summary |
shortlog |
changelog |
graph |
tags |
branches |
files |
changeset |
file |
revisions |
annotate |
diff |
raw

NEWS

author | wenzelm |

Thu Jan 19 21:22:08 2006 +0100 (2006-01-19 ago) | |

changeset 18708 | 4b3dadb4fe33 |

parent 18696 | 60ca2c749782 |

child 18722 | 0888eca0f1be |

permissions | -rw-r--r-- |

setup: theory -> theory;

1 Isabelle NEWS -- history user-relevant changes

2 ==============================================

4 New in this Isabelle release

5 ----------------------------

7 *** General ***

9 * Theory syntax: the header format ``theory A = B + C:'' has been

10 discontinued in favour of ``theory A imports B C begin''. Use isatool

11 fixheaders to convert existing theory files. INCOMPATIBILITY.

13 * Theory syntax: the old non-Isar theory file format has been

14 discontinued altogether. Note that ML proof scripts may still be used

15 with Isar theories; migration is usually quite simple with the ML

16 function use_legacy_bindings. INCOMPATIBILITY.

18 * Legacy goal package: reduced interface to the bare minimum required

19 to keep existing proof scripts running. Most other user-level

20 functions are now part of the OldGoals structure, which is *not* open

21 by default (consider isatool expandshort before open OldGoals).

22 Removed top_sg, prin, printyp, pprint_term/typ altogether, because

23 these tend to cause confusion about the actual goal (!) context being

24 used here, which is not necessarily the same as the_context().

26 * Command 'find_theorems': support "*" wildcard in "name:" criterion.

29 *** Document preparation ***

31 * Added antiquotations @{ML_type text} and @{ML_struct text} which

32 check the given source text as ML type/structure, printing verbatim.

35 *** Pure ***

37 * Isar: improper proof element 'guess' is like 'obtain', but derives

38 the obtained context from the course of reasoning! For example:

40 assume "EX x y. A x & B y" -- "any previous fact"

41 then guess x and y by clarify

43 This technique is potentially adventurous, depending on the facts and

44 proof tools being involved here.

46 * Isar: known facts from the proof context may be specified as literal

47 propositions, using ASCII back-quote syntax. This works wherever

48 named facts used to be allowed so far, in proof commands, proof

49 methods, attributes etc. Literal facts are retrieved from the context

50 according to unification of type and term parameters. For example,

51 provided that "A" and "A ==> B" and "!!x. P x ==> Q x" are known

52 theorems in the current context, then these are valid literal facts:

53 `A` and `A ==> B` and `!!x. P x ==> Q x" as well as `P a ==> Q a` etc.

55 There is also a proof method "fact" which does the same composition

56 for explicit goal states, e.g. the following proof texts coincide with

57 certain special cases of literal facts:

59 have "A" by fact == note `A`

60 have "A ==> B" by fact == note `A ==> B`

61 have "!!x. P x ==> Q x" by fact == note `!!x. P x ==> Q x`

62 have "P a ==> Q a" by fact == note `P a ==> Q a`

64 * Isar: 'def' now admits simultaneous definitions, e.g.:

66 def x == "t" and y == "u"

68 * Isar: added command 'unfolding', which is structurally similar to

69 'using', but affects both the goal state and facts by unfolding given

70 meta-level equations. Thus many occurrences of the 'unfold' method or

71 'unfolded' attribute may be replaced by first-class proof text.

73 * Provers/induct: improved internal context management to support

74 local fixes and defines on-the-fly. Thus explicit meta-level

75 connectives !! and ==> are rarely required anymore in inductive goals

76 (using object-logic connectives for this purpose has been long

77 obsolete anyway). The subsequent proof patterns illustrate advanced

78 techniques of natural induction; general datatypes and inductive sets

79 work analogously (see also src/HOL/Lambda for realistic examples).

81 (1) This is how to ``strengthen'' an inductive goal wrt. certain

82 parameters:

84 lemma

85 fixes n :: nat and x :: 'a

86 assumes a: "A n x"

87 shows "P n x"

88 using a -- {* make induct insert fact a *}

89 proof (induct n fixing: x) -- {* generalize goal to "!!x. A n x ==> P n x" *}

90 case 0

91 show ?case sorry

92 next

93 case (Suc n)

94 note `!!x. A n x ==> P n x` -- {* induction hypothesis, according to induction rule *}

95 note `A (Suc n) x` -- {* induction premise, stemming from fact a *}

96 show ?case sorry

97 qed

99 (2) This is how to perform induction over ``expressions of a certain

100 form'', using a locally defined inductive parameter n == "a x"

101 together with strengthening (the latter is usually required to get

102 sufficiently flexible induction hypotheses):

104 lemma

105 fixes a :: "'a => nat"

106 assumes a: "A (a x)"

107 shows "P (a x)"

108 using a

109 proof (induct n == "a x" fixing: x)

110 ...

112 See also HOL/Isar_examples/Puzzle.thy for an application of the this

113 particular technique.

115 (3) This is how to perform existential reasoning ('obtain') by

116 induction, while avoiding explicit object-logic encodings:

118 fix n :: nat

119 obtain x :: 'a where "P n x" and "Q n x"

120 proof (induct n fixing: thesis)

121 case 0

122 obtain x where "P 0 x" and "Q 0 x" sorry

123 then show thesis by (rule 0)

124 next

125 case (Suc n)

126 obtain x where "P n x" and "Q n x" by (rule Suc.hyps)

127 obtain x where "P (Suc n) x" and "Q (Suc n) x" sorry

128 then show thesis by (rule Suc.prems)

129 qed

131 Here the 'fixing: thesis' specification essentially modifies the scope

132 of the formal thesis parameter, in order to the get the whole

133 existence statement through the induction as expected.

135 * Provers/induct: mutual induction rules are now specified as a list

136 of rule sharing the same induction cases. HOL packages usually

137 provide foo_bar.inducts for mutually defined items foo and bar

138 (e.g. inductive sets or datatypes). INCOMPATIBILITY, users need to

139 specify mutual induction rules differently, i.e. like this:

141 (induct rule: foo_bar.inducts)

142 (induct set: foo bar)

143 (induct type: foo bar)

145 The ML function ProjectRule.projections turns old-style rules into the

146 new format.

148 * Provers/induct: improved handling of simultaneous goals. Instead of

149 introducing object-level conjunction, the statement is now split into

150 several conclusions, while the corresponding symbolic cases are

151 nested accordingly. INCOMPATIBILITY, proofs need to be structured

152 explicitly. For example:

154 lemma

155 fixes n :: nat

156 shows "P n" and "Q n"

157 proof (induct n)

158 case 0 case 1

159 show "P 0" sorry

160 next

161 case 0 case 2

162 show "Q 0" sorry

163 next

164 case (Suc n) case 1

165 note `P n` and `Q n`

166 show "P (Suc n)" sorry

167 next

168 case (Suc n) case 2

169 note `P n` and `Q n`

170 show "Q (Suc n)" sorry

171 qed

173 The split into subcases may be deferred as follows -- this is

174 particularly relevant for goal statements with local premises.

176 lemma

177 fixes n :: nat

178 shows "A n ==> P n" and "B n ==> Q n"

179 proof (induct n)

180 case 0

181 {

182 case 1

183 note `A 0`

184 show "P 0" sorry

185 next

186 case 2

187 note `B 0`

188 show "Q 0" sorry

189 }

190 next

191 case (Suc n)

192 note `A n ==> P n` and `B n ==> Q n`

193 {

194 case 1

195 note `A (Suc n)`

196 show "P (Suc n)" sorry

197 next

198 case 2

199 note `B (Suc n)`

200 show "Q (Suc n)" sorry

201 }

202 qed

204 If simultaneous goals are to be used with mutual rules, the statement

205 needs to be structured carefully as a two-level conjunction, using

206 lists of propositions separated by 'and':

208 lemma

209 shows "a : A ==> P1 a"

210 "a : A ==> P2 a"

211 and "b : B ==> Q1 b"

212 "b : B ==> Q2 b"

213 "b : B ==> Q3 b"

214 proof (induct set: A B)

216 * Provers/induct: support coinduction as well. See

217 src/HOL/Library/Coinductive_List.thy for various examples.

219 * Simplifier: by default the simplifier trace only shows top level rewrites

220 now. That is, trace_simp_depth_limit is set to 1 by default. Thus there is

221 less danger of being flooded by the trace. The trace indicates where parts

222 have been suppressed.

224 * Provers/classical: removed obsolete classical version of elim_format

225 attribute; classical elim/dest rules are now treated uniformly when

226 manipulating the claset.

228 * Provers/classical: stricter checks to ensure that supplied intro,

229 dest and elim rules are well-formed; dest and elim rules must have at

230 least one premise.

232 * Provers/classical: attributes dest/elim/intro take an optional

233 weight argument for the rule (just as the Pure versions). Weights are

234 ignored by automated tools, but determine the search order of single

235 rule steps.

237 * Syntax: input syntax now supports dummy variable binding "%_. b",

238 where the body does not mention the bound variable. Note that dummy

239 patterns implicitly depend on their context of bounds, which makes

240 "{_. _}" match any set comprehension as expected. Potential

241 INCOMPATIBILITY -- parse translations need to cope with syntactic

242 constant "_idtdummy" in the binding position.

244 * Syntax: removed obsolete syntactic constant "_K" and its associated

245 parse translation. INCOMPATIBILITY -- use dummy abstraction instead,

246 for example "A -> B" => "Pi A (%_. B)".

249 *** HOL ***

251 * Prefer ex1I over ex_ex1I in single-step reasoning, e.g. by the

252 'rule' method.

254 * Alternative iff syntax "A <-> B" for equality on bool (with priority

255 25 like -->); output depends on the "iff" print_mode, the default is

256 "A = B" (with priority 50).

258 * "LEAST x:A. P" expands to "LEAST x. x:A & P" (input only).

260 * In the context of the assumption "~(s = t)" the Simplifier rewrites

261 "t = s" to False (by simproc "neq_simproc"). For backward

262 compatibility this can be disabled by ML "reset use_neq_simproc".

264 * Tactics 'sat' and 'satx' reimplemented, several improvements: goals

265 no longer need to be stated as "<prems> ==> False", equivalences (i.e.

266 "=" on type bool) are handled, variable names of the form "lit_<n>"

267 are no longer reserved, significant speedup.

269 * inductive and datatype: provide projections of mutual rules, bundled

270 as foo_bar.inducts;

272 * Library: added theory Coinductive_List of potentially infinite lists

273 as greatest fixed-point.

276 *** ML ***

278 * Pure/library:

280 val burrow: ('a list -> 'b list) -> 'a list list -> 'b list list

281 val fold_burrow: ('a list -> 'c -> 'b list * 'd) -> 'a list list -> 'c -> 'b list list * 'd

283 The semantics of "burrow" is: "take a function with *simulatanously*

284 transforms a list of value, and apply it *simulatanously* to a list of

285 list of values of the appropriate type". Confer this with "map" which

286 would *not* apply its argument function simulatanously but in

287 sequence. "fold_burrow" has an additional context.

289 Both actually avoid the usage of "unflat" since they hide away

290 "unflat" from the user.

292 * Pure/library: functions map2 and fold2 with curried syntax for

293 simultanous mapping and folding:

295 val map2: ('a -> 'b -> 'c) -> 'a list -> 'b list -> 'c list

296 val fold2: ('a -> 'b -> 'c -> 'c) -> 'a list -> 'b list -> 'c -> 'c

298 * Pure/library: indexed lists - some functions in the Isabelle library

299 treating lists over 'a as finite mappings from [0...n] to 'a have been

300 given more convenient names and signatures reminiscent of similar

301 functions for alists, tables, etc:

303 val nth: 'a list -> int -> 'a

304 val nth_update: int * 'a -> 'a list -> 'a list

305 val nth_map: int -> ('a -> 'a) -> 'a list -> 'a list

306 val fold_index: (int * 'a -> 'b -> 'b) -> 'a list -> 'b -> 'b

308 Note that fold_index starts counting at index 0, not 1 like foldln

309 used to.

311 * Pure/General: name_mangler.ML provides a functor for generic name

312 mangling (bijective mapping from any expression values to strings).

314 * Pure/General: rat.ML implements rational numbers.

316 * Pure: datatype Context.generic joins theory/Proof.context and

317 provides some facilities for code that works in either kind of

318 context, notably GenericDataFun for uniform theory and proof data.

320 * Pure: type Context.generic attribute is now the preferred

321 representation for global vs. local attributes while avoiding code

322 duplication; Attrib.theory/context/generic converts between different

323 types of attributes. Various Pure/HOL/ZF packages work with generic

324 attributes now, INCOMPATIBILITY.

326 * Pure: several functions of signature "... -> theory -> theory * ..."

327 have been reoriented to "... -> theory -> ... * theory" in order to

328 allow natural usage in combination with the ||>, ||>>, |-> and

329 fold_map combinators.

331 * Pure: primitive rule lift_rule now takes goal cterm instead of an

332 actual goal state (thm). Use Thm.lift_rule (Thm.cprem_of st i) to

333 achieve the old behaviour.

335 * Pure: the "Goal" constant is now called "prop", supporting a

336 slightly more general idea of ``protecting'' meta-level rule

337 statements.

339 * Pure: structure Goal provides simple interfaces for

340 init/conclude/finish and tactical prove operations (replacing former

341 Tactic.prove). Note that OldGoals.prove_goalw_cterm has long been

342 obsolete, it is ill-behaved in a local proof context (e.g. with local

343 fixes/assumes or in a locale context).

345 * ML/Isar: simplified treatment of user-level errors, using exception

346 ERROR of string uniformly. Function error now merely raises ERROR,

347 without any side effect on output channels. The Isar toplevel takes

348 care of proper display of ERROR exceptions. ML code may use plain

349 handle/can/try; cat_error may be used to concatenate errors like this:

351 ... handle ERROR msg => cat_error msg "..."

353 Toplevel ML code (run directly or through the Isar toplevel) may be

354 embedded into the Isar toplevel with exception display/debug like

355 this:

357 Isar.toplevel (fn () => ...)

359 INCOMPATIBILITY, removed special transform_error facilities, removed

360 obsolete variants of user-level exceptions (ERROR_MESSAGE,

361 Context.PROOF, ProofContext.CONTEXT, Proof.STATE, ProofHistory.FAIL)

362 -- use plain ERROR instead.

364 * Pure/Isar: Toplevel.theory_to_proof admits transactions that modify

365 the theory before entering a proof state. Transactions now always see

366 a quasi-functional intermediate checkpoint, both in interactive and

367 batch mode.

369 * Simplifier: the simpset of a running simplification process now

370 contains a proof context (cf. Simplifier.the_context), which is the

371 very context that the initial simpset has been retrieved from (by

372 simpset_of/local_simpset_of). Consequently, all plug-in components

373 (solver, looper etc.) may depend on arbitrary proof data.

375 * Simplifier.inherit_context inherits the proof context (plus the

376 local bounds) of the current simplification process; any simproc

377 etc. that calls the Simplifier recursively should do this! Removed

378 former Simplifier.inherit_bounds, which is already included here --

379 INCOMPATIBILITY. Tools based on low-level rewriting may even have to

380 specify an explicit context using Simplifier.context/theory_context.

382 * Simplifier/Classical Reasoner: more abstract interfaces

383 change_simpset/claset for modifying the simpset/claset reference of a

384 theory; raw versions simpset/claset_ref etc. have been discontinued --

385 INCOMPATIBILITY.

387 * Provers: more generic wrt. syntax of object-logics, avoid hardwired

388 "Trueprop" etc.

392 New in Isabelle2005 (October 2005)

393 ----------------------------------

395 *** General ***

397 * Theory headers: the new header syntax for Isar theories is

399 theory <name>

400 imports <theory1> ... <theoryN>

401 uses <file1> ... <fileM>

402 begin

404 where the 'uses' part is optional. The previous syntax

406 theory <name> = <theory1> + ... + <theoryN>:

408 will disappear in the next release. Use isatool fixheaders to convert

409 existing theory files. Note that there is no change in ancient

410 non-Isar theories now, but these will disappear soon.

412 * Theory loader: parent theories can now also be referred to via

413 relative and absolute paths.

415 * Command 'find_theorems' searches for a list of criteria instead of a

416 list of constants. Known criteria are: intro, elim, dest, name:string,

417 simp:term, and any term. Criteria can be preceded by '-' to select

418 theorems that do not match. Intro, elim, dest select theorems that

419 match the current goal, name:s selects theorems whose fully qualified

420 name contain s, and simp:term selects all simplification rules whose

421 lhs match term. Any other term is interpreted as pattern and selects

422 all theorems matching the pattern. Available in ProofGeneral under

423 'ProofGeneral -> Find Theorems' or C-c C-f. Example:

425 C-c C-f (100) "(_::nat) + _ + _" intro -name: "HOL."

427 prints the last 100 theorems matching the pattern "(_::nat) + _ + _",

428 matching the current goal as introduction rule and not having "HOL."

429 in their name (i.e. not being defined in theory HOL).

431 * Command 'thms_containing' has been discontinued in favour of

432 'find_theorems'; INCOMPATIBILITY.

434 * Communication with Proof General is now 8bit clean, which means that

435 Unicode text in UTF-8 encoding may be used within theory texts (both

436 formal and informal parts). Cf. option -U of the Isabelle Proof

437 General interface. Here are some simple examples (cf. src/HOL/ex):

439 http://isabelle.in.tum.de/library/HOL/ex/Hebrew.html

440 http://isabelle.in.tum.de/library/HOL/ex/Chinese.html

442 * Improved efficiency of the Simplifier and, to a lesser degree, the

443 Classical Reasoner. Typical big applications run around 2 times

444 faster.

447 *** Document preparation ***

449 * Commands 'display_drafts' and 'print_drafts' perform simple output

450 of raw sources. Only those symbols that do not require additional

451 LaTeX packages (depending on comments in isabellesym.sty) are

452 displayed properly, everything else is left verbatim. isatool display

453 and isatool print are used as front ends (these are subject to the

454 DVI/PDF_VIEWER and PRINT_COMMAND settings, respectively).

456 * Command tags control specific markup of certain regions of text,

457 notably folding and hiding. Predefined tags include "theory" (for

458 theory begin and end), "proof" for proof commands, and "ML" for

459 commands involving ML code; the additional tags "visible" and

460 "invisible" are unused by default. Users may give explicit tag

461 specifications in the text, e.g. ''by %invisible (auto)''. The

462 interpretation of tags is determined by the LaTeX job during document

463 preparation: see option -V of isatool usedir, or options -n and -t of

464 isatool document, or even the LaTeX macros \isakeeptag, \isafoldtag,

465 \isadroptag.

467 Several document versions may be produced at the same time via isatool

468 usedir (the generated index.html will link all of them). Typical

469 specifications include ''-V document=theory,proof,ML'' to present

470 theory/proof/ML parts faithfully, ''-V outline=/proof,/ML'' to fold

471 proof and ML commands, and ''-V mutilated=-theory,-proof,-ML'' to omit

472 these parts without any formal replacement text. The Isabelle site

473 default settings produce ''document'' and ''outline'' versions as

474 specified above.

476 * Several new antiquotations:

478 @{term_type term} prints a term with its type annotated;

480 @{typeof term} prints the type of a term;

482 @{const const} is the same as @{term const}, but checks that the

483 argument is a known logical constant;

485 @{term_style style term} and @{thm_style style thm} print a term or

486 theorem applying a "style" to it

488 @{ML text}

490 Predefined styles are 'lhs' and 'rhs' printing the lhs/rhs of

491 definitions, equations, inequations etc., 'concl' printing only the

492 conclusion of a meta-logical statement theorem, and 'prem1' .. 'prem19'

493 to print the specified premise. TermStyle.add_style provides an ML

494 interface for introducing further styles. See also the "LaTeX Sugar"

495 document practical applications. The ML antiquotation prints

496 type-checked ML expressions verbatim.

498 * Markup commands 'chapter', 'section', 'subsection', 'subsubsection',

499 and 'text' support optional locale specification '(in loc)', which

500 specifies the default context for interpreting antiquotations. For

501 example: 'text (in lattice) {* @{thm inf_assoc}*}'.

503 * Option 'locale=NAME' of antiquotations specifies an alternative

504 context interpreting the subsequent argument. For example: @{thm

505 [locale=lattice] inf_assoc}.

507 * Proper output of proof terms (@{prf ...} and @{full_prf ...}) within

508 a proof context.

510 * Proper output of antiquotations for theory commands involving a

511 proof context (such as 'locale' or 'theorem (in loc) ...').

513 * Delimiters of outer tokens (string etc.) now produce separate LaTeX

514 macros (\isachardoublequoteopen, isachardoublequoteclose etc.).

516 * isatool usedir: new option -C (default true) controls whether option

517 -D should include a copy of the original document directory; -C false

518 prevents unwanted effects such as copying of administrative CVS data.

