files.  For example:
  2464 
  2465   HTML.with_charset "utf-8" use_thy "Hebrew";
  2466   HTML.with_charset "utf-8" use_thy "Chinese";
  2467 
  2468 
  2469 *** System ***
  2470 
  2471 * Allow symlinks to all proper Isabelle executables (Isabelle,
  2472 isabelle, isatool etc.).
  2473 
  2474 * ISABELLE_DOC_FORMAT setting specifies preferred document format (for
  2475 isatool doc, isatool mkdir, display_drafts etc.).
  2476 
  2477 * isatool usedir: option -f allows specification of the ML file to be
  2478 used by Isabelle; default is ROOT.ML.
  2479 
  2480 * New isatool version outputs the version identifier of the Isabelle
  2481 distribution being used.
  2482 
  2483 * HOL: new isatool dimacs2hol converts files in DIMACS CNF format
  2484 (containing Boolean satisfiability problems) into Isabelle/HOL
  2485 theories.
  2486 
  2487 
  2488 
  2489 New in Isabelle2004 (April 2004)
  2490 --------------------------------
  2491 
  2492 *** General ***
  2493 
  2494 * Provers/order.ML:  new efficient reasoner for partial and linear orders.
  2495   Replaces linorder.ML.
  2496 
  2497 * Pure: Greek letters (except small lambda, \<lambda>), as well as Gothic
  2498   (\<aa>...\<zz>\<AA>...\<ZZ>), calligraphic (\<A>...\<Z>), and Euler
  2499   (\<a>...\<z>), are now considered normal letters, and can therefore
  2500   be used anywhere where an ASCII letter (a...zA...Z) has until
  2501   now. COMPATIBILITY: This obviously changes the parsing of some
  2502   terms, especially where a symbol has been used as a binder, say
  2503   '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed
  2504   as an identifier.  Fix it by inserting a space around former
  2505   symbols.  Call 'isatool fixgreek' to try to fix parsing errors in
  2506   existing theory and ML files.
  2507 
  2508 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.
  2509 
  2510 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now
  2511   allowed in identifiers. Similar to Greek letters \<^isub> is now considered
  2512   a normal (but invisible) letter. For multiple letter subscripts repeat
  2513   \<^isub> like this: x\<^isub>1\<^isub>2.
  2514 
  2515 * Pure: There are now sub-/superscripts that can span more than one
  2516   character. Text between \<^bsub> and \<^esub> is set in subscript in
  2517   ProofGeneral and LaTeX, text between \<^bsup> and \<^esup> in
  2518   superscript. The new control characters are not identifier parts.
  2519 
  2520 * Pure: Control-symbols of the form \<^raw:...> will literally print the
  2521   content of "..." to the latex file instead of \isacntrl... . The "..."
  2522   may consist of any printable characters excluding the end bracket >.
  2523 
  2524 * Pure: Using new Isar command "finalconsts" (or the ML functions
  2525   Theory.add_finals or Theory.add_finals_i) it is now possible to
  2526   declare constants "final", which prevents their being given a definition
  2527   later.  It is useful for constants whose behaviour is fixed axiomatically
  2528   rather than definitionally, such as the meta-logic connectives.
  2529 
  2530 * Pure: 'instance' now handles general arities with general sorts
  2531   (i.e. intersections of classes),
  2532 
  2533 * Presentation: generated HTML now uses a CSS style sheet to make layout
  2534   (somewhat) independent of content. It is copied from lib/html/isabelle.css.
  2535   It can be changed to alter the colors/layout of generated pages.
  2536 
  2537 
  2538 *** Isar ***
  2539 
  2540 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,
  2541   cut_tac, subgoal_tac and thin_tac:
  2542   - Now understand static (Isar) contexts.  As a consequence, users of Isar
  2543     locales are no longer forced to write Isar proof scripts.
  2544     For details see Isar Reference Manual, paragraph 4.3.2: Further tactic
  2545     emulations.
  2546   - INCOMPATIBILITY: names of variables to be instantiated may no
  2547     longer be enclosed in quotes.  Instead, precede variable name with `?'.
  2548     This is consistent with the instantiation attribute "where".
  2549 
  2550 * Attributes "where" and "of":
  2551   - Now take type variables of instantiated theorem into account when reading
  2552     the instantiation string.  This fixes a bug that caused instantiated
  2553     theorems to have too special types in some circumstances.
  2554   - "where" permits explicit instantiations of type variables.
  2555 
  2556 * Calculation commands "moreover" and "also" no longer interfere with
  2557   current facts ("this"), admitting arbitrary combinations with "then"
  2558   and derived forms.
  2559 
  2560 * Locales:
  2561   - Goal statements involving the context element "includes" no longer
  2562     generate theorems with internal delta predicates (those ending on
  2563     "_axioms") in the premise.
  2564     Resolve particular premise with <locale>.intro to obtain old form.
  2565   - Fixed bug in type inference ("unify_frozen") that prevented mix of target
  2566     specification and "includes" elements in goal statement.
  2567   - Rule sets <locale>.intro and <locale>.axioms no longer declared as
  2568     [intro?] and [elim?] (respectively) by default.
  2569   - Experimental command for instantiation of locales in proof contexts:
  2570         instantiate <label>[<attrs>]: <loc>
  2571     Instantiates locale <loc> and adds all its theorems to the current context
  2572     taking into account their attributes.  Label and attrs are optional
  2573     modifiers, like in theorem declarations.  If present, names of
  2574     instantiated theorems are qualified with <label>, and the attributes
  2575     <attrs> are applied after any attributes these theorems might have already.
  2576       If the locale has assumptions, a chained fact of the form
  2577     "<loc> t1 ... tn" is expected from which instantiations of the parameters
  2578     are derived.  The command does not support old-style locales declared
  2579     with "locale (open)".
  2580       A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.
  2581 
  2582 * HOL: Tactic emulation methods induct_tac and case_tac understand static
  2583   (Isar) contexts.
  2584 
  2585 
  2586 *** HOL ***
  2587 
  2588 * Proof import: new image HOL4 contains the imported library from
  2589   the HOL4 system with about 2500 theorems. It is imported by
  2590   replaying proof terms produced by HOL4 in Isabelle. The HOL4 image
  2591   can be used like any other Isabelle image.  See
  2592   HOL/Import/HOL/README for more information.
  2593 
  2594 * Simplifier:
  2595   - Much improved handling of linear and partial orders.
  2596     Reasoners for linear and partial orders are set up for type classes
  2597     "linorder" and "order" respectively, and are added to the default simpset
  2598     as solvers.  This means that the simplifier can build transitivity chains
  2599     to solve goals from the assumptions.
  2600   - INCOMPATIBILITY: old proofs break occasionally.  Typically, applications
  2601     of blast or auto after simplification become unnecessary because the goal
  2602     is solved by simplification already.
  2603 
  2604 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,
  2605     all proved in axiomatic type classes for semirings, rings and fields.
  2606 
  2607 * Numerics:
  2608   - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are
  2609     now formalized using the Ring_and_Field theory mentioned above.
  2610   - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently
  2611     than before, because now they are set up once in a generic manner.
  2612   - INCOMPATIBILITY: many type-specific arithmetic laws have gone.
  2613     Look for the general versions in Ring_and_Field (and Power if they concern
  2614     exponentiation).
  2615 
  2616 * Type "rat" of the rational numbers is now available in HOL-Complex.
  2617 
  2618 * Records:
  2619   - Record types are now by default printed with their type abbreviation
  2620     instead of the list of all field types. This can be configured via
  2621     the reference "print_record_type_abbr".
  2622   - Simproc "record_upd_simproc" for simplification of multiple updates added
  2623     (not enabled by default).
  2624   - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.
  2625     EX x. x = sel r to True (not enabled by default).
  2626   - Tactic "record_split_simp_tac" to split and simplify records added.
  2627 
  2628 * 'specification' command added, allowing for definition by
  2629   specification.  There is also an 'ax_specification' command that
  2630   introduces the new constants axiomatically.
  2631 
  2632 * arith(_tac) is now able to generate counterexamples for reals as well.
  2633 
  2634 * HOL-Algebra: new locale "ring" for non-commutative rings.
  2635 
  2636 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function
  2637   definitions, thanks to Sava Krsti\'{c} and John Matthews.
  2638 
  2639 * HOL-Matrix: a first theory for matrices in HOL with an application of
  2640   matrix theory to linear programming.
  2641 
  2642 * Unions and Intersections:
  2643   The latex output syntax of UN and INT has been changed
  2644   from "\Union x \in A. B" to "\Union_{x \in A} B"
  2645   i.e. the index formulae has become a subscript.
  2646   Similarly for "\Union x. B", and for \Inter instead of \Union.
  2647 
  2648 * Unions and Intersections over Intervals:
  2649   There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is
  2650   also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"
  2651   like in normal math, and corresponding versions for < and for intersection.
  2652 
  2653 * HOL/List: Ordering "lexico" is renamed "lenlex" and the standard
  2654   lexicographic dictonary ordering has been added as "lexord".
  2655 
  2656 * ML: the legacy theory structures Int and List have been removed. They had
  2657   conflicted with ML Basis Library structures having the same names.
  2658 
  2659 * 'refute' command added to search for (finite) countermodels.  Only works
  2660   for a fragment of HOL.  The installation of an external SAT solver is
  2661   highly recommended.  See "HOL/Refute.thy" for details.
  2662 
  2663 * 'quickcheck' command: Allows to find counterexamples by evaluating
  2664   formulae under an assignment of free variables to random values.
  2665   In contrast to 'refute', it can deal with inductive datatypes,
  2666   but cannot handle quantifiers. See "HOL/ex/Quickcheck_Examples.thy"
  2667   for examples.
  2668 
  2669 
  2670 *** HOLCF ***
  2671 
  2672 * Streams now come with concatenation and are part of the HOLCF image
  2673 
  2674 
  2675 
  2676 New in Isabelle2003 (May 2003)
  2677 ------------------------------
  2678 
  2679 *** General ***
  2680 
  2681 * Provers/simplifier:
  2682 
  2683   - Completely reimplemented method simp (ML: Asm_full_simp_tac):
  2684     Assumptions are now subject to complete mutual simplification,
  2685     not just from left to right. The simplifier now preserves
  2686     the order of assumptions.
  2687 
  2688     Potential INCOMPATIBILITY:
  2689 
  2690     -- simp sometimes diverges where the old version did
  2691        not, e.g. invoking simp on the goal
  2692 
  2693         [| P (f x); y = x; f x = f y |] ==> Q
  2694 
  2695        now gives rise to the infinite reduction sequence
  2696 
  2697         P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
  2698 
  2699        Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
  2700        kind of problem.
  2701 
  2702     -- Tactics combining classical reasoner and simplification (such as auto)
  2703        are also affected by this change, because many of them rely on
  2704        simp. They may sometimes diverge as well or yield a different numbers
  2705        of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
  2706        in case of problems. Sometimes subsequent calls to the classical
  2707        reasoner will fail because a preceeding call to the simplifier too
  2708        eagerly simplified the goal, e.g. deleted redundant premises.
  2709 
  2710   - The simplifier trace now shows the names of the applied rewrite rules
  2711 
  2712   - You can limit the number of recursive invocations of the simplifier
  2713     during conditional rewriting (where the simplifie tries to solve the
  2714     conditions before applying the rewrite rule):
  2715     ML "simp_depth_limit := n"
  2716     where n is an integer. Thus you can force termination where previously
  2717     the simplifier would diverge.
  2718 
  2719   - Accepts free variables as head terms in congruence rules.  Useful in Isar.
  2720 
  2721   - No longer aborts on failed congruence proof.  Instead, the
  2722     congruence is ignored.
  2723 
  2724 * Pure: New generic framework for extracting programs from constructive
  2725   proofs. See HOL/Extraction.thy for an example instantiation, as well
  2726   as HOL/Extraction for some case studies.
  2727 
  2728 * Pure: The main goal of the proof state is no longer shown by default, only
  2729 the subgoals. This behaviour is controlled by a new flag.
  2730    PG menu: Isabelle/Isar -> Settings -> Show Main Goal
  2731 (ML: Proof.show_main_goal).
  2732 
  2733 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
  2734 rules whose conclusion matches subgoal 1:
  2735       PG menu: Isabelle/Isar -> Show me -> matching rules
  2736 The rules are ordered by how closely they match the subgoal.
  2737 In particular, rules that solve a subgoal outright are displayed first
  2738 (or rather last, the way they are printed).
  2739 (ML: ProofGeneral.print_intros())
  2740 
  2741 * Pure: New flag trace_unify_fail causes unification to print
  2742 diagnostic information (PG: in trace buffer) when it fails. This is
  2743 useful for figuring out why single step proofs like rule, erule or
  2744 assumption failed.
