1 Isabelle NEWS -- history user-relevant changes
2 ==============================================
4 New in this Isabelle release
5 ----------------------------
9 * Provers/order.ML: new efficient reasoner for partial and linear orders.
12 * Pure: Greek letters (except small lambda, \<lambda>), as well as gothic
13 (\<aa>...\<zz>\<AA>...\<ZZ>), caligraphic (\<A>...\<Z>), and euler
14 (\<a>...\<z>), are now considered normal letters, and can therefore
15 be used anywhere where an ASCII letter (a...zA...Z) has until
16 now. COMPATIBILITY: This obviously changes the parsing of some
17 terms, especially where a symbol has been used as a binder, say
18 '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed
19 as an identifier. Fix it by inserting a space around former
20 symbols. Call 'isatool fixgreek' to try to fix parsing errors in
21 existing theory and ML files.
23 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.
25 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now
26 allowed in identifiers. Similar to greek letters \<^isub> is now considered
27 a normal (but invisible) letter. For multiple letter subscripts repeat
28 \<^isub> like this: x\<^isub>1\<^isub>2.
30 * Pure: There are now sub-/superscripts that can span more than one
31 character. Text between \<^bsub> and \<^esub> is set in subscript in
32 PG and LaTeX, text between \<^bsup> and \<^esup> in superscript. The
33 new control characters are not identifier parts.
35 * Pure: Control-symbols of the form \<^raw...> will literally print the
36 content of ... to the latex file instead of \isacntrl... . The ...
37 accepts all printable characters excluding the end bracket >.
39 * Pure: Symbols may only start with one backslash: \<...>. \\<...> is no
40 longer accepted by the scanner.
42 * Pure: Using new Isar command "finalconsts" (or the ML functions
43 Theory.add_finals or Theory.add_finals_i) it is now possible to
44 declare constants "final", which prevents their being given a definition
45 later. It is useful for constants whose behaviour is fixed axiomatically
46 rather than definitionally, such as the meta-logic connectives.
48 * Pure: It is now illegal to specify Theory.ML explicitly as a dependency
49 in the files section of the theory Theory. (This is more of a diagnostic
50 than a restriction, as the theory loader screws up if Theory.ML is manually
55 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,
56 cut_tac, subgoal_tac and thin_tac:
57 - Now understand static (Isar) contexts. As a consequence, users of Isar
58 locales are no longer forced to write Isar proof scripts.
59 For details see Isar Reference Manual, paragraph 4.3.2: Further tactic
61 - INCOMPATIBILITY: names of variables to be instantiated may no
62 longer be enclosed in quotes. Instead, precede variable name with `?'.
63 This is consistent with the instantiation attribute "where".
65 * Attributes "where" and "of":
66 - Now take type variables of instantiated theorem into account when reading
67 the instantiation string. This fixes a bug that caused instantiated
68 theorems to have too special types in some circumstances.
69 - "where" permits explicit instantiations of type variables.
71 * Calculation commands "moreover" and "also":
72 Do not reset facts ("this") any more.
75 - Goal statements involving the context element "includes" no longer
76 generate theorems with internal delta predicates (those ending on
77 "_axioms") in the premise.
78 Resolve particular premise with <locale>.intro to obtain old form.
79 - Fixed bug in type inference ("unify_frozen") that prevented mix of target
80 specification and "includes" elements in goal statement.
81 - Rule sets <locale>.intro and <locale>.axioms no longer declared as
82 [intro?] and [elim?] (respectively) by default.
83 - Experimental command for instantiation of locales in proof contexts:
84 instantiate <label>: <loc>
85 Instantiates locale <loc> and adds all its theorems to the current context
86 taking into account their attributes, and qualifying their names with
87 <label>. If the locale has assumptions, a chained fact of the form
88 "<loc> t1 ... tn" is expected from which instantiations of the parameters
90 A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.
92 * HOL: Tactic emulation methods induct_tac and case_tac understand static
98 - Much improved handling of linear and partial orders.
99 Reasoners for linear and partial orders are set up for type classes
100 "linorder" and "order" respectively, and are added to the default simpset
101 as solvers. This means that the simplifier can build transitivity chains
102 to solve goals from the assumptions.
103 - INCOMPATIBILITY: old proofs break occasionally. Typically, applications
104 of blast or auto after simplification become unnecessary because the goal
105 is solved by simplification already.
107 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,
108 all proved in axiomatic type classes for semirings, rings and fields.
111 - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are
112 now formalized using the Ring_and_Field theory mentioned above.
113 - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently
114 than before, because now they are set up once in a generic manner.
115 - INCOMPATIBILITY: many type-specific arithmetic laws have gone.
116 Look for the general versions in Ring_and_Field (and Power if they concern
119 * Type "rat" of the rational numbers is now available in HOL-Complex.
122 - Record types are now by default printed with their type abbreviation
123 instead of the list of all field types. This can be configured via
124 the reference "print_record_type_abbr".
125 - Simproc "record_upd_simproc" for simplification of multiple updates added
126 (not enabled by default).
127 - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.
128 EX x. x = sel r to True (not enabled by default).
129 - Tactic "record_split_simp_tac" to split and simplify records added.
131 * 'specification' command added, allowing for definition by
132 specification. There is also an 'ax_specification' command that
133 introduces the new constants axiomatically.
135 * arith(_tac) is now able to generate counterexamples for reals as well.
137 * HOL-Algebra: new locale "ring" for non-commutative rings.
139 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function
140 defintions, thanks to Sava Krsti\'{c} and John Matthews.
142 * Unions and Intersections:
143 The x-symbol output syntax of UN and INT has been changed
144 from "\<Union>x \<in> A. B" to "\<Union\<^bsub>x \<in> A\<^esub> B"
145 i.e. the index formulae has become a subscript, like in normal math.
146 Similarly for "\<Union>x. B", and for \<Inter> instead of \<Union>.
147 The subscript version is also accepted as input syntax.
149 * Unions and Intersections over Intervals:
150 There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is
151 also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"
152 like in normal math, and corresponding versions for < and for intersection.
154 * ML: the legacy theory structures Int and List have been removed. They had
155 conflicted with ML Basis Library structures having the same names.
157 * 'refute' command added to search for (finite) countermodels. Only works
158 for a fragment of HOL. The installation of an external SAT solver is
159 highly recommended. See "HOL/Refute.thy" for details.
162 New in Isabelle2003 (May 2003)
163 --------------------------------
167 * Provers/simplifier:
169 - Completely reimplemented method simp (ML: Asm_full_simp_tac):
170 Assumptions are now subject to complete mutual simplification,
171 not just from left to right. The simplifier now preserves
172 the order of assumptions.
174 Potential INCOMPATIBILITY:
176 -- simp sometimes diverges where the old version did
177 not, e.g. invoking simp on the goal
179 [| P (f x); y = x; f x = f y |] ==> Q
181 now gives rise to the infinite reduction sequence
183 P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...
185 Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this
188 -- Tactics combining classical reasoner and simplification (such as auto)
189 are also affected by this change, because many of them rely on
190 simp. They may sometimes diverge as well or yield a different numbers
191 of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto
192 in case of problems. Sometimes subsequent calls to the classical
193 reasoner will fail because a preceeding call to the simplifier too
194 eagerly simplified the goal, e.g. deleted redundant premises.
196 - The simplifier trace now shows the names of the applied rewrite rules
198 - You can limit the number of recursive invocations of the simplifier
199 during conditional rewriting (where the simplifie tries to solve the
200 conditions before applying the rewrite rule):
201 ML "simp_depth_limit := n"
202 where n is an integer. Thus you can force termination where previously
203 the simplifier would diverge.
205 - Accepts free variables as head terms in congruence rules. Useful in Isar.
207 - No longer aborts on failed congruence proof. Instead, the
208 congruence is ignored.
210 * Pure: New generic framework for extracting programs from constructive
211 proofs. See HOL/Extraction.thy for an example instantiation, as well
212 as HOL/Extraction for some case studies.
214 * Pure: The main goal of the proof state is no longer shown by default, only
215 the subgoals. This behaviour is controlled by a new flag.
216 PG menu: Isabelle/Isar -> Settings -> Show Main Goal
217 (ML: Proof.show_main_goal).
219 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all
220 rules whose conclusion matches subgoal 1:
221 PG menu: Isabelle/Isar -> Show me -> matching rules
222 The rules are ordered by how closely they match the subgoal.
223 In particular, rules that solve a subgoal outright are displayed first
224 (or rather last, the way they are printed).
225 (ML: ProofGeneral.print_intros())
227 * Pure: New flag trace_unify_fail causes unification to print
228 diagnostic information (PG: in trace buffer) when it fails. This is
229 useful for figuring out why single step proofs like rule, erule or
232 * Pure: Locale specifications now produce predicate definitions
233 according to the body of text (covering assumptions modulo local
234 definitions); predicate "loc_axioms" covers newly introduced text,
235 while "loc" is cumulative wrt. all included locale expressions; the
236 latter view is presented only on export into the global theory
237 context; potential INCOMPATIBILITY, use "(open)" option to fall back
238 on the old view without predicates;
240 * Pure: predefined locales "var" and "struct" are useful for sharing
241 parameters (as in CASL, for example); just specify something like
242 ``var x + var y + struct M'' as import;
244 * Pure: improved thms_containing: proper indexing of facts instead of
245 raw theorems; check validity of results wrt. current name space;
246 include local facts of proof configuration (also covers active
247 locales), cover fixed variables in index; may use "_" in term
248 specification; an optional limit for the number of printed facts may
249 be given (the default is 40);
251 * Pure: disallow duplicate fact bindings within new-style theory files
254 * Provers: improved induct method: assumptions introduced by case
255 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from
256 the goal statement); "foo" still refers to all facts collectively;
258 * Provers: the function blast.overloaded has been removed: all constants
259 are regarded as potentially overloaded, which improves robustness in exchange
260 for slight decrease in efficiency;
262 * Provers/linorder: New generic prover for transitivity reasoning over
263 linear orders. Note: this prover is not efficient!
265 * Isar: preview of problems to finish 'show' now produce an error
266 rather than just a warning (in interactive mode);
273 - Produces a counter example if it cannot prove a goal.
274 Note that the counter example may be spurious if the goal is not a formula
275 of quantifier-free linear arithmetic.
276 In ProofGeneral the counter example appears in the trace buffer.
278 - Knows about div k and mod k where k is a numeral of type nat or int.