521 *** Pure ***

523 * Considerably improved version of 'constdefs' command. Now performs

524 automatic type-inference of declared constants; additional support for

525 local structure declarations (cf. locales and HOL records), see also

526 isar-ref manual. Potential INCOMPATIBILITY: need to observe strictly

527 sequential dependencies of definitions within a single 'constdefs'

528 section; moreover, the declared name needs to be an identifier. If

529 all fails, consider to fall back on 'consts' and 'defs' separately.

531 * Improved indexed syntax and implicit structures. First of all,

532 indexed syntax provides a notational device for subscripted

533 application, using the new syntax \<^bsub>term\<^esub> for arbitrary

534 expressions. Secondly, in a local context with structure

535 declarations, number indexes \<^sub>n or the empty index (default

536 number 1) refer to a certain fixed variable implicitly; option

537 show_structs controls printing of implicit structures. Typical

538 applications of these concepts involve record types and locales.

540 * New command 'no_syntax' removes grammar declarations (and

541 translations) resulting from the given syntax specification, which is

542 interpreted in the same manner as for the 'syntax' command.

544 * 'Advanced' translation functions (parse_translation etc.) may depend

545 on the signature of the theory context being presently used for

546 parsing/printing, see also isar-ref manual.

548 * Improved 'oracle' command provides a type-safe interface to turn an

549 ML expression of type theory -> T -> term into a primitive rule of

550 type theory -> T -> thm (i.e. the functionality of Thm.invoke_oracle

551 is already included here); see also FOL/ex/IffExample.thy;

552 INCOMPATIBILITY.

554 * axclass: name space prefix for class "c" is now "c_class" (was "c"

555 before); "cI" is no longer bound, use "c.intro" instead.

556 INCOMPATIBILITY. This change avoids clashes of fact bindings for

557 axclasses vs. locales.

559 * Improved internal renaming of symbolic identifiers -- attach primes

560 instead of base 26 numbers.

562 * New flag show_question_marks controls printing of leading question

563 marks in schematic variable names.

565 * In schematic variable names, *any* symbol following \<^isub> or

566 \<^isup> is now treated as part of the base name. For example, the

567 following works without printing of awkward ".0" indexes:

569 lemma "x\<^isub>1 = x\<^isub>2 ==> x\<^isub>2 = x\<^isub>1"

570 by simp

572 * Inner syntax includes (*(*nested*) comments*).

574 * Pretty printer now supports unbreakable blocks, specified in mixfix

575 annotations as "(00...)".

577 * Clear separation of logical types and nonterminals, where the latter

578 may only occur in 'syntax' specifications or type abbreviations.

579 Before that distinction was only partially implemented via type class

580 "logic" vs. "{}". Potential INCOMPATIBILITY in rare cases of improper

581 use of 'types'/'consts' instead of 'nonterminals'/'syntax'. Some very

582 exotic syntax specifications may require further adaption

583 (e.g. Cube/Cube.thy).

585 * Removed obsolete type class "logic", use the top sort {} instead.

586 Note that non-logical types should be declared as 'nonterminals'

587 rather than 'types'. INCOMPATIBILITY for new object-logic

588 specifications.

590 * Attributes 'induct' and 'cases': type or set names may now be

591 locally fixed variables as well.

593 * Simplifier: can now control the depth to which conditional rewriting

594 is traced via the PG menu Isabelle -> Settings -> Trace Simp Depth

595 Limit.

597 * Simplifier: simplification procedures may now take the current

598 simpset into account (cf. Simplifier.simproc(_i) / mk_simproc

599 interface), which is very useful for calling the Simplifier

600 recursively. Minor INCOMPATIBILITY: the 'prems' argument of simprocs

601 is gone -- use prems_of_ss on the simpset instead. Moreover, the

602 low-level mk_simproc no longer applies Logic.varify internally, to

603 allow for use in a context of fixed variables.

605 * thin_tac now works even if the assumption being deleted contains !!

606 or ==>. More generally, erule now works even if the major premise of

607 the elimination rule contains !! or ==>.

609 * Method 'rules' has been renamed to 'iprover'. INCOMPATIBILITY.

611 * Reorganized bootstrapping of the Pure theories; CPure is now derived

612 from Pure, which contains all common declarations already. Both

613 theories are defined via plain Isabelle/Isar .thy files.

614 INCOMPATIBILITY: elements of CPure (such as the CPure.intro /

615 CPure.elim / CPure.dest attributes) now appear in the Pure name space;

616 use isatool fixcpure to adapt your theory and ML sources.

618 * New syntax 'name(i-j, i-, i, ...)' for referring to specific

619 selections of theorems in named facts via index ranges.

621 * 'print_theorems': in theory mode, really print the difference

622 wrt. the last state (works for interactive theory development only),

623 in proof mode print all local facts (cf. 'print_facts');

625 * 'hide': option '(open)' hides only base names.

627 * More efficient treatment of intermediate checkpoints in interactive

628 theory development.

630 * Code generator is now invoked via code_module (incremental code

631 generation) and code_library (modular code generation, ML structures

632 for each theory). INCOMPATIBILITY: new keywords 'file' and 'contains'

633 must be quoted when used as identifiers.

635 * New 'value' command for reading, evaluating and printing terms using

636 the code generator. INCOMPATIBILITY: command keyword 'value' must be

637 quoted when used as identifier.

640 *** Locales ***

642 * New commands for the interpretation of locale expressions in

643 theories (1), locales (2) and proof contexts (3). These generate

644 proof obligations from the expression specification. After the

645 obligations have been discharged, theorems of the expression are added

646 to the theory, target locale or proof context. The synopsis of the

647 commands is a follows:

649 (1) interpretation expr inst

650 (2) interpretation target < expr

651 (3) interpret expr inst

653 Interpretation in theories and proof contexts require a parameter

654 instantiation of terms from the current context. This is applied to

655 specifications and theorems of the interpreted expression.

656 Interpretation in locales only permits parameter renaming through the

657 locale expression. Interpretation is smart in that interpretations

658 that are active already do not occur in proof obligations, neither are

659 instantiated theorems stored in duplicate. Use 'print_interps' to

660 inspect active interpretations of a particular locale. For details,

661 see the Isar Reference manual. Examples can be found in

662 HOL/Finite_Set.thy and HOL/Algebra/UnivPoly.thy.

664 INCOMPATIBILITY: former 'instantiate' has been withdrawn, use

665 'interpret' instead.

667 * New context element 'constrains' for adding type constraints to

668 parameters.

670 * Context expressions: renaming of parameters with syntax

671 redeclaration.

673 * Locale declaration: 'includes' disallowed.

675 * Proper static binding of attribute syntax -- i.e. types / terms /

676 facts mentioned as arguments are always those of the locale definition

677 context, independently of the context of later invocations. Moreover,

678 locale operations (renaming and type / term instantiation) are applied

679 to attribute arguments as expected.

681 INCOMPATIBILITY of the ML interface: always pass Attrib.src instead of

682 actual attributes; rare situations may require Attrib.attribute to

683 embed those attributes into Attrib.src that lack concrete syntax.

684 Attribute implementations need to cooperate properly with the static

685 binding mechanism. Basic parsers Args.XXX_typ/term/prop and

686 Attrib.XXX_thm etc. already do the right thing without further

687 intervention. Only unusual applications -- such as "where" or "of"

688 (cf. src/Pure/Isar/attrib.ML), which process arguments depending both

689 on the context and the facts involved -- may have to assign parsed

690 values to argument tokens explicitly.

692 * Changed parameter management in theorem generation for long goal

693 statements with 'includes'. INCOMPATIBILITY: produces a different

694 theorem statement in rare situations.

696 * Locale inspection command 'print_locale' omits notes elements. Use

697 'print_locale!' to have them included in the output.

700 *** Provers ***

702 * Provers/hypsubst.ML: improved version of the subst method, for

703 single-step rewriting: it now works in bound variable contexts. New is

704 'subst (asm)', for rewriting an assumption. INCOMPATIBILITY: may

705 rewrite a different subterm than the original subst method, which is

706 still available as 'simplesubst'.

708 * Provers/quasi.ML: new transitivity reasoners for transitivity only

709 and quasi orders.

711 * Provers/trancl.ML: new transitivity reasoner for transitive and

712 reflexive-transitive closure of relations.

714 * Provers/blast.ML: new reference depth_limit to make blast's depth

715 limit (previously hard-coded with a value of 20) user-definable.

717 * Provers/simplifier.ML has been moved to Pure, where Simplifier.setup

718 is peformed already. Object-logics merely need to finish their

719 initial simpset configuration as before. INCOMPATIBILITY.

722 *** HOL ***

724 * Symbolic syntax of Hilbert Choice Operator is now as follows:

726 syntax (epsilon)

727 "_Eps" :: "[pttrn, bool] => 'a" ("(3\<some>_./ _)" [0, 10] 10)

729 The symbol \<some> is displayed as the alternative epsilon of LaTeX

730 and x-symbol; use option '-m epsilon' to get it actually printed.

731 Moreover, the mathematically important symbolic identifier \<epsilon>

732 becomes available as variable, constant etc. INCOMPATIBILITY,

734 * "x > y" abbreviates "y < x" and "x >= y" abbreviates "y <= x".

735 Similarly for all quantifiers: "ALL x > y" etc. The x-symbol for >=

736 is \<ge>. New transitivity rules have been added to HOL/Orderings.thy to

737 support corresponding Isar calculations.

739 * "{x:A. P}" abbreviates "{x. x:A & P}", and similarly for "\<in>"

740 instead of ":".

742 * theory SetInterval: changed the syntax for open intervals:

744 Old New

745 {..n(} {..<n}

746 {)n..} {n<..}

747 {m..n(} {m..<n}

748 {)m..n} {m<..n}

749 {)m..n(} {m<..<n}

751 The old syntax is still supported but will disappear in the next

752 release. For conversion use the following Emacs search and replace

753 patterns (these are not perfect but work quite well):

755 {)\([^\.]*\)\.\. -> {\1<\.\.}

756 \.\.\([^(}]*\)(} -> \.\.<\1}

758 * Theory Commutative_Ring (in Library): method comm_ring for proving

759 equalities in commutative rings; method 'algebra' provides a generic

760 interface.

762 * Theory Finite_Set: changed the syntax for 'setsum', summation over

763 finite sets: "setsum (%x. e) A", which used to be "\<Sum>x:A. e", is

764 now either "SUM x:A. e" or "\<Sum>x \<in> A. e". The bound variable can

765 be a tuple pattern.

767 Some new syntax forms are available:

769 "\<Sum>x | P. e" for "setsum (%x. e) {x. P}"

770 "\<Sum>x = a..b. e" for "setsum (%x. e) {a..b}"

771 "\<Sum>x = a..<b. e" for "setsum (%x. e) {a..<b}"

772 "\<Sum>x < k. e" for "setsum (%x. e) {..<k}"

774 The latter form "\<Sum>x < k. e" used to be based on a separate

775 function "Summation", which has been discontinued.

777 * theory Finite_Set: in structured induction proofs, the insert case

778 is now 'case (insert x F)' instead of the old counterintuitive 'case

779 (insert F x)'.

781 * The 'refute' command has been extended to support a much larger

782 fragment of HOL, including axiomatic type classes, constdefs and

783 typedefs, inductive datatypes and recursion.

785 * New tactics 'sat' and 'satx' to prove propositional tautologies.

786 Requires zChaff with proof generation to be installed. See

787 HOL/ex/SAT_Examples.thy for examples.

789 * Datatype induction via method 'induct' now preserves the name of the

790 induction variable. For example, when proving P(xs::'a list) by

791 induction on xs, the induction step is now P(xs) ==> P(a#xs) rather

792 than P(list) ==> P(a#list) as previously. Potential INCOMPATIBILITY

793 in unstructured proof scripts.

795 * Reworked implementation of records. Improved scalability for

796 records with many fields, avoiding performance problems for type

797 inference. Records are no longer composed of nested field types, but

798 of nested extension types. Therefore the record type only grows linear

799 in the number of extensions and not in the number of fields. The

800 top-level (users) view on records is preserved. Potential

801 INCOMPATIBILITY only in strange cases, where the theory depends on the

802 old record representation. The type generated for a record is called

803 <record_name>_ext_type.

805 Flag record_quick_and_dirty_sensitive can be enabled to skip the

806 proofs triggered by a record definition or a simproc (if

807 quick_and_dirty is enabled). Definitions of large records can take

808 quite long.

810 New simproc record_upd_simproc for simplification of multiple record

811 updates enabled by default. Moreover, trivial updates are also

812 removed: r(|x := x r|) = r. INCOMPATIBILITY: old proofs break

813 occasionally, since simplification is more powerful by default.

815 * typedef: proper support for polymorphic sets, which contain extra

816 type-variables in the term.

818 * Simplifier: automatically reasons about transitivity chains

819 involving "trancl" (r^+) and "rtrancl" (r^*) by setting up tactics

820 provided by Provers/trancl.ML as additional solvers. INCOMPATIBILITY:

821 old proofs break occasionally as simplification may now solve more

822 goals than previously.

824 * Simplifier: converts x <= y into x = y if assumption y <= x is

825 present. Works for all partial orders (class "order"), in particular

826 numbers and sets. For linear orders (e.g. numbers) it treats ~ x < y

827 just like y <= x.

829 * Simplifier: new simproc for "let x = a in f x". If a is a free or

830 bound variable or a constant then the let is unfolded. Otherwise

831 first a is simplified to b, and then f b is simplified to g. If

832 possible we abstract b from g arriving at "let x = b in h x",

833 otherwise we unfold the let and arrive at g. The simproc can be

834 enabled/disabled by the reference use_let_simproc. Potential

835 INCOMPATIBILITY since simplification is more powerful by default.

837 * Classical reasoning: the meson method now accepts theorems as arguments.

839 * Prover support: pre-release of the Isabelle-ATP linkup, which runs background

840 jobs to provide advice on the provability of subgoals.

842 * Theory OrderedGroup and Ring_and_Field: various additions and

843 improvements to faciliate calculations involving equalities and

844 inequalities.

846 The following theorems have been eliminated or modified

847 (INCOMPATIBILITY):

849 abs_eq now named abs_of_nonneg

850 abs_of_ge_0 now named abs_of_nonneg

851 abs_minus_eq now named abs_of_nonpos

852 imp_abs_id now named abs_of_nonneg

853 imp_abs_neg_id now named abs_of_nonpos

854 mult_pos now named mult_pos_pos

855 mult_pos_le now named mult_nonneg_nonneg

856 mult_pos_neg_le now named mult_nonneg_nonpos

857 mult_pos_neg2_le now named mult_nonneg_nonpos2

858 mult_neg now named mult_neg_neg

859 mult_neg_le now named mult_nonpos_nonpos

861 * Theory Parity: added rules for simplifying exponents.

863 * Theory List:

865 The following theorems have been eliminated or modified

866 (INCOMPATIBILITY):

868 list_all_Nil now named list_all.simps(1)

869 list_all_Cons now named list_all.simps(2)

870 list_all_conv now named list_all_iff

871 set_mem_eq now named mem_iff

873 * Theories SetsAndFunctions and BigO (see HOL/Library) support

874 asymptotic "big O" calculations. See the notes in BigO.thy.

877 *** HOL-Complex ***

879 * Theory RealDef: better support for embedding natural numbers and

880 integers in the reals.

882 The following theorems have been eliminated or modified

883 (INCOMPATIBILITY):

885 exp_ge_add_one_self now requires no hypotheses

886 real_of_int_add reversed direction of equality (use [symmetric])

887 real_of_int_minus reversed direction of equality (use [symmetric])

888 real_of_int_diff reversed direction of equality (use [symmetric])

889 real_of_int_mult reversed direction of equality (use [symmetric])

891 * Theory RComplete: expanded support for floor and ceiling functions.

893 * Theory Ln is new, with properties of the natural logarithm

895 * Hyperreal: There is a new type constructor "star" for making

896 nonstandard types. The old type names are now type synonyms:

898 hypreal = real star

899 hypnat = nat star

900 hcomplex = complex star

902 * Hyperreal: Many groups of similarly-defined constants have been

903 replaced by polymorphic versions (INCOMPATIBILITY):

905 star_of <-- hypreal_of_real, hypnat_of_nat, hcomplex_of_complex

907 starset <-- starsetNat, starsetC

908 *s* <-- *sNat*, *sc*

909 starset_n <-- starsetNat_n, starsetC_n

910 *sn* <-- *sNatn*, *scn*

911 InternalSets <-- InternalNatSets, InternalCSets

913 starfun <-- starfun{Nat,Nat2,C,RC,CR}

914 *f* <-- *fNat*, *fNat2*, *fc*, *fRc*, *fcR*

915 starfun_n <-- starfun{Nat,Nat2,C,RC,CR}_n

916 *fn* <-- *fNatn*, *fNat2n*, *fcn*, *fRcn*, *fcRn*

917 InternalFuns <-- InternalNatFuns, InternalNatFuns2, Internal{C,RC,CR}Funs

919 * Hyperreal: Many type-specific theorems have been removed in favor of

920 theorems specific to various axiomatic type classes (INCOMPATIBILITY):

922 add_commute <-- {hypreal,hypnat,hcomplex}_add_commute

923 add_assoc <-- {hypreal,hypnat,hcomplex}_add_assocs

924 OrderedGroup.add_0 <-- {hypreal,hypnat,hcomplex}_add_zero_left

925 OrderedGroup.add_0_right <-- {hypreal,hcomplex}_add_zero_right

926 right_minus <-- hypreal_add_minus

927 left_minus <-- {hypreal,hcomplex}_add_minus_left

928 mult_commute <-- {hypreal,hypnat,hcomplex}_mult_commute

929 mult_assoc <-- {hypreal,hypnat,hcomplex}_mult_assoc

930 mult_1_left <-- {hypreal,hypnat}_mult_1, hcomplex_mult_one_left

931 mult_1_right <-- hcomplex_mult_one_right

932 mult_zero_left <-- hcomplex_mult_zero_left

933 left_distrib <-- {hypreal,hypnat,hcomplex}_add_mult_distrib

934 right_distrib <-- hypnat_add_mult_distrib2

935 zero_neq_one <-- {hypreal,hypnat,hcomplex}_zero_not_eq_one

936 right_inverse <-- hypreal_mult_inverse

937 left_inverse <-- hypreal_mult_inverse_left, hcomplex_mult_inv_left

938 order_refl <-- {hypreal,hypnat}_le_refl

939 order_trans <-- {hypreal,hypnat}_le_trans

940 order_antisym <-- {hypreal,hypnat}_le_anti_sym

941 order_less_le <-- {hypreal,hypnat}_less_le

942 linorder_linear <-- {hypreal,hypnat}_le_linear

943 add_left_mono <-- {hypreal,hypnat}_add_left_mono

944 mult_strict_left_mono <-- {hypreal,hypnat}_mult_less_mono2

945 add_nonneg_nonneg <-- hypreal_le_add_order

947 * Hyperreal: Separate theorems having to do with type-specific

948 versions of constants have been merged into theorems that apply to the

949 new polymorphic constants (INCOMPATIBILITY):