  2745 
  2746 * Pure: Locale specifications now produce predicate definitions
  2747 according to the body of text (covering assumptions modulo local
  2748 definitions); predicate "loc_axioms" covers newly introduced text,
  2749 while "loc" is cumulative wrt. all included locale expressions; the
  2750 latter view is presented only on export into the global theory
  2751 context; potential INCOMPATIBILITY, use "(open)" option to fall back
  2752 on the old view without predicates;
  2753 
  2754 * Pure: predefined locales "var" and "struct" are useful for sharing
  2755 parameters (as in CASL, for example); just specify something like
  2756 ``var x + var y + struct M'' as import;
  2757 
  2758 * Pure: improved thms_containing: proper indexing of facts instead of
  2759 raw theorems; check validity of results wrt. current name space;
  2760 include local facts of proof configuration (also covers active
  2761 locales), cover fixed variables in index; may use "_" in term
  2762 specification; an optional limit for the number of printed facts may
  2763 be given (the default is 40);
  2764 
  2765 * Pure: disallow duplicate fact bindings within new-style theory files
  2766 (batch-mode only);
  2767 
  2768 * Provers: improved induct method: assumptions introduced by case
  2769 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
  2770 the goal statement); "foo" still refers to all facts collectively;
  2771 
  2772 * Provers: the function blast.overloaded has been removed: all constants
  2773 are regarded as potentially overloaded, which improves robustness in exchange
  2774 for slight decrease in efficiency;
  2775 
  2776 * Provers/linorder: New generic prover for transitivity reasoning over
  2777 linear orders.  Note: this prover is not efficient!
  2778 
  2779 * Isar: preview of problems to finish 'show' now produce an error
  2780 rather than just a warning (in interactive mode);
  2781 
  2782 
  2783 *** HOL ***
  2784 
  2785 * arith(_tac)
  2786 
  2787  - Produces a counter example if it cannot prove a goal.
  2788    Note that the counter example may be spurious if the goal is not a formula
  2789    of quantifier-free linear arithmetic.
  2790    In ProofGeneral the counter example appears in the trace buffer.
  2791 
  2792  - Knows about div k and mod k where k is a numeral of type nat or int.
  2793 
  2794  - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
  2795    linear arithmetic fails. This takes account of quantifiers and divisibility.
  2796    Presburger arithmetic can also be called explicitly via presburger(_tac).
  2797 
  2798 * simp's arithmetic capabilities have been enhanced a bit: it now
  2799 takes ~= in premises into account (by performing a case split);
  2800 
  2801 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
  2802 are distributed over a sum of terms;
  2803 
  2804 * New tactic "trans_tac" and method "trans" instantiate
  2805 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
  2806 "<=", "<" and "=").
  2807 
  2808 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main
  2809 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
  2810 
  2811 * 'typedef' command has new option "open" to suppress the set
  2812 definition;
  2813 
  2814 * functions Min and Max on finite sets have been introduced (theory
  2815 Finite_Set);
  2816 
  2817 * attribute [symmetric] now works for relations as well; it turns
  2818 (x,y) : R^-1 into (y,x) : R, and vice versa;
  2819 
  2820 * induct over a !!-quantified statement (say !!x1..xn):
  2821   each "case" automatically performs "fix x1 .. xn" with exactly those names.
  2822 
  2823 * Map: `empty' is no longer a constant but a syntactic abbreviation for
  2824 %x. None. Warning: empty_def now refers to the previously hidden definition
  2825 of the empty set.
  2826 
  2827 * Algebra: formalization of classical algebra.  Intended as base for
  2828 any algebraic development in Isabelle.  Currently covers group theory
  2829 (up to Sylow's theorem) and ring theory (Universal Property of
  2830 Univariate Polynomials).  Contributions welcome;
  2831 
  2832 * GroupTheory: deleted, since its material has been moved to Algebra;
  2833 
  2834 * Complex: new directory of the complex numbers with numeric constants,
  2835 nonstandard complex numbers, and some complex analysis, standard and
  2836 nonstandard (Jacques Fleuriot);
  2837 
  2838 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
  2839 
  2840 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques
  2841 Fleuriot);
  2842 
  2843 * Real/HahnBanach: updated and adapted to locales;
  2844 
  2845 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
  2846 Gray and Kramer);
  2847 
  2848 * UNITY: added the Meier-Sanders theory of progress sets;
  2849 
  2850 * MicroJava: bytecode verifier and lightweight bytecode verifier
  2851 as abstract algorithms, instantiated to the JVM;
  2852 
  2853 * Bali: Java source language formalization. Type system, operational
  2854 semantics, axiomatic semantics. Supported language features:
  2855 classes, interfaces, objects,virtual methods, static methods,
  2856 static/instance fields, arrays, access modifiers, definite
  2857 assignment, exceptions.
  2858 
  2859 
  2860 *** ZF ***
  2861 
  2862 * ZF/Constructible: consistency proof for AC (Gdel's constructible
  2863 universe, etc.);
  2864 
  2865 * Main ZF: virtually all theories converted to new-style format;
  2866 
  2867 
  2868 *** ML ***
  2869 
  2870 * Pure: Tactic.prove provides sane interface for internal proofs;
  2871 omits the infamous "standard" operation, so this is more appropriate
  2872 than prove_goalw_cterm in many situations (e.g. in simprocs);
  2873 
  2874 * Pure: improved error reporting of simprocs;
  2875 
  2876 * Provers: Simplifier.simproc(_i) provides sane interface for setting
  2877 up simprocs;
  2878 
  2879 
  2880 *** Document preparation ***
  2881 
  2882 * uses \par instead of \\ for line breaks in theory text. This may
  2883 shift some page breaks in large documents. To get the old behaviour
  2884 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
  2885 
  2886 * minimized dependencies of isabelle.sty and isabellesym.sty on
  2887 other packages
  2888 
  2889 * \<euro> now needs package babel/greek instead of marvosym (which
  2890 broke \Rightarrow)
  2891 
  2892 * normal size for \<zero>...\<nine> (uses \mathbf instead of
  2893 textcomp package)
  2894 
  2895 
  2896 
  2897 New in Isabelle2002 (March 2002)
  2898 --------------------------------
  2899 
  2900 *** Document preparation ***
  2901 
  2902 * greatly simplified document preparation setup, including more
  2903 graceful interpretation of isatool usedir -i/-d/-D options, and more
  2904 instructive isatool mkdir; users should basically be able to get
  2905 started with "isatool mkdir HOL Test && isatool make"; alternatively,
  2906 users may run a separate document processing stage manually like this:
  2907 "isatool usedir -D output HOL Test && isatool document Test/output";
  2908 
  2909 * theory dependency graph may now be incorporated into documents;
  2910 isatool usedir -g true will produce session_graph.eps/.pdf for use
  2911 with \includegraphics of LaTeX;
  2912 
  2913 * proper spacing of consecutive markup elements, especially text
  2914 blocks after section headings;
  2915 
  2916 * support bold style (for single symbols only), input syntax is like
  2917 this: "\<^bold>\<alpha>" or "\<^bold>A";
  2918 
  2919 * \<bullet> is now output as bold \cdot by default, which looks much
  2920 better in printed text;
  2921 
  2922 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
  2923 note that these symbols are currently unavailable in Proof General /
  2924 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
  2925 
  2926 * isatool latex no longer depends on changed TEXINPUTS, instead
  2927 isatool document copies the Isabelle style files to the target
  2928 location;
  2929 
  2930 
  2931 *** Isar ***
  2932 
  2933 * Pure/Provers: improved proof by cases and induction;
  2934   - 'case' command admits impromptu naming of parameters (such as
  2935     "case (Suc n)");
  2936   - 'induct' method divinates rule instantiation from the inductive
  2937     claim; no longer requires excessive ?P bindings for proper
  2938     instantiation of cases;
  2939   - 'induct' method properly enumerates all possibilities of set/type
  2940     rules; as a consequence facts may be also passed through *type*
  2941     rules without further ado;
  2942   - 'induct' method now derives symbolic cases from the *rulified*
  2943     rule (before it used to rulify cases stemming from the internal
  2944     atomized version); this means that the context of a non-atomic
  2945     statement becomes is included in the hypothesis, avoiding the
  2946     slightly cumbersome show "PROP ?case" form;
  2947   - 'induct' may now use elim-style induction rules without chaining
  2948     facts, using ``missing'' premises from the goal state; this allows
  2949     rules stemming from inductive sets to be applied in unstructured
  2950     scripts, while still benefitting from proper handling of non-atomic
  2951     statements; NB: major inductive premises need to be put first, all
  2952     the rest of the goal is passed through the induction;
  2953   - 'induct' proper support for mutual induction involving non-atomic
  2954     rule statements (uses the new concept of simultaneous goals, see
  2955     below);
  2956   - append all possible rule selections, but only use the first
  2957     success (no backtracking);
  2958   - removed obsolete "(simplified)" and "(stripped)" options of methods;
  2959   - undeclared rule case names default to numbers 1, 2, 3, ...;
  2960   - added 'print_induct_rules' (covered by help item in recent Proof
  2961     General versions);
  2962   - moved induct/cases attributes to Pure, methods to Provers;
  2963   - generic method setup instantiated for FOL and HOL;
  2964 
  2965 * Pure: support multiple simultaneous goal statements, for example
  2966 "have a: A and b: B" (same for 'theorem' etc.); being a pure
  2967 meta-level mechanism, this acts as if several individual goals had
  2968 been stated separately; in particular common proof methods need to be
  2969 repeated in order to cover all claims; note that a single elimination
  2970 step is *not* sufficient to establish the two conjunctions, so this
  2971 fails:
  2972 
  2973   assume "A & B" then have A and B ..   (*".." fails*)
  2974 
  2975 better use "obtain" in situations as above; alternative refer to
  2976 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
  2977 
  2978 * Pure: proper integration with ``locales''; unlike the original
  2979 version by Florian Kammller, Isar locales package high-level proof
  2980 contexts rather than raw logical ones (e.g. we admit to include
  2981 attributes everywhere); operations on locales include merge and
  2982 rename; support for implicit arguments (``structures''); simultaneous
  2983 type-inference over imports and text; see also HOL/ex/Locales.thy for
  2984 some examples;
  2985 
  2986 * Pure: the following commands have been ``localized'', supporting a
  2987 target locale specification "(in name)": 'lemma', 'theorem',
  2988 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
  2989 stored both within the locale and at the theory level (exported and
  2990 qualified by the locale name);
  2991 
  2992 * Pure: theory goals may now be specified in ``long'' form, with
  2993 ad-hoc contexts consisting of arbitrary locale elements. for example
  2994 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
  2995 definitions may be given, too); the result is a meta-level rule with
  2996 the context elements being discharged in the obvious way;
  2997 
  2998 * Pure: new proof command 'using' allows to augment currently used
  2999 facts after a goal statement ('using' is syntactically analogous to
  3000 'apply', but acts on the goal's facts only); this allows chained facts
  3001 to be separated into parts given before and after a claim, as in
  3002 ``from a and b have C using d and e <proof>'';
  3003 
  3004 * Pure: renamed "antecedent" case to "rule_context";
  3005 
  3006 * Pure: new 'judgment' command records explicit information about the
  3007 object-logic embedding (used by several tools internally); no longer
  3008 use hard-wired "Trueprop";
  3009 
  3010 * Pure: added 'corollary' command;
  3011 
  3012 * Pure: fixed 'token_translation' command;
  3013 
  3014 * Pure: removed obsolete 'exported' attribute;
  3015 
  3016 * Pure: dummy pattern "_" in is/let is now automatically lifted over
  3017 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
  3018 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
  3019 
  3020 * Pure: method 'atomize' presents local goal premises as object-level
  3021 statements (atomic meta-level propositions); setup controlled via
  3022 rewrite rules declarations of 'atomize' attribute; example
  3023 application: 'induct' method with proper rule statements in improper
  3024 proof *scripts*;
  3025 
  3026 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
  3027 now consider the syntactic context of assumptions, giving a better
  3028 chance to get type-inference of the arguments right (this is
  3029 especially important for locales);
  3030 
  3031 * Pure: "sorry" no longer requires quick_and_dirty in interactive
  3032 mode;
  3033 
  3034 * Pure/obtain: the formal conclusion "thesis", being marked as
  3035 ``internal'', may no longer be reference directly in the text;
  3036 potential INCOMPATIBILITY, may need to use "?thesis" in rare
  3037 situations;
  3038 
  3039 * Pure: generic 'sym' attribute which declares a rule both as pure
  3040 'elim?' and for the 'symmetric' operation;
  3041 
  3042 * Pure: marginal comments ``--'' may now occur just anywhere in the
  3043 text; the fixed correlation with particular command syntax has been
  3044 discontinued;
  3045 
  3046 * Pure: new method 'rules' is particularly well-suited for proof
  3047 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
  3048 but often produces more compact proof terms with less detours;
  3049 
  3050 * Pure/Provers/classical: simplified integration with pure rule
  3051 attributes and methods; the classical "intro?/elim?/dest?"