280 - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free
281 linear arithmetic fails. This takes account of quantifiers and divisibility.
282 Presburger arithmetic can also be called explicitly via presburger(_tac).
284 * simp's arithmetic capabilities have been enhanced a bit: it now
285 takes ~= in premises into account (by performing a case split);
287 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands
288 are distributed over a sum of terms;
290 * New tactic "trans_tac" and method "trans" instantiate
291 Provers/linorder.ML for axclasses "order" and "linorder" (predicates
294 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main
295 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";
297 * 'typedef' command has new option "open" to suppress the set
300 * functions Min and Max on finite sets have been introduced (theory
303 * attribute [symmetric] now works for relations as well; it turns
304 (x,y) : R^-1 into (y,x) : R, and vice versa;
306 * induct over a !!-quantified statement (say !!x1..xn):
307 each "case" automatically performs "fix x1 .. xn" with exactly those names.
309 * Map: `empty' is no longer a constant but a syntactic abbreviation for
310 %x. None. Warning: empty_def now refers to the previously hidden definition
313 * Algebra: formalization of classical algebra. Intended as base for
314 any algebraic development in Isabelle. Currently covers group theory
315 (up to Sylow's theorem) and ring theory (Universal Property of
316 Univariate Polynomials). Contributions welcome;
318 * GroupTheory: deleted, since its material has been moved to Algebra;
320 * Complex: new directory of the complex numbers with numeric constants,
321 nonstandard complex numbers, and some complex analysis, standard and
322 nonstandard (Jacques Fleuriot);
324 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;
326 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques
329 * Real/HahnBanach: updated and adapted to locales;
331 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,
334 * UNITY: added the Meier-Sanders theory of progress sets;
336 * MicroJava: bytecode verifier and lightweight bytecode verifier
337 as abstract algorithms, instantiated to the JVM;
339 * Bali: Java source language formalization. Type system, operational
340 semantics, axiomatic semantics. Supported language features:
341 classes, interfaces, objects,virtual methods, static methods,
342 static/instance fields, arrays, access modifiers, definite
343 assignment, exceptions.
348 * ZF/Constructible: consistency proof for AC (Gödel's constructible
351 * Main ZF: virtually all theories converted to new-style format;
356 * Pure: Tactic.prove provides sane interface for internal proofs;
357 omits the infamous "standard" operation, so this is more appropriate
358 than prove_goalw_cterm in many situations (e.g. in simprocs);
360 * Pure: improved error reporting of simprocs;
362 * Provers: Simplifier.simproc(_i) provides sane interface for setting
366 *** Document preparation ***
368 * uses \par instead of \\ for line breaks in theory text. This may
369 shift some page breaks in large documents. To get the old behaviour
370 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.
372 * minimized dependencies of isabelle.sty and isabellesym.sty on
375 * \<euro> now needs package babel/greek instead of marvosym (which
378 * normal size for \<zero>...\<nine> (uses \mathbf instead of
382 New in Isabelle2002 (March 2002)
383 --------------------------------
385 *** Document preparation ***
387 * greatly simplified document preparation setup, including more
388 graceful interpretation of isatool usedir -i/-d/-D options, and more
389 instructive isatool mkdir; users should basically be able to get
390 started with "isatool mkdir HOL Test && isatool make"; alternatively,
391 users may run a separate document processing stage manually like this:
392 "isatool usedir -D output HOL Test && isatool document Test/output";
394 * theory dependency graph may now be incorporated into documents;
395 isatool usedir -g true will produce session_graph.eps/.pdf for use
396 with \includegraphics of LaTeX;
398 * proper spacing of consecutive markup elements, especially text
399 blocks after section headings;
401 * support bold style (for single symbols only), input syntax is like
402 this: "\<^bold>\<alpha>" or "\<^bold>A";
404 * \<bullet> is now output as bold \cdot by default, which looks much
405 better in printed text;
407 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;
408 note that these symbols are currently unavailable in Proof General /
409 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;
411 * isatool latex no longer depends on changed TEXINPUTS, instead
412 isatool document copies the Isabelle style files to the target
418 * Pure/Provers: improved proof by cases and induction;
419 - 'case' command admits impromptu naming of parameters (such as
421 - 'induct' method divinates rule instantiation from the inductive
422 claim; no longer requires excessive ?P bindings for proper
423 instantiation of cases;
424 - 'induct' method properly enumerates all possibilities of set/type
425 rules; as a consequence facts may be also passed through *type*
426 rules without further ado;
427 - 'induct' method now derives symbolic cases from the *rulified*
428 rule (before it used to rulify cases stemming from the internal
429 atomized version); this means that the context of a non-atomic
430 statement becomes is included in the hypothesis, avoiding the
431 slightly cumbersome show "PROP ?case" form;
432 - 'induct' may now use elim-style induction rules without chaining
433 facts, using ``missing'' premises from the goal state; this allows
434 rules stemming from inductive sets to be applied in unstructured
435 scripts, while still benefitting from proper handling of non-atomic
436 statements; NB: major inductive premises need to be put first, all
437 the rest of the goal is passed through the induction;
438 - 'induct' proper support for mutual induction involving non-atomic
439 rule statements (uses the new concept of simultaneous goals, see
441 - append all possible rule selections, but only use the first
442 success (no backtracking);
443 - removed obsolete "(simplified)" and "(stripped)" options of methods;
444 - undeclared rule case names default to numbers 1, 2, 3, ...;
445 - added 'print_induct_rules' (covered by help item in recent Proof
447 - moved induct/cases attributes to Pure, methods to Provers;
448 - generic method setup instantiated for FOL and HOL;
450 * Pure: support multiple simultaneous goal statements, for example
451 "have a: A and b: B" (same for 'theorem' etc.); being a pure
452 meta-level mechanism, this acts as if several individual goals had
453 been stated separately; in particular common proof methods need to be
454 repeated in order to cover all claims; note that a single elimination
455 step is *not* sufficient to establish the two conjunctions, so this
458 assume "A & B" then have A and B .. (*".." fails*)
460 better use "obtain" in situations as above; alternative refer to
461 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;
463 * Pure: proper integration with ``locales''; unlike the original
464 version by Florian Kammüller, Isar locales package high-level proof
465 contexts rather than raw logical ones (e.g. we admit to include
466 attributes everywhere); operations on locales include merge and
467 rename; support for implicit arguments (``structures''); simultaneous
468 type-inference over imports and text; see also HOL/ex/Locales.thy for
471 * Pure: the following commands have been ``localized'', supporting a
472 target locale specification "(in name)": 'lemma', 'theorem',
473 'corollary', 'lemmas', 'theorems', 'declare'; the results will be
474 stored both within the locale and at the theory level (exported and
475 qualified by the locale name);
477 * Pure: theory goals may now be specified in ``long'' form, with
478 ad-hoc contexts consisting of arbitrary locale elements. for example
479 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and
480 definitions may be given, too); the result is a meta-level rule with
481 the context elements being discharged in the obvious way;
483 * Pure: new proof command 'using' allows to augment currently used
484 facts after a goal statement ('using' is syntactically analogous to
485 'apply', but acts on the goal's facts only); this allows chained facts
486 to be separated into parts given before and after a claim, as in
487 ``from a and b have C using d and e <proof>'';
489 * Pure: renamed "antecedent" case to "rule_context";
491 * Pure: new 'judgment' command records explicit information about the
492 object-logic embedding (used by several tools internally); no longer
493 use hard-wired "Trueprop";
495 * Pure: added 'corollary' command;
497 * Pure: fixed 'token_translation' command;
499 * Pure: removed obsolete 'exported' attribute;
501 * Pure: dummy pattern "_" in is/let is now automatically lifted over
502 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")
503 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");
505 * Pure: method 'atomize' presents local goal premises as object-level
506 statements (atomic meta-level propositions); setup controlled via
507 rewrite rules declarations of 'atomize' attribute; example
508 application: 'induct' method with proper rule statements in improper
511 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)
512 now consider the syntactic context of assumptions, giving a better
513 chance to get type-inference of the arguments right (this is
514 especially important for locales);
516 * Pure: "sorry" no longer requires quick_and_dirty in interactive
519 * Pure/obtain: the formal conclusion "thesis", being marked as
520 ``internal'', may no longer be reference directly in the text;
521 potential INCOMPATIBILITY, may need to use "?thesis" in rare
524 * Pure: generic 'sym' attribute which declares a rule both as pure
525 'elim?' and for the 'symmetric' operation;
527 * Pure: marginal comments ``--'' may now occur just anywhere in the
528 text; the fixed correlation with particular command syntax has been
531 * Pure: new method 'rules' is particularly well-suited for proof
532 search in intuitionistic logic; a bit slower than 'blast' or 'fast',
533 but often produces more compact proof terms with less detours;
535 * Pure/Provers/classical: simplified integration with pure rule
536 attributes and methods; the classical "intro?/elim?/dest?"