951 STAR_UNIV_set <-- {STAR_real,NatStar_real,STARC_complex}_set

952 STAR_empty_set <-- {STAR,NatStar,STARC}_empty_set

953 STAR_Un <-- {STAR,NatStar,STARC}_Un

954 STAR_Int <-- {STAR,NatStar,STARC}_Int

955 STAR_Compl <-- {STAR,NatStar,STARC}_Compl

956 STAR_subset <-- {STAR,NatStar,STARC}_subset

957 STAR_mem <-- {STAR,NatStar,STARC}_mem

958 STAR_mem_Compl <-- {STAR,STARC}_mem_Compl

959 STAR_diff <-- {STAR,STARC}_diff

960 STAR_star_of_image_subset <-- {STAR_hypreal_of_real, NatStar_hypreal_of_real,

961 STARC_hcomplex_of_complex}_image_subset

962 starset_n_Un <-- starset{Nat,C}_n_Un

963 starset_n_Int <-- starset{Nat,C}_n_Int

964 starset_n_Compl <-- starset{Nat,C}_n_Compl

965 starset_n_diff <-- starset{Nat,C}_n_diff

966 InternalSets_Un <-- Internal{Nat,C}Sets_Un

967 InternalSets_Int <-- Internal{Nat,C}Sets_Int

968 InternalSets_Compl <-- Internal{Nat,C}Sets_Compl

969 InternalSets_diff <-- Internal{Nat,C}Sets_diff

970 InternalSets_UNIV_diff <-- Internal{Nat,C}Sets_UNIV_diff

971 InternalSets_starset_n <-- Internal{Nat,C}Sets_starset{Nat,C}_n

972 starset_starset_n_eq <-- starset{Nat,C}_starset{Nat,C}_n_eq

973 starset_n_starset <-- starset{Nat,C}_n_starset{Nat,C}

974 starfun_n_starfun <-- starfun{Nat,Nat2,C,RC,CR}_n_starfun{Nat,Nat2,C,RC,CR}

975 starfun <-- starfun{Nat,Nat2,C,RC,CR}

976 starfun_mult <-- starfun{Nat,Nat2,C,RC,CR}_mult

977 starfun_add <-- starfun{Nat,Nat2,C,RC,CR}_add

978 starfun_minus <-- starfun{Nat,Nat2,C,RC,CR}_minus

979 starfun_diff <-- starfun{C,RC,CR}_diff

980 starfun_o <-- starfun{NatNat2,Nat2,_stafunNat,C,C_starfunRC,_starfunCR}_o

981 starfun_o2 <-- starfun{NatNat2,_stafunNat,C,C_starfunRC,_starfunCR}_o2

982 starfun_const_fun <-- starfun{Nat,Nat2,C,RC,CR}_const_fun

983 starfun_inverse <-- starfun{Nat,C,RC,CR}_inverse

984 starfun_eq <-- starfun{Nat,Nat2,C,RC,CR}_eq

985 starfun_eq_iff <-- starfun{C,RC,CR}_eq_iff

986 starfun_Id <-- starfunC_Id

987 starfun_approx <-- starfun{Nat,CR}_approx

988 starfun_capprox <-- starfun{C,RC}_capprox

989 starfun_abs <-- starfunNat_rabs

990 starfun_lambda_cancel <-- starfun{C,CR,RC}_lambda_cancel

991 starfun_lambda_cancel2 <-- starfun{C,CR,RC}_lambda_cancel2

992 starfun_mult_HFinite_approx <-- starfunCR_mult_HFinite_capprox

993 starfun_mult_CFinite_capprox <-- starfun{C,RC}_mult_CFinite_capprox

994 starfun_add_capprox <-- starfun{C,RC}_add_capprox

995 starfun_add_approx <-- starfunCR_add_approx

996 starfun_inverse_inverse <-- starfunC_inverse_inverse

997 starfun_divide <-- starfun{C,CR,RC}_divide

998 starfun_n <-- starfun{Nat,C}_n

999 starfun_n_mult <-- starfun{Nat,C}_n_mult

1000 starfun_n_add <-- starfun{Nat,C}_n_add

1001 starfun_n_add_minus <-- starfunNat_n_add_minus

1002 starfun_n_const_fun <-- starfun{Nat,C}_n_const_fun

1003 starfun_n_minus <-- starfun{Nat,C}_n_minus

1004 starfun_n_eq <-- starfun{Nat,C}_n_eq

1006 star_n_add <-- {hypreal,hypnat,hcomplex}_add

1007 star_n_minus <-- {hypreal,hcomplex}_minus

1008 star_n_diff <-- {hypreal,hcomplex}_diff

1009 star_n_mult <-- {hypreal,hcomplex}_mult

1010 star_n_inverse <-- {hypreal,hcomplex}_inverse

1011 star_n_le <-- {hypreal,hypnat}_le

1012 star_n_less <-- {hypreal,hypnat}_less

1013 star_n_zero_num <-- {hypreal,hypnat,hcomplex}_zero_num

1014 star_n_one_num <-- {hypreal,hypnat,hcomplex}_one_num

1015 star_n_abs <-- hypreal_hrabs

1016 star_n_divide <-- hcomplex_divide

1018 star_of_add <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_add

1019 star_of_minus <-- {hypreal_of_real,hcomplex_of_complex}_minus

1020 star_of_diff <-- hypreal_of_real_diff

1021 star_of_mult <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_mult

1022 star_of_one <-- {hypreal_of_real,hcomplex_of_complex}_one

1023 star_of_zero <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_zero

1024 star_of_le <-- {hypreal_of_real,hypnat_of_nat}_le_iff

1025 star_of_less <-- {hypreal_of_real,hypnat_of_nat}_less_iff

1026 star_of_eq <-- {hypreal_of_real,hypnat_of_nat,hcomplex_of_complex}_eq_iff

1027 star_of_inverse <-- {hypreal_of_real,hcomplex_of_complex}_inverse

1028 star_of_divide <-- {hypreal_of_real,hcomplex_of_complex}_divide

1029 star_of_of_nat <-- {hypreal_of_real,hcomplex_of_complex}_of_nat

1030 star_of_of_int <-- {hypreal_of_real,hcomplex_of_complex}_of_int

1031 star_of_number_of <-- {hypreal,hcomplex}_number_of

1032 star_of_number_less <-- number_of_less_hypreal_of_real_iff

1033 star_of_number_le <-- number_of_le_hypreal_of_real_iff

1034 star_of_eq_number <-- hypreal_of_real_eq_number_of_iff

1035 star_of_less_number <-- hypreal_of_real_less_number_of_iff

1036 star_of_le_number <-- hypreal_of_real_le_number_of_iff

1037 star_of_power <-- hypreal_of_real_power

1038 star_of_eq_0 <-- hcomplex_of_complex_zero_iff

1040 * Hyperreal: new method "transfer" that implements the transfer

1041 principle of nonstandard analysis. With a subgoal that mentions

1042 nonstandard types like "'a star", the command "apply transfer"

1043 replaces it with an equivalent one that mentions only standard types.

1044 To be successful, all free variables must have standard types; non-

1045 standard variables must have explicit universal quantifiers.

1047 * Hyperreal: A theory of Taylor series.

1050 *** HOLCF ***

1052 * Discontinued special version of 'constdefs' (which used to support

1053 continuous functions) in favor of the general Pure one with full

1054 type-inference.

1056 * New simplification procedure for solving continuity conditions; it

1057 is much faster on terms with many nested lambda abstractions (cubic

1058 instead of exponential time).

1060 * New syntax for domain package: selector names are now optional.

1061 Parentheses should be omitted unless argument is lazy, for example:

1063 domain 'a stream = cons "'a" (lazy "'a stream")

1065 * New command 'fixrec' for defining recursive functions with pattern

1066 matching; defining multiple functions with mutual recursion is also

1067 supported. Patterns may include the constants cpair, spair, up, sinl,

1068 sinr, or any data constructor defined by the domain package. The given

1069 equations are proven as rewrite rules. See HOLCF/ex/Fixrec_ex.thy for

1070 syntax and examples.

1072 * New commands 'cpodef' and 'pcpodef' for defining predicate subtypes

1073 of cpo and pcpo types. Syntax is exactly like the 'typedef' command,

1074 but the proof obligation additionally includes an admissibility

1075 requirement. The packages generate instances of class cpo or pcpo,

1076 with continuity and strictness theorems for Rep and Abs.

1078 * HOLCF: Many theorems have been renamed according to a more standard naming

1079 scheme (INCOMPATIBILITY):

1081 foo_inject: "foo$x = foo$y ==> x = y"

1082 foo_eq: "(foo$x = foo$y) = (x = y)"

1083 foo_less: "(foo$x << foo$y) = (x << y)"

1084 foo_strict: "foo$UU = UU"

1085 foo_defined: "... ==> foo$x ~= UU"

1086 foo_defined_iff: "(foo$x = UU) = (x = UU)"

1089 *** ZF ***

1091 * ZF/ex: theories Group and Ring provide examples in abstract algebra,

1092 including the First Isomorphism Theorem (on quotienting by the kernel

1093 of a homomorphism).

1095 * ZF/Simplifier: install second copy of type solver that actually

1096 makes use of TC rules declared to Isar proof contexts (or locales);

1097 the old version is still required for ML proof scripts.

1100 *** Cube ***

1102 * Converted to Isar theory format; use locales instead of axiomatic

1103 theories.

1106 *** ML ***

1108 * Pure/library.ML no longer defines its own option datatype, but uses

1109 that of the SML basis, which has constructors NONE and SOME instead of

1110 None and Some, as well as exception Option.Option instead of OPTION.

1111 The functions the, if_none, is_some, is_none have been adapted

1112 accordingly, while Option.map replaces apsome.

1114 * Pure/library.ML: the exception LIST has been given up in favour of

1115 the standard exceptions Empty and Subscript, as well as

1116 Library.UnequalLengths. Function like Library.hd and Library.tl are

1117 superceded by the standard hd and tl functions etc.

1119 A number of basic list functions are no longer exported to the ML

1120 toplevel, as they are variants of predefined functions. The following

1121 suggests how one can translate existing code:

1123 rev_append xs ys = List.revAppend (xs, ys)

1124 nth_elem (i, xs) = List.nth (xs, i)

1125 last_elem xs = List.last xs

1126 flat xss = List.concat xss

1127 seq fs = List.app fs

1128 partition P xs = List.partition P xs

1129 mapfilter f xs = List.mapPartial f xs

1131 * Pure/library.ML: several combinators for linear functional

1132 transformations, notably reverse application and composition:

1134 x |> f f #> g

1135 (x, y) |-> f f #-> g

1137 * Pure/library.ML: introduced/changed precedence of infix operators:

1139 infix 1 |> |-> ||> ||>> |>> |>>> #> #->;

1140 infix 2 ?;

1141 infix 3 o oo ooo oooo;

1142 infix 4 ~~ upto downto;

1144 Maybe INCOMPATIBILITY when any of those is used in conjunction with other

1145 infix operators.

1147 * Pure/library.ML: natural list combinators fold, fold_rev, and

1148 fold_map support linear functional transformations and nesting. For

1149 example:

1151 fold f [x1, ..., xN] y =

1152 y |> f x1 |> ... |> f xN

1154 (fold o fold) f [xs1, ..., xsN] y =

1155 y |> fold f xs1 |> ... |> fold f xsN

1157 fold f [x1, ..., xN] =

1158 f x1 #> ... #> f xN

1160 (fold o fold) f [xs1, ..., xsN] =

1161 fold f xs1 #> ... #> fold f xsN

1163 * Pure/library.ML: the following selectors on type 'a option are

1164 available:

1166 the: 'a option -> 'a (*partial*)

1167 these: 'a option -> 'a where 'a = 'b list

1168 the_default: 'a -> 'a option -> 'a

1169 the_list: 'a option -> 'a list

1171 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides

1172 basic operations for association lists, following natural argument

1173 order; moreover the explicit equality predicate passed here avoids

1174 potentially expensive polymorphic runtime equality checks.

1175 The old functions may be expressed as follows:

1177 assoc = uncurry (AList.lookup (op =))

1178 assocs = these oo AList.lookup (op =)

1179 overwrite = uncurry (AList.update (op =)) o swap

1181 * Pure/General: structure AList (cf. Pure/General/alist.ML) provides

1183 val make: ('a -> 'b) -> 'a list -> ('a * 'b) list

1184 val find: ('a * 'b -> bool) -> ('c * 'b) list -> 'a -> 'c list

1186 replacing make_keylist and keyfilter (occassionally used)

1187 Naive rewrites:

1189 make_keylist = AList.make

1190 keyfilter = AList.find (op =)

1192 * eq_fst and eq_snd now take explicit equality parameter, thus

1193 avoiding eqtypes. Naive rewrites:

1195 eq_fst = eq_fst (op =)

1196 eq_snd = eq_snd (op =)

1198 * Removed deprecated apl and apr (rarely used).

1199 Naive rewrites:

1201 apl (n, op) =>>= curry op n

1202 apr (op, m) =>>= fn n => op (n, m)

1204 * Pure/General: structure OrdList (cf. Pure/General/ord_list.ML)

1205 provides a reasonably efficient light-weight implementation of sets as

1206 lists.

1208 * Pure/General: generic tables (cf. Pure/General/table.ML) provide a

1209 few new operations; existing lookup and update are now curried to

1210 follow natural argument order (for use with fold etc.);

1211 INCOMPATIBILITY, use (uncurry Symtab.lookup) etc. as last resort.

1213 * Pure/General: output via the Isabelle channels of

1214 writeln/warning/error etc. is now passed through Output.output, with a

1215 hook for arbitrary transformations depending on the print_mode

1216 (cf. Output.add_mode -- the first active mode that provides a output

1217 function wins). Already formatted output may be embedded into further

1218 text via Output.raw; the result of Pretty.string_of/str_of and derived

1219 functions (string_of_term/cterm/thm etc.) is already marked raw to

1220 accommodate easy composition of diagnostic messages etc. Programmers

1221 rarely need to care about Output.output or Output.raw at all, with

1222 some notable exceptions: Output.output is required when bypassing the

1223 standard channels (writeln etc.), or in token translations to produce

1224 properly formatted results; Output.raw is required when capturing

1225 already output material that will eventually be presented to the user

1226 a second time. For the default print mode, both Output.output and

1227 Output.raw have no effect.

1229 * Pure/General: Output.time_accumulator NAME creates an operator ('a

1230 -> 'b) -> 'a -> 'b to measure runtime and count invocations; the

1231 cumulative results are displayed at the end of a batch session.

1233 * Pure/General: File.sysify_path and File.quote_sysify path have been

1234 replaced by File.platform_path and File.shell_path (with appropriate

1235 hooks). This provides a clean interface for unusual systems where the

1236 internal and external process view of file names are different.

1238 * Pure: more efficient orders for basic syntactic entities: added

1239 fast_string_ord, fast_indexname_ord, fast_term_ord; changed sort_ord

1240 and typ_ord to use fast_string_ord and fast_indexname_ord (term_ord is

1241 NOT affected); structures Symtab, Vartab, Typtab, Termtab use the fast

1242 orders now -- potential INCOMPATIBILITY for code that depends on a

1243 particular order for Symtab.keys, Symtab.dest, etc. (consider using

1244 Library.sort_strings on result).

1246 * Pure/term.ML: combinators fold_atyps, fold_aterms, fold_term_types,

1247 fold_types traverse types/terms from left to right, observing natural

1248 argument order. Supercedes previous foldl_XXX versions, add_frees,

1249 add_vars etc. have been adapted as well: INCOMPATIBILITY.

1251 * Pure: name spaces have been refined, with significant changes of the

1252 internal interfaces -- INCOMPATIBILITY. Renamed cond_extern(_table)

1253 to extern(_table). The plain name entry path is superceded by a

1254 general 'naming' context, which also includes the 'policy' to produce

1255 a fully qualified name and external accesses of a fully qualified

1256 name; NameSpace.extend is superceded by context dependent

1257 Sign.declare_name. Several theory and proof context operations modify

1258 the naming context. Especially note Theory.restore_naming and

1259 ProofContext.restore_naming to get back to a sane state; note that

1260 Theory.add_path is no longer sufficient to recover from

1261 Theory.absolute_path in particular.

1263 * Pure: new flags short_names (default false) and unique_names

1264 (default true) for controlling output of qualified names. If

1265 short_names is set, names are printed unqualified. If unique_names is

1266 reset, the name prefix is reduced to the minimum required to achieve

1267 the original result when interning again, even if there is an overlap

1268 with earlier declarations.

1270 * Pure/TheoryDataFun: change of the argument structure; 'prep_ext' is

1271 now 'extend', and 'merge' gets an additional Pretty.pp argument

1272 (useful for printing error messages). INCOMPATIBILITY.

1274 * Pure: major reorganization of the theory context. Type Sign.sg and

1275 Theory.theory are now identified, referring to the universal

1276 Context.theory (see Pure/context.ML). Actual signature and theory

1277 content is managed as theory data. The old code and interfaces were

1278 spread over many files and structures; the new arrangement introduces

1279 considerable INCOMPATIBILITY to gain more clarity:

1281 Context -- theory management operations (name, identity, inclusion,

1282 parents, ancestors, merge, etc.), plus generic theory data;

1284 Sign -- logical signature and syntax operations (declaring consts,

1285 types, etc.), plus certify/read for common entities;

1287 Theory -- logical theory operations (stating axioms, definitions,

1288 oracles), plus a copy of logical signature operations (consts,

1289 types, etc.); also a few basic management operations (Theory.copy,

1290 Theory.merge, etc.)

1292 The most basic sign_of operations (Theory.sign_of, Thm.sign_of_thm

1293 etc.) as well as the sign field in Thm.rep_thm etc. have been retained

1294 for convenience -- they merely return the theory.

1296 * Pure: type Type.tsig is superceded by theory in most interfaces.

1298 * Pure: the Isar proof context type is already defined early in Pure

1299 as Context.proof (note that ProofContext.context and Proof.context are

1300 aliases, where the latter is the preferred name). This enables other

1301 Isabelle components to refer to that type even before Isar is present.

1303 * Pure/sign/theory: discontinued named name spaces (i.e. classK,

1304 typeK, constK, axiomK, oracleK), but provide explicit operations for

1305 any of these kinds. For example, Sign.intern typeK is now

1306 Sign.intern_type, Theory.hide_space Sign.typeK is now

1307 Theory.hide_types. Also note that former

1308 Theory.hide_classes/types/consts are now

1309 Theory.hide_classes_i/types_i/consts_i, while the non '_i' versions

1310 internalize their arguments! INCOMPATIBILITY.

1312 * Pure: get_thm interface (of PureThy and ProofContext) expects

1313 datatype thmref (with constructors Name and NameSelection) instead of

1314 plain string -- INCOMPATIBILITY;

1316 * Pure: cases produced by proof methods specify options, where NONE

1317 means to remove case bindings -- INCOMPATIBILITY in

1318 (RAW_)METHOD_CASES.

1320 * Pure: the following operations retrieve axioms or theorems from a

1321 theory node or theory hierarchy, respectively:

1323 Theory.axioms_of: theory -> (string * term) list

1324 Theory.all_axioms_of: theory -> (string * term) list

1325 PureThy.thms_of: theory -> (string * thm) list

1326 PureThy.all_thms_of: theory -> (string * thm) list

1328 * Pure: print_tac now outputs the goal through the trace channel.

1330 * Isar toplevel: improved diagnostics, mostly for Poly/ML only.

1331 Reference Toplevel.debug (default false) controls detailed printing

1332 and tracing of low-level exceptions; Toplevel.profiling (default 0)

1333 controls execution profiling -- set to 1 for time and 2 for space

1334 (both increase the runtime).

1336 * Isar session: The initial use of ROOT.ML is now always timed,

1337 i.e. the log will show the actual process times, in contrast to the

1338 elapsed wall-clock time that the outer shell wrapper produces.

1340 * Simplifier: improved handling of bound variables (nameless

1341 representation, avoid allocating new strings). Simprocs that invoke

1342 the Simplifier recursively should use Simplifier.inherit_bounds to

1343 avoid local name clashes. Failure to do so produces warnings

1344 "Simplifier: renamed bound variable ..."; set Simplifier.debug_bounds

1345 for further details.

1347 * ML functions legacy_bindings and use_legacy_bindings produce ML fact

1348 bindings for all theorems stored within a given theory; this may help

1349 in porting non-Isar theories to Isar ones, while keeping ML proof

1350 scripts for the time being.

1352 * ML operator HTML.with_charset specifies the charset begin used for

1353 generated HTML files. For example:

1355 HTML.with_charset "utf-8" use_thy "Hebrew";

1356 HTML.with_charset "utf-8" use_thy "Chinese";

1359 *** System ***

1361 * Allow symlinks to all proper Isabelle executables (Isabelle,

1362 isabelle, isatool etc.).

1364 * ISABELLE_DOC_FORMAT setting specifies preferred document format (for

1365 isatool doc, isatool mkdir, display_drafts etc.).

1367 * isatool usedir: option -f allows specification of the ML file to be

1368 used by Isabelle; default is ROOT.ML.

1370 * New isatool version outputs the version identifier of the Isabelle

1371 distribution being used.

1373 * HOL: new isatool dimacs2hol converts files in DIMACS CNF format

1374 (containing Boolean satisfiability problems) into Isabelle/HOL

1375 theories.

1379 New in Isabelle2004 (April 2004)

1380 --------------------------------

1382 *** General ***

1384 * Provers/order.ML: new efficient reasoner for partial and linear orders.

1385 Replaces linorder.ML.

1387 * Pure: Greek letters (except small lambda, \<lambda>), as well as Gothic

1388 (\<aa>...\<zz>\<AA>...\<ZZ>), calligraphic (\<A>...\<Z>), and Euler

1389 (\<a>...\<z>), are now considered normal letters, and can therefore

1390 be used anywhere where an ASCII letter (a...zA...Z) has until

1391 now. COMPATIBILITY: This obviously changes the parsing of some

1392 terms, especially where a symbol has been used as a binder, say

1393 '\<Pi>x. ...', which is now a type error since \<Pi>x will be parsed

1394 as an identifier. Fix it by inserting a space around former

1395 symbols. Call 'isatool fixgreek' to try to fix parsing errors in

1396 existing theory and ML files.

1398 * Pure: Macintosh and Windows line-breaks are now allowed in theory files.

1400 * Pure: single letter sub/superscripts (\<^isub> and \<^isup>) are now

1401 allowed in identifiers. Similar to Greek letters \<^isub> is now considered

1402 a normal (but invisible) letter. For multiple letter subscripts repeat

1403 \<^isub> like this: x\<^isub>1\<^isub>2.