  3052 declarations coincide with the pure ones; the "rule" method no longer
  3053 includes classically swapped intros; "intro" and "elim" methods no
  3054 longer pick rules from the context; also got rid of ML declarations
  3055 AddXIs/AddXEs/AddXDs; all of this has some potential for
  3056 INCOMPATIBILITY;
  3057 
  3058 * Provers/classical: attribute 'swapped' produces classical inversions
  3059 of introduction rules;
  3060 
  3061 * Provers/simplifier: 'simplified' attribute may refer to explicit
  3062 rules instead of full simplifier context; 'iff' attribute handles
  3063 conditional rules;
  3064 
  3065 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
  3066 
  3067 * HOL: 'recdef' now fails on unfinished automated proofs, use
  3068 "(permissive)" option to recover old behavior;
  3069 
  3070 * HOL: 'inductive' no longer features separate (collective) attributes
  3071 for 'intros' (was found too confusing);
  3072 
  3073 * HOL: properly declared induction rules less_induct and
  3074 wf_induct_rule;
  3075 
  3076 
  3077 *** HOL ***
  3078 
  3079 * HOL: moved over to sane numeral syntax; the new policy is as
  3080 follows:
  3081 
  3082   - 0 and 1 are polymorphic constants, which are defined on any
  3083   numeric type (nat, int, real etc.);
  3084 
  3085   - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
  3086   binary representation internally;
  3087 
  3088   - type nat has special constructor Suc, and generally prefers Suc 0
  3089   over 1::nat and Suc (Suc 0) over 2::nat;
  3090 
  3091 This change may cause significant problems of INCOMPATIBILITY; here
  3092 are some hints on converting existing sources:
  3093 
  3094   - due to the new "num" token, "-0" and "-1" etc. are now atomic
  3095   entities, so expressions involving "-" (unary or binary minus) need
  3096   to be spaced properly;
  3097 
  3098   - existing occurrences of "1" may need to be constraint "1::nat" or
  3099   even replaced by Suc 0; similar for old "2";
  3100 
  3101   - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
  3102 
  3103   - remove all special provisions on numerals in proofs;
  3104 
  3105 * HOL: simp rules nat_number expand numerals on nat to Suc/0
  3106 representation (depends on bin_arith_simps in the default context);
  3107 
  3108 * HOL: symbolic syntax for x^2 (numeral 2);
  3109 
  3110 * HOL: the class of all HOL types is now called "type" rather than
  3111 "term"; INCOMPATIBILITY, need to adapt references to this type class
  3112 in axclass/classes, instance/arities, and (usually rare) occurrences
  3113 in typings (of consts etc.); internally the class is called
  3114 "HOL.type", ML programs should refer to HOLogic.typeS;
  3115 
  3116 * HOL/record package improvements:
  3117   - new derived operations "fields" to build a partial record section,
  3118     "extend" to promote a fixed record to a record scheme, and
  3119     "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
  3120     declared as simp by default;
  3121   - shared operations ("more", "fields", etc.) now need to be always
  3122     qualified) --- potential INCOMPATIBILITY;
  3123   - removed "make_scheme" operations (use "make" with "extend") --
  3124     INCOMPATIBILITY;
  3125   - removed "more" class (simply use "term") -- INCOMPATIBILITY;
  3126   - provides cases/induct rules for use with corresponding Isar
  3127     methods (for concrete records, record schemes, concrete more
  3128     parts, and schematic more parts -- in that order);
  3129   - internal definitions directly based on a light-weight abstract
  3130     theory of product types over typedef rather than datatype;
  3131 
  3132 * HOL: generic code generator for generating executable ML code from
  3133 specifications; specific support for HOL constructs such as inductive
  3134 datatypes and sets, as well as recursive functions; can be invoked
  3135 via 'generate_code' theory section;
  3136 
  3137 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
  3138 
  3139 * HOL: consolidated and renamed several theories.  In particular:
  3140         Ord.thy has been absorbed into HOL.thy
  3141         String.thy has been absorbed into List.thy
  3142 
  3143 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
  3144 (beware of argument permutation!);
  3145 
  3146 * HOL: linorder_less_split superseded by linorder_cases;
  3147 
  3148 * HOL/List: "nodups" renamed to "distinct";
  3149 
  3150 * HOL: added "The" definite description operator; move Hilbert's "Eps"
  3151 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
  3152   - Ex_def has changed, now need to use some_eq_ex
  3153 
  3154 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
  3155 in this (rare) case use:
  3156 
  3157   delSWrapper "split_all_tac"
  3158   addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
  3159 
  3160 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
  3161 MAY FAIL;
  3162 
  3163 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
  3164 Isabelle's type classes, ^ on functions and relations has too general
  3165 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
  3166 necessary to attach explicit type constraints;
  3167 
  3168 * HOL/Relation: the prefix name of the infix "O" has been changed from
  3169 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
  3170 renamed accordingly (eg "compI" -> "rel_compI").
  3171 
  3172 * HOL: syntax translations now work properly with numerals and records
  3173 expressions;
  3174 
  3175 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
  3176 of "lam" -- INCOMPATIBILITY;
  3177 
  3178 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
  3179 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
  3180 renamed "Product_Type.unit";
  3181 
  3182 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
  3183 
  3184 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
  3185 the "cases" method);
  3186 
  3187 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
  3188 Florian Kammller);
  3189 
  3190 * HOL/IMP: updated and converted to new-style theory format; several
  3191 parts turned into readable document, with proper Isar proof texts and
  3192 some explanations (by Gerwin Klein);
  3193 
  3194 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
  3195 
  3196 * HOL-Hyperreal is now a logic image;
  3197 
  3198 
  3199 *** HOLCF ***
  3200 
  3201 * Isar: consts/constdefs supports mixfix syntax for continuous
  3202 operations;
  3203 
  3204 * Isar: domain package adapted to new-style theory format, e.g. see
  3205 HOLCF/ex/Dnat.thy;
  3206 
  3207 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
  3208 potential INCOMPATIBILITY; now use plain induct_tac instead of former
  3209 lift.induct_tac, always use UU instead of Undef;
  3210 
  3211 * HOLCF/IMP: updated and converted to new-style theory;
  3212 
  3213 
  3214 *** ZF ***
  3215 
  3216 * Isar: proper integration of logic-specific tools and packages,
  3217 including theory commands '(co)inductive', '(co)datatype',
  3218 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
  3219 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
  3220 
  3221 * theory Main no longer includes AC; for the Axiom of Choice, base
  3222 your theory on Main_ZFC;
  3223 
  3224 * the integer library now covers quotients and remainders, with many
  3225 laws relating division to addition, multiplication, etc.;
  3226 
  3227 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
  3228 typeless version of the formalism;
  3229 
  3230 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
  3231 format;
  3232 
  3233 * ZF/Induct: new directory for examples of inductive definitions,
  3234 including theory Multiset for multiset orderings; converted to
  3235 new-style theory format;
  3236 
  3237 * ZF: many new theorems about lists, ordinals, etc.;
  3238 
  3239 
  3240 *** General ***
  3241 
  3242 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
  3243 variable proof controls level of detail: 0 = no proofs (only oracle
  3244 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
  3245 also ref manual for further ML interfaces;
  3246 
  3247 * Pure/axclass: removed obsolete ML interface
  3248 goal_subclass/goal_arity;
  3249 
  3250 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
  3251 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
  3252 separate tokens, so expressions involving minus need to be spaced
  3253 properly;
  3254 
  3255 * Pure/syntax: support non-oriented infixes, using keyword "infix"
  3256 rather than "infixl" or "infixr";
  3257 
  3258 * Pure/syntax: concrete syntax for dummy type variables admits genuine
  3259 sort constraint specifications in type inference; e.g. "x::_::foo"
  3260 ensures that the type of "x" is of sort "foo" (but not necessarily a
  3261 type variable);
  3262 
  3263 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
  3264 control output of nested => (types); the default behavior is
  3265 "type_brackets";
  3266 
  3267 * Pure/syntax: builtin parse translation for "_constify" turns valued
  3268 tokens into AST constants;
  3269 
  3270 * Pure/syntax: prefer later declarations of translations and print
  3271 translation functions; potential INCOMPATIBILITY: need to reverse
  3272 multiple declarations for same syntax element constant;
  3273 
  3274 * Pure/show_hyps reset by default (in accordance to existing Isar
  3275 practice);
  3276 
  3277 * Provers/classical: renamed addaltern to addafter, addSaltern to
  3278 addSafter;
  3279 
  3280 * Provers/clasimp: ``iff'' declarations now handle conditional rules
  3281 as well;
  3282 
  3283 * system: tested support for MacOS X; should be able to get Isabelle +
  3284 Proof General to work in a plain Terminal after installing Poly/ML
  3285 (e.g. from the Isabelle distribution area) and GNU bash alone
  3286 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
  3287 support requires further installations, e.g. from
  3288 http://fink.sourceforge.net/);
  3289 
  3290 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
  3291 
  3292 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
  3293 of 40 MB), cf. ML_OPTIONS;
  3294 
  3295 * system: Proof General keywords specification is now part of the
  3296 Isabelle distribution (see etc/isar-keywords.el);
  3297 
  3298 * system: support for persistent Proof General sessions (refrain from
  3299 outdating all loaded theories on startup); user may create writable
  3300 logic images like this: ``isabelle -q HOL Test'';
  3301 
  3302 * system: smart selection of Isabelle process versus Isabelle
  3303 interface, accommodates case-insensitive file systems (e.g. HFS+); may
  3304 run both "isabelle" and "Isabelle" even if file names are badly
  3305 damaged (executable inspects the case of the first letter of its own
  3306 name); added separate "isabelle-process" and "isabelle-interface";
  3307 
  3308 * system: refrain from any attempt at filtering input streams; no
  3309 longer support ``8bit'' encoding of old isabelle font, instead proper
  3310 iso-latin characters may now be used; the related isatools
  3311 "symbolinput" and "nonascii" have disappeared as well;
  3312 
  3313 * system: removed old "xterm" interface (the print modes "xterm" and
  3314 "xterm_color" are still available for direct use in a suitable
  3315 terminal);
  3316 
  3317 
  3318 
  3319 New in Isabelle99-2 (February 2001)
  3320 -----------------------------------
  3321 
  3322 *** Overview of INCOMPATIBILITIES ***
  3323 
  3324 * HOL: please note that theories in the Library and elsewhere often use the
  3325 new-style (Isar) format; to refer to their theorems in an ML script you must
  3326 bind them to ML identifers by e.g.      val thm_name = thm "thm_name";
  3327 
  3328 * HOL: inductive package no longer splits induction rule aggressively,
  3329 but only as far as specified by the introductions given; the old
  3330 format may be recovered via ML function complete_split_rule or attribute
  3331 'split_rule (complete)';
  3332 
  3333 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
  3334 gfp_Tarski to gfp_unfold;
  3335 
  3336 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
  3337 
  3338 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
  3339 relation); infix "^^" has been renamed "``"; infix "``" has been
  3340 renamed "`"; "univalent" has been renamed "single_valued";
  3341 
  3342 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
  3343 operation;
  3344 
  3345 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
  3346 
  3347 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
  3348 
  3349 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
  3350 consequence, it is no longer monotonic wrt. the local goal context
  3351 (which is now passed through the inductive cases);
  3352 
  3353 * Document preparation: renamed standard symbols \<ll> to \<lless> and
  3354 \<gg> to \<ggreater>;
  3355 
  3356 
  3357 *** Document preparation ***
  3358 
  3359 * \isabellestyle{NAME} selects version of Isabelle output (currently
  3360 available: are "it" for near math-mode best-style output, "sl" for
  3361 slanted text style, and "tt" for plain type-writer; if no
  3362 \isabellestyle command is given, output is according to slanted
  3363 type-writer);
  3364 
  3365 * support sub/super scripts (for single symbols only), input syntax is
  3366 like this: "A\<^sup>*" or "A\<^sup>\<star>";
  3367 
  3368 * some more standard symbols; see Appendix A of the system manual for
  3369 the complete list of symbols defined in isabellesym.sty;
  3370 
  3371 * improved isabelle style files; more abstract symbol implementation
  3372 (should now use \isamath{...} and \isatext{...} in custom symbol
  3373 definitions);
  3374 
  3375 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
  3376 state; Note that presentation of goal states does not conform to
  3377 actual human-readable proof documents.  Please do not include goal
  3378 states into document output unless you really know what you are doing!