537 declarations coincide with the pure ones; the "rule" method no longer
538 includes classically swapped intros; "intro" and "elim" methods no
539 longer pick rules from the context; also got rid of ML declarations
540 AddXIs/AddXEs/AddXDs; all of this has some potential for
543 * Provers/classical: attribute 'swapped' produces classical inversions
544 of introduction rules;
546 * Provers/simplifier: 'simplified' attribute may refer to explicit
547 rules instead of full simplifier context; 'iff' attribute handles
550 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;
552 * HOL: 'recdef' now fails on unfinished automated proofs, use
553 "(permissive)" option to recover old behavior;
555 * HOL: 'inductive' no longer features separate (collective) attributes
556 for 'intros' (was found too confusing);
558 * HOL: properly declared induction rules less_induct and
564 * HOL: moved over to sane numeral syntax; the new policy is as
567 - 0 and 1 are polymorphic constants, which are defined on any
568 numeric type (nat, int, real etc.);
570 - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based
571 binary representation internally;
573 - type nat has special constructor Suc, and generally prefers Suc 0
574 over 1::nat and Suc (Suc 0) over 2::nat;
576 This change may cause significant problems of INCOMPATIBILITY; here
577 are some hints on converting existing sources:
579 - due to the new "num" token, "-0" and "-1" etc. are now atomic
580 entities, so expressions involving "-" (unary or binary minus) need
581 to be spaced properly;
583 - existing occurrences of "1" may need to be constraint "1::nat" or
584 even replaced by Suc 0; similar for old "2";
586 - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";
588 - remove all special provisions on numerals in proofs;
590 * HOL: simp rules nat_number expand numerals on nat to Suc/0
591 representation (depends on bin_arith_simps in the default context);
593 * HOL: symbolic syntax for x^2 (numeral 2);
595 * HOL: the class of all HOL types is now called "type" rather than
596 "term"; INCOMPATIBILITY, need to adapt references to this type class
597 in axclass/classes, instance/arities, and (usually rare) occurrences
598 in typings (of consts etc.); internally the class is called
599 "HOL.type", ML programs should refer to HOLogic.typeS;
601 * HOL/record package improvements:
602 - new derived operations "fields" to build a partial record section,
603 "extend" to promote a fixed record to a record scheme, and
604 "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*
605 declared as simp by default;
606 - shared operations ("more", "fields", etc.) now need to be always
607 qualified) --- potential INCOMPATIBILITY;
608 - removed "make_scheme" operations (use "make" with "extend") --
610 - removed "more" class (simply use "term") -- INCOMPATIBILITY;
611 - provides cases/induct rules for use with corresponding Isar
612 methods (for concrete records, record schemes, concrete more
613 parts, and schematic more parts -- in that order);
614 - internal definitions directly based on a light-weight abstract
615 theory of product types over typedef rather than datatype;
617 * HOL: generic code generator for generating executable ML code from
618 specifications; specific support for HOL constructs such as inductive
619 datatypes and sets, as well as recursive functions; can be invoked
620 via 'generate_code' theory section;
622 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);
624 * HOL: consolidated and renamed several theories. In particular:
625 Ord.thy has been absorbed into HOL.thy
626 String.thy has been absorbed into List.thy
628 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"
629 (beware of argument permutation!);
631 * HOL: linorder_less_split superseded by linorder_cases;
633 * HOL/List: "nodups" renamed to "distinct";
635 * HOL: added "The" definite description operator; move Hilbert's "Eps"
636 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:
637 - Ex_def has changed, now need to use some_eq_ex
639 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so
640 in this (rare) case use:
642 delSWrapper "split_all_tac"
643 addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)
645 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS
648 * HOL: introduced f^n = f o ... o f; warning: due to the limits of
649 Isabelle's type classes, ^ on functions and relations has too general
650 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be
651 necessary to attach explicit type constraints;
653 * HOL/Relation: the prefix name of the infix "O" has been changed from
654 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been
655 renamed accordingly (eg "compI" -> "rel_compI").
657 * HOL: syntax translations now work properly with numerals and records
660 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead
661 of "lam" -- INCOMPATIBILITY;
663 * HOL: got rid of some global declarations (potential INCOMPATIBILITY
664 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"
665 renamed "Product_Type.unit";
667 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl
669 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or
672 * HOL/GroupTheory: group theory examples including Sylow's theorem (by
675 * HOL/IMP: updated and converted to new-style theory format; several
676 parts turned into readable document, with proper Isar proof texts and
677 some explanations (by Gerwin Klein);
679 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);
681 * HOL-Hyperreal is now a logic image;
686 * Isar: consts/constdefs supports mixfix syntax for continuous
689 * Isar: domain package adapted to new-style theory format, e.g. see
692 * theory Lift: proper use of rep_datatype lift instead of ML hacks --
693 potential INCOMPATIBILITY; now use plain induct_tac instead of former
694 lift.induct_tac, always use UU instead of Undef;
696 * HOLCF/IMP: updated and converted to new-style theory;
701 * Isar: proper integration of logic-specific tools and packages,
702 including theory commands '(co)inductive', '(co)datatype',
703 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',
704 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');
706 * theory Main no longer includes AC; for the Axiom of Choice, base
707 your theory on Main_ZFC;
709 * the integer library now covers quotients and remainders, with many
710 laws relating division to addition, multiplication, etc.;
712 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a
713 typeless version of the formalism;
715 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory
718 * ZF/Induct: new directory for examples of inductive definitions,
719 including theory Multiset for multiset orderings; converted to
720 new-style theory format;
722 * ZF: many new theorems about lists, ordinals, etc.;
727 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference
728 variable proof controls level of detail: 0 = no proofs (only oracle
729 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see
730 also ref manual for further ML interfaces;
732 * Pure/axclass: removed obsolete ML interface
733 goal_subclass/goal_arity;
735 * Pure/syntax: new token syntax "num" for plain numerals (without "#"
736 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now
737 separate tokens, so expressions involving minus need to be spaced
740 * Pure/syntax: support non-oriented infixes, using keyword "infix"
741 rather than "infixl" or "infixr";
743 * Pure/syntax: concrete syntax for dummy type variables admits genuine
744 sort constraint specifications in type inference; e.g. "x::_::foo"
745 ensures that the type of "x" is of sort "foo" (but not necessarily a
748 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"
749 control output of nested => (types); the default behavior is
752 * Pure/syntax: builtin parse translation for "_constify" turns valued
753 tokens into AST constants;
755 * Pure/syntax: prefer later declarations of translations and print
756 translation functions; potential INCOMPATIBILITY: need to reverse
757 multiple declarations for same syntax element constant;
759 * Pure/show_hyps reset by default (in accordance to existing Isar
762 * Provers/classical: renamed addaltern to addafter, addSaltern to
765 * Provers/clasimp: ``iff'' declarations now handle conditional rules
768 * system: tested support for MacOS X; should be able to get Isabelle +
769 Proof General to work in a plain Terminal after installing Poly/ML
770 (e.g. from the Isabelle distribution area) and GNU bash alone
771 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol
772 support requires further installations, e.g. from
773 http://fink.sourceforge.net/);
775 * system: support Poly/ML 4.1.1 (able to manage larger heaps);
777 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead
778 of 40 MB), cf. ML_OPTIONS;
780 * system: Proof General keywords specification is now part of the
781 Isabelle distribution (see etc/isar-keywords.el);
783 * system: support for persistent Proof General sessions (refrain from
784 outdating all loaded theories on startup); user may create writable
785 logic images like this: ``isabelle -q HOL Test'';
787 * system: smart selection of Isabelle process versus Isabelle
788 interface, accommodates case-insensitive file systems (e.g. HFS+); may
789 run both "isabelle" and "Isabelle" even if file names are badly
790 damaged (executable inspects the case of the first letter of its own
791 name); added separate "isabelle-process" and "isabelle-interface";
793 * system: refrain from any attempt at filtering input streams; no
794 longer support ``8bit'' encoding of old isabelle font, instead proper
795 iso-latin characters may now be used; the related isatools
796 "symbolinput" and "nonascii" have disappeared as well;
798 * system: removed old "xterm" interface (the print modes "xterm" and
799 "xterm_color" are still available for direct use in a suitable
804 New in Isabelle99-2 (February 2001)
805 -----------------------------------
807 *** Overview of INCOMPATIBILITIES ***
809 * HOL: please note that theories in the Library and elsewhere often use the
810 new-style (Isar) format; to refer to their theorems in an ML script you must
811 bind them to ML identifers by e.g. val thm_name = thm "thm_name";
813 * HOL: inductive package no longer splits induction rule aggressively,
814 but only as far as specified by the introductions given; the old
815 format may be recovered via ML function complete_split_rule or attribute
816 'split_rule (complete)';
818 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,
819 gfp_Tarski to gfp_unfold;
821 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;
823 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a
824 relation); infix "^^" has been renamed "``"; infix "``" has been
825 renamed "`"; "univalent" has been renamed "single_valued";
827 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"
830 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;
832 * Isar: 'obtain' no longer declares "that" fact as simp/intro;
834 * Isar/HOL: method 'induct' now handles non-atomic goals; as a
835 consequence, it is no longer monotonic wrt. the local goal context
836 (which is now passed through the inductive cases);
838 * Document preparation: renamed standard symbols \<ll> to \<lless> and
839 \<gg> to \<ggreater>;
842 *** Document preparation ***
844 * \isabellestyle{NAME} selects version of Isabelle output (currently
845 available: are "it" for near math-mode best-style output, "sl" for
846 slanted text style, and "tt" for plain type-writer; if no
847 \isabellestyle command is given, output is according to slanted
850 * support sub/super scripts (for single symbols only), input syntax is
851 like this: "A\<^sup>*" or "A\<^sup>\<star>";
853 * some more standard symbols; see Appendix A of the system manual for
854 the complete list of symbols defined in isabellesym.sty;
856 * improved isabelle style files; more abstract symbol implementation
857 (should now use \isamath{...} and \isatext{...} in custom symbol
860 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals
861 state; Note that presentation of goal states does not conform to
862 actual human-readable proof documents. Please do not include goal
863 states into document output unless you really know what you are doing!
865 * proper indentation of antiquoted output with proportional LaTeX
868 * no_document ML operator temporarily disables LaTeX document
871 * isatool unsymbolize tunes sources for plain ASCII communication;
876 * Pure: Isar now suffers initial goal statements to contain unbound
877 schematic variables (this does not conform to actual readable proof
878 documents, due to unpredictable outcome and non-compositional proof
879 checking); users who know what they are doing may use schematic goals
880 for Prolog-style synthesis of proven results;
882 * Pure: assumption method (an implicit finishing) now handles actual
885 * Pure: improved 'obtain' --- moved to Pure, insert "that" into
886 initial goal, declare "that" only as Pure intro (only for single
887 steps); the "that" rule assumption may now be involved in implicit
888 finishing, thus ".." becomes a feasible for trivial obtains;
890 * Pure: default proof step now includes 'intro_classes'; thus trivial
891 instance proofs may be performed by "..";
893 * Pure: ?thesis / ?this / "..." now work for pure meta-level
896 * Pure: more robust selection of calculational rules;
898 * Pure: the builtin notion of 'finished' goal now includes the ==-refl
899 rule (as well as the assumption rule);
901 * Pure: 'thm_deps' command visualizes dependencies of theorems and
902 lemmas, using the graph browser tool;
904 * Pure: predict failure of "show" in interactive mode;
906 * Pure: 'thms_containing' now takes actual terms as arguments;
908 * HOL: improved method 'induct' --- now handles non-atomic goals
909 (potential INCOMPATIBILITY); tuned error handling;
911 * HOL: cases and induct rules now provide explicit hints about the
912 number of facts to be consumed (0 for "type" and 1 for "set" rules);
913 any remaining facts are inserted into the goal verbatim;
915 * HOL: local contexts (aka cases) may now contain term bindings as
916 well; the 'cases' and 'induct' methods new provide a ?case binding for
917 the result to be shown in each case;
919 * HOL: added 'recdef_tc' command;
921 * isatool convert assists in eliminating legacy ML scripts;
926 * HOL/Library: a collection of generic theories to be used together
927 with main HOL; the theory loader path already includes this directory
928 by default; the following existing theories have been moved here:
929 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While
930 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);
932 * HOL/Unix: "Some aspects of Unix file-system security", a typical
933 modelling and verification task performed in Isabelle/HOL +
934 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).