1405 * Pure: There are now sub-/superscripts that can span more than one

1406 character. Text between \<^bsub> and \<^esub> is set in subscript in

1407 ProofGeneral and LaTeX, text between \<^bsup> and \<^esup> in

1408 superscript. The new control characters are not identifier parts.

1410 * Pure: Control-symbols of the form \<^raw:...> will literally print the

1411 content of "..." to the latex file instead of \isacntrl... . The "..."

1412 may consist of any printable characters excluding the end bracket >.

1414 * Pure: Using new Isar command "finalconsts" (or the ML functions

1415 Theory.add_finals or Theory.add_finals_i) it is now possible to

1416 declare constants "final", which prevents their being given a definition

1417 later. It is useful for constants whose behaviour is fixed axiomatically

1418 rather than definitionally, such as the meta-logic connectives.

1420 * Pure: 'instance' now handles general arities with general sorts

1421 (i.e. intersections of classes),

1423 * Presentation: generated HTML now uses a CSS style sheet to make layout

1424 (somewhat) independent of content. It is copied from lib/html/isabelle.css.

1425 It can be changed to alter the colors/layout of generated pages.

1428 *** Isar ***

1430 * Tactic emulation methods rule_tac, erule_tac, drule_tac, frule_tac,

1431 cut_tac, subgoal_tac and thin_tac:

1432 - Now understand static (Isar) contexts. As a consequence, users of Isar

1433 locales are no longer forced to write Isar proof scripts.

1434 For details see Isar Reference Manual, paragraph 4.3.2: Further tactic

1435 emulations.

1436 - INCOMPATIBILITY: names of variables to be instantiated may no

1437 longer be enclosed in quotes. Instead, precede variable name with `?'.

1438 This is consistent with the instantiation attribute "where".

1440 * Attributes "where" and "of":

1441 - Now take type variables of instantiated theorem into account when reading

1442 the instantiation string. This fixes a bug that caused instantiated

1443 theorems to have too special types in some circumstances.

1444 - "where" permits explicit instantiations of type variables.

1446 * Calculation commands "moreover" and "also" no longer interfere with

1447 current facts ("this"), admitting arbitrary combinations with "then"

1448 and derived forms.

1450 * Locales:

1451 - Goal statements involving the context element "includes" no longer

1452 generate theorems with internal delta predicates (those ending on

1453 "_axioms") in the premise.

1454 Resolve particular premise with <locale>.intro to obtain old form.

1455 - Fixed bug in type inference ("unify_frozen") that prevented mix of target

1456 specification and "includes" elements in goal statement.

1457 - Rule sets <locale>.intro and <locale>.axioms no longer declared as

1458 [intro?] and [elim?] (respectively) by default.

1459 - Experimental command for instantiation of locales in proof contexts:

1460 instantiate <label>[<attrs>]: <loc>

1461 Instantiates locale <loc> and adds all its theorems to the current context

1462 taking into account their attributes. Label and attrs are optional

1463 modifiers, like in theorem declarations. If present, names of

1464 instantiated theorems are qualified with <label>, and the attributes

1465 <attrs> are applied after any attributes these theorems might have already.

1466 If the locale has assumptions, a chained fact of the form

1467 "<loc> t1 ... tn" is expected from which instantiations of the parameters

1468 are derived. The command does not support old-style locales declared

1469 with "locale (open)".

1470 A few (very simple) examples can be found in FOL/ex/LocaleInst.thy.

1472 * HOL: Tactic emulation methods induct_tac and case_tac understand static

1473 (Isar) contexts.

1476 *** HOL ***

1478 * Proof import: new image HOL4 contains the imported library from

1479 the HOL4 system with about 2500 theorems. It is imported by

1480 replaying proof terms produced by HOL4 in Isabelle. The HOL4 image

1481 can be used like any other Isabelle image. See

1482 HOL/Import/HOL/README for more information.

1484 * Simplifier:

1485 - Much improved handling of linear and partial orders.

1486 Reasoners for linear and partial orders are set up for type classes

1487 "linorder" and "order" respectively, and are added to the default simpset

1488 as solvers. This means that the simplifier can build transitivity chains

1489 to solve goals from the assumptions.

1490 - INCOMPATIBILITY: old proofs break occasionally. Typically, applications

1491 of blast or auto after simplification become unnecessary because the goal

1492 is solved by simplification already.

1494 * Numerics: new theory Ring_and_Field contains over 250 basic numerical laws,

1495 all proved in axiomatic type classes for semirings, rings and fields.

1497 * Numerics:

1498 - Numeric types (nat, int, and in HOL-Complex rat, real, complex, etc.) are

1499 now formalized using the Ring_and_Field theory mentioned above.

1500 - INCOMPATIBILITY: simplification and arithmetic behaves somewhat differently

1501 than before, because now they are set up once in a generic manner.

1502 - INCOMPATIBILITY: many type-specific arithmetic laws have gone.

1503 Look for the general versions in Ring_and_Field (and Power if they concern

1504 exponentiation).

1506 * Type "rat" of the rational numbers is now available in HOL-Complex.

1508 * Records:

1509 - Record types are now by default printed with their type abbreviation

1510 instead of the list of all field types. This can be configured via

1511 the reference "print_record_type_abbr".

1512 - Simproc "record_upd_simproc" for simplification of multiple updates added

1513 (not enabled by default).

1514 - Simproc "record_ex_sel_eq_simproc" to simplify EX x. sel r = x resp.

1515 EX x. x = sel r to True (not enabled by default).

1516 - Tactic "record_split_simp_tac" to split and simplify records added.

1518 * 'specification' command added, allowing for definition by

1519 specification. There is also an 'ax_specification' command that

1520 introduces the new constants axiomatically.

1522 * arith(_tac) is now able to generate counterexamples for reals as well.

1524 * HOL-Algebra: new locale "ring" for non-commutative rings.

1526 * HOL-ex: InductiveInvariant_examples illustrates advanced recursive function

1527 definitions, thanks to Sava Krsti\'{c} and John Matthews.

1529 * HOL-Matrix: a first theory for matrices in HOL with an application of

1530 matrix theory to linear programming.

1532 * Unions and Intersections:

1533 The latex output syntax of UN and INT has been changed

1534 from "\Union x \in A. B" to "\Union_{x \in A} B"

1535 i.e. the index formulae has become a subscript.

1536 Similarly for "\Union x. B", and for \Inter instead of \Union.

1538 * Unions and Intersections over Intervals:

1539 There is new short syntax "UN i<=n. A" for "UN i:{0..n}. A". There is

1540 also an x-symbol version with subscripts "\<Union>\<^bsub>i <= n\<^esub>. A"

1541 like in normal math, and corresponding versions for < and for intersection.

1543 * HOL/List: Ordering "lexico" is renamed "lenlex" and the standard

1544 lexicographic dictonary ordering has been added as "lexord".

1546 * ML: the legacy theory structures Int and List have been removed. They had

1547 conflicted with ML Basis Library structures having the same names.

1549 * 'refute' command added to search for (finite) countermodels. Only works

1550 for a fragment of HOL. The installation of an external SAT solver is

1551 highly recommended. See "HOL/Refute.thy" for details.

1553 * 'quickcheck' command: Allows to find counterexamples by evaluating

1554 formulae under an assignment of free variables to random values.

1555 In contrast to 'refute', it can deal with inductive datatypes,

1556 but cannot handle quantifiers. See "HOL/ex/Quickcheck_Examples.thy"

1557 for examples.

1560 *** HOLCF ***

1562 * Streams now come with concatenation and are part of the HOLCF image

1566 New in Isabelle2003 (May 2003)

1567 ------------------------------

1569 *** General ***

1571 * Provers/simplifier:

1573 - Completely reimplemented method simp (ML: Asm_full_simp_tac):

1574 Assumptions are now subject to complete mutual simplification,

1575 not just from left to right. The simplifier now preserves

1576 the order of assumptions.

1578 Potential INCOMPATIBILITY:

1580 -- simp sometimes diverges where the old version did

1581 not, e.g. invoking simp on the goal

1583 [| P (f x); y = x; f x = f y |] ==> Q

1585 now gives rise to the infinite reduction sequence

1587 P(f x) --(f x = f y)--> P(f y) --(y = x)--> P(f x) --(f x = f y)--> ...

1589 Using "simp (asm_lr)" (ML: Asm_lr_simp_tac) instead often solves this

1590 kind of problem.

1592 -- Tactics combining classical reasoner and simplification (such as auto)

1593 are also affected by this change, because many of them rely on

1594 simp. They may sometimes diverge as well or yield a different numbers

1595 of subgoals. Try to use e.g. force, fastsimp, or safe instead of auto

1596 in case of problems. Sometimes subsequent calls to the classical

1597 reasoner will fail because a preceeding call to the simplifier too

1598 eagerly simplified the goal, e.g. deleted redundant premises.

1600 - The simplifier trace now shows the names of the applied rewrite rules

1602 - You can limit the number of recursive invocations of the simplifier

1603 during conditional rewriting (where the simplifie tries to solve the

1604 conditions before applying the rewrite rule):

1605 ML "simp_depth_limit := n"

1606 where n is an integer. Thus you can force termination where previously

1607 the simplifier would diverge.

1609 - Accepts free variables as head terms in congruence rules. Useful in Isar.

1611 - No longer aborts on failed congruence proof. Instead, the

1612 congruence is ignored.

1614 * Pure: New generic framework for extracting programs from constructive

1615 proofs. See HOL/Extraction.thy for an example instantiation, as well

1616 as HOL/Extraction for some case studies.

1618 * Pure: The main goal of the proof state is no longer shown by default, only

1619 the subgoals. This behaviour is controlled by a new flag.

1620 PG menu: Isabelle/Isar -> Settings -> Show Main Goal

1621 (ML: Proof.show_main_goal).

1623 * Pure: You can find all matching introduction rules for subgoal 1, i.e. all

1624 rules whose conclusion matches subgoal 1:

1625 PG menu: Isabelle/Isar -> Show me -> matching rules

1626 The rules are ordered by how closely they match the subgoal.

1627 In particular, rules that solve a subgoal outright are displayed first

1628 (or rather last, the way they are printed).

1629 (ML: ProofGeneral.print_intros())

1631 * Pure: New flag trace_unify_fail causes unification to print

1632 diagnostic information (PG: in trace buffer) when it fails. This is

1633 useful for figuring out why single step proofs like rule, erule or

1634 assumption failed.

1636 * Pure: Locale specifications now produce predicate definitions

1637 according to the body of text (covering assumptions modulo local

1638 definitions); predicate "loc_axioms" covers newly introduced text,

1639 while "loc" is cumulative wrt. all included locale expressions; the

1640 latter view is presented only on export into the global theory

1641 context; potential INCOMPATIBILITY, use "(open)" option to fall back

1642 on the old view without predicates;

1644 * Pure: predefined locales "var" and "struct" are useful for sharing

1645 parameters (as in CASL, for example); just specify something like

1646 ``var x + var y + struct M'' as import;

1648 * Pure: improved thms_containing: proper indexing of facts instead of

1649 raw theorems; check validity of results wrt. current name space;

1650 include local facts of proof configuration (also covers active

1651 locales), cover fixed variables in index; may use "_" in term

1652 specification; an optional limit for the number of printed facts may

1653 be given (the default is 40);

1655 * Pure: disallow duplicate fact bindings within new-style theory files

1656 (batch-mode only);

1658 * Provers: improved induct method: assumptions introduced by case

1659 "foo" are split into "foo.hyps" (from the rule) and "foo.prems" (from

1660 the goal statement); "foo" still refers to all facts collectively;

1662 * Provers: the function blast.overloaded has been removed: all constants

1663 are regarded as potentially overloaded, which improves robustness in exchange

1664 for slight decrease in efficiency;

1666 * Provers/linorder: New generic prover for transitivity reasoning over

1667 linear orders. Note: this prover is not efficient!

1669 * Isar: preview of problems to finish 'show' now produce an error

1670 rather than just a warning (in interactive mode);

1673 *** HOL ***

1675 * arith(_tac)

1677 - Produces a counter example if it cannot prove a goal.

1678 Note that the counter example may be spurious if the goal is not a formula

1679 of quantifier-free linear arithmetic.

1680 In ProofGeneral the counter example appears in the trace buffer.

1682 - Knows about div k and mod k where k is a numeral of type nat or int.

1684 - Calls full Presburger arithmetic (by Amine Chaieb) if quantifier-free

1685 linear arithmetic fails. This takes account of quantifiers and divisibility.

1686 Presburger arithmetic can also be called explicitly via presburger(_tac).

1688 * simp's arithmetic capabilities have been enhanced a bit: it now

1689 takes ~= in premises into account (by performing a case split);

1691 * simp reduces "m*(n div m) + n mod m" to n, even if the two summands

1692 are distributed over a sum of terms;

1694 * New tactic "trans_tac" and method "trans" instantiate

1695 Provers/linorder.ML for axclasses "order" and "linorder" (predicates

1696 "<=", "<" and "=").

1698 * function INCOMPATIBILITIES: Pi-sets have been redefined and moved from main

1699 HOL to Library/FuncSet; constant "Fun.op o" is now called "Fun.comp";

1701 * 'typedef' command has new option "open" to suppress the set

1702 definition;

1704 * functions Min and Max on finite sets have been introduced (theory

1705 Finite_Set);

1707 * attribute [symmetric] now works for relations as well; it turns

1708 (x,y) : R^-1 into (y,x) : R, and vice versa;

1710 * induct over a !!-quantified statement (say !!x1..xn):

1711 each "case" automatically performs "fix x1 .. xn" with exactly those names.

1713 * Map: `empty' is no longer a constant but a syntactic abbreviation for

1714 %x. None. Warning: empty_def now refers to the previously hidden definition

1715 of the empty set.

1717 * Algebra: formalization of classical algebra. Intended as base for

1718 any algebraic development in Isabelle. Currently covers group theory

1719 (up to Sylow's theorem) and ring theory (Universal Property of

1720 Univariate Polynomials). Contributions welcome;

1722 * GroupTheory: deleted, since its material has been moved to Algebra;

1724 * Complex: new directory of the complex numbers with numeric constants,

1725 nonstandard complex numbers, and some complex analysis, standard and

1726 nonstandard (Jacques Fleuriot);

1728 * HOL-Complex: new image for analysis, replacing HOL-Real and HOL-Hyperreal;

1730 * Hyperreal: introduced Gauge integration and hyperreal logarithms (Jacques

1731 Fleuriot);

1733 * Real/HahnBanach: updated and adapted to locales;

1735 * NumberTheory: added Gauss's law of quadratic reciprocity (by Avigad,

1736 Gray and Kramer);

1738 * UNITY: added the Meier-Sanders theory of progress sets;

1740 * MicroJava: bytecode verifier and lightweight bytecode verifier

1741 as abstract algorithms, instantiated to the JVM;

1743 * Bali: Java source language formalization. Type system, operational

1744 semantics, axiomatic semantics. Supported language features:

1745 classes, interfaces, objects,virtual methods, static methods,

1746 static/instance fields, arrays, access modifiers, definite

1747 assignment, exceptions.

1750 *** ZF ***

1752 * ZF/Constructible: consistency proof for AC (Gdel's constructible

1753 universe, etc.);

1755 * Main ZF: virtually all theories converted to new-style format;

1758 *** ML ***

1760 * Pure: Tactic.prove provides sane interface for internal proofs;

1761 omits the infamous "standard" operation, so this is more appropriate

1762 than prove_goalw_cterm in many situations (e.g. in simprocs);

1764 * Pure: improved error reporting of simprocs;

1766 * Provers: Simplifier.simproc(_i) provides sane interface for setting

1767 up simprocs;

1770 *** Document preparation ***

1772 * uses \par instead of \\ for line breaks in theory text. This may

1773 shift some page breaks in large documents. To get the old behaviour

1774 use \renewcommand{\isanewline}{\mbox{}\\\mbox{}} in root.tex.

1776 * minimized dependencies of isabelle.sty and isabellesym.sty on

1777 other packages

1779 * \<euro> now needs package babel/greek instead of marvosym (which

1780 broke \Rightarrow)

1782 * normal size for \<zero>...\<nine> (uses \mathbf instead of

1783 textcomp package)

1787 New in Isabelle2002 (March 2002)

1788 --------------------------------

1790 *** Document preparation ***

1792 * greatly simplified document preparation setup, including more

1793 graceful interpretation of isatool usedir -i/-d/-D options, and more

1794 instructive isatool mkdir; users should basically be able to get

1795 started with "isatool mkdir HOL Test && isatool make"; alternatively,

1796 users may run a separate document processing stage manually like this:

1797 "isatool usedir -D output HOL Test && isatool document Test/output";

1799 * theory dependency graph may now be incorporated into documents;

1800 isatool usedir -g true will produce session_graph.eps/.pdf for use

1801 with \includegraphics of LaTeX;

1803 * proper spacing of consecutive markup elements, especially text

1804 blocks after section headings;

1806 * support bold style (for single symbols only), input syntax is like

1807 this: "\<^bold>\<alpha>" or "\<^bold>A";

1809 * \<bullet> is now output as bold \cdot by default, which looks much

1810 better in printed text;

1812 * added default LaTeX bindings for \<tturnstile> and \<TTurnstile>;

1813 note that these symbols are currently unavailable in Proof General /

1814 X-Symbol; new symbols \<zero>, \<one>, ..., \<nine>, and \<euro>;

1816 * isatool latex no longer depends on changed TEXINPUTS, instead

1817 isatool document copies the Isabelle style files to the target

1818 location;

1821 *** Isar ***

1823 * Pure/Provers: improved proof by cases and induction;

1824 - 'case' command admits impromptu naming of parameters (such as

1825 "case (Suc n)");

1826 - 'induct' method divinates rule instantiation from the inductive

1827 claim; no longer requires excessive ?P bindings for proper

1828 instantiation of cases;

1829 - 'induct' method properly enumerates all possibilities of set/type

1830 rules; as a consequence facts may be also passed through *type*

1831 rules without further ado;

1832 - 'induct' method now derives symbolic cases from the *rulified*

1833 rule (before it used to rulify cases stemming from the internal

1834 atomized version); this means that the context of a non-atomic

1835 statement becomes is included in the hypothesis, avoiding the

1836 slightly cumbersome show "PROP ?case" form;

1837 - 'induct' may now use elim-style induction rules without chaining

1838 facts, using ``missing'' premises from the goal state; this allows

1839 rules stemming from inductive sets to be applied in unstructured

1840 scripts, while still benefitting from proper handling of non-atomic

1841 statements; NB: major inductive premises need to be put first, all

1842 the rest of the goal is passed through the induction;

1843 - 'induct' proper support for mutual induction involving non-atomic

1844 rule statements (uses the new concept of simultaneous goals, see

1845 below);

1846 - append all possible rule selections, but only use the first

1847 success (no backtracking);

1848 - removed obsolete "(simplified)" and "(stripped)" options of methods;

1849 - undeclared rule case names default to numbers 1, 2, 3, ...;

1850 - added 'print_induct_rules' (covered by help item in recent Proof

1851 General versions);

1852 - moved induct/cases attributes to Pure, methods to Provers;

1853 - generic method setup instantiated for FOL and HOL;

1855 * Pure: support multiple simultaneous goal statements, for example

1856 "have a: A and b: B" (same for 'theorem' etc.); being a pure

1857 meta-level mechanism, this acts as if several individual goals had

1858 been stated separately; in particular common proof methods need to be

1859 repeated in order to cover all claims; note that a single elimination

1860 step is *not* sufficient to establish the two conjunctions, so this

1861 fails:

1863 assume "A & B" then have A and B .. (*".." fails*)

1865 better use "obtain" in situations as above; alternative refer to

1866 multi-step methods like 'auto', 'simp_all', 'blast+' etc.;

1868 * Pure: proper integration with ``locales''; unlike the original

1869 version by Florian Kammller, Isar locales package high-level proof

1870 contexts rather than raw logical ones (e.g. we admit to include

1871 attributes everywhere); operations on locales include merge and

1872 rename; support for implicit arguments (``structures''); simultaneous

1873 type-inference over imports and text; see also HOL/ex/Locales.thy for

1874 some examples;

1876 * Pure: the following commands have been ``localized'', supporting a

1877 target locale specification "(in name)": 'lemma', 'theorem',

1878 'corollary', 'lemmas', 'theorems', 'declare'; the results will be

1879 stored both within the locale and at the theory level (exported and

1880 qualified by the locale name);

1882 * Pure: theory goals may now be specified in ``long'' form, with

1883 ad-hoc contexts consisting of arbitrary locale elements. for example

1884 ``lemma foo: fixes x assumes "A x" shows "B x"'' (local syntax and

1885 definitions may be given, too); the result is a meta-level rule with

1886 the context elements being discharged in the obvious way;

1888 * Pure: new proof command 'using' allows to augment currently used

1889 facts after a goal statement ('using' is syntactically analogous to

1890 'apply', but acts on the goal's facts only); this allows chained facts

1891 to be separated into parts given before and after a claim, as in

1892 ``from a and b have C using d and e <proof>'';

1894 * Pure: renamed "antecedent" case to "rule_context";

1896 * Pure: new 'judgment' command records explicit information about the

1897 object-logic embedding (used by several tools internally); no longer

1898 use hard-wired "Trueprop";

1900 * Pure: added 'corollary' command;

1902 * Pure: fixed 'token_translation' command;

1904 * Pure: removed obsolete 'exported' attribute;

1906 * Pure: dummy pattern "_" in is/let is now automatically lifted over

1907 bound variables: "ALL x. P x --> Q x" (is "ALL x. _ --> ?C x")

1908 supersedes more cumbersome ... (is "ALL x. _ x --> ?C x");

1910 * Pure: method 'atomize' presents local goal premises as object-level

1911 statements (atomic meta-level propositions); setup controlled via

1912 rewrite rules declarations of 'atomize' attribute; example

1913 application: 'induct' method with proper rule statements in improper

1914 proof *scripts*;

1916 * Pure: emulation of instantiation tactics (rule_tac, cut_tac, etc.)