  3379 
  3380 * proper indentation of antiquoted output with proportional LaTeX
  3381 fonts;
  3382 
  3383 * no_document ML operator temporarily disables LaTeX document
  3384 generation;
  3385 
  3386 * isatool unsymbolize tunes sources for plain ASCII communication;
  3387 
  3388 
  3389 *** Isar ***
  3390 
  3391 * Pure: Isar now suffers initial goal statements to contain unbound
  3392 schematic variables (this does not conform to actual readable proof
  3393 documents, due to unpredictable outcome and non-compositional proof
  3394 checking); users who know what they are doing may use schematic goals
  3395 for Prolog-style synthesis of proven results;
  3396 
  3397 * Pure: assumption method (an implicit finishing) now handles actual
  3398 rules as well;
  3399 
  3400 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
  3401 initial goal, declare "that" only as Pure intro (only for single
  3402 steps); the "that" rule assumption may now be involved in implicit
  3403 finishing, thus ".." becomes a feasible for trivial obtains;
  3404 
  3405 * Pure: default proof step now includes 'intro_classes'; thus trivial
  3406 instance proofs may be performed by "..";
  3407 
  3408 * Pure: ?thesis / ?this / "..." now work for pure meta-level
  3409 statements as well;
  3410 
  3411 * Pure: more robust selection of calculational rules;
  3412 
  3413 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
  3414 rule (as well as the assumption rule);
  3415 
  3416 * Pure: 'thm_deps' command visualizes dependencies of theorems and
  3417 lemmas, using the graph browser tool;
  3418 
  3419 * Pure: predict failure of "show" in interactive mode;
  3420 
  3421 * Pure: 'thms_containing' now takes actual terms as arguments;
  3422 
  3423 * HOL: improved method 'induct' --- now handles non-atomic goals
  3424 (potential INCOMPATIBILITY); tuned error handling;
  3425 
  3426 * HOL: cases and induct rules now provide explicit hints about the
  3427 number of facts to be consumed (0 for "type" and 1 for "set" rules);
  3428 any remaining facts are inserted into the goal verbatim;
  3429 
  3430 * HOL: local contexts (aka cases) may now contain term bindings as
  3431 well; the 'cases' and 'induct' methods new provide a ?case binding for
  3432 the result to be shown in each case;
  3433 
  3434 * HOL: added 'recdef_tc' command;
  3435 
  3436 * isatool convert assists in eliminating legacy ML scripts;
  3437 
  3438 
  3439 *** HOL ***
  3440 
  3441 * HOL/Library: a collection of generic theories to be used together
  3442 with main HOL; the theory loader path already includes this directory
  3443 by default; the following existing theories have been moved here:
  3444 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
  3445 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
  3446 
  3447 * HOL/Unix: "Some aspects of Unix file-system security", a typical
  3448 modelling and verification task performed in Isabelle/HOL +
  3449 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
  3450 
  3451 * HOL/Algebra: special summation operator SUM no longer exists, it has
  3452 been replaced by setsum; infix 'assoc' now has priority 50 (like
  3453 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
  3454 'domain', this makes the theory consistent with mathematical
  3455 literature;
  3456 
  3457 * HOL basics: added overloaded operations "inverse" and "divide"
  3458 (infix "/"), syntax for generic "abs" operation, generic summation
  3459 operator \<Sum>;
  3460 
  3461 * HOL/typedef: simplified package, provide more useful rules (see also
  3462 HOL/subset.thy);
  3463 
  3464 * HOL/datatype: induction rule for arbitrarily branching datatypes is
  3465 now expressed as a proper nested rule (old-style tactic scripts may
  3466 require atomize_strip_tac to cope with non-atomic premises);
  3467 
  3468 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
  3469 to "split_conv" (old name still available for compatibility);
  3470 
  3471 * HOL: improved concrete syntax for strings (e.g. allows translation
  3472 rules with string literals);
  3473 
  3474 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
  3475  and Fleuriot's mechanization of analysis, including the transcendental
  3476  functions for the reals;
  3477 
  3478 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
  3479 
  3480 
  3481 *** CTT ***
  3482 
  3483 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
  3484 "lam" is displayed as TWO lambda-symbols
  3485 
  3486 * CTT: theory Main now available, containing everything (that is, Bool
  3487 and Arith);
  3488 
  3489 
  3490 *** General ***
  3491 
  3492 * Pure: the Simplifier has been implemented properly as a derived rule
  3493 outside of the actual kernel (at last!); the overall performance
  3494 penalty in practical applications is about 50%, while reliability of
  3495 the Isabelle inference kernel has been greatly improved;
  3496 
  3497 * print modes "brackets" and "no_brackets" control output of nested =>
  3498 (types) and ==> (props); the default behaviour is "brackets";
  3499 
  3500 * Provers: fast_tac (and friends) now handle actual object-logic rules
  3501 as assumptions as well;
  3502 
  3503 * system: support Poly/ML 4.0;
  3504 
  3505 * system: isatool install handles KDE version 1 or 2;
  3506 
  3507 
  3508 
  3509 New in Isabelle99-1 (October 2000)
  3510 ----------------------------------
  3511 
  3512 *** Overview of INCOMPATIBILITIES ***
  3513 
  3514 * HOL: simplification of natural numbers is much changed; to partly
  3515 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
  3516 issue the following ML commands:
  3517 
  3518   Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
  3519   Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
  3520 
  3521 * HOL: simplification no longer dives into case-expressions; this is
  3522 controlled by "t.weak_case_cong" for each datatype t;
  3523 
  3524 * HOL: nat_less_induct renamed to less_induct;
  3525 
  3526 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
  3527 fixsome to patch .thy and .ML sources automatically;
  3528 
  3529   select_equality  -> some_equality
  3530   select_eq_Ex     -> some_eq_ex
  3531   selectI2EX       -> someI2_ex
  3532   selectI2         -> someI2
  3533   selectI          -> someI
  3534   select1_equality -> some1_equality
  3535   Eps_sym_eq       -> some_sym_eq_trivial
  3536   Eps_eq           -> some_eq_trivial
  3537 
  3538 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
  3539 
  3540 * HOL: removed obsolete theorem binding expand_if (refer to split_if
  3541 instead);
  3542 
  3543 * HOL: the recursion equations generated by 'recdef' are now called
  3544 f.simps instead of f.rules;
  3545 
  3546 * HOL: qed_spec_mp now also handles bounded ALL as well;
  3547 
  3548 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
  3549 sometimes be needed;
  3550 
  3551 * HOL: the constant for "f``x" is now "image" rather than "op ``";
  3552 
  3553 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
  3554 
  3555 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
  3556 product is now "<*>" instead of "Times"; the lexicographic product is
  3557 now "<*lex*>" instead of "**";
  3558 
  3559 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
  3560 of main HOL, but was unused); better use HOL's datatype package;
  3561 
  3562 * HOL: removed "symbols" syntax for constant "override" of theory Map;
  3563 the old syntax may be recovered as follows:
  3564 
  3565   syntax (symbols)
  3566     override  :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
  3567       (infixl "\\<oplus>" 100)
  3568 
  3569 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3570 
  3571 * HOL/ML: even fewer consts are declared as global (see theories Ord,
  3572 Lfp, Gfp, WF); this only affects ML packages that refer to const names
  3573 internally;
  3574 
  3575 * HOL and ZF: syntax for quotienting wrt an equivalence relation
  3576 changed from A/r to A//r;
  3577 
  3578 * ZF: new treatment of arithmetic (nat & int) may break some old
  3579 proofs;
  3580 
  3581 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
  3582 rulify -> rule_format, elimify -> elim_format, ...);
  3583 
  3584 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3585 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
  3586 should have to change intro!! to intro? only); replaced "delrule" by
  3587 "rule del";
  3588 
  3589 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
  3590 
  3591 * Provers: strengthened force_tac by using new first_best_tac;
  3592 
  3593 * LaTeX document preparation: several changes of isabelle.sty (see
  3594 lib/texinputs);
  3595 
  3596 
  3597 *** Document preparation ***
  3598 
  3599 * formal comments (text blocks etc.) in new-style theories may now
  3600 contain antiquotations of thm/prop/term/typ/text to be presented
  3601 according to latex print mode; concrete syntax is like this:
  3602 @{term[show_types] "f(x) = a + x"};
  3603 
  3604 * isatool mkdir provides easy setup of Isabelle session directories,
  3605 including proper document sources;
  3606 
  3607 * generated LaTeX sources are now deleted after successful run
  3608 (isatool document -c); may retain a copy somewhere else via -D option
  3609 of isatool usedir;
  3610 
  3611 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
  3612 style files, achieving self-contained LaTeX sources and simplifying
  3613 LaTeX debugging;
  3614 
  3615 * old-style theories now produce (crude) LaTeX output as well;
  3616 
  3617 * browser info session directories are now self-contained (may be put
  3618 on WWW server seperately); improved graphs of nested sessions; removed
  3619 graph for 'all sessions';
  3620 
  3621 * several improvements in isabelle style files; \isabellestyle{it}
  3622 produces fake math mode output; \isamarkupheader is now \section by
  3623 default; see lib/texinputs/isabelle.sty etc.;
  3624 
  3625 
  3626 *** Isar ***
  3627 
  3628 * Isar/Pure: local results and corresponding term bindings are now
  3629 subject to Hindley-Milner polymorphism (similar to ML); this
  3630 accommodates incremental type-inference very nicely;
  3631 
  3632 * Isar/Pure: new derived language element 'obtain' supports
  3633 generalized existence reasoning;
  3634 
  3635 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
  3636 support accumulation of results, without applying any rules yet;
  3637 useful to collect intermediate results without explicit name
  3638 references, and for use with transitivity rules with more than 2
  3639 premises;
  3640 
  3641 * Isar/Pure: scalable support for case-analysis type proofs: new
  3642 'case' language element refers to local contexts symbolically, as
  3643 produced by certain proof methods; internally, case names are attached
  3644 to theorems as "tags";
  3645 
  3646 * Isar/Pure: theory command 'hide' removes declarations from
  3647 class/type/const name spaces;
  3648 
  3649 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
  3650 indicate potential overloading;
  3651 
  3652 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
  3653 
  3654 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
  3655 "{a,b,c}" instead of {a,b,c};
  3656 
  3657 * Isar/Pure now provides its own version of intro/elim/dest
  3658 attributes; useful for building new logics, but beware of confusion
  3659 with the version in Provers/classical;
  3660 
  3661 * Isar/Pure: the local context of (non-atomic) goals is provided via
  3662 case name 'antecedent';
  3663 
  3664 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
  3665 to the current context is now done automatically);
  3666 
  3667 * Isar/Pure: theory command 'method_setup' provides a simple interface
  3668 for definining proof methods in ML;
  3669 
  3670 * Isar/Provers: intro/elim/dest attributes changed; renamed
  3671 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
  3672 most cases, one should have to change intro!! to intro? only);
  3673 replaced "delrule" by "rule del";
  3674 
  3675 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
  3676 'symmetric' attribute (the latter supercedes [RS sym]);
  3677 
  3678 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
  3679 method modifier); 'simp' method: 'only:' modifier removes loopers as
  3680 well (including splits);
  3681 
  3682 * Isar/Provers: Simplifier and Classical methods now support all kind
  3683 of modifiers used in the past, including 'cong', 'iff', etc.