936 * HOL/Algebra: special summation operator SUM no longer exists, it has
937 been replaced by setsum; infix 'assoc' now has priority 50 (like
938 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to
939 'domain', this makes the theory consistent with mathematical
942 * HOL basics: added overloaded operations "inverse" and "divide"
943 (infix "/"), syntax for generic "abs" operation, generic summation
946 * HOL/typedef: simplified package, provide more useful rules (see also
949 * HOL/datatype: induction rule for arbitrarily branching datatypes is
950 now expressed as a proper nested rule (old-style tactic scripts may
951 require atomize_strip_tac to cope with non-atomic premises);
953 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule
954 to "split_conv" (old name still available for compatibility);
956 * HOL: improved concrete syntax for strings (e.g. allows translation
957 rules with string literals);
959 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals
960 and Fleuriot's mechanization of analysis, including the transcendental
961 functions for the reals;
963 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;
968 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that
969 "lam" is displayed as TWO lambda-symbols
971 * CTT: theory Main now available, containing everything (that is, Bool
977 * Pure: the Simplifier has been implemented properly as a derived rule
978 outside of the actual kernel (at last!); the overall performance
979 penalty in practical applications is about 50%, while reliability of
980 the Isabelle inference kernel has been greatly improved;
982 * print modes "brackets" and "no_brackets" control output of nested =>
983 (types) and ==> (props); the default behaviour is "brackets";
985 * Provers: fast_tac (and friends) now handle actual object-logic rules
986 as assumptions as well;
988 * system: support Poly/ML 4.0;
990 * system: isatool install handles KDE version 1 or 2;
994 New in Isabelle99-1 (October 2000)
995 ----------------------------------
997 *** Overview of INCOMPATIBILITIES ***
999 * HOL: simplification of natural numbers is much changed; to partly
1000 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)
1001 issue the following ML commands:
1003 Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;
1004 Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];
1006 * HOL: simplification no longer dives into case-expressions; this is
1007 controlled by "t.weak_case_cong" for each datatype t;
1009 * HOL: nat_less_induct renamed to less_induct;
1011 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool
1012 fixsome to patch .thy and .ML sources automatically;
1014 select_equality -> some_equality
1015 select_eq_Ex -> some_eq_ex
1016 selectI2EX -> someI2_ex
1019 select1_equality -> some1_equality
1020 Eps_sym_eq -> some_sym_eq_trivial
1021 Eps_eq -> some_eq_trivial
1023 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;
1025 * HOL: removed obsolete theorem binding expand_if (refer to split_if
1028 * HOL: the recursion equations generated by 'recdef' are now called
1029 f.simps instead of f.rules;
1031 * HOL: qed_spec_mp now also handles bounded ALL as well;
1033 * HOL: 0 is now overloaded, so the type constraint ":: nat" may
1034 sometimes be needed;
1036 * HOL: the constant for "f``x" is now "image" rather than "op ``";
1038 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";
1040 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian
1041 product is now "<*>" instead of "Times"; the lexicographic product is
1042 now "<*lex*>" instead of "**";
1044 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part
1045 of main HOL, but was unused); better use HOL's datatype package;
1047 * HOL: removed "symbols" syntax for constant "override" of theory Map;
1048 the old syntax may be recovered as follows:
1051 override :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"
1052 (infixl "\\<oplus>" 100)
1054 * HOL/Real: "rabs" replaced by overloaded "abs" function;
1056 * HOL/ML: even fewer consts are declared as global (see theories Ord,
1057 Lfp, Gfp, WF); this only affects ML packages that refer to const names
1060 * HOL and ZF: syntax for quotienting wrt an equivalence relation
1061 changed from A/r to A//r;
1063 * ZF: new treatment of arithmetic (nat & int) may break some old
1066 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,
1067 rulify -> rule_format, elimify -> elim_format, ...);
1069 * Isar/Provers: intro/elim/dest attributes changed; renamed
1070 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one
1071 should have to change intro!! to intro? only); replaced "delrule" by
1074 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;
1076 * Provers: strengthened force_tac by using new first_best_tac;
1078 * LaTeX document preparation: several changes of isabelle.sty (see
1082 *** Document preparation ***
1084 * formal comments (text blocks etc.) in new-style theories may now
1085 contain antiquotations of thm/prop/term/typ/text to be presented
1086 according to latex print mode; concrete syntax is like this:
1087 @{term[show_types] "f(x) = a + x"};
1089 * isatool mkdir provides easy setup of Isabelle session directories,
1090 including proper document sources;
1092 * generated LaTeX sources are now deleted after successful run
1093 (isatool document -c); may retain a copy somewhere else via -D option
1096 * isatool usedir -D now lets isatool latex -o sty update the Isabelle
1097 style files, achieving self-contained LaTeX sources and simplifying
1100 * old-style theories now produce (crude) LaTeX output as well;
1102 * browser info session directories are now self-contained (may be put
1103 on WWW server seperately); improved graphs of nested sessions; removed
1104 graph for 'all sessions';
1106 * several improvements in isabelle style files; \isabellestyle{it}
1107 produces fake math mode output; \isamarkupheader is now \section by
1108 default; see lib/texinputs/isabelle.sty etc.;
1113 * Isar/Pure: local results and corresponding term bindings are now
1114 subject to Hindley-Milner polymorphism (similar to ML); this
1115 accommodates incremental type-inference very nicely;
1117 * Isar/Pure: new derived language element 'obtain' supports
1118 generalized existence reasoning;
1120 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'
1121 support accumulation of results, without applying any rules yet;
1122 useful to collect intermediate results without explicit name
1123 references, and for use with transitivity rules with more than 2
1126 * Isar/Pure: scalable support for case-analysis type proofs: new
1127 'case' language element refers to local contexts symbolically, as
1128 produced by certain proof methods; internally, case names are attached
1129 to theorems as "tags";
1131 * Isar/Pure: theory command 'hide' removes declarations from
1132 class/type/const name spaces;
1134 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to
1135 indicate potential overloading;
1137 * Isar/Pure: changed syntax of local blocks from {{ }} to { };
1139 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write
1140 "{a,b,c}" instead of {a,b,c};
1142 * Isar/Pure now provides its own version of intro/elim/dest
1143 attributes; useful for building new logics, but beware of confusion
1144 with the version in Provers/classical;
1146 * Isar/Pure: the local context of (non-atomic) goals is provided via
1147 case name 'antecedent';
1149 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms
1150 to the current context is now done automatically);
1152 * Isar/Pure: theory command 'method_setup' provides a simple interface
1153 for definining proof methods in ML;
1155 * Isar/Provers: intro/elim/dest attributes changed; renamed
1156 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in
1157 most cases, one should have to change intro!! to intro? only);
1158 replaced "delrule" by "rule del";
1160 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and
1161 'symmetric' attribute (the latter supercedes [RS sym]);
1163 * Isar/Provers: splitter support (via 'split' attribute and 'simp'
1164 method modifier); 'simp' method: 'only:' modifier removes loopers as
1165 well (including splits);
1167 * Isar/Provers: Simplifier and Classical methods now support all kind
1168 of modifiers used in the past, including 'cong', 'iff', etc.
1170 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination
1171 of Simplifier and Classical reasoner);
1173 * Isar/HOL: new proof method 'cases' and improved version of 'induct'
1174 now support named cases; major packages (inductive, datatype, primrec,
1175 recdef) support case names and properly name parameters;
1177 * Isar/HOL: new transitivity rules for substitution in inequalities --
1178 monotonicity conditions are extracted to be proven at end of
1181 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof
1184 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =
1185 split_if split_if_asm; datatype package provides theorems foo.splits =
1186 foo.split foo.split_asm for each datatype;
1188 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"
1189 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof
1190 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:
1191 use "(cases (simplified))" method in proper proof texts);
1193 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;
1195 * Isar: names of theorems etc. may be natural numbers as well;
1197 * Isar: 'pr' command: optional arguments for goals_limit and
1198 ProofContext.prems_limit; no longer prints theory contexts, but only
1201 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit
1202 additional print modes to be specified; e.g. "pr(latex)" will print
1203 proof state according to the Isabelle LaTeX style;
1205 * Isar: improved support for emulating tactic scripts, including proof
1206 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',
1207 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'
1208 (for HOL datatypes);
1210 * Isar: simplified (more robust) goal selection of proof methods: 1st
1211 goal, all goals, or explicit goal specifier (tactic emulation); thus
1212 'proof method scripts' have to be in depth-first order;
1214 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';
1216 * Isar: removed 'help' command, which hasn't been too helpful anyway;
1217 should instead use individual commands for printing items
1218 (print_commands, print_methods etc.);
1220 * Isar: added 'nothing' --- the empty list of theorems;
1225 * HOL/MicroJava: formalization of a fragment of Java, together with a
1226 corresponding virtual machine and a specification of its bytecode
1227 verifier and a lightweight bytecode verifier, including proofs of
1228 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and
1229 Cornelia Pusch (see also the homepage of project Bali at
1230 http://isabelle.in.tum.de/Bali/);
1232 * HOL/Algebra: new theory of rings and univariate polynomials, by
1235 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese
1236 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M
1239 * HOL/Lattice: fundamental concepts of lattice theory and order
1240 structures, including duals, properties of bounds versus algebraic
1241 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski
1242 Theorem for complete lattices etc.; may also serve as a demonstration
1243 for abstract algebraic reasoning using axiomatic type classes, and
1244 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;
1246 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David
1249 * HOL/IMPP: extension of IMP with local variables and mutually
1250 recursive procedures, by David von Oheimb;
1252 * HOL/Lambda: converted into new-style theory and document;
1254 * HOL/ex/Multiquote: example of multiple nested quotations and
1255 anti-quotations -- basically a generalized version of de-Bruijn
1256 representation; very useful in avoiding lifting of operations;
1258 * HOL/record: added general record equality rule to simpset; fixed
1259 select-update simplification procedure to handle extended records as
1260 well; admit "r" as field name;
1262 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with
1263 other numeric types and also as the identity of groups, rings, etc.;
1265 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.