1917 now consider the syntactic context of assumptions, giving a better

1918 chance to get type-inference of the arguments right (this is

1919 especially important for locales);

1921 * Pure: "sorry" no longer requires quick_and_dirty in interactive

1922 mode;

1924 * Pure/obtain: the formal conclusion "thesis", being marked as

1925 ``internal'', may no longer be reference directly in the text;

1926 potential INCOMPATIBILITY, may need to use "?thesis" in rare

1927 situations;

1929 * Pure: generic 'sym' attribute which declares a rule both as pure

1930 'elim?' and for the 'symmetric' operation;

1932 * Pure: marginal comments ``--'' may now occur just anywhere in the

1933 text; the fixed correlation with particular command syntax has been

1934 discontinued;

1936 * Pure: new method 'rules' is particularly well-suited for proof

1937 search in intuitionistic logic; a bit slower than 'blast' or 'fast',

1938 but often produces more compact proof terms with less detours;

1940 * Pure/Provers/classical: simplified integration with pure rule

1941 attributes and methods; the classical "intro?/elim?/dest?"

1942 declarations coincide with the pure ones; the "rule" method no longer

1943 includes classically swapped intros; "intro" and "elim" methods no

1944 longer pick rules from the context; also got rid of ML declarations

1945 AddXIs/AddXEs/AddXDs; all of this has some potential for

1946 INCOMPATIBILITY;

1948 * Provers/classical: attribute 'swapped' produces classical inversions

1949 of introduction rules;

1951 * Provers/simplifier: 'simplified' attribute may refer to explicit

1952 rules instead of full simplifier context; 'iff' attribute handles

1953 conditional rules;

1955 * HOL: 'typedef' now allows alternative names for Rep/Abs morphisms;

1957 * HOL: 'recdef' now fails on unfinished automated proofs, use

1958 "(permissive)" option to recover old behavior;

1960 * HOL: 'inductive' no longer features separate (collective) attributes

1961 for 'intros' (was found too confusing);

1963 * HOL: properly declared induction rules less_induct and

1964 wf_induct_rule;

1967 *** HOL ***

1969 * HOL: moved over to sane numeral syntax; the new policy is as

1970 follows:

1972 - 0 and 1 are polymorphic constants, which are defined on any

1973 numeric type (nat, int, real etc.);

1975 - 2, 3, 4, ... and -1, -2, -3, ... are polymorphic numerals, based

1976 binary representation internally;

1978 - type nat has special constructor Suc, and generally prefers Suc 0

1979 over 1::nat and Suc (Suc 0) over 2::nat;

1981 This change may cause significant problems of INCOMPATIBILITY; here

1982 are some hints on converting existing sources:

1984 - due to the new "num" token, "-0" and "-1" etc. are now atomic

1985 entities, so expressions involving "-" (unary or binary minus) need

1986 to be spaced properly;

1988 - existing occurrences of "1" may need to be constraint "1::nat" or

1989 even replaced by Suc 0; similar for old "2";

1991 - replace "#nnn" by "nnn", and "#-nnn" by "-nnn";

1993 - remove all special provisions on numerals in proofs;

1995 * HOL: simp rules nat_number expand numerals on nat to Suc/0

1996 representation (depends on bin_arith_simps in the default context);

1998 * HOL: symbolic syntax for x^2 (numeral 2);

2000 * HOL: the class of all HOL types is now called "type" rather than

2001 "term"; INCOMPATIBILITY, need to adapt references to this type class

2002 in axclass/classes, instance/arities, and (usually rare) occurrences

2003 in typings (of consts etc.); internally the class is called

2004 "HOL.type", ML programs should refer to HOLogic.typeS;

2006 * HOL/record package improvements:

2007 - new derived operations "fields" to build a partial record section,

2008 "extend" to promote a fixed record to a record scheme, and

2009 "truncate" for the reverse; cf. theorems "xxx.defs", which are *not*

2010 declared as simp by default;

2011 - shared operations ("more", "fields", etc.) now need to be always

2012 qualified) --- potential INCOMPATIBILITY;

2013 - removed "make_scheme" operations (use "make" with "extend") --

2014 INCOMPATIBILITY;

2015 - removed "more" class (simply use "term") -- INCOMPATIBILITY;

2016 - provides cases/induct rules for use with corresponding Isar

2017 methods (for concrete records, record schemes, concrete more

2018 parts, and schematic more parts -- in that order);

2019 - internal definitions directly based on a light-weight abstract

2020 theory of product types over typedef rather than datatype;

2022 * HOL: generic code generator for generating executable ML code from

2023 specifications; specific support for HOL constructs such as inductive

2024 datatypes and sets, as well as recursive functions; can be invoked

2025 via 'generate_code' theory section;

2027 * HOL: canonical cases/induct rules for n-tuples (n = 3..7);

2029 * HOL: consolidated and renamed several theories. In particular:

2030 Ord.thy has been absorbed into HOL.thy

2031 String.thy has been absorbed into List.thy

2033 * HOL: concrete setsum syntax "\<Sum>i:A. b" == "setsum (%i. b) A"

2034 (beware of argument permutation!);

2036 * HOL: linorder_less_split superseded by linorder_cases;

2038 * HOL/List: "nodups" renamed to "distinct";

2040 * HOL: added "The" definite description operator; move Hilbert's "Eps"

2041 to peripheral theory "Hilbert_Choice"; some INCOMPATIBILITIES:

2042 - Ex_def has changed, now need to use some_eq_ex

2044 * HOL: made split_all_tac safe; EXISTING PROOFS MAY FAIL OR LOOP, so

2045 in this (rare) case use:

2047 delSWrapper "split_all_tac"

2048 addSbefore ("unsafe_split_all_tac", unsafe_split_all_tac)

2050 * HOL: added safe wrapper "split_conv_tac" to claset; EXISTING PROOFS

2051 MAY FAIL;

2053 * HOL: introduced f^n = f o ... o f; warning: due to the limits of

2054 Isabelle's type classes, ^ on functions and relations has too general

2055 a domain, namely ('a * 'b) set and 'a => 'b; this means that it may be

2056 necessary to attach explicit type constraints;

2058 * HOL/Relation: the prefix name of the infix "O" has been changed from

2059 "comp" to "rel_comp"; INCOMPATIBILITY: a few theorems have been

2060 renamed accordingly (eg "compI" -> "rel_compI").

2062 * HOL: syntax translations now work properly with numerals and records

2063 expressions;

2065 * HOL: bounded abstraction now uses syntax "%" / "\<lambda>" instead

2066 of "lam" -- INCOMPATIBILITY;

2068 * HOL: got rid of some global declarations (potential INCOMPATIBILITY

2069 for ML tools): const "()" renamed "Product_Type.Unity", type "unit"

2070 renamed "Product_Type.unit";

2072 * HOL: renamed rtrancl_into_rtrancl2 to converse_rtrancl_into_rtrancl

2074 * HOL: removed obsolete theorem "optionE" (use "option.exhaust", or

2075 the "cases" method);

2077 * HOL/GroupTheory: group theory examples including Sylow's theorem (by

2078 Florian Kammller);

2080 * HOL/IMP: updated and converted to new-style theory format; several

2081 parts turned into readable document, with proper Isar proof texts and

2082 some explanations (by Gerwin Klein);

2084 * HOL-Real: added Complex_Numbers (by Gertrud Bauer);

2086 * HOL-Hyperreal is now a logic image;

2089 *** HOLCF ***

2091 * Isar: consts/constdefs supports mixfix syntax for continuous

2092 operations;

2094 * Isar: domain package adapted to new-style theory format, e.g. see

2095 HOLCF/ex/Dnat.thy;

2097 * theory Lift: proper use of rep_datatype lift instead of ML hacks --

2098 potential INCOMPATIBILITY; now use plain induct_tac instead of former

2099 lift.induct_tac, always use UU instead of Undef;

2101 * HOLCF/IMP: updated and converted to new-style theory;

2104 *** ZF ***

2106 * Isar: proper integration of logic-specific tools and packages,

2107 including theory commands '(co)inductive', '(co)datatype',

2108 'rep_datatype', 'inductive_cases', as well as methods 'ind_cases',

2109 'induct_tac', 'case_tac', and 'typecheck' (with attribute 'TC');

2111 * theory Main no longer includes AC; for the Axiom of Choice, base

2112 your theory on Main_ZFC;

2114 * the integer library now covers quotients and remainders, with many

2115 laws relating division to addition, multiplication, etc.;

2117 * ZF/UNITY: Chandy and Misra's UNITY is now available in ZF, giving a

2118 typeless version of the formalism;

2120 * ZF/AC, Coind, IMP, Resid: updated and converted to new-style theory

2121 format;

2123 * ZF/Induct: new directory for examples of inductive definitions,

2124 including theory Multiset for multiset orderings; converted to

2125 new-style theory format;

2127 * ZF: many new theorems about lists, ordinals, etc.;

2130 *** General ***

2132 * Pure/kernel: meta-level proof terms (by Stefan Berghofer); reference

2133 variable proof controls level of detail: 0 = no proofs (only oracle

2134 dependencies), 1 = lemma dependencies, 2 = compact proof terms; see

2135 also ref manual for further ML interfaces;

2137 * Pure/axclass: removed obsolete ML interface

2138 goal_subclass/goal_arity;

2140 * Pure/syntax: new token syntax "num" for plain numerals (without "#"

2141 of "xnum"); potential INCOMPATIBILITY, since -0, -1 etc. are now

2142 separate tokens, so expressions involving minus need to be spaced

2143 properly;

2145 * Pure/syntax: support non-oriented infixes, using keyword "infix"

2146 rather than "infixl" or "infixr";

2148 * Pure/syntax: concrete syntax for dummy type variables admits genuine

2149 sort constraint specifications in type inference; e.g. "x::_::foo"

2150 ensures that the type of "x" is of sort "foo" (but not necessarily a

2151 type variable);

2153 * Pure/syntax: print modes "type_brackets" and "no_type_brackets"

2154 control output of nested => (types); the default behavior is

2155 "type_brackets";

2157 * Pure/syntax: builtin parse translation for "_constify" turns valued

2158 tokens into AST constants;

2160 * Pure/syntax: prefer later declarations of translations and print

2161 translation functions; potential INCOMPATIBILITY: need to reverse

2162 multiple declarations for same syntax element constant;

2164 * Pure/show_hyps reset by default (in accordance to existing Isar

2165 practice);

2167 * Provers/classical: renamed addaltern to addafter, addSaltern to

2168 addSafter;

2170 * Provers/clasimp: ``iff'' declarations now handle conditional rules

2171 as well;

2173 * system: tested support for MacOS X; should be able to get Isabelle +

2174 Proof General to work in a plain Terminal after installing Poly/ML

2175 (e.g. from the Isabelle distribution area) and GNU bash alone

2176 (e.g. from http://www.apple.com); full X11, XEmacs and X-Symbol

2177 support requires further installations, e.g. from

2178 http://fink.sourceforge.net/);

2180 * system: support Poly/ML 4.1.1 (able to manage larger heaps);

2182 * system: reduced base memory usage by Poly/ML (approx. 20 MB instead

2183 of 40 MB), cf. ML_OPTIONS;

2185 * system: Proof General keywords specification is now part of the

2186 Isabelle distribution (see etc/isar-keywords.el);

2188 * system: support for persistent Proof General sessions (refrain from

2189 outdating all loaded theories on startup); user may create writable

2190 logic images like this: ``isabelle -q HOL Test'';

2192 * system: smart selection of Isabelle process versus Isabelle

2193 interface, accommodates case-insensitive file systems (e.g. HFS+); may

2194 run both "isabelle" and "Isabelle" even if file names are badly

2195 damaged (executable inspects the case of the first letter of its own

2196 name); added separate "isabelle-process" and "isabelle-interface";

2198 * system: refrain from any attempt at filtering input streams; no

2199 longer support ``8bit'' encoding of old isabelle font, instead proper

2200 iso-latin characters may now be used; the related isatools

2201 "symbolinput" and "nonascii" have disappeared as well;

2203 * system: removed old "xterm" interface (the print modes "xterm" and

2204 "xterm_color" are still available for direct use in a suitable

2205 terminal);

2209 New in Isabelle99-2 (February 2001)

2210 -----------------------------------

2212 *** Overview of INCOMPATIBILITIES ***

2214 * HOL: please note that theories in the Library and elsewhere often use the

2215 new-style (Isar) format; to refer to their theorems in an ML script you must

2216 bind them to ML identifers by e.g. val thm_name = thm "thm_name";

2218 * HOL: inductive package no longer splits induction rule aggressively,

2219 but only as far as specified by the introductions given; the old

2220 format may be recovered via ML function complete_split_rule or attribute

2221 'split_rule (complete)';

2223 * HOL: induct renamed to lfp_induct, lfp_Tarski to lfp_unfold,

2224 gfp_Tarski to gfp_unfold;

2226 * HOL: contrapos, contrapos2 renamed to contrapos_nn, contrapos_pp;

2228 * HOL: infix "dvd" now has priority 50 rather than 70 (because it is a

2229 relation); infix "^^" has been renamed "``"; infix "``" has been

2230 renamed "`"; "univalent" has been renamed "single_valued";

2232 * HOL/Real: "rinv" and "hrinv" replaced by overloaded "inverse"

2233 operation;

2235 * HOLCF: infix "`" has been renamed "$"; the symbol syntax is \<cdot>;

2237 * Isar: 'obtain' no longer declares "that" fact as simp/intro;

2239 * Isar/HOL: method 'induct' now handles non-atomic goals; as a

2240 consequence, it is no longer monotonic wrt. the local goal context

2241 (which is now passed through the inductive cases);

2243 * Document preparation: renamed standard symbols \<ll> to \<lless> and

2244 \<gg> to \<ggreater>;

2247 *** Document preparation ***

2249 * \isabellestyle{NAME} selects version of Isabelle output (currently

2250 available: are "it" for near math-mode best-style output, "sl" for

2251 slanted text style, and "tt" for plain type-writer; if no

2252 \isabellestyle command is given, output is according to slanted

2253 type-writer);

2255 * support sub/super scripts (for single symbols only), input syntax is

2256 like this: "A\<^sup>*" or "A\<^sup>\<star>";

2258 * some more standard symbols; see Appendix A of the system manual for

2259 the complete list of symbols defined in isabellesym.sty;

2261 * improved isabelle style files; more abstract symbol implementation

2262 (should now use \isamath{...} and \isatext{...} in custom symbol

2263 definitions);

2265 * antiquotation @{goals} and @{subgoals} for output of *dynamic* goals

2266 state; Note that presentation of goal states does not conform to

2267 actual human-readable proof documents. Please do not include goal

2268 states into document output unless you really know what you are doing!

2270 * proper indentation of antiquoted output with proportional LaTeX

2271 fonts;

2273 * no_document ML operator temporarily disables LaTeX document

2274 generation;

2276 * isatool unsymbolize tunes sources for plain ASCII communication;

2279 *** Isar ***

2281 * Pure: Isar now suffers initial goal statements to contain unbound

2282 schematic variables (this does not conform to actual readable proof

2283 documents, due to unpredictable outcome and non-compositional proof

2284 checking); users who know what they are doing may use schematic goals

2285 for Prolog-style synthesis of proven results;

2287 * Pure: assumption method (an implicit finishing) now handles actual

2288 rules as well;

2290 * Pure: improved 'obtain' --- moved to Pure, insert "that" into

2291 initial goal, declare "that" only as Pure intro (only for single

2292 steps); the "that" rule assumption may now be involved in implicit

2293 finishing, thus ".." becomes a feasible for trivial obtains;

2295 * Pure: default proof step now includes 'intro_classes'; thus trivial

2296 instance proofs may be performed by "..";

2298 * Pure: ?thesis / ?this / "..." now work for pure meta-level

2299 statements as well;

2301 * Pure: more robust selection of calculational rules;

2303 * Pure: the builtin notion of 'finished' goal now includes the ==-refl

2304 rule (as well as the assumption rule);

2306 * Pure: 'thm_deps' command visualizes dependencies of theorems and

2307 lemmas, using the graph browser tool;

2309 * Pure: predict failure of "show" in interactive mode;

2311 * Pure: 'thms_containing' now takes actual terms as arguments;

2313 * HOL: improved method 'induct' --- now handles non-atomic goals

2314 (potential INCOMPATIBILITY); tuned error handling;

2316 * HOL: cases and induct rules now provide explicit hints about the

2317 number of facts to be consumed (0 for "type" and 1 for "set" rules);

2318 any remaining facts are inserted into the goal verbatim;

2320 * HOL: local contexts (aka cases) may now contain term bindings as

2321 well; the 'cases' and 'induct' methods new provide a ?case binding for

2322 the result to be shown in each case;

2324 * HOL: added 'recdef_tc' command;

2326 * isatool convert assists in eliminating legacy ML scripts;

2329 *** HOL ***

2331 * HOL/Library: a collection of generic theories to be used together

2332 with main HOL; the theory loader path already includes this directory

2333 by default; the following existing theories have been moved here:

2334 HOL/Induct/Multiset, HOL/Induct/Acc (as Accessible_Part), HOL/While

2335 (as While_Combinator), HOL/Lex/Prefix (as List_Prefix);

2337 * HOL/Unix: "Some aspects of Unix file-system security", a typical

2338 modelling and verification task performed in Isabelle/HOL +

2339 Isabelle/Isar + Isabelle document preparation (by Markus Wenzel).

2341 * HOL/Algebra: special summation operator SUM no longer exists, it has

2342 been replaced by setsum; infix 'assoc' now has priority 50 (like

2343 'dvd'); axiom 'one_not_zero' has been moved from axclass 'ring' to

2344 'domain', this makes the theory consistent with mathematical

2345 literature;

2347 * HOL basics: added overloaded operations "inverse" and "divide"

2348 (infix "/"), syntax for generic "abs" operation, generic summation

2349 operator \<Sum>;

2351 * HOL/typedef: simplified package, provide more useful rules (see also

2352 HOL/subset.thy);

2354 * HOL/datatype: induction rule for arbitrarily branching datatypes is

2355 now expressed as a proper nested rule (old-style tactic scripts may

2356 require atomize_strip_tac to cope with non-atomic premises);

2358 * HOL: renamed theory "Prod" to "Product_Type", renamed "split" rule

2359 to "split_conv" (old name still available for compatibility);

2361 * HOL: improved concrete syntax for strings (e.g. allows translation

2362 rules with string literals);

2364 * HOL-Real-Hyperreal: this extends HOL-Real with the hyperreals

2365 and Fleuriot's mechanization of analysis, including the transcendental

2366 functions for the reals;

2368 * HOL/Real, HOL/Hyperreal: improved arithmetic simplification;

2371 *** CTT ***

2373 * CTT: x-symbol support for Pi, Sigma, -->, : (membership); note that

2374 "lam" is displayed as TWO lambda-symbols

2376 * CTT: theory Main now available, containing everything (that is, Bool

2377 and Arith);

2380 *** General ***

2382 * Pure: the Simplifier has been implemented properly as a derived rule

2383 outside of the actual kernel (at last!); the overall performance

2384 penalty in practical applications is about 50%, while reliability of

2385 the Isabelle inference kernel has been greatly improved;

2387 * print modes "brackets" and "no_brackets" control output of nested =>

2388 (types) and ==> (props); the default behaviour is "brackets";

2390 * Provers: fast_tac (and friends) now handle actual object-logic rules

2391 as assumptions as well;

2393 * system: support Poly/ML 4.0;

2395 * system: isatool install handles KDE version 1 or 2;

2399 New in Isabelle99-1 (October 2000)

2400 ----------------------------------

2402 *** Overview of INCOMPATIBILITIES ***

2404 * HOL: simplification of natural numbers is much changed; to partly

2405 recover the old behaviour (e.g. to prevent n+n rewriting to #2*n)

2406 issue the following ML commands:

2408 Delsimprocs Nat_Numeral_Simprocs.cancel_numerals;

2409 Delsimprocs [Nat_Numeral_Simprocs.combine_numerals];

2411 * HOL: simplification no longer dives into case-expressions; this is

2412 controlled by "t.weak_case_cong" for each datatype t;

2414 * HOL: nat_less_induct renamed to less_induct;

2416 * HOL: systematic renaming of the SOME (Eps) rules, may use isatool

2417 fixsome to patch .thy and .ML sources automatically;

2419 select_equality -> some_equality

2420 select_eq_Ex -> some_eq_ex

2421 selectI2EX -> someI2_ex

2422 selectI2 -> someI2

2423 selectI -> someI

2424 select1_equality -> some1_equality

2425 Eps_sym_eq -> some_sym_eq_trivial

2426 Eps_eq -> some_eq_trivial

2428 * HOL: exhaust_tac on datatypes superceded by new generic case_tac;

2430 * HOL: removed obsolete theorem binding expand_if (refer to split_if

2431 instead);

2433 * HOL: the recursion equations generated by 'recdef' are now called

2434 f.simps instead of f.rules;

2436 * HOL: qed_spec_mp now also handles bounded ALL as well;

2438 * HOL: 0 is now overloaded, so the type constraint ":: nat" may

2439 sometimes be needed;

2441 * HOL: the constant for "f``x" is now "image" rather than "op ``";

2443 * HOL: the constant for "f-``x" is now "vimage" rather than "op -``";

2445 * HOL: the disjoint sum is now "<+>" instead of "Plus"; the cartesian

2446 product is now "<*>" instead of "Times"; the lexicographic product is

2447 now "<*lex*>" instead of "**";

2449 * HOL: theory Sexp is now in HOL/Induct examples (it used to be part

2450 of main HOL, but was unused); better use HOL's datatype package;

2452 * HOL: removed "symbols" syntax for constant "override" of theory Map;

2453 the old syntax may be recovered as follows:

2455 syntax (symbols)

2456 override :: "('a ~=> 'b) => ('a ~=> 'b) => ('a ~=> 'b)"

2457 (infixl "\\<oplus>" 100)

2459 * HOL/Real: "rabs" replaced by overloaded "abs" function;

2461 * HOL/ML: even fewer consts are declared as global (see theories Ord,

2462 Lfp, Gfp, WF); this only affects ML packages that refer to const names

2463 internally;

2465 * HOL and ZF: syntax for quotienting wrt an equivalence relation

2466 changed from A/r to A//r;

2468 * ZF: new treatment of arithmetic (nat & int) may break some old

2469 proofs;

2471 * Isar: renamed some attributes (RS -> THEN, simplify -> simplified,

2472 rulify -> rule_format, elimify -> elim_format, ...);

2474 * Isar/Provers: intro/elim/dest attributes changed; renamed

2475 intro/intro!/intro!! flags to intro!/intro/intro? (in most cases, one

2476 should have to change intro!! to intro? only); replaced "delrule" by

2477 "rule del";

2479 * Isar/HOL: renamed "intrs" to "intros" in inductive definitions;

2481 * Provers: strengthened force_tac by using new first_best_tac;

2483 * LaTeX document preparation: several changes of isabelle.sty (see

2484 lib/texinputs);

2487 *** Document preparation ***

2489 * formal comments (text blocks etc.) in new-style theories may now

2490 contain antiquotations of thm/prop/term/typ/text to be presented

2491 according to latex print mode; concrete syntax is like this:

2492 @{term[show_types] "f(x) = a + x"};

2494 * isatool mkdir provides easy setup of Isabelle session directories,

2495 including proper document sources;

2497 * generated LaTeX sources are now deleted after successful run

2498 (isatool document -c); may retain a copy somewhere else via -D option

2499 of isatool usedir;

2501 * isatool usedir -D now lets isatool latex -o sty update the Isabelle

2502 style files, achieving self-contained LaTeX sources and simplifying

2503 LaTeX debugging;

2505 * old-style theories now produce (crude) LaTeX output as well;

2507 * browser info session directories are now self-contained (may be put

2508 on WWW server seperately); improved graphs of nested sessions; removed

2509 graph for 'all sessions';

2511 * several improvements in isabelle style files; \isabellestyle{it}

2512 produces fake math mode output; \isamarkupheader is now \section by

2513 default; see lib/texinputs/isabelle.sty etc.;

2516 *** Isar ***

2518 * Isar/Pure: local results and corresponding term bindings are now

2519 subject to Hindley-Milner polymorphism (similar to ML); this

2520 accommodates incremental type-inference very nicely;

2522 * Isar/Pure: new derived language element 'obtain' supports

2523 generalized existence reasoning;

2525 * Isar/Pure: new calculational elements 'moreover' and 'ultimately'

2526 support accumulation of results, without applying any rules yet;

2527 useful to collect intermediate results without explicit name

2528 references, and for use with transitivity rules with more than 2

2529 premises;

2531 * Isar/Pure: scalable support for case-analysis type proofs: new

2532 'case' language element refers to local contexts symbolically, as

2533 produced by certain proof methods; internally, case names are attached

2534 to theorems as "tags";

2536 * Isar/Pure: theory command 'hide' removes declarations from

2537 class/type/const name spaces;

2539 * Isar/Pure: theory command 'defs' supports option "(overloaded)" to

2540 indicate potential overloading;

2542 * Isar/Pure: changed syntax of local blocks from {{ }} to { };

2544 * Isar/Pure: syntax of sorts made 'inner', i.e. have to write

2545 "{a,b,c}" instead of {a,b,c};

2547 * Isar/Pure now provides its own version of intro/elim/dest

2548 attributes; useful for building new logics, but beware of confusion

2549 with the version in Provers/classical;

2551 * Isar/Pure: the local context of (non-atomic) goals is provided via

2552 case name 'antecedent';

2554 * Isar/Pure: removed obsolete 'transfer' attribute (transfer of thms

2555 to the current context is now done automatically);

2557 * Isar/Pure: theory command 'method_setup' provides a simple interface

2558 for definining proof methods in ML;

2560 * Isar/Provers: intro/elim/dest attributes changed; renamed

2561 intro/intro!/intro!! flags to intro!/intro/intro? (INCOMPATIBILITY, in

2562 most cases, one should have to change intro!! to intro? only);

2563 replaced "delrule" by "rule del";

2565 * Isar/Provers: new 'hypsubst' method, plain 'subst' method and

2566 'symmetric' attribute (the latter supercedes [RS sym]);

2568 * Isar/Provers: splitter support (via 'split' attribute and 'simp'

2569 method modifier); 'simp' method: 'only:' modifier removes loopers as

2570 well (including splits);

2572 * Isar/Provers: Simplifier and Classical methods now support all kind

2573 of modifiers used in the past, including 'cong', 'iff', etc.

2575 * Isar/Provers: added 'fastsimp' and 'clarsimp' methods (combination

2576 of Simplifier and Classical reasoner);

2578 * Isar/HOL: new proof method 'cases' and improved version of 'induct'

2579 now support named cases; major packages (inductive, datatype, primrec,

2580 recdef) support case names and properly name parameters;

2582 * Isar/HOL: new transitivity rules for substitution in inequalities --

2583 monotonicity conditions are extracted to be proven at end of

2584 calculations;

2586 * Isar/HOL: removed 'case_split' thm binding, should use 'cases' proof

2587 method anyway;

2589 * Isar/HOL: removed old expand_if = split_if; theorems if_splits =

2590 split_if split_if_asm; datatype package provides theorems foo.splits =

2591 foo.split foo.split_asm for each datatype;

2593 * Isar/HOL: tuned inductive package, rename "intrs" to "intros"

2594 (potential INCOMPATIBILITY), emulation of mk_cases feature for proof

2595 scripts: new 'inductive_cases' command and 'ind_cases' method; (Note:

2596 use "(cases (simplified))" method in proper proof texts);

2598 * Isar/HOL: added global 'arith_split' attribute for 'arith' method;

2600 * Isar: names of theorems etc. may be natural numbers as well;

2602 * Isar: 'pr' command: optional arguments for goals_limit and

2603 ProofContext.prems_limit; no longer prints theory contexts, but only

2604 proof states;

2606 * Isar: diagnostic commands 'pr', 'thm', 'prop', 'term', 'typ' admit

2607 additional print modes to be specified; e.g. "pr(latex)" will print

2608 proof state according to the Isabelle LaTeX style;

2610 * Isar: improved support for emulating tactic scripts, including proof

2611 methods 'rule_tac' etc., 'cut_tac', 'thin_tac', 'subgoal_tac',

2612 'rename_tac', 'rotate_tac', 'tactic', and 'case_tac' / 'induct_tac'

2613 (for HOL datatypes);

2615 * Isar: simplified (more robust) goal selection of proof methods: 1st

2616 goal, all goals, or explicit goal specifier (tactic emulation); thus

2617 'proof method scripts' have to be in depth-first order;

2619 * Isar: tuned 'let' syntax: replaced 'as' keyword by 'and';

2621 * Isar: removed 'help' command, which hasn't been too helpful anyway;

2622 should instead use individual commands for printing items

2623 (print_commands, print_methods etc.);

2625 * Isar: added 'nothing' --- the empty list of theorems;

2628 *** HOL ***

2630 * HOL/MicroJava: formalization of a fragment of Java, together with a

2631 corresponding virtual machine and a specification of its bytecode

2632 verifier and a lightweight bytecode verifier, including proofs of

2633 type-safety; by Gerwin Klein, Tobias Nipkow, David von Oheimb, and

2634 Cornelia Pusch (see also the homepage of project Bali at

2635 http://isabelle.in.tum.de/Bali/);

2637 * HOL/Algebra: new theory of rings and univariate polynomials, by

2638 Clemens Ballarin;

2640 * HOL/NumberTheory: fundamental Theorem of Arithmetic, Chinese

2641 Remainder Theorem, Fermat/Euler Theorem, Wilson's Theorem, by Thomas M

2642 Rasmussen;

2644 * HOL/Lattice: fundamental concepts of lattice theory and order

2645 structures, including duals, properties of bounds versus algebraic

2646 laws, lattice operations versus set-theoretic ones, the Knaster-Tarski

2647 Theorem for complete lattices etc.; may also serve as a demonstration

2648 for abstract algebraic reasoning using axiomatic type classes, and

2649 mathematics-style proof in Isabelle/Isar; by Markus Wenzel;

2651 * HOL/Prolog: a (bare-bones) implementation of Lambda-Prolog, by David

2652 von Oheimb;

2654 * HOL/IMPP: extension of IMP with local variables and mutually

2655 recursive procedures, by David von Oheimb;

2657 * HOL/Lambda: converted into new-style theory and document;

2659 * HOL/ex/Multiquote: example of multiple nested quotations and

2660 anti-quotations -- basically a generalized version of de-Bruijn

2661 representation; very useful in avoiding lifting of operations;

2663 * HOL/record: added general record equality rule to simpset; fixed

2664 select-update simplification procedure to handle extended records as

2665 well; admit "r" as field name;

2667 * HOL: 0 is now overloaded over the new sort "zero", allowing its use with

2668 other numeric types and also as the identity of groups, rings, etc.;

2670 * HOL: new axclass plus_ac0 for addition with the AC-laws and 0 as identity.

2671 Types nat and int belong to this axclass;

2673 * HOL: greatly improved simplification involving numerals of type nat, int, real:

2674 (i + #8 + j) = Suc k simplifies to #7 + (i + j) = k

2675 i*j + k + j*#3*i simplifies to #4*(i*j) + k

2676 two terms #m*u and #n*u are replaced by #(m+n)*u

2677 (where #m, #n and u can implicitly be 1; this is simproc combine_numerals)

2678 and the term/formula #m*u+x ~~ #n*u+y simplifies simplifies to #(m-n)+x ~~ y

2679 or x ~~ #(n-m)+y, where ~~ is one of = < <= or - (simproc cancel_numerals);

2681 * HOL: meson_tac is available (previously in ex/meson.ML); it is a

2682 powerful prover for predicate logic but knows nothing of clasets; see

2683 ex/mesontest.ML and ex/mesontest2.ML for example applications;

2685 * HOL: new version of "case_tac" subsumes both boolean case split and

2686 "exhaust_tac" on datatypes; INCOMPATIBILITY: exhaust_tac no longer

2687 exists, may define val exhaust_tac = case_tac for ad-hoc portability;

2689 * HOL: simplification no longer dives into case-expressions: only the

2690 selector expression is simplified, but not the remaining arms; to

2691 enable full simplification of case-expressions for datatype t, you may

2692 remove t.weak_case_cong from the simpset, either globally (Delcongs

2693 [thm"t.weak_case_cong"];) or locally (delcongs [...]).

2695 * HOL/recdef: the recursion equations generated by 'recdef' for

2696 function 'f' are now called f.simps instead of f.rules; if all

2697 termination conditions are proved automatically, these simplification

2698 rules are added to the simpset, as in primrec; rules may be named

2699 individually as well, resulting in a separate list of theorems for

2700 each equation;

2702 * HOL/While is a new theory that provides a while-combinator. It

2703 permits the definition of tail-recursive functions without the

2704 provision of a termination measure. The latter is necessary once the

2705 invariant proof rule for while is applied.

2707 * HOL: new (overloaded) notation for the set of elements below/above

2708 some element: {..u}, {..u(}, {l..}, {)l..}. See theory SetInterval.

2710 * HOL: theorems impI, allI, ballI bound as "strip";

2712 * HOL: new tactic induct_thm_tac: thm -> string -> int -> tactic

2713 induct_tac th "x1 ... xn" expects th to have a conclusion of the form

2714 P v1 ... vn and abbreviates res_inst_tac [("v1","x1"),...,("vn","xn")] th;

2716 * HOL/Real: "rabs" replaced by overloaded "abs" function;

2718 * HOL: theory Sexp now in HOL/Induct examples (it used to be part of

2719 main HOL, but was unused);

2721 * HOL: fewer consts declared as global (e.g. have to refer to

2722 "Lfp.lfp" instead of "lfp" internally; affects ML packages only);

2724 * HOL: tuned AST representation of nested pairs, avoiding bogus output

2725 in case of overlap with user translations (e.g. judgements over

2726 tuples); (note that the underlying logical represenation is still

2727 bogus);

2730 *** ZF ***

2732 * ZF: simplification automatically cancels common terms in arithmetic

2733 expressions over nat and int;

2735 * ZF: new treatment of nat to minimize type-checking: all operators

2736 coerce their operands to a natural number using the function natify,

2737 making the algebraic laws unconditional;

2739 * ZF: as above, for int: operators coerce their operands to an integer

2740 using the function intify;

2742 * ZF: the integer library now contains many of the usual laws for the

2743 orderings, including $<=, and monotonicity laws for $+ and $*;

2745 * ZF: new example ZF/ex/NatSum to demonstrate integer arithmetic

2746 simplification;

2748 * FOL and ZF: AddIffs now available, giving theorems of the form P<->Q

2749 to the simplifier and classical reasoner simultaneously;

2752 *** General ***

2754 * Provers: blast_tac now handles actual object-logic rules as

2755 assumptions; note that auto_tac uses blast_tac internally as well;

2757 * Provers: new functions rulify/rulify_no_asm: thm -> thm for turning

2758 outer -->/All/Ball into ==>/!!; qed_spec_mp now uses rulify_no_asm;

2760 * Provers: delrules now handles destruct rules as well (no longer need

2761 explicit make_elim);

2763 * Provers: Blast_tac now warns of and ignores "weak elimination rules" e.g.

2764 [| inj ?f; ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W

2765 use instead the strong form,

2766 [| inj ?f; ~ ?W ==> ?f ?x = ?f ?y; ?x = ?y ==> ?W |] ==> ?W

2767 in HOL, FOL and ZF the function cla_make_elim will create such rules

2768 from destruct-rules;

2770 * Provers: Simplifier.easy_setup provides a fast path to basic

2771 Simplifier setup for new object-logics;

2773 * Pure: AST translation rules no longer require constant head on LHS;

2775 * Pure: improved name spaces: ambiguous output is qualified; support

2776 for hiding of names;

2778 * system: smart setup of canonical ML_HOME, ISABELLE_INTERFACE, and

2779 XSYMBOL_HOME; no longer need to do manual configuration in most

2780 situations;

2782 * system: compression of ML heaps images may now be controlled via -c

2783 option of isabelle and isatool usedir (currently only observed by

2784 Poly/ML);

2786 * system: isatool installfonts may handle X-Symbol fonts as well (very

2787 useful for remote X11);

2789 * system: provide TAGS file for Isabelle sources;

2791 * ML: infix 'OF' is a version of 'MRS' with more appropriate argument

2792 order;

2794 * ML: renamed flags Syntax.trace_norm_ast to Syntax.trace_ast; global

2795 timing flag supersedes proof_timing and Toplevel.trace;

2797 * ML: new combinators |>> and |>>> for incremental transformations

2798 with secondary results (e.g. certain theory extensions):

2800 * ML: PureThy.add_defs gets additional argument to indicate potential

2801 overloading (usually false);

2803 * ML: PureThy.add_thms/add_axioms/add_defs now return theorems as

2804 results;

2808 New in Isabelle99 (October 1999)

2809 --------------------------------

2811 *** Overview of INCOMPATIBILITIES (see below for more details) ***

2813 * HOL: The THEN and ELSE parts of conditional expressions (if P then x else y)

2814 are no longer simplified. (This allows the simplifier to unfold recursive

2815 functional programs.) To restore the old behaviour, declare

2817 Delcongs [if_weak_cong];

2819 * HOL: Removed the obsolete syntax "Compl A"; use -A for set

2820 complement;

2822 * HOL: the predicate "inj" is now defined by translation to "inj_on";

2824 * HOL/datatype: mutual_induct_tac no longer exists --

2825 use induct_tac "x_1 ... x_n" instead of mutual_induct_tac ["x_1", ..., "x_n"]

2827 * HOL/typedef: fixed type inference for representing set; type

2828 arguments now have to occur explicitly on the rhs as type constraints;

2830 * ZF: The con_defs part of an inductive definition may no longer refer

2831 to constants declared in the same theory;

2833 * HOL, ZF: the function mk_cases, generated by the inductive

2834 definition package, has lost an argument. To simplify its result, it

2835 uses the default simpset instead of a supplied list of theorems.

2837 * HOL/List: the constructors of type list are now Nil and Cons;

2839 * Simplifier: the type of the infix ML functions

2840 setSSolver addSSolver setSolver addSolver

2841 is now simpset * solver -> simpset where `solver' is a new abstract type

2842 for packaging solvers. A solver is created via

2843 mk_solver: string -> (thm list -> int -> tactic) -> solver

2844 where the string argument is only a comment.

2847 *** Proof tools ***

2849 * Provers/Arith/fast_lin_arith.ML contains a functor for creating a

2850 decision procedure for linear arithmetic. Currently it is used for

2851 types `nat', `int', and `real' in HOL (see below); it can, should and

2852 will be instantiated for other types and logics as well.

2854 * The simplifier now accepts rewrite rules with flexible heads, eg

2855 hom ?f ==> ?f(?x+?y) = ?f ?x + ?f ?y

2856 They are applied like any rule with a non-pattern lhs, i.e. by first-order

2857 matching.