  3684 
  3685 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
  3686 of Simplifier and Classical reasoner);
  3687 
  3688 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
  3689 now support named cases; major packages (inductive, datatype, primrec,
  3690 recdef) support case names and properly name parameters;
  3691 
  3692 * Isar/HOL: new transitivity rules for substitution in inequalities --
  3693 monotonicity conditions are extracted to be proven at end of
  3694 calculations;
  3695 
  3696 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
  3697 method anyway;
  3698 
  3699 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
  3700 split_if split_if_asm; datatype package provides theorems foo.splits =
  3701 foo.split foo.split_asm for each datatype;
  3702 
  3703 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
  3704 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
  3705 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
  3706 use "(cases (simplified))" method in proper proof texts);
  3707 
  3708 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
  3709 
  3710 * Isar: names of theorems etc. may be natural numbers as well;
  3711 
  3712 * Isar: 'pr' command: optional arguments for goals_limit and
  3713 ProofContext.prems_limit; no longer prints theory contexts, but only
  3714 proof states;
  3715 
  3716 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
  3717 additional print modes to be specified; e.g. "pr(latex)" will print
  3718 proof state according to the Isabelle LaTeX style;
  3719 
  3720 * Isar: improved support for emulating tactic scripts, including proof
  3721 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
  3722 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
  3723 (for HOL datatypes);
  3724 
  3725 * Isar: simplified (more robust) goal selection of proof methods: 1st
  3726 goal, all goals, or explicit goal specifier (tactic emulation); thus
  3727 'proof method scripts' have to be in depth-first order;
  3728 
  3729 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
  3730 
  3731 * Isar: removed 'help' command, which hasn't been too helpful anyway;
  3732 should instead use individual commands for printing items
  3733 (print_commands, print_methods etc.);
  3734 
  3735 * Isar: added 'nothing' --- the empty list of theorems;
  3736 
  3737 
  3738 *** HOL ***
  3739 
  3740 * HOL/MicroJava: formalization of a fragment of Java, together with a
  3741 corresponding virtual machine and a specification of its bytecode
  3742 verifier and a lightweight bytecode verifier, including proofs of
  3743 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
  3744 Cornelia Pusch (see also the homepage of project Bali at
  3745 http://isabelle.in.tum.de/Bali/);
  3746 
  3747 * HOL/Algebra: new theory of rings and univariate polynomials, by
  3748 Clemens Ballarin;
  3749 
  3750 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
  3751 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
  3752 Rasmussen;
  3753 
  3754 * HOL/Lattice: fundamental concepts of lattice theory and order
  3755 structures, including duals, properties of bounds versus algebraic
  3756 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
  3757 Theorem for complete lattices etc.; may also serve as a demonstration
  3758 for abstract algebraic reasoning using axiomatic type classes, and
  3759 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
  3760 
  3761 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
  3762 von Oheimb;
  3763 
  3764 * HOL/IMPP: extension of IMP with local variables and mutually
  3765 recursive procedures, by David von Oheimb;
  3766 
  3767 * HOL/Lambda: converted into new-style theory and document;
  3768 
  3769 * HOL/ex/Multiquote: example of multiple nested quotations and
  3770 anti-quotations -- basically a generalized version of de-Bruijn
  3771 representation; very useful in avoiding lifting of operations;
  3772 
  3773 * HOL/record: added general record equality rule to simpset; fixed
  3774 select-update simplification procedure to handle extended records as
  3775 well; admit "r" as field name;
  3776 
  3777 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
  3778 other numeric types and also as the identity of groups, rings, etc.;
  3779 
  3780 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
  3781 Types nat and int belong to this axclass;
  3782 
  3783 * HOL: greatly improved simplification involving numerals of type nat, int, real:
  3784    (i + #8 + j) = Suc k simplifies to  #7 + (i + j) = k
  3785    i*j + k + j*#3*i     simplifies to  #4*(i*j) + k
  3786   two terms #m*u and #n*u are replaced by #(m+n)*u
  3787     (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
  3788   and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
  3789     or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
  3790 
  3791 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
  3792 powerful prover for predicate logic but knows nothing of clasets; see
  3793 ex/mesontest.ML and ex/mesontest2.ML for example applications;
  3794 
  3795 * HOL: new version of "case_tac" subsumes both boolean case split and
  3796 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
  3797 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
  3798 
  3799 * HOL: simplification no longer dives into case-expressions: only the
  3800 selector expression is simplified, but not the remaining arms; to
  3801 enable full simplification of case-expressions for datatype t, you may
  3802 remove t.weak_case_cong from the simpset, either globally (Delcongs
  3803 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
  3804 
  3805 * HOL/recdef: the recursion equations generated by 'recdef' for
  3806 function 'f' are now called f.simps instead of f.rules; if all
  3807 termination conditions are proved automatically, these simplification
  3808 rules are added to the simpset, as in primrec; rules may be named
  3809 individually as well, resulting in a separate list of theorems for
  3810 each equation;
  3811 
  3812 * HOL/While is a new theory that provides a while-combinator. It
  3813 permits the definition of tail-recursive functions without the
  3814 provision of a termination measure. The latter is necessary once the
  3815 invariant proof rule for while is applied.
  3816 
  3817 * HOL: new (overloaded) notation for the set of elements below/above
  3818 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
  3819 
  3820 * HOL: theorems impI, allI, ballI bound as "strip";
  3821 
  3822 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
  3823 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
  3824 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
  3825 
  3826 * HOL/Real: "rabs" replaced by overloaded "abs" function;
  3827 
  3828 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
  3829 main HOL, but was unused);
  3830 
  3831 * HOL: fewer consts declared as global (e.g. have to refer to
  3832 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
  3833 
  3834 * HOL: tuned AST representation of nested pairs, avoiding bogus output
  3835 in case of overlap with user translations (e.g. judgements over
  3836 tuples); (note that the underlying logical represenation is still
  3837 bogus);
  3838 
  3839 
  3840 *** ZF ***
  3841 
  3842 * ZF: simplification automatically cancels common terms in arithmetic
  3843 expressions over nat and int;
  3844 
  3845 * ZF: new treatment of nat to minimize type-checking: all operators
  3846 coerce their operands to a natural number using the function natify,
  3847 making the algebraic laws unconditional;
  3848 
  3849 * ZF: as above, for int: operators coerce their operands to an integer
  3850 using the function intify;
  3851 
  3852 * ZF: the integer library now contains many of the usual laws for the
  3853 orderings, including $<=, and monotonicity laws for $+ and $*;
  3854 
  3855 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
  3856 simplification;
  3857 
  3858 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
  3859 to the simplifier and classical reasoner simultaneously;
  3860 
  3861 
  3862 *** General ***
  3863 
  3864 * Provers: blast_tac now handles actual object-logic rules as
  3865 assumptions; note that auto_tac uses blast_tac internally as well;
  3866 
  3867 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
  3868 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
  3869 
  3870 * Provers: delrules now handles destruct rules as well (no longer need
  3871 explicit make_elim);
  3872 
  3873 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
  3874   [| inj ?f;          ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3875 use instead the strong form,
  3876   [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
  3877 in HOL, FOL and ZF the function cla_make_elim will create such rules
  3878 from destruct-rules;
  3879 
  3880 * Provers: Simplifier.easy_setup provides a fast path to basic
  3881 Simplifier setup for new object-logics;
  3882 
  3883 * Pure: AST translation rules no longer require constant head on LHS;
  3884 
  3885 * Pure: improved name spaces: ambiguous output is qualified; support
  3886 for hiding of names;
  3887 
  3888 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
  3889 XSYMBOL_HOME; no longer need to do manual configuration in most
  3890 situations;
  3891 
  3892 * system: compression of ML heaps images may now be controlled via -c
  3893 option of isabelle and isatool usedir (currently only observed by
  3894 Poly/ML);
  3895 
  3896 * system: isatool installfonts may handle X-Symbol fonts as well (very
  3897 useful for remote X11);
  3898 
  3899 * system: provide TAGS file for Isabelle sources;
  3900 
  3901 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
  3902 order;
  3903 
  3904 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
  3905 timing flag supersedes proof_timing and Toplevel.trace;
  3906 
  3907 * ML: new combinators |>> and |>>> for incremental transformations
  3908 with secondary results (e.g. certain theory extensions):
  3909 
  3910 * ML: PureThy.add_defs gets additional argument to indicate potential
  3911 overloading (usually false);
  3912 
  3913 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
  3914 results;
  3915 
  3916 
  3917 
  3918 New in Isabelle99 (October 1999)
  3919 --------------------------------
  3920 
  3921 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  3922 
  3923 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
  3924 are no longer simplified.  (This allows the simplifier to unfold recursive
  3925 functional programs.)  To restore the old behaviour, declare
  3926 
  3927     Delcongs [if_weak_cong];
  3928 
  3929 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
  3930 complement;
  3931 
  3932 * HOL: the predicate "inj" is now defined by translation to "inj_on";
  3933 
  3934 * HOL/datatype: mutual_induct_tac no longer exists --
  3935   use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
  3936 
  3937 * HOL/typedef: fixed type inference for representing set; type
  3938 arguments now have to occur explicitly on the rhs as type constraints;
  3939 
  3940 * ZF: The con_defs part of an inductive definition may no longer refer
  3941 to constants declared in the same theory;
  3942 
  3943 * HOL, ZF: the function mk_cases, generated by the inductive
  3944 definition package, has lost an argument.  To simplify its result, it
  3945 uses the default simpset instead of a supplied list of theorems.
  3946 
  3947 * HOL/List: the constructors of type list are now Nil and Cons;
  3948 
  3949 * Simplifier: the type of the infix ML functions
  3950         setSSolver addSSolver setSolver addSolver
  3951 is now  simpset * solver -> simpset  where `solver' is a new abstract type
  3952 for packaging solvers. A solver is created via
  3953         mk_solver: string -> (thm list -> int -> tactic) -> solver
  3954 where the string argument is only a comment.
  3955 
  3956 
  3957 *** Proof tools ***
  3958 
  3959 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
  3960 decision procedure for linear arithmetic. Currently it is used for
  3961 types `nat', `int', and `real' in HOL (see below); it can, should and
  3962 will be instantiated for other types and logics as well.
  3963 
  3964 * The simplifier now accepts rewrite rules with flexible heads, eg
  3965      hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
  3966   They are applied like any rule with a non-pattern lhs, i.e. by first-order
  3967   matching.
  3968 
  3969 
  3970 *** General ***
  3971 
  3972 * New Isabelle/Isar subsystem provides an alternative to traditional
  3973 tactical theorem proving; together with the ProofGeneral/isar user
  3974 interface it offers an interactive environment for developing human
  3975 readable proof documents (Isar == Intelligible semi-automated
  3976 reasoning); for further information see isatool doc isar-ref,
  3977 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
  3978 
  3979 * improved and simplified presentation of theories: better HTML markup
  3980 (including colors), graph views in several sizes; isatool usedir now
  3981 provides a proper interface for user theories (via -P option); actual
  3982 document preparation based on (PDF)LaTeX is available as well (for
  3983 new-style theories only); see isatool doc system for more information;
  3984 
  3985 * native support for Proof General, both for classic Isabelle and
  3986 Isabelle/Isar;
  3987 
  3988 * ML function thm_deps visualizes dependencies of theorems and lemmas,
  3989 using the graph browser tool;
  3990 
  3991 * Isabelle manuals now also available as PDF;
  3992 
  3993 * theory loader rewritten from scratch (may not be fully
  3994 bug-compatible); old loadpath variable has been replaced by show_path,
  3995 add_path, del_path, reset_path functions; new operations such as
  3996 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
  3997 isatool doc ref);
  3998 
  3999 * improved isatool install: option -k creates KDE application icon,
  4000 option -p DIR installs standalone binaries;
  4001 
  4002 * added ML_PLATFORM setting (useful for cross-platform installations);
  4003 more robust handling of platform specific ML images for SML/NJ;
  4004 
  4005 * the settings environment is now statically scoped, i.e. it is never
  4006 created again in sub-processes invoked from isabelle, isatool, or
  4007 Isabelle;
  4008 
  4009 * path element specification '~~' refers to '$ISABELLE_HOME';
  4010 
  4011 * in locales, the "assumes" and "defines" parts may be omitted if
  4012 empty;
  4013 
  4014 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
  4015 long arrows);
  4016 
  4017 * new print_mode "HTML";
  4018 
  4019 * new flag show_tags controls display of tags of theorems (which are
  4020 basically just comments that may be attached by some tools);
  4021 
  4022 * Isamode 2.6 requires patch to accomodate change of Isabelle font
  4023 mode and goal output format:
  4024 
  4025 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
  4026 244c244
  4027 <       (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
  4028 ---
  4029 >       (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
  4030 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
  4031 181c181
  4032 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
  4033 ---
  4034 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
  4035 
  4036 * function bind_thms stores lists of theorems (cf. bind_thm);
  4037 
  4038 * new shorthand tactics ftac, eatac, datac, fatac;
  4039 
  4040 * qed (and friends) now accept "" as result name; in that case the
  4041 theorem is not stored, but proper checks and presentation of the
  4042 result still apply;
  4043 
  4044 * theorem database now also indexes constants "Trueprop", "all",
  4045 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
  4046 
  4047 
  4048 *** HOL ***
  4049 
  4050 ** HOL arithmetic **
  4051 
  4052 * There are now decision procedures for linear arithmetic over nat and
  4053 int:
  4054 
  4055 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
  4056 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
  4057 are treated as atomic; subformulae not involving type `nat' or `int'
  4058 are ignored; quantified subformulae are ignored unless they are
  4059 positive universal or negative existential. The tactic has to be
  4060 invoked by hand and can be a little bit slow. In particular, the
  4061 running time is exponential in the number of occurrences of `min' and
  4062 `max', and `-' on `nat'.
  4063 
  4064 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
  4065 (negated) (in)equalities among the premises and the conclusion into
  4066 account (i.e. no compound formulae) and does not know about `min' and
  4067 `max', and `-' on `nat'. It is fast and is used automatically by the
  4068 simplifier.
  4069 
  4070 NB: At the moment, these decision procedures do not cope with mixed
  4071 nat/int formulae where the two parts interact, such as `m < n ==>
  4072 int(m) < int(n)'.