1266 Types nat and int belong to this axclass;
1268 * HOL: greatly improved simplification involving numerals of type nat, int, real:
1269 (i + #8 + j) = Suc k simplifies to #7 + (i + j) = k
1270 i*j + k + j*#3*i simplifies to #4*(i*j) + k
1271 two terms #m*u and #n*u are replaced by #(m+n)*u
1272 (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)
1273 and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y
1274 or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);
1276 * HOL: meson_tac is available (previously in ex/meson.ML); it is a
1277 powerful prover for predicate logic but knows nothing of clasets; see
1278 ex/mesontest.ML and ex/mesontest2.ML for example applications;
1280 * HOL: new version of "case_tac" subsumes both boolean case split and
1281 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer
1282 exists, may define val exhaust_tac = case_tac for ad-hoc portability;
1284 * HOL: simplification no longer dives into case-expressions: only the
1285 selector expression is simplified, but not the remaining arms; to
1286 enable full simplification of case-expressions for datatype t, you may
1287 remove t.weak_case_cong from the simpset, either globally (Delcongs
1288 [thm"t.weak_case_cong"];) or locally (delcongs [...]).
1290 * HOL/recdef: the recursion equations generated by 'recdef' for
1291 function 'f' are now called f.simps instead of f.rules; if all
1292 termination conditions are proved automatically, these simplification
1293 rules are added to the simpset, as in primrec; rules may be named
1294 individually as well, resulting in a separate list of theorems for
1297 * HOL/While is a new theory that provides a while-combinator. It
1298 permits the definition of tail-recursive functions without the
1299 provision of a termination measure. The latter is necessary once the
1300 invariant proof rule for while is applied.
1302 * HOL: new (overloaded) notation for the set of elements below/above
1303 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.
1305 * HOL: theorems impI, allI, ballI bound as "strip";
1307 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic
1308 induct_tac th "x1 ... xn" expects th to have a conclusion of the form
1309 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;
1311 * HOL/Real: "rabs" replaced by overloaded "abs" function;
1313 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of
1314 main HOL, but was unused);
1316 * HOL: fewer consts declared as global (e.g. have to refer to
1317 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);
1319 * HOL: tuned AST representation of nested pairs, avoiding bogus output
1320 in case of overlap with user translations (e.g. judgements over
1321 tuples); (note that the underlying logical represenation is still
1327 * ZF: simplification automatically cancels common terms in arithmetic
1328 expressions over nat and int;
1330 * ZF: new treatment of nat to minimize type-checking: all operators
1331 coerce their operands to a natural number using the function natify,
1332 making the algebraic laws unconditional;
1334 * ZF: as above, for int: operators coerce their operands to an integer
1335 using the function intify;
1337 * ZF: the integer library now contains many of the usual laws for the
1338 orderings, including $<=, and monotonicity laws for $+ and $*;
1340 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic
1343 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q
1344 to the simplifier and classical reasoner simultaneously;
1349 * Provers: blast_tac now handles actual object-logic rules as
1350 assumptions; note that auto_tac uses blast_tac internally as well;
1352 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning
1353 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;
1355 * Provers: delrules now handles destruct rules as well (no longer need
1356 explicit make_elim);
1358 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.
1359 [| inj ?f; ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
1360 use instead the strong form,
1361 [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W
1362 in HOL, FOL and ZF the function cla_make_elim will create such rules
1363 from destruct-rules;
1365 * Provers: Simplifier.easy_setup provides a fast path to basic
1366 Simplifier setup for new object-logics;
1368 * Pure: AST translation rules no longer require constant head on LHS;
1370 * Pure: improved name spaces: ambiguous output is qualified; support
1371 for hiding of names;
1373 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and
1374 XSYMBOL_HOME; no longer need to do manual configuration in most
1377 * system: compression of ML heaps images may now be controlled via -c
1378 option of isabelle and isatool usedir (currently only observed by
1381 * system: isatool installfonts may handle X-Symbol fonts as well (very
1382 useful for remote X11);
1384 * system: provide TAGS file for Isabelle sources;
1386 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument
1389 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global
1390 timing flag supersedes proof_timing and Toplevel.trace;
1392 * ML: new combinators |>> and |>>> for incremental transformations
1393 with secondary results (e.g. certain theory extensions):
1395 * ML: PureThy.add_defs gets additional argument to indicate potential
1396 overloading (usually false);
1398 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as
1403 New in Isabelle99 (October 1999)
1404 --------------------------------
1406 *** Overview of INCOMPATIBILITIES (see below for more details) ***
1408 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)
1409 are no longer simplified. (This allows the simplifier to unfold recursive
1410 functional programs.) To restore the old behaviour, declare
1412 Delcongs [if_weak_cong];
1414 * HOL: Removed the obsolete syntax "Compl A"; use -A for set
1417 * HOL: the predicate "inj" is now defined by translation to "inj_on";
1419 * HOL/datatype: mutual_induct_tac no longer exists --
1420 use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]
1422 * HOL/typedef: fixed type inference for representing set; type
1423 arguments now have to occur explicitly on the rhs as type constraints;
1425 * ZF: The con_defs part of an inductive definition may no longer refer
1426 to constants declared in the same theory;
1428 * HOL, ZF: the function mk_cases, generated by the inductive
1429 definition package, has lost an argument. To simplify its result, it
1430 uses the default simpset instead of a supplied list of theorems.
1432 * HOL/List: the constructors of type list are now Nil and Cons;
1434 * Simplifier: the type of the infix ML functions
1435 setSSolver addSSolver setSolver addSolver
1436 is now simpset * solver -> simpset where `solver' is a new abstract type
1437 for packaging solvers. A solver is created via
1438 mk_solver: string -> (thm list -> int -> tactic) -> solver
1439 where the string argument is only a comment.
1444 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a
1445 decision procedure for linear arithmetic. Currently it is used for
1446 types `nat', `int', and `real' in HOL (see below); it can, should and
1447 will be instantiated for other types and logics as well.
1449 * The simplifier now accepts rewrite rules with flexible heads, eg
1450 hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y
1451 They are applied like any rule with a non-pattern lhs, i.e. by first-order
1457 * New Isabelle/Isar subsystem provides an alternative to traditional
1458 tactical theorem proving; together with the ProofGeneral/isar user
1459 interface it offers an interactive environment for developing human
1460 readable proof documents (Isar == Intelligible semi-automated
1461 reasoning); for further information see isatool doc isar-ref,
1462 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/
1464 * improved and simplified presentation of theories: better HTML markup
1465 (including colors), graph views in several sizes; isatool usedir now
1466 provides a proper interface for user theories (via -P option); actual
1467 document preparation based on (PDF)LaTeX is available as well (for
1468 new-style theories only); see isatool doc system for more information;
1470 * native support for Proof General, both for classic Isabelle and
1473 * ML function thm_deps visualizes dependencies of theorems and lemmas,
1474 using the graph browser tool;
1476 * Isabelle manuals now also available as PDF;
1478 * theory loader rewritten from scratch (may not be fully
1479 bug-compatible); old loadpath variable has been replaced by show_path,
1480 add_path, del_path, reset_path functions; new operations such as
1481 update_thy, touch_thy, remove_thy, use/update_thy_only (see also
1484 * improved isatool install: option -k creates KDE application icon,
1485 option -p DIR installs standalone binaries;
1487 * added ML_PLATFORM setting (useful for cross-platform installations);
1488 more robust handling of platform specific ML images for SML/NJ;
1490 * the settings environment is now statically scoped, i.e. it is never
1491 created again in sub-processes invoked from isabelle, isatool, or
1494 * path element specification '~~' refers to '$ISABELLE_HOME';
1496 * in locales, the "assumes" and "defines" parts may be omitted if
1499 * new print_mode "xsymbols" for extended symbol support (e.g. genuine
1502 * new print_mode "HTML";
1504 * new flag show_tags controls display of tags of theorems (which are
1505 basically just comments that may be attached by some tools);
1507 * Isamode 2.6 requires patch to accomodate change of Isabelle font
1508 mode and goal output format:
1510 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el
1512 < (list (isa-getenv "ISABELLE") "-msymbols" logic-name)
1514 > (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)
1515 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el
1517 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"
1519 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"
1521 * function bind_thms stores lists of theorems (cf. bind_thm);
1523 * new shorthand tactics ftac, eatac, datac, fatac;
1525 * qed (and friends) now accept "" as result name; in that case the
1526 theorem is not stored, but proper checks and presentation of the
1529 * theorem database now also indexes constants "Trueprop", "all",
1530 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;
1535 ** HOL arithmetic **
1537 * There are now decision procedures for linear arithmetic over nat and
1540 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',
1541 `+', `-', `Suc', `min', `max' and numerical constants; other subterms
1542 are treated as atomic; subformulae not involving type `nat' or `int'
1543 are ignored; quantified subformulae are ignored unless they are
1544 positive universal or negative existential. The tactic has to be
1545 invoked by hand and can be a little bit slow. In particular, the
1546 running time is exponential in the number of occurrences of `min' and
1547 `max', and `-' on `nat'.
1549 2. fast_arith_tac is a cut-down version of arith_tac: it only takes
1550 (negated) (in)equalities among the premises and the conclusion into
1551 account (i.e. no compound formulae) and does not know about `min' and
1552 `max', and `-' on `nat'. It is fast and is used automatically by the
1555 NB: At the moment, these decision procedures do not cope with mixed
1556 nat/int formulae where the two parts interact, such as `m < n ==>
1559 * HOL/Numeral provides a generic theory of numerals (encoded
1560 efficiently as bit strings); setup for types nat/int/real is in place;
1561 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than
1562 int, existing theories and proof scripts may require a few additional
1565 * integer division and remainder can now be performed on constant
1568 * many properties of integer multiplication, division and remainder
1571 * An interface to the Stanford Validity Checker (SVC) is available through the
1572 tactic svc_tac. Propositional tautologies and theorems of linear arithmetic
1573 are proved automatically. SVC must be installed separately, and its results
1574 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any
1575 invocation of the underlying oracle). For SVC see
1576 http://verify.stanford.edu/SVC
1578 * IsaMakefile: the HOL-Real target now builds an actual image;
1583 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces
1584 (in Isabelle/Isar) -- by Gertrud Bauer;
1586 * HOL/BCV: generic model of bytecode verification, i.e. data-flow
1587 analysis for assembly languages with subtypes;
1589 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization
1590 -- avoids syntactic ambiguities and treats state, transition, and
1591 temporal levels more uniformly; introduces INCOMPATIBILITIES due to
1592 changed syntax and (many) tactics;
1594 * HOL/inductive: Now also handles more general introduction rules such
1595 as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity
1596 theorems are now maintained within the theory (maintained via the
1599 * HOL/datatype: Now also handles arbitrarily branching datatypes
1600 (using function types) such as
1602 datatype 'a tree = Atom 'a | Branch "nat => 'a tree"
1604 * HOL/record: record_simproc (part of the default simpset) takes care
1605 of selectors applied to updated records; record_split_tac is no longer
1606 part of the default claset; update_defs may now be removed from the
1607 simpset in many cases; COMPATIBILITY: old behavior achieved by
1609 claset_ref () := claset() addSWrapper record_split_wrapper;
1610 Delsimprocs [record_simproc]
1612 * HOL/typedef: fixed type inference for representing set; type
1613 arguments now have to occur explicitly on the rhs as type constraints;
1615 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem
1616 names rather than an ML expression;
1618 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be
1619 supplied later. Program schemes can be defined, such as
1620 "While B C s = (if B s then While B C (C s) else s)"
1621 where the well-founded relation can be chosen after B and C have been given.