2860 *** General ***

2862 * New Isabelle/Isar subsystem provides an alternative to traditional

2863 tactical theorem proving; together with the ProofGeneral/isar user

2864 interface it offers an interactive environment for developing human

2865 readable proof documents (Isar == Intelligible semi-automated

2866 reasoning); for further information see isatool doc isar-ref,

2867 src/HOL/Isar_examples and http://isabelle.in.tum.de/Isar/

2869 * improved and simplified presentation of theories: better HTML markup

2870 (including colors), graph views in several sizes; isatool usedir now

2871 provides a proper interface for user theories (via -P option); actual

2872 document preparation based on (PDF)LaTeX is available as well (for

2873 new-style theories only); see isatool doc system for more information;

2875 * native support for Proof General, both for classic Isabelle and

2876 Isabelle/Isar;

2878 * ML function thm_deps visualizes dependencies of theorems and lemmas,

2879 using the graph browser tool;

2881 * Isabelle manuals now also available as PDF;

2883 * theory loader rewritten from scratch (may not be fully

2884 bug-compatible); old loadpath variable has been replaced by show_path,

2885 add_path, del_path, reset_path functions; new operations such as

2886 update_thy, touch_thy, remove_thy, use/update_thy_only (see also

2887 isatool doc ref);

2889 * improved isatool install: option -k creates KDE application icon,

2890 option -p DIR installs standalone binaries;

2892 * added ML_PLATFORM setting (useful for cross-platform installations);

2893 more robust handling of platform specific ML images for SML/NJ;

2895 * the settings environment is now statically scoped, i.e. it is never

2896 created again in sub-processes invoked from isabelle, isatool, or

2897 Isabelle;

2899 * path element specification '~~' refers to '$ISABELLE_HOME';

2901 * in locales, the "assumes" and "defines" parts may be omitted if

2902 empty;

2904 * new print_mode "xsymbols" for extended symbol support (e.g. genuine

2905 long arrows);

2907 * new print_mode "HTML";

2909 * new flag show_tags controls display of tags of theorems (which are

2910 basically just comments that may be attached by some tools);

2912 * Isamode 2.6 requires patch to accomodate change of Isabelle font

2913 mode and goal output format:

2915 diff -r Isamode-2.6/elisp/isa-load.el Isamode/elisp/isa-load.el

2916 244c244

2917 < (list (isa-getenv "ISABELLE") "-msymbols" logic-name)

2918 ---

2919 > (list (isa-getenv "ISABELLE") "-misabelle_font" "-msymbols" logic-name)

2920 diff -r Isabelle-2.6/elisp/isa-proofstate.el Isamode/elisp/isa-proofstate.el

2921 181c181

2922 < (defconst proofstate-proofstart-regexp "^Level [0-9]+$"

2923 ---

2924 > (defconst proofstate-proofstart-regexp "^Level [0-9]+"

2926 * function bind_thms stores lists of theorems (cf. bind_thm);

2928 * new shorthand tactics ftac, eatac, datac, fatac;

2930 * qed (and friends) now accept "" as result name; in that case the

2931 theorem is not stored, but proper checks and presentation of the

2932 result still apply;

2934 * theorem database now also indexes constants "Trueprop", "all",

2935 "==>", "=="; thus thms_containing, findI etc. may retrieve more rules;

2938 *** HOL ***

2940 ** HOL arithmetic **

2942 * There are now decision procedures for linear arithmetic over nat and

2943 int:

2945 1. arith_tac copes with arbitrary formulae involving `=', `<', `<=',

2946 `+', `-', `Suc', `min', `max' and numerical constants; other subterms

2947 are treated as atomic; subformulae not involving type `nat' or `int'

2948 are ignored; quantified subformulae are ignored unless they are

2949 positive universal or negative existential. The tactic has to be

2950 invoked by hand and can be a little bit slow. In particular, the

2951 running time is exponential in the number of occurrences of `min' and

2952 `max', and `-' on `nat'.

2954 2. fast_arith_tac is a cut-down version of arith_tac: it only takes

2955 (negated) (in)equalities among the premises and the conclusion into

2956 account (i.e. no compound formulae) and does not know about `min' and

2957 `max', and `-' on `nat'. It is fast and is used automatically by the

2958 simplifier.

2960 NB: At the moment, these decision procedures do not cope with mixed

2961 nat/int formulae where the two parts interact, such as `m < n ==>

2962 int(m) < int(n)'.

2964 * HOL/Numeral provides a generic theory of numerals (encoded

2965 efficiently as bit strings); setup for types nat/int/real is in place;

2966 INCOMPATIBILITY: since numeral syntax is now polymorphic, rather than

2967 int, existing theories and proof scripts may require a few additional

2968 type constraints;

2970 * integer division and remainder can now be performed on constant

2971 arguments;

2973 * many properties of integer multiplication, division and remainder

2974 are now available;

2976 * An interface to the Stanford Validity Checker (SVC) is available through the

2977 tactic svc_tac. Propositional tautologies and theorems of linear arithmetic

2978 are proved automatically. SVC must be installed separately, and its results

2979 must be TAKEN ON TRUST (Isabelle does not check the proofs, but tags any

2980 invocation of the underlying oracle). For SVC see

2981 http://verify.stanford.edu/SVC

2983 * IsaMakefile: the HOL-Real target now builds an actual image;

2986 ** HOL misc **

2988 * HOL/Real/HahnBanach: the Hahn-Banach theorem for real vector spaces

2989 (in Isabelle/Isar) -- by Gertrud Bauer;

2991 * HOL/BCV: generic model of bytecode verification, i.e. data-flow

2992 analysis for assembly languages with subtypes;

2994 * HOL/TLA (Lamport's Temporal Logic of Actions): major reorganization

2995 -- avoids syntactic ambiguities and treats state, transition, and

2996 temporal levels more uniformly; introduces INCOMPATIBILITIES due to

2997 changed syntax and (many) tactics;

2999 * HOL/inductive: Now also handles more general introduction rules such

3000 as "ALL y. (y, x) : r --> y : acc r ==> x : acc r"; monotonicity

3001 theorems are now maintained within the theory (maintained via the

3002 "mono" attribute);

3004 * HOL/datatype: Now also handles arbitrarily branching datatypes

3005 (using function types) such as

3007 datatype 'a tree = Atom 'a | Branch "nat => 'a tree"

3009 * HOL/record: record_simproc (part of the default simpset) takes care

3010 of selectors applied to updated records; record_split_tac is no longer

3011 part of the default claset; update_defs may now be removed from the

3012 simpset in many cases; COMPATIBILITY: old behavior achieved by

3014 claset_ref () := claset() addSWrapper record_split_wrapper;

3015 Delsimprocs [record_simproc]

3017 * HOL/typedef: fixed type inference for representing set; type

3018 arguments now have to occur explicitly on the rhs as type constraints;

3020 * HOL/recdef (TFL): 'congs' syntax now expects comma separated list of theorem

3021 names rather than an ML expression;

3023 * HOL/defer_recdef (TFL): like recdef but the well-founded relation can be

3024 supplied later. Program schemes can be defined, such as

3025 "While B C s = (if B s then While B C (C s) else s)"

3026 where the well-founded relation can be chosen after B and C have been given.

3028 * HOL/List: the constructors of type list are now Nil and Cons;

3029 INCOMPATIBILITY: while [] and infix # syntax is still there, of

3030 course, ML tools referring to List.list.op # etc. have to be adapted;

3032 * HOL_quantifiers flag superseded by "HOL" print mode, which is

3033 disabled by default; run isabelle with option -m HOL to get back to

3034 the original Gordon/HOL-style output;

3036 * HOL/Ord.thy: new bounded quantifier syntax (input only): ALL x<y. P,

3037 ALL x<=y. P, EX x<y. P, EX x<=y. P;

3039 * HOL basic syntax simplified (more orthogonal): all variants of

3040 All/Ex now support plain / symbolic / HOL notation; plain syntax for

3041 Eps operator is provided as well: "SOME x. P[x]";

3043 * HOL/Sum.thy: sum_case has been moved to HOL/Datatype;

3045 * HOL/Univ.thy: infix syntax <*>, <+>, <**>, <+> eliminated and made

3046 thus available for user theories;

3048 * HOLCF/IOA/Sequents: renamed 'Cons' to 'Consq' to avoid clash with

3049 HOL/List; hardly an INCOMPATIBILITY since '>>' syntax is used all the

3050 time;

3052 * HOL: new tactic smp_tac: int -> int -> tactic, which applies spec

3053 several times and then mp;

3056 *** LK ***

3058 * the notation <<...>> is now available as a notation for sequences of

3059 formulas;

3061 * the simplifier is now installed

3063 * the axiom system has been generalized (thanks to Soren Heilmann)

3065 * the classical reasoner now has a default rule database

3068 *** ZF ***

3070 * new primrec section allows primitive recursive functions to be given

3071 directly (as in HOL) over datatypes and the natural numbers;

3073 * new tactics induct_tac and exhaust_tac for induction (or case

3074 analysis) over datatypes and the natural numbers;

3076 * the datatype declaration of type T now defines the recursor T_rec;

3078 * simplification automatically does freeness reasoning for datatype

3079 constructors;

3081 * automatic type-inference, with AddTCs command to insert new

3082 type-checking rules;

3084 * datatype introduction rules are now added as Safe Introduction rules

3085 to the claset;

3087 * the syntax "if P then x else y" is now available in addition to

3088 if(P,x,y);

3091 *** Internal programming interfaces ***

3093 * tuned simplifier trace output; new flag debug_simp;

3095 * structures Vartab / Termtab (instances of TableFun) offer efficient

3096 tables indexed by indexname_ord / term_ord (compatible with aconv);

3098 * AxClass.axclass_tac lost the theory argument;

3100 * tuned current_goals_markers semantics: begin / end goal avoids

3101 printing empty lines;

3103 * removed prs and prs_fn hook, which was broken because it did not

3104 include \n in its semantics, forcing writeln to add one

3105 uncoditionally; replaced prs_fn by writeln_fn; consider std_output:

3106 string -> unit if you really want to output text without newline;

3108 * Symbol.output subject to print mode; INCOMPATIBILITY: defaults to

3109 plain output, interface builders may have to enable 'isabelle_font'

3110 mode to get Isabelle font glyphs as before;

3112 * refined token_translation interface; INCOMPATIBILITY: output length

3113 now of type real instead of int;

3115 * theory loader actions may be traced via new ThyInfo.add_hook

3116 interface (see src/Pure/Thy/thy_info.ML); example application: keep

3117 your own database of information attached to *whole* theories -- as

3118 opposed to intra-theory data slots offered via TheoryDataFun;

3120 * proper handling of dangling sort hypotheses (at last!);

3121 Thm.strip_shyps and Drule.strip_shyps_warning take care of removing

3122 extra sort hypotheses that can be witnessed from the type signature;

3123 the force_strip_shyps flag is gone, any remaining shyps are simply

3124 left in the theorem (with a warning issued by strip_shyps_warning);

3128 New in Isabelle98-1 (October 1998)

3129 ----------------------------------

3131 *** Overview of INCOMPATIBILITIES (see below for more details) ***

3133 * several changes of automated proof tools;

3135 * HOL: major changes to the inductive and datatype packages, including

3136 some minor incompatibilities of theory syntax;

3138 * HOL: renamed r^-1 to 'converse' from 'inverse'; 'inj_onto' is now

3139 called `inj_on';

3141 * HOL: removed duplicate thms in Arith:

3142 less_imp_add_less should be replaced by trans_less_add1

3143 le_imp_add_le should be replaced by trans_le_add1

3145 * HOL: unary minus is now overloaded (new type constraints may be

3146 required);

3148 * HOL and ZF: unary minus for integers is now #- instead of #~. In

3149 ZF, expressions such as n#-1 must be changed to n#- 1, since #-1 is

3150 now taken as an integer constant.

3152 * Pure: ML function 'theory_of' renamed to 'theory';

3155 *** Proof tools ***

3157 * Simplifier:

3158 1. Asm_full_simp_tac is now more aggressive.

3159 1. It will sometimes reorient premises if that increases their power to

3160 simplify.

3161 2. It does no longer proceed strictly from left to right but may also

3162 rotate premises to achieve further simplification.

3163 For compatibility reasons there is now Asm_lr_simp_tac which is like the

3164 old Asm_full_simp_tac in that it does not rotate premises.

3165 2. The simplifier now knows a little bit about nat-arithmetic.

3167 * Classical reasoner: wrapper mechanism for the classical reasoner now

3168 allows for selected deletion of wrappers, by introduction of names for

3169 wrapper functionals. This implies that addbefore, addSbefore,

3170 addaltern, and addSaltern now take a pair (name, tactic) as argument,

3171 and that adding two tactics with the same name overwrites the first

3172 one (emitting a warning).

3173 type wrapper = (int -> tactic) -> (int -> tactic)

3174 setWrapper, setSWrapper, compWrapper and compSWrapper are replaced by

3175 addWrapper, addSWrapper: claset * (string * wrapper) -> claset

3176 delWrapper, delSWrapper: claset * string -> claset

3177 getWrapper is renamed to appWrappers, getSWrapper to appSWrappers;

3179 * Classical reasoner: addbefore/addSbefore now have APPEND/ORELSE

3180 semantics; addbefore now affects only the unsafe part of step_tac

3181 etc.; this affects addss/auto_tac/force_tac, so EXISTING PROOFS MAY

3182 FAIL, but proofs should be fixable easily, e.g. by replacing Auto_tac

3183 by Force_tac;

3185 * Classical reasoner: setwrapper to setWrapper and compwrapper to

3186 compWrapper; added safe wrapper (and access functions for it);

3188 * HOL/split_all_tac is now much faster and fails if there is nothing

3189 to split. Some EXISTING PROOFS MAY REQUIRE ADAPTION because the order

3190 and the names of the automatically generated variables have changed.

3191 split_all_tac has moved within claset() from unsafe wrappers to safe

3192 wrappers, which means that !!-bound variables are split much more

3193 aggressively, and safe_tac and clarify_tac now split such variables.

3194 If this splitting is not appropriate, use delSWrapper "split_all_tac".

3195 Note: the same holds for record_split_tac, which does the job of

3196 split_all_tac for record fields.

3198 * HOL/Simplifier: Rewrite rules for case distinctions can now be added

3199 permanently to the default simpset using Addsplits just like

3200 Addsimps. They can be removed via Delsplits just like

3201 Delsimps. Lower-case versions are also available.

3203 * HOL/Simplifier: The rule split_if is now part of the default

3204 simpset. This means that the simplifier will eliminate all occurrences

3205 of if-then-else in the conclusion of a goal. To prevent this, you can

3206 either remove split_if completely from the default simpset by

3207 `Delsplits [split_if]' or remove it in a specific call of the

3208 simplifier using `... delsplits [split_if]'. You can also add/delete

3209 other case splitting rules to/from the default simpset: every datatype

3210 generates suitable rules `split_t_case' and `split_t_case_asm' (where

3211 t is the name of the datatype).

3213 * Classical reasoner / Simplifier combination: new force_tac (and

3214 derivatives Force_tac, force) combines rewriting and classical

3215 reasoning (and whatever other tools) similarly to auto_tac, but is

3216 aimed to solve the given subgoal completely.

3219 *** General ***

3221 * new top-level commands `Goal' and `Goalw' that improve upon `goal'

3222 and `goalw': the theory is no longer needed as an explicit argument -

3223 the current theory context is used; assumptions are no longer returned

3224 at the ML-level unless one of them starts with ==> or !!; it is

3225 recommended to convert to these new commands using isatool fixgoal

3226 (backup your sources first!);

3228 * new top-level commands 'thm' and 'thms' for retrieving theorems from

3229 the current theory context, and 'theory' to lookup stored theories;

3231 * new theory section 'locale' for declaring constants, assumptions and

3232 definitions that have local scope;

3234 * new theory section 'nonterminals' for purely syntactic types;

3236 * new theory section 'setup' for generic ML setup functions

3237 (e.g. package initialization);

3239 * the distribution now includes Isabelle icons: see

3240 lib/logo/isabelle-{small,tiny}.xpm;

3242 * isatool install - install binaries with absolute references to

3243 ISABELLE_HOME/bin;

3245 * isatool logo -- create instances of the Isabelle logo (as EPS);

3247 * print mode 'emacs' reserved for Isamode;

3249 * support multiple print (ast) translations per constant name;

3251 * theorems involving oracles are now printed with a suffixed [!];

3254 *** HOL ***

3256 * there is now a tutorial on Isabelle/HOL (do 'isatool doc tutorial');

3258 * HOL/inductive package reorganized and improved: now supports mutual

3259 definitions such as

3261 inductive EVEN ODD

3262 intrs

3263 null "0 : EVEN"

3264 oddI "n : EVEN ==> Suc n : ODD"

3265 evenI "n : ODD ==> Suc n : EVEN"

3267 new theorem list "elims" contains an elimination rule for each of the

3268 recursive sets; inductive definitions now handle disjunctive premises

3269 correctly (also ZF);

3271 INCOMPATIBILITIES: requires Inductive as an ancestor; component

3272 "mutual_induct" no longer exists - the induction rule is always

3273 contained in "induct";

3276 * HOL/datatype package re-implemented and greatly improved: now

3277 supports mutually recursive datatypes such as

3279 datatype

3280 'a aexp = IF_THEN_ELSE ('a bexp) ('a aexp) ('a aexp)

3281 | SUM ('a aexp) ('a aexp)

3282 | DIFF ('a aexp) ('a aexp)

3283 | NUM 'a

3284 and

3285 'a bexp = LESS ('a aexp) ('a aexp)

3286 | AND ('a bexp) ('a bexp)

3287 | OR ('a bexp) ('a bexp)

3289 as well as indirectly recursive datatypes such as

3291 datatype

3292 ('a, 'b) term = Var 'a

3293 | App 'b ((('a, 'b) term) list)

3295 The new tactic mutual_induct_tac [<var_1>, ..., <var_n>] i performs

3296 induction on mutually / indirectly recursive datatypes.

3298 Primrec equations are now stored in theory and can be accessed via

3299 <function_name>.simps.

3301 INCOMPATIBILITIES:

3303 - Theories using datatypes must now have theory Datatype as an

3304 ancestor.

3305 - The specific <typename>.induct_tac no longer exists - use the

3306 generic induct_tac instead.

3307 - natE has been renamed to nat.exhaust - use exhaust_tac

3308 instead of res_inst_tac ... natE. Note that the variable

3309 names in nat.exhaust differ from the names in natE, this

3310 may cause some "fragile" proofs to fail.

3311 - The theorems split_<typename>_case and split_<typename>_case_asm

3312 have been renamed to <typename>.split and <typename>.split_asm.

3313 - Since default sorts of type variables are now handled correctly,

3314 some datatype definitions may have to be annotated with explicit

3315 sort constraints.

3316 - Primrec definitions no longer require function name and type

3317 of recursive argument.

3319 Consider using isatool fixdatatype to adapt your theories and proof

3320 scripts to the new package (backup your sources first!).

3323 * HOL/record package: considerably improved implementation; now

3324 includes concrete syntax for record types, terms, updates; theorems

3325 for surjective pairing and splitting !!-bound record variables; proof

3326 support is as follows:

3328 1) standard conversions (selectors or updates applied to record

3329 constructor terms) are part of the standard simpset;

3331 2) inject equations of the form ((x, y) = (x', y')) == x=x' & y=y' are

3332 made part of standard simpset and claset via addIffs;

3334 3) a tactic for record field splitting (record_split_tac) is part of

3335 the standard claset (addSWrapper);

3337 To get a better idea about these rules you may retrieve them via

3338 something like 'thms "foo.simps"' or 'thms "foo.iffs"', where "foo" is

3339 the name of your record type.

3341 The split tactic 3) conceptually simplifies by the following rule:

3343 "(!!x. PROP ?P x) == (!!a b. PROP ?P (a, b))"

3345 Thus any record variable that is bound by meta-all will automatically

3346 blow up into some record constructor term, consequently the

3347 simplifications of 1), 2) apply. Thus force_tac, auto_tac etc. shall

3348 solve record problems automatically.

3351 * reorganized the main HOL image: HOL/Integ and String loaded by

3352 default; theory Main includes everything;

3354 * automatic simplification of integer sums and comparisons, using cancellation;

3356 * added option_map_eq_Some and not_Some_eq to the default simpset and claset;

3358 * added disj_not1 = "(~P | Q) = (P --> Q)" to the default simpset;

3360 * many new identities for unions, intersections, set difference, etc.;

3362 * expand_if, expand_split, expand_sum_case and expand_nat_case are now

3363 called split_if, split_split, split_sum_case and split_nat_case (to go

3364 with add/delsplits);

3366 * HOL/Prod introduces simplification procedure unit_eq_proc rewriting

3367 (?x::unit) = (); this is made part of the default simpset, which COULD

3368 MAKE EXISTING PROOFS FAIL under rare circumstances (consider

3369 'Delsimprocs [unit_eq_proc];' as last resort); also note that

3370 unit_abs_eta_conv is added in order to counter the effect of

3371 unit_eq_proc on (%u::unit. f u), replacing it by f rather than by

3372 %u.f();

3374 * HOL/Fun INCOMPATIBILITY: `inj_onto' is now called `inj_on' (which

3375 makes more sense);

3377 * HOL/Set INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;

3378 It and 'sym RS equals0D' are now in the default claset, giving automatic

3379 disjointness reasoning but breaking a few old proofs.

3381 * HOL/Relation INCOMPATIBILITY: renamed the relational operator r^-1

3382 to 'converse' from 'inverse' (for compatibility with ZF and some

3383 literature);

3385 * HOL/recdef can now declare non-recursive functions, with {} supplied as

3386 the well-founded relation;

3388 * HOL/Set INCOMPATIBILITY: the complement of set A is now written -A instead of

3389 Compl A. The "Compl" syntax remains available as input syntax for this

3390 release ONLY.

3392 * HOL/Update: new theory of function updates:

3393 f(a:=b) == %x. if x=a then b else f x

3394 may also be iterated as in f(a:=b,c:=d,...);

3396 * HOL/Vimage: new theory for inverse image of a function, syntax f-``B;

3398 * HOL/List:

3399 - new function list_update written xs[i:=v] that updates the i-th

3400 list position. May also be iterated as in xs[i:=a,j:=b,...].

3401 - new function `upt' written [i..j(] which generates the list

3402 [i,i+1,...,j-1], i.e. the upper bound is excluded. To include the upper

3403 bound write [i..j], which is a shorthand for [i..j+1(].