  4073 
  4074 * HOL/Numeral provides a generic theory of numerals (encoded
  4075 efficiently as bit strings); setup for types nat/int/real is in place;
  4076 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
  4077 int, existing theories and proof scripts may require a few additional
  4078 type constraints;
  4079 
  4080 * integer division and remainder can now be performed on constant
  4081 arguments;
  4082 
  4083 * many properties of integer multiplication, division and remainder
  4084 are now available;
  4085 
  4086 * An interface to the Stanford Validity Checker (SVC) is available through the
  4087 tactic svc_tac.  Propositional tautologies and theorems of linear arithmetic
  4088 are proved automatically.  SVC must be installed separately, and its results
  4089 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
  4090 invocation of the underlying oracle).  For SVC see
  4091   http://verify.stanford.edu/SVC
  4092 
  4093 * IsaMakefile: the HOL-Real target now builds an actual image;
  4094 
  4095 
  4096 ** HOL misc **
  4097 
  4098 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
  4099 (in Isabelle/Isar) -- by Gertrud Bauer;
  4100 
  4101 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
  4102 analysis for assembly languages with subtypes;
  4103 
  4104 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
  4105 -- avoids syntactic ambiguities and treats state, transition, and
  4106 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
  4107 changed syntax and (many) tactics;
  4108 
  4109 * HOL/inductive: Now also handles more general introduction rules such
  4110   as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
  4111   theorems are now maintained within the theory (maintained via the
  4112   "mono" attribute);
  4113 
  4114 * HOL/datatype: Now also handles arbitrarily branching datatypes
  4115   (using function types) such as
  4116 
  4117   datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
  4118 
  4119 * HOL/record: record_simproc (part of the default simpset) takes care
  4120 of selectors applied to updated records; record_split_tac is no longer
  4121 part of the default claset; update_defs may now be removed from the
  4122 simpset in many cases; COMPATIBILITY: old behavior achieved by
  4123 
  4124   claset_ref () := claset() addSWrapper record_split_wrapper;
  4125   Delsimprocs [record_simproc]
  4126 
  4127 * HOL/typedef: fixed type inference for representing set; type
  4128 arguments now have to occur explicitly on the rhs as type constraints;
  4129 
  4130 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
  4131 names rather than an ML expression;
  4132 
  4133 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
  4134 supplied later.  Program schemes can be defined, such as
  4135     "While B C s = (if B s then While B C (C s) else s)"
  4136 where the well-founded relation can be chosen after B and C have been given.
  4137 
  4138 * HOL/List: the constructors of type list are now Nil and Cons;
  4139 INCOMPATIBILITY: while [] and infix # syntax is still there, of
  4140 course, ML tools referring to List.list.op # etc. have to be adapted;
  4141 
  4142 * HOL_quantifiers flag superseded by "HOL" print mode, which is
  4143 disabled by default; run isabelle with option -m HOL to get back to
  4144 the original Gordon/HOL-style output;
  4145 
  4146 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
  4147 ALL x<=y. P, EX x<y. P, EX x<=y. P;
  4148 
  4149 * HOL basic syntax simplified (more orthogonal): all variants of
  4150 All/Ex now support plain / symbolic / HOL notation; plain syntax for
  4151 Eps operator is provided as well: "SOME x. P[x]";
  4152 
  4153 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
  4154 
  4155 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
  4156 thus available for user theories;
  4157 
  4158 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
  4159 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
  4160 time;
  4161 
  4162 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
  4163 several times and then mp;
  4164 
  4165 
  4166 *** LK ***
  4167 
  4168 * the notation <<...>> is now available as a notation for sequences of
  4169 formulas;
  4170 
  4171 * the simplifier is now installed
  4172 
  4173 * the axiom system has been generalized (thanks to Soren Heilmann)
  4174 
  4175 * the classical reasoner now has a default rule database
  4176 
  4177 
  4178 *** ZF ***
  4179 
  4180 * new primrec section allows primitive recursive functions to be given
  4181 directly (as in HOL) over datatypes and the natural numbers;
  4182 
  4183 * new tactics induct_tac and exhaust_tac for induction (or case
  4184 analysis) over datatypes and the natural numbers;
  4185 
  4186 * the datatype declaration of type T now defines the recursor T_rec;
  4187 
  4188 * simplification automatically does freeness reasoning for datatype
  4189 constructors;
  4190 
  4191 * automatic type-inference, with AddTCs command to insert new
  4192 type-checking rules;
  4193 
  4194 * datatype introduction rules are now added as Safe Introduction rules
  4195 to the claset;
  4196 
  4197 * the syntax "if P then x else y" is now available in addition to
  4198 if(P,x,y);
  4199 
  4200 
  4201 *** Internal programming interfaces ***
  4202 
  4203 * tuned simplifier trace output; new flag debug_simp;
  4204 
  4205 * structures Vartab / Termtab (instances of TableFun) offer efficient
  4206 tables indexed by indexname_ord / term_ord (compatible with aconv);
  4207 
  4208 * AxClass.axclass_tac lost the theory argument;
  4209 
  4210 * tuned current_goals_markers semantics: begin / end goal avoids
  4211 printing empty lines;
  4212 
  4213 * removed prs and prs_fn hook, which was broken because it did not
  4214 include \n in its semantics, forcing writeln to add one
  4215 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
  4216 string -> unit if you really want to output text without newline;
  4217 
  4218 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
  4219 plain output, interface builders may have to enable 'isabelle_font'
  4220 mode to get Isabelle font glyphs as before;
  4221 
  4222 * refined token_translation interface; INCOMPATIBILITY: output length
  4223 now of type real instead of int;
  4224 
  4225 * theory loader actions may be traced via new ThyInfo.add_hook
  4226 interface (see src/Pure/Thy/thy_info.ML); example application: keep
  4227 your own database of information attached to *whole* theories -- as
  4228 opposed to intra-theory data slots offered via TheoryDataFun;
  4229 
  4230 * proper handling of dangling sort hypotheses (at last!);
  4231 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
  4232 extra sort hypotheses that can be witnessed from the type signature;
  4233 the force_strip_shyps flag is gone, any remaining shyps are simply
  4234 left in the theorem (with a warning issued by strip_shyps_warning);
  4235 
  4236 
  4237 
  4238 New in Isabelle98-1 (October 1998)
  4239 ----------------------------------
  4240 
  4241 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  4242 
  4243 * several changes of automated proof tools;
  4244 
  4245 * HOL: major changes to the inductive and datatype packages, including
  4246 some minor incompatibilities of theory syntax;
  4247 
  4248 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
  4249 called `inj_on';
  4250 
  4251 * HOL: removed duplicate thms in Arith:
  4252   less_imp_add_less  should be replaced by  trans_less_add1
  4253   le_imp_add_le      should be replaced by  trans_le_add1
  4254 
  4255 * HOL: unary minus is now overloaded (new type constraints may be
  4256 required);
  4257 
  4258 * HOL and ZF: unary minus for integers is now #- instead of #~.  In
  4259 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
  4260 now taken as an integer constant.
  4261 
  4262 * Pure: ML function 'theory_of' renamed to 'theory';
  4263 
  4264 
  4265 *** Proof tools ***
  4266 
  4267 * Simplifier:
  4268   1. Asm_full_simp_tac is now more aggressive.
  4269      1. It will sometimes reorient premises if that increases their power to
  4270         simplify.
  4271      2. It does no longer proceed strictly from left to right but may also
  4272         rotate premises to achieve further simplification.
  4273      For compatibility reasons there is now Asm_lr_simp_tac which is like the
  4274      old Asm_full_simp_tac in that it does not rotate premises.
  4275   2. The simplifier now knows a little bit about nat-arithmetic.
  4276 
  4277 * Classical reasoner: wrapper mechanism for the classical reasoner now
  4278 allows for selected deletion of wrappers, by introduction of names for
  4279 wrapper functionals.  This implies that addbefore, addSbefore,
  4280 addaltern, and addSaltern now take a pair (name, tactic) as argument,
  4281 and that adding two tactics with the same name overwrites the first
  4282 one (emitting a warning).
  4283   type wrapper = (int -> tactic) -> (int -> tactic)
  4284   setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
  4285   addWrapper, addSWrapper: claset * (string * wrapper) -> claset
  4286   delWrapper, delSWrapper: claset *  string            -> claset
  4287   getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
  4288 
  4289 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
  4290 semantics; addbefore now affects only the unsafe part of step_tac
  4291 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
  4292 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
  4293 by Force_tac;
  4294 
  4295 * Classical reasoner: setwrapper to setWrapper and compwrapper to
  4296 compWrapper; added safe wrapper (and access functions for it);
  4297 
  4298 * HOL/split_all_tac is now much faster and fails if there is nothing
  4299 to split.  Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
  4300 and the names of the automatically generated variables have changed.
  4301 split_all_tac has moved within claset() from unsafe wrappers to safe
  4302 wrappers, which means that !!-bound variables are split much more
  4303 aggressively, and safe_tac and clarify_tac now split such variables.
  4304 If this splitting is not appropriate, use delSWrapper "split_all_tac".
  4305 Note: the same holds for record_split_tac, which does the job of
  4306 split_all_tac for record fields.
  4307 
  4308 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
  4309 permanently to the default simpset using Addsplits just like
  4310 Addsimps. They can be removed via Delsplits just like
  4311 Delsimps. Lower-case versions are also available.
  4312 
  4313 * HOL/Simplifier: The rule split_if is now part of the default
  4314 simpset. This means that the simplifier will eliminate all occurrences
  4315 of if-then-else in the conclusion of a goal. To prevent this, you can
  4316 either remove split_if completely from the default simpset by
  4317 `Delsplits [split_if]' or remove it in a specific call of the
  4318 simplifier using `... delsplits [split_if]'.  You can also add/delete
  4319 other case splitting rules to/from the default simpset: every datatype
  4320 generates suitable rules `split_t_case' and `split_t_case_asm' (where
  4321 t is the name of the datatype).
  4322 
  4323 * Classical reasoner / Simplifier combination: new force_tac (and
  4324 derivatives Force_tac, force) combines rewriting and classical
  4325 reasoning (and whatever other tools) similarly to auto_tac, but is
  4326 aimed to solve the given subgoal completely.
  4327 
  4328 
  4329 *** General ***
  4330 
  4331 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
  4332 and `goalw': the theory is no longer needed as an explicit argument -
  4333 the current theory context is used; assumptions are no longer returned
  4334 at the ML-level unless one of them starts with ==> or !!; it is
  4335 recommended to convert to these new commands using isatool fixgoal
  4336 (backup your sources first!);
  4337 
  4338 * new top-level commands 'thm' and 'thms' for retrieving theorems from
  4339 the current theory context, and 'theory' to lookup stored theories;
  4340 
  4341 * new theory section 'locale' for declaring constants, assumptions and
  4342 definitions that have local scope;
  4343 
  4344 * new theory section 'nonterminals' for purely syntactic types;
  4345 
  4346 * new theory section 'setup' for generic ML setup functions
  4347 (e.g. package initialization);
  4348 
  4349 * the distribution now includes Isabelle icons: see
  4350 lib/logo/isabelle-{small,tiny}.xpm;
  4351 
  4352 * isatool install - install binaries with absolute references to
  4353 ISABELLE_HOME/bin;
  4354 
  4355 * isatool logo -- create instances of the Isabelle logo (as EPS);
  4356 
  4357 * print mode 'emacs' reserved for Isamode;
  4358 
  4359 * support multiple print (ast) translations per constant name;
  4360 
  4361 * theorems involving oracles are now printed with a suffixed [!];
  4362 
  4363 
  4364 *** HOL ***
  4365 
  4366 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
  4367 
  4368 * HOL/inductive package reorganized and improved: now supports mutual
  4369 definitions such as
  4370 
  4371   inductive EVEN ODD
  4372     intrs
  4373       null "0 : EVEN"
  4374       oddI "n : EVEN ==> Suc n : ODD"
  4375       evenI "n : ODD ==> Suc n : EVEN"
  4376 
  4377 new theorem list "elims" contains an elimination rule for each of the
  4378 recursive sets; inductive definitions now handle disjunctive premises
  4379 correctly (also ZF);
  4380 
  4381 INCOMPATIBILITIES: requires Inductive as an ancestor; component
  4382 "mutual_induct" no longer exists - the induction rule is always
  4383 contained in "induct";
  4384 
  4385 
  4386 * HOL/datatype package re-implemented and greatly improved: now
  4387 supports mutually recursive datatypes such as
  4388 
  4389   datatype
  4390     'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
  4391             | SUM ('a aexp) ('a aexp)
  4392             | DIFF ('a aexp) ('a aexp)
  4393             | NUM 'a
  4394   and
  4395     'a bexp = LESS ('a aexp) ('a aexp)
  4396             | AND ('a bexp) ('a bexp)
  4397             | OR ('a bexp) ('a bexp)
  4398 
  4399 as well as indirectly recursive datatypes such as
  4400 
  4401   datatype
  4402     ('a, 'b) term = Var 'a
  4403                   | App 'b ((('a, 'b) term) list)
  4404 
  4405 The new tactic  mutual_induct_tac [<var_1>, ..., <var_n>] i  performs
  4406 induction on mutually / indirectly recursive datatypes.
  4407 
  4408 Primrec equations are now stored in theory and can be accessed via
  4409 <function_name>.simps.
  4410 
  4411 INCOMPATIBILITIES:
  4412 
  4413   - Theories using datatypes must now have theory Datatype as an
  4414     ancestor.