1623 * HOL/List: the constructors of type list are now Nil and Cons;
1624 INCOMPATIBILITY: while [] and infix # syntax is still there, of
1625 course, ML tools referring to List.list.op # etc. have to be adapted;
1627 * HOL_quantifiers flag superseded by "HOL" print mode, which is
1628 disabled by default; run isabelle with option -m HOL to get back to
1629 the original Gordon/HOL-style output;
1631 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,
1632 ALL x<=y. P, EX x<y. P, EX x<=y. P;
1634 * HOL basic syntax simplified (more orthogonal): all variants of
1635 All/Ex now support plain / symbolic / HOL notation; plain syntax for
1636 Eps operator is provided as well: "SOME x. P[x]";
1638 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;
1640 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made
1641 thus available for user theories;
1643 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with
1644 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the
1647 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec
1648 several times and then mp;
1653 * the notation <<...>> is now available as a notation for sequences of
1656 * the simplifier is now installed
1658 * the axiom system has been generalized (thanks to Soren Heilmann)
1660 * the classical reasoner now has a default rule database
1665 * new primrec section allows primitive recursive functions to be given
1666 directly (as in HOL) over datatypes and the natural numbers;
1668 * new tactics induct_tac and exhaust_tac for induction (or case
1669 analysis) over datatypes and the natural numbers;
1671 * the datatype declaration of type T now defines the recursor T_rec;
1673 * simplification automatically does freeness reasoning for datatype
1676 * automatic type-inference, with AddTCs command to insert new
1677 type-checking rules;
1679 * datatype introduction rules are now added as Safe Introduction rules
1682 * the syntax "if P then x else y" is now available in addition to
1686 *** Internal programming interfaces ***
1688 * tuned simplifier trace output; new flag debug_simp;
1690 * structures Vartab / Termtab (instances of TableFun) offer efficient
1691 tables indexed by indexname_ord / term_ord (compatible with aconv);
1693 * AxClass.axclass_tac lost the theory argument;
1695 * tuned current_goals_markers semantics: begin / end goal avoids
1696 printing empty lines;
1698 * removed prs and prs_fn hook, which was broken because it did not
1699 include \n in its semantics, forcing writeln to add one
1700 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:
1701 string -> unit if you really want to output text without newline;
1703 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to
1704 plain output, interface builders may have to enable 'isabelle_font'
1705 mode to get Isabelle font glyphs as before;
1707 * refined token_translation interface; INCOMPATIBILITY: output length
1708 now of type real instead of int;
1710 * theory loader actions may be traced via new ThyInfo.add_hook
1711 interface (see src/Pure/Thy/thy_info.ML); example application: keep
1712 your own database of information attached to *whole* theories -- as
1713 opposed to intra-theory data slots offered via TheoryDataFun;
1715 * proper handling of dangling sort hypotheses (at last!);
1716 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing
1717 extra sort hypotheses that can be witnessed from the type signature;
1718 the force_strip_shyps flag is gone, any remaining shyps are simply
1719 left in the theorem (with a warning issued by strip_shyps_warning);
1723 New in Isabelle98-1 (October 1998)
1724 ----------------------------------
1726 *** Overview of INCOMPATIBILITIES (see below for more details) ***
1728 * several changes of automated proof tools;
1730 * HOL: major changes to the inductive and datatype packages, including
1731 some minor incompatibilities of theory syntax;
1733 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now
1736 * HOL: removed duplicate thms in Arith:
1737 less_imp_add_less should be replaced by trans_less_add1
1738 le_imp_add_le should be replaced by trans_le_add1
1740 * HOL: unary minus is now overloaded (new type constraints may be
1743 * HOL and ZF: unary minus for integers is now #- instead of #~. In
1744 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is
1745 now taken as an integer constant.
1747 * Pure: ML function 'theory_of' renamed to 'theory';
1753 1. Asm_full_simp_tac is now more aggressive.
1754 1. It will sometimes reorient premises if that increases their power to
1756 2. It does no longer proceed strictly from left to right but may also
1757 rotate premises to achieve further simplification.
1758 For compatibility reasons there is now Asm_lr_simp_tac which is like the
1759 old Asm_full_simp_tac in that it does not rotate premises.
1760 2. The simplifier now knows a little bit about nat-arithmetic.
1762 * Classical reasoner: wrapper mechanism for the classical reasoner now
1763 allows for selected deletion of wrappers, by introduction of names for
1764 wrapper functionals. This implies that addbefore, addSbefore,
1765 addaltern, and addSaltern now take a pair (name, tactic) as argument,
1766 and that adding two tactics with the same name overwrites the first
1767 one (emitting a warning).
1768 type wrapper = (int -> tactic) -> (int -> tactic)
1769 setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by
1770 addWrapper, addSWrapper: claset * (string * wrapper) -> claset
1771 delWrapper, delSWrapper: claset * string -> claset
1772 getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;
1774 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE
1775 semantics; addbefore now affects only the unsafe part of step_tac
1776 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY
1777 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac
1780 * Classical reasoner: setwrapper to setWrapper and compwrapper to
1781 compWrapper; added safe wrapper (and access functions for it);
1783 * HOL/split_all_tac is now much faster and fails if there is nothing
1784 to split. Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order
1785 and the names of the automatically generated variables have changed.
1786 split_all_tac has moved within claset() from unsafe wrappers to safe
1787 wrappers, which means that !!-bound variables are split much more
1788 aggressively, and safe_tac and clarify_tac now split such variables.
1789 If this splitting is not appropriate, use delSWrapper "split_all_tac".
1790 Note: the same holds for record_split_tac, which does the job of
1791 split_all_tac for record fields.
1793 * HOL/Simplifier: Rewrite rules for case distinctions can now be added
1794 permanently to the default simpset using Addsplits just like
1795 Addsimps. They can be removed via Delsplits just like
1796 Delsimps. Lower-case versions are also available.
1798 * HOL/Simplifier: The rule split_if is now part of the default
1799 simpset. This means that the simplifier will eliminate all occurrences
1800 of if-then-else in the conclusion of a goal. To prevent this, you can
1801 either remove split_if completely from the default simpset by
1802 `Delsplits [split_if]' or remove it in a specific call of the
1803 simplifier using `... delsplits [split_if]'. You can also add/delete
1804 other case splitting rules to/from the default simpset: every datatype
1805 generates suitable rules `split_t_case' and `split_t_case_asm' (where
1806 t is the name of the datatype).
1808 * Classical reasoner / Simplifier combination: new force_tac (and
1809 derivatives Force_tac, force) combines rewriting and classical
1810 reasoning (and whatever other tools) similarly to auto_tac, but is
1811 aimed to solve the given subgoal completely.
1816 * new top-level commands `Goal' and `Goalw' that improve upon `goal'
1817 and `goalw': the theory is no longer needed as an explicit argument -
1818 the current theory context is used; assumptions are no longer returned
1819 at the ML-level unless one of them starts with ==> or !!; it is
1820 recommended to convert to these new commands using isatool fixgoal
1821 (backup your sources first!);
1823 * new top-level commands 'thm' and 'thms' for retrieving theorems from
1824 the current theory context, and 'theory' to lookup stored theories;
1826 * new theory section 'locale' for declaring constants, assumptions and
1827 definitions that have local scope;
1829 * new theory section 'nonterminals' for purely syntactic types;
1831 * new theory section 'setup' for generic ML setup functions
1832 (e.g. package initialization);
1834 * the distribution now includes Isabelle icons: see
1835 lib/logo/isabelle-{small,tiny}.xpm;
1837 * isatool install - install binaries with absolute references to
1840 * isatool logo -- create instances of the Isabelle logo (as EPS);
1842 * print mode 'emacs' reserved for Isamode;
1844 * support multiple print (ast) translations per constant name;
1846 * theorems involving oracles are now printed with a suffixed [!];
1851 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');
1853 * HOL/inductive package reorganized and improved: now supports mutual
1859 oddI "n : EVEN ==> Suc n : ODD"
1860 evenI "n : ODD ==> Suc n : EVEN"
1862 new theorem list "elims" contains an elimination rule for each of the
1863 recursive sets; inductive definitions now handle disjunctive premises
1864 correctly (also ZF);
1866 INCOMPATIBILITIES: requires Inductive as an ancestor; component
1867 "mutual_induct" no longer exists - the induction rule is always
1868 contained in "induct";
1871 * HOL/datatype package re-implemented and greatly improved: now
1872 supports mutually recursive datatypes such as
1875 'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)
1876 | SUM ('a aexp) ('a aexp)
1877 | DIFF ('a aexp) ('a aexp)
1880 'a bexp = LESS ('a aexp) ('a aexp)
1881 | AND ('a bexp) ('a bexp)
1882 | OR ('a bexp) ('a bexp)
1884 as well as indirectly recursive datatypes such as
1887 ('a, 'b) term = Var 'a
1888 | App 'b ((('a, 'b) term) list)
1890 The new tactic mutual_induct_tac [<var_1>, ..., <var_n>] i performs
1891 induction on mutually / indirectly recursive datatypes.
1893 Primrec equations are now stored in theory and can be accessed via
1894 <function_name>.simps.
1898 - Theories using datatypes must now have theory Datatype as an
1900 - The specific <typename>.induct_tac no longer exists - use the
1901 generic induct_tac instead.