  4415   - The specific <typename>.induct_tac no longer exists - use the
  4416     generic induct_tac instead.
  4417   - natE has been renamed to nat.exhaust - use exhaust_tac
  4418     instead of res_inst_tac ... natE. Note that the variable
  4419     names in nat.exhaust differ from the names in natE, this
  4420     may cause some "fragile" proofs to fail.
  4421   - The theorems split_<typename>_case and split_<typename>_case_asm
  4422     have been renamed to <typename>.split and <typename>.split_asm.
  4423   - Since default sorts of type variables are now handled correctly,
  4424     some datatype definitions may have to be annotated with explicit
  4425     sort constraints.
  4426   - Primrec definitions no longer require function name and type
  4427     of recursive argument.
  4428 
  4429 Consider using isatool fixdatatype to adapt your theories and proof
  4430 scripts to the new package (backup your sources first!).
  4431 
  4432 
  4433 * HOL/record package: considerably improved implementation; now
  4434 includes concrete syntax for record types, terms, updates; theorems
  4435 for surjective pairing and splitting !!-bound record variables; proof
  4436 support is as follows:
  4437 
  4438   1) standard conversions (selectors or updates applied to record
  4439 constructor terms) are part of the standard simpset;
  4440 
  4441   2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
  4442 made part of standard simpset and claset via addIffs;
  4443 
  4444   3) a tactic for record field splitting (record_split_tac) is part of
  4445 the standard claset (addSWrapper);
  4446 
  4447 To get a better idea about these rules you may retrieve them via
  4448 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
  4449 the name of your record type.
  4450 
  4451 The split tactic 3) conceptually simplifies by the following rule:
  4452 
  4453   "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
  4454 
  4455 Thus any record variable that is bound by meta-all will automatically
  4456 blow up into some record constructor term, consequently the
  4457 simplifications of 1), 2) apply.  Thus force_tac, auto_tac etc. shall
  4458 solve record problems automatically.
  4459 
  4460 
  4461 * reorganized the main HOL image: HOL/Integ and String loaded by
  4462 default; theory Main includes everything;
  4463 
  4464 * automatic simplification of integer sums and comparisons, using cancellation;
  4465 
  4466 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
  4467 
  4468 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
  4469 
  4470 * many new identities for unions, intersections, set difference, etc.;
  4471 
  4472 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
  4473 called split_if, split_split, split_sum_case and split_nat_case (to go
  4474 with add/delsplits);
  4475 
  4476 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
  4477 (?x::unit) = (); this is made part of the default simpset, which COULD
  4478 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
  4479 'Delsimprocs [unit_eq_proc];' as last resort); also note that
  4480 unit_abs_eta_conv is added in order to counter the effect of
  4481 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
  4482 %u.f();
  4483 
  4484 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
  4485 makes more sense);
  4486 
  4487 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  4488   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  4489   disjointness reasoning but breaking a few old proofs.
  4490 
  4491 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
  4492 to 'converse' from 'inverse' (for compatibility with ZF and some
  4493 literature);
  4494 
  4495 * HOL/recdef can now declare non-recursive functions, with {} supplied as
  4496 the well-founded relation;
  4497 
  4498 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
  4499     Compl A.  The "Compl" syntax remains available as input syntax for this
  4500     release ONLY.
  4501 
  4502 * HOL/Update: new theory of function updates:
  4503     f(a:=b) == %x. if x=a then b else f x
  4504 may also be iterated as in f(a:=b,c:=d,...);
  4505 
  4506 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
  4507 
  4508 * HOL/List:
  4509   - new function list_update written xs[i:=v] that updates the i-th
  4510     list position. May also be iterated as in xs[i:=a,j:=b,...].
  4511   - new function `upt' written [i..j(] which generates the list
  4512     [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
  4513     bound write [i..j], which is a shorthand for [i..j+1(].
  4514   - new lexicographic orderings and corresponding wellfoundedness theorems.
  4515 
  4516 * HOL/Arith:
  4517   - removed 'pred' (predecessor) function;
  4518   - generalized some theorems about n-1;
  4519   - many new laws about "div" and "mod";
  4520   - new laws about greatest common divisors (see theory ex/Primes);
  4521 
  4522 * HOL/Relation: renamed the relational operator r^-1 "converse"
  4523 instead of "inverse";
  4524 
  4525 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
  4526   of the multiset ordering;
  4527 
  4528 * directory HOL/Real: a construction of the reals using Dedekind cuts
  4529   (not included by default);
  4530 
  4531 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
  4532 
  4533 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
  4534   programs, i.e. different program variables may have different types.
  4535 
  4536 * calling (stac rew i) now fails if "rew" has no effect on the goal
  4537   [previously, this check worked only if the rewrite rule was unconditional]
  4538   Now rew can involve either definitions or equalities (either == or =).
  4539 
  4540 
  4541 *** ZF ***
  4542 
  4543 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
  4544   only the theorems proved on ZF.ML;
  4545 
  4546 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
  4547   It and 'sym RS equals0D' are now in the default  claset, giving automatic
  4548   disjointness reasoning but breaking a few old proofs.
  4549 
  4550 * ZF/Update: new theory of function updates
  4551     with default rewrite rule  f(x:=y) ` z = if(z=x, y, f`z)
  4552   may also be iterated as in f(a:=b,c:=d,...);
  4553 
  4554 * in  let x=t in u(x), neither t nor u(x) has to be an FOL term.
  4555 
  4556 * calling (stac rew i) now fails if "rew" has no effect on the goal
  4557   [previously, this check worked only if the rewrite rule was unconditional]
  4558   Now rew can involve either definitions or equalities (either == or =).
  4559 
  4560 * case_tac provided for compatibility with HOL
  4561     (like the old excluded_middle_tac, but with subgoals swapped)
  4562 
  4563 
  4564 *** Internal programming interfaces ***
  4565 
  4566 * Pure: several new basic modules made available for general use, see
  4567 also src/Pure/README;
  4568 
  4569 * improved the theory data mechanism to support encapsulation (data
  4570 kind name replaced by private Object.kind, acting as authorization
  4571 key); new type-safe user interface via functor TheoryDataFun; generic
  4572 print_data function becomes basically useless;
  4573 
  4574 * removed global_names compatibility flag -- all theory declarations
  4575 are qualified by default;
  4576 
  4577 * module Pure/Syntax now offers quote / antiquote translation
  4578 functions (useful for Hoare logic etc. with implicit dependencies);
  4579 see HOL/ex/Antiquote for an example use;
  4580 
  4581 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
  4582 cterm -> thm;
  4583 
  4584 * new tactical CHANGED_GOAL for checking that a tactic modifies a
  4585 subgoal;
  4586 
  4587 * Display.print_goals function moved to Locale.print_goals;
  4588 
  4589 * standard print function for goals supports current_goals_markers
  4590 variable for marking begin of proof, end of proof, start of goal; the
  4591 default is ("", "", ""); setting current_goals_markers := ("<proof>",
  4592 "</proof>", "<goal>") causes SGML like tagged proof state printing,
  4593 for example;
  4594 
  4595 
  4596 
  4597 New in Isabelle98 (January 1998)
  4598 --------------------------------
  4599 
  4600 *** Overview of INCOMPATIBILITIES (see below for more details) ***
  4601 
  4602 * changed lexical syntax of terms / types: dots made part of long
  4603 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
  4604 
  4605 * simpset (and claset) reference variable replaced by functions
  4606 simpset / simpset_ref;
  4607 
  4608 * no longer supports theory aliases (via merge) and non-trivial
  4609 implicit merge of thms' signatures;
  4610 
  4611 * most internal names of constants changed due to qualified names;
  4612 
  4613 * changed Pure/Sequence interface (see Pure/seq.ML);
  4614 
  4615 
  4616 *** General Changes ***
  4617 
  4618 * hierachically structured name spaces (for consts, types, axms, thms
  4619 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
  4620 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
  4621 isatool fixdots ensures space after dots (e.g. "%x. x"); set
  4622 long_names for fully qualified output names; NOTE: ML programs
  4623 (special tactics, packages etc.) referring to internal names may have
  4624 to be adapted to cope with fully qualified names; in case of severe
  4625 backward campatibility problems try setting 'global_names' at compile
  4626 time to have enrything declared within a flat name space; one may also
  4627 fine tune name declarations in theories via the 'global' and 'local'
  4628 section;
  4629 
  4630 * reimplemented the implicit simpset and claset using the new anytype
  4631 data filed in signatures; references simpset:simpset ref etc. are
  4632 replaced by functions simpset:unit->simpset and
  4633 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
  4634 to patch your ML files accordingly;
  4635 
  4636 * HTML output now includes theory graph data for display with Java
  4637 applet or isatool browser; data generated automatically via isatool
  4638 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
  4639 
  4640 * defs may now be conditional; improved rewrite_goals_tac to handle
  4641 conditional equations;
  4642 
  4643 * defs now admits additional type arguments, using TYPE('a) syntax;
  4644 
  4645 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
  4646 creates a new theory node; implicit merge of thms' signatures is
  4647 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
  4648 transfer:theory->thm->thm in (rare) cases;
  4649 
  4650 * improved handling of draft signatures / theories; draft thms (and
  4651 ctyps, cterms) are automatically promoted to real ones;
  4652 
  4653 * slightly changed interfaces for oracles: admit many per theory, named
  4654 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
  4655 
  4656 * print_goals: optional output of const types (set show_consts and
  4657 show_types);
  4658 
  4659 * improved output of warnings (###) and errors (***);
  4660 
  4661 * subgoal_tac displays a warning if the new subgoal has type variables;
  4662 
  4663 * removed old README and Makefiles;
  4664 
  4665 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
  4666 
  4667 * removed obsolete init_pps and init_database;
  4668 
  4669 * deleted the obsolete tactical STATE, which was declared by
  4670     fun STATE tacfun st = tacfun st st;
  4671 
  4672 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
  4673 (which abbreviates $HOME);
  4674 
  4675 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
  4676 use isatool fixseq to adapt your ML programs (this works for fully
  4677 qualified references to the Sequence structure only!);
  4678 
  4679 * use_thy no longer requires writable current directory; it always
  4680 reloads .ML *and* .thy file, if either one is out of date;
  4681 
  4682 
  4683 *** Classical Reasoner ***
  4684 
  4685 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
  4686 tactics that use classical reasoning to simplify a subgoal without
  4687 splitting it into several subgoals;
  4688 
  4689 * Safe_tac: like safe_tac but uses the default claset;
  4690 
  4691 
  4692 *** Simplifier ***
  4693 
  4694 * added simplification meta rules:
  4695     (asm_)(full_)simplify: simpset -> thm -> thm;
  4696 
  4697 * simplifier.ML no longer part of Pure -- has to be loaded by object
  4698 logics (again);
  4699 
  4700 * added prems argument to simplification procedures;
  4701 
  4702 * HOL, FOL, ZF: added infix function `addsplits':
  4703   instead of `<simpset> setloop (split_tac <thms>)'
  4704   you can simply write `<simpset> addsplits <thms>'
  4705 
  4706 
  4707 *** Syntax ***
  4708 
  4709 * TYPE('a) syntax for type reflection terms;
  4710 
  4711 * no longer handles consts with name "" -- declare as 'syntax' instead;
  4712 
  4713 * pretty printer: changed order of mixfix annotation preference (again!);
  4714 
  4715 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
  4716 
  4717 
  4718 *** HOL ***
  4719 
  4720 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4721   with `addloop' of the simplifier to faciliate case splitting in premises.
  4722 
  4723 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
  4724 
  4725 * HOL/Auth: new protocol proofs including some for the Internet
  4726   protocol TLS;
  4727 
  4728 * HOL/Map: new theory of `maps' a la VDM;
  4729 
  4730 * HOL/simplifier: simplification procedures nat_cancel_sums for
  4731 cancelling out common nat summands from =, <, <= (in)equalities, or
  4732 differences; simplification procedures nat_cancel_factor for
  4733 cancelling common factor from =, <, <= (in)equalities over natural
  4734 sums; nat_cancel contains both kinds of procedures, it is installed by
  4735 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
  4736 
  4737 * HOL/simplifier: terms of the form
  4738   `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)'  (or t=x)
  4739   are rewritten to
  4740   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
  4741   and those of the form
  4742   `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)'  (or t=x)
  4743   are rewritten to
  4744   `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
  4745 
  4746 * HOL/datatype
  4747   Each datatype `t' now comes with a theorem `split_t_case' of the form
  4748 
  4749   P(t_case f1 ... fn x) =
  4750      ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
  4751         ...
  4752        (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
  4753      )
  4754 
  4755   and a theorem `split_t_case_asm' of the form
  4756 
  4757   P(t_case f1 ... fn x) =
  4758     ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
  4759         ...
  4760        (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
  4761      )
  4762   which can be added to a simpset via `addsplits'. The existing theorems
  4763   expand_list_case and expand_option_case have been renamed to
  4764   split_list_case and split_option_case.