1902 - natE has been renamed to nat.exhaust - use exhaust_tac
1903 instead of res_inst_tac ... natE. Note that the variable
1904 names in nat.exhaust differ from the names in natE, this
1905 may cause some "fragile" proofs to fail.
1906 - The theorems split_<typename>_case and split_<typename>_case_asm
1907 have been renamed to <typename>.split and <typename>.split_asm.
1908 - Since default sorts of type variables are now handled correctly,
1909 some datatype definitions may have to be annotated with explicit
1911 - Primrec definitions no longer require function name and type
1912 of recursive argument.
1914 Consider using isatool fixdatatype to adapt your theories and proof
1915 scripts to the new package (backup your sources first!).
1918 * HOL/record package: considerably improved implementation; now
1919 includes concrete syntax for record types, terms, updates; theorems
1920 for surjective pairing and splitting !!-bound record variables; proof
1921 support is as follows:
1923 1) standard conversions (selectors or updates applied to record
1924 constructor terms) are part of the standard simpset;
1926 2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are
1927 made part of standard simpset and claset via addIffs;
1929 3) a tactic for record field splitting (record_split_tac) is part of
1930 the standard claset (addSWrapper);
1932 To get a better idea about these rules you may retrieve them via
1933 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is
1934 the name of your record type.
1936 The split tactic 3) conceptually simplifies by the following rule:
1938 "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"
1940 Thus any record variable that is bound by meta-all will automatically
1941 blow up into some record constructor term, consequently the
1942 simplifications of 1), 2) apply. Thus force_tac, auto_tac etc. shall
1943 solve record problems automatically.
1946 * reorganized the main HOL image: HOL/Integ and String loaded by
1947 default; theory Main includes everything;
1949 * automatic simplification of integer sums and comparisons, using cancellation;
1951 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;
1953 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;
1955 * many new identities for unions, intersections, set difference, etc.;
1957 * expand_if, expand_split, expand_sum_case and expand_nat_case are now
1958 called split_if, split_split, split_sum_case and split_nat_case (to go
1959 with add/delsplits);
1961 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting
1962 (?x::unit) = (); this is made part of the default simpset, which COULD
1963 MAKE EXISTING PROOFS FAIL under rare circumstances (consider
1964 'Delsimprocs [unit_eq_proc];' as last resort); also note that
1965 unit_abs_eta_conv is added in order to counter the effect of
1966 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by
1969 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which
1972 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
1973 It and 'sym RS equals0D' are now in the default claset, giving automatic
1974 disjointness reasoning but breaking a few old proofs.
1976 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1
1977 to 'converse' from 'inverse' (for compatibility with ZF and some
1980 * HOL/recdef can now declare non-recursive functions, with {} supplied as
1981 the well-founded relation;
1983 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of
1984 Compl A. The "Compl" syntax remains available as input syntax for this
1987 * HOL/Update: new theory of function updates:
1988 f(a:=b) == %x. if x=a then b else f x
1989 may also be iterated as in f(a:=b,c:=d,...);
1991 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;
1994 - new function list_update written xs[i:=v] that updates the i-th
1995 list position. May also be iterated as in xs[i:=a,j:=b,...].
1996 - new function `upt' written [i..j(] which generates the list
1997 [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper
1998 bound write [i..j], which is a shorthand for [i..j+1(].
1999 - new lexicographic orderings and corresponding wellfoundedness theorems.
2002 - removed 'pred' (predecessor) function;
2003 - generalized some theorems about n-1;
2004 - many new laws about "div" and "mod";
2005 - new laws about greatest common divisors (see theory ex/Primes);
2007 * HOL/Relation: renamed the relational operator r^-1 "converse"
2008 instead of "inverse";
2010 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness
2011 of the multiset ordering;
2013 * directory HOL/Real: a construction of the reals using Dedekind cuts
2014 (not included by default);
2016 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;
2018 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted
2019 programs, i.e. different program variables may have different types.
2021 * calling (stac rew i) now fails if "rew" has no effect on the goal
2022 [previously, this check worked only if the rewrite rule was unconditional]
2023 Now rew can involve either definitions or equalities (either == or =).
2028 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains
2029 only the theorems proved on ZF.ML;
2031 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;
2032 It and 'sym RS equals0D' are now in the default claset, giving automatic
2033 disjointness reasoning but breaking a few old proofs.
2035 * ZF/Update: new theory of function updates
2036 with default rewrite rule f(x:=y) ` z = if(z=x, y, f`z)
2037 may also be iterated as in f(a:=b,c:=d,...);
2039 * in let x=t in u(x), neither t nor u(x) has to be an FOL term.
2041 * calling (stac rew i) now fails if "rew" has no effect on the goal
2042 [previously, this check worked only if the rewrite rule was unconditional]
2043 Now rew can involve either definitions or equalities (either == or =).
2045 * case_tac provided for compatibility with HOL
2046 (like the old excluded_middle_tac, but with subgoals swapped)
2049 *** Internal programming interfaces ***
2051 * Pure: several new basic modules made available for general use, see
2052 also src/Pure/README;
2054 * improved the theory data mechanism to support encapsulation (data
2055 kind name replaced by private Object.kind, acting as authorization
2056 key); new type-safe user interface via functor TheoryDataFun; generic
2057 print_data function becomes basically useless;
2059 * removed global_names compatibility flag -- all theory declarations
2060 are qualified by default;
2062 * module Pure/Syntax now offers quote / antiquote translation
2063 functions (useful for Hoare logic etc. with implicit dependencies);
2064 see HOL/ex/Antiquote for an example use;
2066 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->
2069 * new tactical CHANGED_GOAL for checking that a tactic modifies a
2072 * Display.print_goals function moved to Locale.print_goals;
2074 * standard print function for goals supports current_goals_markers
2075 variable for marking begin of proof, end of proof, start of goal; the
2076 default is ("", "", ""); setting current_goals_markers := ("<proof>",
2077 "</proof>", "<goal>") causes SGML like tagged proof state printing,
2082 New in Isabelle98 (January 1998)
2083 --------------------------------
2085 *** Overview of INCOMPATIBILITIES (see below for more details) ***
2087 * changed lexical syntax of terms / types: dots made part of long
2088 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";
2090 * simpset (and claset) reference variable replaced by functions
2091 simpset / simpset_ref;
2093 * no longer supports theory aliases (via merge) and non-trivial
2094 implicit merge of thms' signatures;
2096 * most internal names of constants changed due to qualified names;
2098 * changed Pure/Sequence interface (see Pure/seq.ML);
2101 *** General Changes ***
2103 * hierachically structured name spaces (for consts, types, axms, thms
2104 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of
2105 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:
2106 isatool fixdots ensures space after dots (e.g. "%x. x"); set
2107 long_names for fully qualified output names; NOTE: ML programs
2108 (special tactics, packages etc.) referring to internal names may have
2109 to be adapted to cope with fully qualified names; in case of severe
2110 backward campatibility problems try setting 'global_names' at compile
2111 time to have enrything declared within a flat name space; one may also
2112 fine tune name declarations in theories via the 'global' and 'local'
2115 * reimplemented the implicit simpset and claset using the new anytype
2116 data filed in signatures; references simpset:simpset ref etc. are
2117 replaced by functions simpset:unit->simpset and
2118 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp
2119 to patch your ML files accordingly;
2121 * HTML output now includes theory graph data for display with Java
2122 applet or isatool browser; data generated automatically via isatool
2123 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);
2125 * defs may now be conditional; improved rewrite_goals_tac to handle
2126 conditional equations;
2128 * defs now admits additional type arguments, using TYPE('a) syntax;
2130 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always
2131 creates a new theory node; implicit merge of thms' signatures is
2132 restricted to 'trivial' ones; COMPATIBILITY: one may have to use
2133 transfer:theory->thm->thm in (rare) cases;
2135 * improved handling of draft signatures / theories; draft thms (and
2136 ctyps, cterms) are automatically promoted to real ones;
2138 * slightly changed interfaces for oracles: admit many per theory, named
2139 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;
2141 * print_goals: optional output of const types (set show_consts and
2144 * improved output of warnings (###) and errors (***);
2146 * subgoal_tac displays a warning if the new subgoal has type variables;
2148 * removed old README and Makefiles;
2150 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;
2152 * removed obsolete init_pps and init_database;
2154 * deleted the obsolete tactical STATE, which was declared by
2155 fun STATE tacfun st = tacfun st st;
2157 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~
2158 (which abbreviates $HOME);
2160 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:
2161 use isatool fixseq to adapt your ML programs (this works for fully
2162 qualified references to the Sequence structure only!);
2164 * use_thy no longer requires writable current directory; it always
2165 reloads .ML *and* .thy file, if either one is out of date;
2168 *** Classical Reasoner ***
2170 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new
2171 tactics that use classical reasoning to simplify a subgoal without
2172 splitting it into several subgoals;
2174 * Safe_tac: like safe_tac but uses the default claset;
2179 * added simplification meta rules:
2180 (asm_)(full_)simplify: simpset -> thm -> thm;
2182 * simplifier.ML no longer part of Pure -- has to be loaded by object
2185 * added prems argument to simplification procedures;
2187 * HOL, FOL, ZF: added infix function `addsplits':
2188 instead of `<simpset> setloop (split_tac <thms>)'
2189 you can simply write `<simpset> addsplits <thms>'
2194 * TYPE('a) syntax for type reflection terms;
2196 * no longer handles consts with name "" -- declare as 'syntax' instead;
2198 * pretty printer: changed order of mixfix annotation preference (again!);
2200 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;
2205 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.
2206 with `addloop' of the simplifier to faciliate case splitting in premises.
2208 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;
2210 * HOL/Auth: new protocol proofs including some for the Internet
2213 * HOL/Map: new theory of `maps' a la VDM;
2215 * HOL/simplifier: simplification procedures nat_cancel_sums for
2216 cancelling out common nat summands from =, <, <= (in)equalities, or
2217 differences; simplification procedures nat_cancel_factor for
2218 cancelling common factor from =, <, <= (in)equalities over natural
2219 sums; nat_cancel contains both kinds of procedures, it is installed by
2220 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;
2222 * HOL/simplifier: terms of the form
2223 `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)' (or t=x)
2225 `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',
2226 and those of the form
2227 `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)' (or t=x)
2229 `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',
2232 Each datatype `t' now comes with a theorem `split_t_case' of the form
2234 P(t_case f1 ... fn x) =
2235 ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &
2237 (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))
2240 and a theorem `split_t_case_asm' of the form
2242 P(t_case f1 ... fn x) =
2243 ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |
2245 (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))
2247 which can be added to a simpset via `addsplits'. The existing theorems
2248 expand_list_case and expand_option_case have been renamed to
2249 split_list_case and split_option_case.