  4765 
  4766 * HOL/Arithmetic:
  4767   - `pred n' is automatically converted to `n-1'.
  4768     Users are strongly encouraged not to use `pred' any longer,
  4769     because it will disappear altogether at some point.
  4770   - Users are strongly encouraged to write "0 < n" rather than
  4771     "n ~= 0". Theorems and proof tools have been modified towards this
  4772     `standard'.
  4773 
  4774 * HOL/Lists:
  4775   the function "set_of_list" has been renamed "set" (and its theorems too);
  4776   the function "nth" now takes its arguments in the reverse order and
  4777   has acquired the infix notation "!" as in "xs!n".
  4778 
  4779 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
  4780 
  4781 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
  4782   specialist theorems (like UN1_I) are gone.  Similarly for (INT x.B x);
  4783 
  4784 * HOL/record: extensible records with schematic structural subtyping
  4785 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
  4786 still lacks various theorems and concrete record syntax;
  4787 
  4788 
  4789 *** HOLCF ***
  4790 
  4791 * removed "axioms" and "generated by" sections;
  4792 
  4793 * replaced "ops" section by extended "consts" section, which is capable of
  4794   handling the continuous function space "->" directly;
  4795 
  4796 * domain package:
  4797   . proves theorems immediately and stores them in the theory,
  4798   . creates hierachical name space,
  4799   . now uses normal mixfix annotations (instead of cinfix...),
  4800   . minor changes to some names and values (for consistency),
  4801   . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
  4802   . separator between mutual domain defs: changed "," to "and",
  4803   . improved handling of sort constraints;  now they have to
  4804     appear on the left-hand side of the equations only;
  4805 
  4806 * fixed LAM <x,y,zs>.b syntax;
  4807 
  4808 * added extended adm_tac to simplifier in HOLCF -- can now discharge
  4809 adm (%x. P (t x)), where P is chainfinite and t continuous;
  4810 
  4811 
  4812 *** FOL and ZF ***
  4813 
  4814 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
  4815   with `addloop' of the simplifier to faciliate case splitting in premises.
  4816 
  4817 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
  4818 in HOL, they strip ALL and --> from proved theorems;
  4819 
  4820 
  4821 
  4822 New in Isabelle94-8 (May 1997)
  4823 ------------------------------
  4824 
  4825 *** General Changes ***
  4826 
  4827 * new utilities to build / run / maintain Isabelle etc. (in parts
  4828 still somewhat experimental); old Makefiles etc. still functional;
  4829 
  4830 * new 'Isabelle System Manual';
  4831 
  4832 * INSTALL text, together with ./configure and ./build scripts;
  4833 
  4834 * reimplemented type inference for greater efficiency, better error
  4835 messages and clean internal interface;
  4836 
  4837 * prlim command for dealing with lots of subgoals (an easier way of
  4838 setting goals_limit);
  4839 
  4840 
  4841 *** Syntax ***
  4842 
  4843 * supports alternative (named) syntax tables (parser and pretty
  4844 printer); internal interface is provided by add_modesyntax(_i);
  4845 
  4846 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
  4847 be used in conjunction with the Isabelle symbol font; uses the
  4848 "symbols" syntax table;
  4849 
  4850 * added token_translation interface (may translate name tokens in
  4851 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
  4852 the current print_mode); IMPORTANT: user print translation functions
  4853 are responsible for marking newly introduced bounds
  4854 (Syntax.mark_boundT);
  4855 
  4856 * token translations for modes "xterm" and "xterm_color" that display
  4857 names in bold, underline etc. or colors (which requires a color
  4858 version of xterm);
  4859 
  4860 * infixes may now be declared with names independent of their syntax;
  4861 
  4862 * added typed_print_translation (like print_translation, but may
  4863 access type of constant);
  4864 
  4865 
  4866 *** Classical Reasoner ***
  4867 
  4868 Blast_tac: a new tactic!  It is often more powerful than fast_tac, but has
  4869 some limitations.  Blast_tac...
  4870   + ignores addss, addbefore, addafter; this restriction is intrinsic
  4871   + ignores elimination rules that don't have the correct format
  4872         (the conclusion MUST be a formula variable)
  4873   + ignores types, which can make HOL proofs fail
  4874   + rules must not require higher-order unification, e.g. apply_type in ZF
  4875     [message "Function Var's argument not a bound variable" relates to this]
  4876   + its proof strategy is more general but can actually be slower
  4877 
  4878 * substitution with equality assumptions no longer permutes other
  4879 assumptions;
  4880 
  4881 * minor changes in semantics of addafter (now called addaltern); renamed
  4882 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
  4883 (and access functions for it);
  4884 
  4885 * improved combination of classical reasoner and simplifier:
  4886   + functions for handling clasimpsets
  4887   + improvement of addss: now the simplifier is called _after_ the
  4888     safe steps.
  4889   + safe variant of addss called addSss: uses safe simplifications
  4890     _during_ the safe steps. It is more complete as it allows multiple
  4891     instantiations of unknowns (e.g. with slow_tac).
  4892 
  4893 *** Simplifier ***
  4894 
  4895 * added interface for simplification procedures (functions that
  4896 produce *proven* rewrite rules on the fly, depending on current
  4897 redex);
  4898 
  4899 * ordering on terms as parameter (used for ordered rewriting);
  4900 
  4901 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
  4902 
  4903 * the solver is now split into a safe and an unsafe part.
  4904 This should be invisible for the normal user, except that the
  4905 functions setsolver and addsolver have been renamed to setSolver and
  4906 addSolver; added safe_asm_full_simp_tac;
  4907 
  4908 
  4909 *** HOL ***
  4910 
  4911 * a generic induction tactic `induct_tac' which works for all datatypes and
  4912 also for type `nat';
  4913 
  4914 * a generic case distinction tactic `exhaust_tac' which works for all
  4915 datatypes and also for type `nat';
  4916 
  4917 * each datatype comes with a function `size';
  4918 
  4919 * patterns in case expressions allow tuple patterns as arguments to
  4920 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
  4921 
  4922 * primrec now also works with type nat;
  4923 
  4924 * recdef: a new declaration form, allows general recursive functions to be
  4925 defined in theory files.  See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
  4926 
  4927 * the constant for negation has been renamed from "not" to "Not" to
  4928 harmonize with FOL, ZF, LK, etc.;
  4929 
  4930 * HOL/ex/LFilter theory of a corecursive "filter" functional for
  4931 infinite lists;
  4932 
  4933 * HOL/Modelcheck demonstrates invocation of model checker oracle;
  4934 
  4935 * HOL/ex/Ring.thy declares cring_simp, which solves equational
  4936 problems in commutative rings, using axiomatic type classes for + and *;
  4937 
  4938 * more examples in HOL/MiniML and HOL/Auth;
  4939 
  4940 * more default rewrite rules for quantifiers, union/intersection;
  4941 
  4942 * a new constant `arbitrary == @x.False';
  4943 
  4944 * HOLCF/IOA replaces old HOL/IOA;
  4945 
  4946 * HOLCF changes: derived all rules and arities
  4947   + axiomatic type classes instead of classes
  4948   + typedef instead of faking type definitions
  4949   + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
  4950   + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
  4951   + eliminated the types void, one, tr
  4952   + use unit lift and bool lift (with translations) instead of one and tr
  4953   + eliminated blift from Lift3.thy (use Def instead of blift)
  4954   all eliminated rules are derived as theorems --> no visible changes ;
  4955 
  4956 
  4957 *** ZF ***
  4958 
  4959 * ZF now has Fast_tac, Simp_tac and Auto_tac.  Union_iff is a now a default
  4960 rewrite rule; this may affect some proofs.  eq_cs is gone but can be put back
  4961 as ZF_cs addSIs [equalityI];
  4962 
  4963 
  4964 
  4965 New in Isabelle94-7 (November 96)
  4966 ---------------------------------
  4967 
  4968 * allowing negative levels (as offsets) in prlev and choplev;
  4969 
  4970 * super-linear speedup for large simplifications;
  4971 
  4972 * FOL, ZF and HOL now use miniscoping: rewriting pushes
  4973 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
  4974 FAIL); can suppress it using the command Delsimps (ex_simps @
  4975 all_simps); De Morgan laws are also now included, by default;
  4976 
  4977 * improved printing of ==>  :  ~:
  4978 
  4979 * new object-logic "Sequents" adds linear logic, while replacing LK
  4980 and Modal (thanks to Sara Kalvala);
  4981 
  4982 * HOL/Auth: correctness proofs for authentication protocols;
  4983 
  4984 * HOL: new auto_tac combines rewriting and classical reasoning (many
  4985 examples on HOL/Auth);
  4986 
  4987 * HOL: new command AddIffs for declaring theorems of the form P=Q to
  4988 the rewriter and classical reasoner simultaneously;
  4989 
  4990 * function uresult no longer returns theorems in "standard" format;
  4991 regain previous version by: val uresult = standard o uresult;
  4992 
  4993 
  4994 
  4995 New in Isabelle94-6
  4996 -------------------
  4997 
  4998 * oracles -- these establish an interface between Isabelle and trusted
  4999 external reasoners, which may deliver results as theorems;
  5000 
  5001 * proof objects (in particular record all uses of oracles);
  5002 
  5003 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
  5004 
  5005 * "constdefs" section in theory files;
  5006 
  5007 * "primrec" section (HOL) no longer requires names;
  5008 
  5009 * internal type "tactic" now simply "thm -> thm Sequence.seq";
  5010 
  5011 
  5012 
  5013 New in Isabelle94-5
  5014 -------------------
  5015 
  5016 * reduced space requirements;
  5017 
  5018 * automatic HTML generation from theories;
  5019 
  5020 * theory files no longer require "..." (quotes) around most types;
  5021 
  5022 * new examples, including two proofs of the Church-Rosser theorem;
  5023 
  5024 * non-curried (1994) version of HOL is no longer distributed;
  5025 
  5026 
  5027 
  5028 New in Isabelle94-4
  5029 -------------------
  5030 
  5031 * greatly reduced space requirements;
  5032 
  5033 * theory files (.thy) no longer require \...\ escapes at line breaks;
  5034 
  5035 * searchable theorem database (see the section "Retrieving theorems" on
  5036 page 8 of the Reference Manual);
  5037 
  5038 * new examples, including Grabczewski's monumental case study of the
  5039 Axiom of Choice;
  5040 
  5041 * The previous version of HOL renamed to Old_HOL;
  5042 
  5043 * The new version of HOL (previously called CHOL) uses a curried syntax
  5044 for functions.  Application looks like f a b instead of f(a,b);
  5045 
  5046 * Mutually recursive inductive definitions finally work in HOL;
  5047 
  5048 * In ZF, pattern-matching on tuples is now available in all abstractions and
  5049 translates to the operator "split";
  5050 
  5051 
  5052 
  5053 New in Isabelle94-3
  5054 -------------------
  5055 
  5056 * new infix operator, addss, allowing the classical reasoner to
  5057 perform simplification at each step of its search.  Example:
  5058         fast_tac (cs addss ss)
  5059 
  5060 * a new logic, CHOL, the same as HOL, but with a curried syntax
  5061 for functions.  Application looks like f a b instead of f(a,b).  Also pairs
  5062 look like (a,b) instead of <a,b>;
  5063 
  5064 * PLEASE NOTE: CHOL will eventually replace HOL!
  5065 
  5066 * In CHOL, pattern-matching on tuples is now available in all abstractions.
  5067 It translates to the operator "split".  A new theory of integers is available;
  5068 
  5069 * In ZF, integer numerals now denote two's-complement binary integers.
  5070 Arithmetic operations can be performed by rewriting.  See ZF/ex/Bin.ML;
  5071 
  5072 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
  5073 of the Axiom of Choice;
  5074 
  5075 
  5076 
  5077 New in Isabelle94-2
  5078 -------------------
  5079 
  5080 * Significantly faster resolution;
  5081 
  5082 * the different sections in a .thy file can now be mixed and repeated
  5083 freely;
  5084 
  5085 * Database of theorems for FOL, HOL and ZF.  New
  5086 commands including qed, qed_goal and bind_thm store theorems in the database.
  5087 
  5088 * Simple database queries: return a named theorem (get_thm) or all theorems of
  5089 a given theory (thms_of), or find out what theory a theorem was proved in
  5090 (theory_of_thm);
  5091 
  5092 * Bugs fixed in the inductive definition and datatype packages;
  5093 
  5094 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
  5095 and HOL_dup_cs obsolete;
  5096 
  5097 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
  5098 have been removed;
  5099 
  5100 * Simpler definition of function space in ZF;
  5101 
  5102 * new results about cardinal and ordinal arithmetic in ZF;
  5103 
  5104 * 'subtype' facility in HOL for introducing new types as subsets of existing
  5105 types;
  5106 
  5107 :mode=text:wrap=hard:maxLineLen=72:
  5108 $Id$