2252 - `pred n' is automatically converted to `n-1'.
2253 Users are strongly encouraged not to use `pred' any longer,
2254 because it will disappear altogether at some point.
2255 - Users are strongly encouraged to write "0 < n" rather than
2256 "n ~= 0". Theorems and proof tools have been modified towards this
2260 the function "set_of_list" has been renamed "set" (and its theorems too);
2261 the function "nth" now takes its arguments in the reverse order and
2262 has acquired the infix notation "!" as in "xs!n".
2264 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};
2266 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its
2267 specialist theorems (like UN1_I) are gone. Similarly for (INT x.B x);
2269 * HOL/record: extensible records with schematic structural subtyping
2270 (single inheritance); EXPERIMENTAL version demonstrating the encoding,
2271 still lacks various theorems and concrete record syntax;
2276 * removed "axioms" and "generated by" sections;
2278 * replaced "ops" section by extended "consts" section, which is capable of
2279 handling the continuous function space "->" directly;
2282 . proves theorems immediately and stores them in the theory,
2283 . creates hierachical name space,
2284 . now uses normal mixfix annotations (instead of cinfix...),
2285 . minor changes to some names and values (for consistency),
2286 . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,
2287 . separator between mutual domain defs: changed "," to "and",
2288 . improved handling of sort constraints; now they have to
2289 appear on the left-hand side of the equations only;
2291 * fixed LAM <x,y,zs>.b syntax;
2293 * added extended adm_tac to simplifier in HOLCF -- can now discharge
2294 adm (%x. P (t x)), where P is chainfinite and t continuous;
2299 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.
2300 with `addloop' of the simplifier to faciliate case splitting in premises.
2302 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as
2303 in HOL, they strip ALL and --> from proved theorems;
2307 New in Isabelle94-8 (May 1997)
2308 ------------------------------
2310 *** General Changes ***
2312 * new utilities to build / run / maintain Isabelle etc. (in parts
2313 still somewhat experimental); old Makefiles etc. still functional;
2315 * new 'Isabelle System Manual';
2317 * INSTALL text, together with ./configure and ./build scripts;
2319 * reimplemented type inference for greater efficiency, better error
2320 messages and clean internal interface;
2322 * prlim command for dealing with lots of subgoals (an easier way of
2323 setting goals_limit);
2328 * supports alternative (named) syntax tables (parser and pretty
2329 printer); internal interface is provided by add_modesyntax(_i);
2331 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to
2332 be used in conjunction with the Isabelle symbol font; uses the
2333 "symbols" syntax table;
2335 * added token_translation interface (may translate name tokens in
2336 arbitrary ways, dependent on their type (free, bound, tfree, ...) and
2337 the current print_mode); IMPORTANT: user print translation functions
2338 are responsible for marking newly introduced bounds
2339 (Syntax.mark_boundT);
2341 * token translations for modes "xterm" and "xterm_color" that display
2342 names in bold, underline etc. or colors (which requires a color
2345 * infixes may now be declared with names independent of their syntax;
2347 * added typed_print_translation (like print_translation, but may
2348 access type of constant);
2351 *** Classical Reasoner ***
2353 Blast_tac: a new tactic! It is often more powerful than fast_tac, but has
2354 some limitations. Blast_tac...
2355 + ignores addss, addbefore, addafter; this restriction is intrinsic
2356 + ignores elimination rules that don't have the correct format
2357 (the conclusion MUST be a formula variable)
2358 + ignores types, which can make HOL proofs fail
2359 + rules must not require higher-order unification, e.g. apply_type in ZF
2360 [message "Function Var's argument not a bound variable" relates to this]
2361 + its proof strategy is more general but can actually be slower
2363 * substitution with equality assumptions no longer permutes other
2366 * minor changes in semantics of addafter (now called addaltern); renamed
2367 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper
2368 (and access functions for it);
2370 * improved combination of classical reasoner and simplifier:
2371 + functions for handling clasimpsets
2372 + improvement of addss: now the simplifier is called _after_ the
2374 + safe variant of addss called addSss: uses safe simplifications
2375 _during_ the safe steps. It is more complete as it allows multiple
2376 instantiations of unknowns (e.g. with slow_tac).
2380 * added interface for simplification procedures (functions that
2381 produce *proven* rewrite rules on the fly, depending on current
2384 * ordering on terms as parameter (used for ordered rewriting);
2386 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;
2388 * the solver is now split into a safe and an unsafe part.
2389 This should be invisible for the normal user, except that the
2390 functions setsolver and addsolver have been renamed to setSolver and
2391 addSolver; added safe_asm_full_simp_tac;
2396 * a generic induction tactic `induct_tac' which works for all datatypes and
2397 also for type `nat';
2399 * a generic case distinction tactic `exhaust_tac' which works for all
2400 datatypes and also for type `nat';
2402 * each datatype comes with a function `size';
2404 * patterns in case expressions allow tuple patterns as arguments to
2405 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';
2407 * primrec now also works with type nat;
2409 * recdef: a new declaration form, allows general recursive functions to be
2410 defined in theory files. See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.
2412 * the constant for negation has been renamed from "not" to "Not" to
2413 harmonize with FOL, ZF, LK, etc.;
2415 * HOL/ex/LFilter theory of a corecursive "filter" functional for
2418 * HOL/Modelcheck demonstrates invocation of model checker oracle;
2420 * HOL/ex/Ring.thy declares cring_simp, which solves equational
2421 problems in commutative rings, using axiomatic type classes for + and *;
2423 * more examples in HOL/MiniML and HOL/Auth;
2425 * more default rewrite rules for quantifiers, union/intersection;
2427 * a new constant `arbitrary == @x.False';
2429 * HOLCF/IOA replaces old HOL/IOA;
2431 * HOLCF changes: derived all rules and arities
2432 + axiomatic type classes instead of classes
2433 + typedef instead of faking type definitions
2434 + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.
2435 + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po
2436 + eliminated the types void, one, tr
2437 + use unit lift and bool lift (with translations) instead of one and tr
2438 + eliminated blift from Lift3.thy (use Def instead of blift)
2439 all eliminated rules are derived as theorems --> no visible changes ;
2444 * ZF now has Fast_tac, Simp_tac and Auto_tac. Union_iff is a now a default
2445 rewrite rule; this may affect some proofs. eq_cs is gone but can be put back
2446 as ZF_cs addSIs [equalityI];
2450 New in Isabelle94-7 (November 96)
2451 ---------------------------------
2453 * allowing negative levels (as offsets) in prlev and choplev;
2455 * super-linear speedup for large simplifications;
2457 * FOL, ZF and HOL now use miniscoping: rewriting pushes
2458 quantifications in as far as possible (COULD MAKE EXISTING PROOFS
2459 FAIL); can suppress it using the command Delsimps (ex_simps @
2460 all_simps); De Morgan laws are also now included, by default;
2462 * improved printing of ==> : ~:
2464 * new object-logic "Sequents" adds linear logic, while replacing LK
2465 and Modal (thanks to Sara Kalvala);
2467 * HOL/Auth: correctness proofs for authentication protocols;
2469 * HOL: new auto_tac combines rewriting and classical reasoning (many
2470 examples on HOL/Auth);
2472 * HOL: new command AddIffs for declaring theorems of the form P=Q to
2473 the rewriter and classical reasoner simultaneously;
2475 * function uresult no longer returns theorems in "standard" format;
2476 regain previous version by: val uresult = standard o uresult;
2483 * oracles -- these establish an interface between Isabelle and trusted
2484 external reasoners, which may deliver results as theorems;
2486 * proof objects (in particular record all uses of oracles);
2488 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;
2490 * "constdefs" section in theory files;
2492 * "primrec" section (HOL) no longer requires names;
2494 * internal type "tactic" now simply "thm -> thm Sequence.seq";
2501 * reduced space requirements;
2503 * automatic HTML generation from theories;
2505 * theory files no longer require "..." (quotes) around most types;
2507 * new examples, including two proofs of the Church-Rosser theorem;
2509 * non-curried (1994) version of HOL is no longer distributed;
2516 * greatly reduced space requirements;
2518 * theory files (.thy) no longer require \...\ escapes at line breaks;
2520 * searchable theorem database (see the section "Retrieving theorems" on
2521 page 8 of the Reference Manual);
2523 * new examples, including Grabczewski's monumental case study of the
2526 * The previous version of HOL renamed to Old_HOL;
2528 * The new version of HOL (previously called CHOL) uses a curried syntax
2529 for functions. Application looks like f a b instead of f(a,b);
2531 * Mutually recursive inductive definitions finally work in HOL;
2533 * In ZF, pattern-matching on tuples is now available in all abstractions and
2534 translates to the operator "split";
2541 * new infix operator, addss, allowing the classical reasoner to
2542 perform simplification at each step of its search. Example:
2543 fast_tac (cs addss ss)
2545 * a new logic, CHOL, the same as HOL, but with a curried syntax
2546 for functions. Application looks like f a b instead of f(a,b). Also pairs
2547 look like (a,b) instead of <a,b>;
2549 * PLEASE NOTE: CHOL will eventually replace HOL!
2551 * In CHOL, pattern-matching on tuples is now available in all abstractions.
2552 It translates to the operator "split". A new theory of integers is available;
2554 * In ZF, integer numerals now denote two's-complement binary integers.
2555 Arithmetic operations can be performed by rewriting. See ZF/ex/Bin.ML;
2557 * Many new examples: I/O automata, Church-Rosser theorem, equivalents
2558 of the Axiom of Choice;
2565 * Significantly faster resolution;
2567 * the different sections in a .thy file can now be mixed and repeated
2570 * Database of theorems for FOL, HOL and ZF. New
2571 commands including qed, qed_goal and bind_thm store theorems in the database.
2573 * Simple database queries: return a named theorem (get_thm) or all theorems of
2574 a given theory (thms_of), or find out what theory a theorem was proved in
2577 * Bugs fixed in the inductive definition and datatype packages;
2579 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs
2580 and HOL_dup_cs obsolete;
2582 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1
2585 * Simpler definition of function space in ZF;
2587 * new results about cardinal and ordinal arithmetic in ZF;
2589 * 'subtype' facility in HOL for introducing new types as subsets of existing