3404 - new lexicographic orderings and corresponding wellfoundedness theorems.

3406 * HOL/Arith:

3407 - removed 'pred' (predecessor) function;

3408 - generalized some theorems about n-1;

3409 - many new laws about "div" and "mod";

3410 - new laws about greatest common divisors (see theory ex/Primes);

3412 * HOL/Relation: renamed the relational operator r^-1 "converse"

3413 instead of "inverse";

3415 * HOL/Induct/Multiset: a theory of multisets, including the wellfoundedness

3416 of the multiset ordering;

3418 * directory HOL/Real: a construction of the reals using Dedekind cuts

3419 (not included by default);

3421 * directory HOL/UNITY: Chandy and Misra's UNITY formalism;

3423 * directory HOL/Hoare: a new version of Hoare logic which permits many-sorted

3424 programs, i.e. different program variables may have different types.

3426 * calling (stac rew i) now fails if "rew" has no effect on the goal

3427 [previously, this check worked only if the rewrite rule was unconditional]

3428 Now rew can involve either definitions or equalities (either == or =).

3431 *** ZF ***

3433 * theory Main includes everything; INCOMPATIBILITY: theory ZF.thy contains

3434 only the theorems proved on ZF.ML;

3436 * ZF INCOMPATIBILITY: rule `equals0D' is now a well-formed destruct rule;

3437 It and 'sym RS equals0D' are now in the default claset, giving automatic

3438 disjointness reasoning but breaking a few old proofs.

3440 * ZF/Update: new theory of function updates

3441 with default rewrite rule f(x:=y) ` z = if(z=x, y, f`z)

3442 may also be iterated as in f(a:=b,c:=d,...);

3444 * in let x=t in u(x), neither t nor u(x) has to be an FOL term.

3446 * calling (stac rew i) now fails if "rew" has no effect on the goal

3447 [previously, this check worked only if the rewrite rule was unconditional]

3448 Now rew can involve either definitions or equalities (either == or =).

3450 * case_tac provided for compatibility with HOL

3451 (like the old excluded_middle_tac, but with subgoals swapped)

3454 *** Internal programming interfaces ***

3456 * Pure: several new basic modules made available for general use, see

3457 also src/Pure/README;

3459 * improved the theory data mechanism to support encapsulation (data

3460 kind name replaced by private Object.kind, acting as authorization

3461 key); new type-safe user interface via functor TheoryDataFun; generic

3462 print_data function becomes basically useless;

3464 * removed global_names compatibility flag -- all theory declarations

3465 are qualified by default;

3467 * module Pure/Syntax now offers quote / antiquote translation

3468 functions (useful for Hoare logic etc. with implicit dependencies);

3469 see HOL/ex/Antiquote for an example use;

3471 * Simplifier now offers conversions (asm_)(full_)rewrite: simpset ->

3472 cterm -> thm;

3474 * new tactical CHANGED_GOAL for checking that a tactic modifies a

3475 subgoal;

3477 * Display.print_goals function moved to Locale.print_goals;

3479 * standard print function for goals supports current_goals_markers

3480 variable for marking begin of proof, end of proof, start of goal; the

3481 default is ("", "", ""); setting current_goals_markers := ("<proof>",

3482 "</proof>", "<goal>") causes SGML like tagged proof state printing,

3483 for example;

3487 New in Isabelle98 (January 1998)

3488 --------------------------------

3490 *** Overview of INCOMPATIBILITIES (see below for more details) ***

3492 * changed lexical syntax of terms / types: dots made part of long

3493 identifiers, e.g. "%x.x" no longer possible, should be "%x. x";

3495 * simpset (and claset) reference variable replaced by functions

3496 simpset / simpset_ref;

3498 * no longer supports theory aliases (via merge) and non-trivial

3499 implicit merge of thms' signatures;

3501 * most internal names of constants changed due to qualified names;

3503 * changed Pure/Sequence interface (see Pure/seq.ML);

3506 *** General Changes ***

3508 * hierachically structured name spaces (for consts, types, axms, thms

3509 etc.); new lexical class 'longid' (e.g. Foo.bar.x) may render much of

3510 old input syntactically incorrect (e.g. "%x.x"); COMPATIBILITY:

3511 isatool fixdots ensures space after dots (e.g. "%x. x"); set

3512 long_names for fully qualified output names; NOTE: ML programs

3513 (special tactics, packages etc.) referring to internal names may have

3514 to be adapted to cope with fully qualified names; in case of severe

3515 backward campatibility problems try setting 'global_names' at compile

3516 time to have enrything declared within a flat name space; one may also

3517 fine tune name declarations in theories via the 'global' and 'local'

3518 section;

3520 * reimplemented the implicit simpset and claset using the new anytype

3521 data filed in signatures; references simpset:simpset ref etc. are

3522 replaced by functions simpset:unit->simpset and

3523 simpset_ref:unit->simpset ref; COMPATIBILITY: use isatool fixclasimp

3524 to patch your ML files accordingly;

3526 * HTML output now includes theory graph data for display with Java

3527 applet or isatool browser; data generated automatically via isatool

3528 usedir (see -i option, ISABELLE_USEDIR_OPTIONS);

3530 * defs may now be conditional; improved rewrite_goals_tac to handle

3531 conditional equations;

3533 * defs now admits additional type arguments, using TYPE('a) syntax;

3535 * theory aliases via merge (e.g. M=A+B+C) no longer supported, always

3536 creates a new theory node; implicit merge of thms' signatures is

3537 restricted to 'trivial' ones; COMPATIBILITY: one may have to use

3538 transfer:theory->thm->thm in (rare) cases;

3540 * improved handling of draft signatures / theories; draft thms (and

3541 ctyps, cterms) are automatically promoted to real ones;

3543 * slightly changed interfaces for oracles: admit many per theory, named

3544 (e.g. oracle foo = mlfun), additional name argument for invoke_oracle;

3546 * print_goals: optional output of const types (set show_consts and

3547 show_types);

3549 * improved output of warnings (###) and errors (***);

3551 * subgoal_tac displays a warning if the new subgoal has type variables;

3553 * removed old README and Makefiles;

3555 * replaced print_goals_ref hook by print_current_goals_fn and result_error_fn;

3557 * removed obsolete init_pps and init_database;

3559 * deleted the obsolete tactical STATE, which was declared by

3560 fun STATE tacfun st = tacfun st st;

3562 * cd and use now support path variables, e.g. $ISABELLE_HOME, or ~

3563 (which abbreviates $HOME);

3565 * changed Pure/Sequence interface (see Pure/seq.ML); COMPATIBILITY:

3566 use isatool fixseq to adapt your ML programs (this works for fully

3567 qualified references to the Sequence structure only!);

3569 * use_thy no longer requires writable current directory; it always

3570 reloads .ML *and* .thy file, if either one is out of date;

3573 *** Classical Reasoner ***

3575 * Clarify_tac, clarify_tac, clarify_step_tac, Clarify_step_tac: new

3576 tactics that use classical reasoning to simplify a subgoal without

3577 splitting it into several subgoals;

3579 * Safe_tac: like safe_tac but uses the default claset;

3582 *** Simplifier ***

3584 * added simplification meta rules:

3585 (asm_)(full_)simplify: simpset -> thm -> thm;

3587 * simplifier.ML no longer part of Pure -- has to be loaded by object

3588 logics (again);

3590 * added prems argument to simplification procedures;

3592 * HOL, FOL, ZF: added infix function `addsplits':

3593 instead of `<simpset> setloop (split_tac <thms>)'

3594 you can simply write `<simpset> addsplits <thms>'

3597 *** Syntax ***

3599 * TYPE('a) syntax for type reflection terms;

3601 * no longer handles consts with name "" -- declare as 'syntax' instead;

3603 * pretty printer: changed order of mixfix annotation preference (again!);

3605 * Pure: fixed idt/idts vs. pttrn/pttrns syntactic categories;

3608 *** HOL ***

3610 * HOL: there is a new splitter `split_asm_tac' that can be used e.g.

3611 with `addloop' of the simplifier to faciliate case splitting in premises.

3613 * HOL/TLA: Stephan Merz's formalization of Lamport's Temporal Logic of Actions;

3615 * HOL/Auth: new protocol proofs including some for the Internet

3616 protocol TLS;

3618 * HOL/Map: new theory of `maps' a la VDM;

3620 * HOL/simplifier: simplification procedures nat_cancel_sums for

3621 cancelling out common nat summands from =, <, <= (in)equalities, or

3622 differences; simplification procedures nat_cancel_factor for

3623 cancelling common factor from =, <, <= (in)equalities over natural

3624 sums; nat_cancel contains both kinds of procedures, it is installed by

3625 default in Arith.thy -- this COULD MAKE EXISTING PROOFS FAIL;

3627 * HOL/simplifier: terms of the form

3628 `? x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x)' (or t=x)

3629 are rewritten to

3630 `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t)',

3631 and those of the form

3632 `! x. P1(x) & ... & Pn(x) & x=t & Q1(x) & ... Qn(x) --> R(x)' (or t=x)

3633 are rewritten to

3634 `P1(t) & ... & Pn(t) & Q1(t) & ... Qn(t) --> R(t)',

3636 * HOL/datatype

3637 Each datatype `t' now comes with a theorem `split_t_case' of the form

3639 P(t_case f1 ... fn x) =

3640 ( (!y1 ... ym1. x = C1 y1 ... ym1 --> P(f1 y1 ... ym1)) &

3641 ...

3642 (!y1 ... ymn. x = Cn y1 ... ymn --> P(f1 y1 ... ymn))

3643 )

3645 and a theorem `split_t_case_asm' of the form

3647 P(t_case f1 ... fn x) =

3648 ~( (? y1 ... ym1. x = C1 y1 ... ym1 & ~P(f1 y1 ... ym1)) |

3649 ...

3650 (? y1 ... ymn. x = Cn y1 ... ymn & ~P(f1 y1 ... ymn))

3651 )

3652 which can be added to a simpset via `addsplits'. The existing theorems

3653 expand_list_case and expand_option_case have been renamed to

3654 split_list_case and split_option_case.

3656 * HOL/Arithmetic:

3657 - `pred n' is automatically converted to `n-1'.

3658 Users are strongly encouraged not to use `pred' any longer,

3659 because it will disappear altogether at some point.

3660 - Users are strongly encouraged to write "0 < n" rather than

3661 "n ~= 0". Theorems and proof tools have been modified towards this

3662 `standard'.

3664 * HOL/Lists:

3665 the function "set_of_list" has been renamed "set" (and its theorems too);

3666 the function "nth" now takes its arguments in the reverse order and

3667 has acquired the infix notation "!" as in "xs!n".

3669 * HOL/Set: UNIV is now a constant and is no longer translated to Compl{};

3671 * HOL/Set: The operator (UN x.B x) now abbreviates (UN x:UNIV. B x) and its

3672 specialist theorems (like UN1_I) are gone. Similarly for (INT x.B x);

3674 * HOL/record: extensible records with schematic structural subtyping

3675 (single inheritance); EXPERIMENTAL version demonstrating the encoding,

3676 still lacks various theorems and concrete record syntax;

3679 *** HOLCF ***

3681 * removed "axioms" and "generated by" sections;

3683 * replaced "ops" section by extended "consts" section, which is capable of

3684 handling the continuous function space "->" directly;

3686 * domain package:

3687 . proves theorems immediately and stores them in the theory,

3688 . creates hierachical name space,

3689 . now uses normal mixfix annotations (instead of cinfix...),

3690 . minor changes to some names and values (for consistency),

3691 . e.g. cases -> casedist, dists_eq -> dist_eqs, [take_lemma] -> take_lemmas,

3692 . separator between mutual domain defs: changed "," to "and",

3693 . improved handling of sort constraints; now they have to

3694 appear on the left-hand side of the equations only;

3696 * fixed LAM <x,y,zs>.b syntax;

3698 * added extended adm_tac to simplifier in HOLCF -- can now discharge

3699 adm (%x. P (t x)), where P is chainfinite and t continuous;

3702 *** FOL and ZF ***

3704 * FOL: there is a new splitter `split_asm_tac' that can be used e.g.

3705 with `addloop' of the simplifier to faciliate case splitting in premises.

3707 * qed_spec_mp, qed_goal_spec_mp, qed_goalw_spec_mp are available, as

3708 in HOL, they strip ALL and --> from proved theorems;

3712 New in Isabelle94-8 (May 1997)

3713 ------------------------------

3715 *** General Changes ***

3717 * new utilities to build / run / maintain Isabelle etc. (in parts

3718 still somewhat experimental); old Makefiles etc. still functional;

3720 * new 'Isabelle System Manual';

3722 * INSTALL text, together with ./configure and ./build scripts;

3724 * reimplemented type inference for greater efficiency, better error

3725 messages and clean internal interface;

3727 * prlim command for dealing with lots of subgoals (an easier way of

3728 setting goals_limit);

3731 *** Syntax ***

3733 * supports alternative (named) syntax tables (parser and pretty

3734 printer); internal interface is provided by add_modesyntax(_i);

3736 * Pure, FOL, ZF, HOL, HOLCF now support symbolic input and output; to

3737 be used in conjunction with the Isabelle symbol font; uses the

3738 "symbols" syntax table;

3740 * added token_translation interface (may translate name tokens in

3741 arbitrary ways, dependent on their type (free, bound, tfree, ...) and

3742 the current print_mode); IMPORTANT: user print translation functions

3743 are responsible for marking newly introduced bounds

3744 (Syntax.mark_boundT);

3746 * token translations for modes "xterm" and "xterm_color" that display

3747 names in bold, underline etc. or colors (which requires a color

3748 version of xterm);

3750 * infixes may now be declared with names independent of their syntax;

3752 * added typed_print_translation (like print_translation, but may

3753 access type of constant);

3756 *** Classical Reasoner ***

3758 Blast_tac: a new tactic! It is often more powerful than fast_tac, but has

3759 some limitations. Blast_tac...

3760 + ignores addss, addbefore, addafter; this restriction is intrinsic

3761 + ignores elimination rules that don't have the correct format

3762 (the conclusion MUST be a formula variable)

3763 + ignores types, which can make HOL proofs fail

3764 + rules must not require higher-order unification, e.g. apply_type in ZF

3765 [message "Function Var's argument not a bound variable" relates to this]

3766 + its proof strategy is more general but can actually be slower

3768 * substitution with equality assumptions no longer permutes other

3769 assumptions;

3771 * minor changes in semantics of addafter (now called addaltern); renamed

3772 setwrapper to setWrapper and compwrapper to compWrapper; added safe wrapper

3773 (and access functions for it);

3775 * improved combination of classical reasoner and simplifier:

3776 + functions for handling clasimpsets

3777 + improvement of addss: now the simplifier is called _after_ the

3778 safe steps.

3779 + safe variant of addss called addSss: uses safe simplifications

3780 _during_ the safe steps. It is more complete as it allows multiple

3781 instantiations of unknowns (e.g. with slow_tac).

3783 *** Simplifier ***

3785 * added interface for simplification procedures (functions that

3786 produce *proven* rewrite rules on the fly, depending on current

3787 redex);

3789 * ordering on terms as parameter (used for ordered rewriting);

3791 * new functions delcongs, deleqcongs, and Delcongs. richer rep_ss;

3793 * the solver is now split into a safe and an unsafe part.

3794 This should be invisible for the normal user, except that the

3795 functions setsolver and addsolver have been renamed to setSolver and

3796 addSolver; added safe_asm_full_simp_tac;

3799 *** HOL ***

3801 * a generic induction tactic `induct_tac' which works for all datatypes and

3802 also for type `nat';

3804 * a generic case distinction tactic `exhaust_tac' which works for all

3805 datatypes and also for type `nat';

3807 * each datatype comes with a function `size';

3809 * patterns in case expressions allow tuple patterns as arguments to

3810 constructors, for example `case x of [] => ... | (x,y,z)#ps => ...';

3812 * primrec now also works with type nat;

3814 * recdef: a new declaration form, allows general recursive functions to be

3815 defined in theory files. See HOL/ex/Fib, HOL/ex/Primes, HOL/Subst/Unify.

3817 * the constant for negation has been renamed from "not" to "Not" to

3818 harmonize with FOL, ZF, LK, etc.;

3820 * HOL/ex/LFilter theory of a corecursive "filter" functional for

3821 infinite lists;

3823 * HOL/Modelcheck demonstrates invocation of model checker oracle;

3825 * HOL/ex/Ring.thy declares cring_simp, which solves equational

3826 problems in commutative rings, using axiomatic type classes for + and *;

3828 * more examples in HOL/MiniML and HOL/Auth;

3830 * more default rewrite rules for quantifiers, union/intersection;

3832 * a new constant `arbitrary == @x.False';

3834 * HOLCF/IOA replaces old HOL/IOA;

3836 * HOLCF changes: derived all rules and arities

3837 + axiomatic type classes instead of classes

3838 + typedef instead of faking type definitions

3839 + eliminated the internal constants less_fun, less_cfun, UU_fun, UU_cfun etc.

3840 + new axclasses cpo, chfin, flat with flat < chfin < pcpo < cpo < po

3841 + eliminated the types void, one, tr

3842 + use unit lift and bool lift (with translations) instead of one and tr

3843 + eliminated blift from Lift3.thy (use Def instead of blift)

3844 all eliminated rules are derived as theorems --> no visible changes ;

3847 *** ZF ***

3849 * ZF now has Fast_tac, Simp_tac and Auto_tac. Union_iff is a now a default

3850 rewrite rule; this may affect some proofs. eq_cs is gone but can be put back

3851 as ZF_cs addSIs [equalityI];

3855 New in Isabelle94-7 (November 96)

3856 ---------------------------------

3858 * allowing negative levels (as offsets) in prlev and choplev;

3860 * super-linear speedup for large simplifications;

3862 * FOL, ZF and HOL now use miniscoping: rewriting pushes

3863 quantifications in as far as possible (COULD MAKE EXISTING PROOFS

3864 FAIL); can suppress it using the command Delsimps (ex_simps @

3865 all_simps); De Morgan laws are also now included, by default;

3867 * improved printing of ==> : ~:

3869 * new object-logic "Sequents" adds linear logic, while replacing LK

3870 and Modal (thanks to Sara Kalvala);

3872 * HOL/Auth: correctness proofs for authentication protocols;

3874 * HOL: new auto_tac combines rewriting and classical reasoning (many

3875 examples on HOL/Auth);

3877 * HOL: new command AddIffs for declaring theorems of the form P=Q to

3878 the rewriter and classical reasoner simultaneously;

3880 * function uresult no longer returns theorems in "standard" format;

3881 regain previous version by: val uresult = standard o uresult;

3885 New in Isabelle94-6

3886 -------------------

3888 * oracles -- these establish an interface between Isabelle and trusted

3889 external reasoners, which may deliver results as theorems;

3891 * proof objects (in particular record all uses of oracles);

3893 * Simp_tac, Fast_tac, etc. that refer to implicit simpset / claset;

3895 * "constdefs" section in theory files;

3897 * "primrec" section (HOL) no longer requires names;

3899 * internal type "tactic" now simply "thm -> thm Sequence.seq";

3903 New in Isabelle94-5

3904 -------------------

3906 * reduced space requirements;

3908 * automatic HTML generation from theories;

3910 * theory files no longer require "..." (quotes) around most types;

3912 * new examples, including two proofs of the Church-Rosser theorem;

3914 * non-curried (1994) version of HOL is no longer distributed;

3918 New in Isabelle94-4

3919 -------------------

3921 * greatly reduced space requirements;

3923 * theory files (.thy) no longer require \...\ escapes at line breaks;

3925 * searchable theorem database (see the section "Retrieving theorems" on

3926 page 8 of the Reference Manual);

3928 * new examples, including Grabczewski's monumental case study of the

3929 Axiom of Choice;

3931 * The previous version of HOL renamed to Old_HOL;

3933 * The new version of HOL (previously called CHOL) uses a curried syntax

3934 for functions. Application looks like f a b instead of f(a,b);

3936 * Mutually recursive inductive definitions finally work in HOL;

3938 * In ZF, pattern-matching on tuples is now available in all abstractions and

3939 translates to the operator "split";

3943 New in Isabelle94-3

3944 -------------------

3946 * new infix operator, addss, allowing the classical reasoner to

3947 perform simplification at each step of its search. Example:

3948 fast_tac (cs addss ss)

3950 * a new logic, CHOL, the same as HOL, but with a curried syntax

3951 for functions. Application looks like f a b instead of f(a,b). Also pairs

3952 look like (a,b) instead of <a,b>;

3954 * PLEASE NOTE: CHOL will eventually replace HOL!

3956 * In CHOL, pattern-matching on tuples is now available in all abstractions.

3957 It translates to the operator "split". A new theory of integers is available;

3959 * In ZF, integer numerals now denote two's-complement binary integers.

3960 Arithmetic operations can be performed by rewriting. See ZF/ex/Bin.ML;

3962 * Many new examples: I/O automata, Church-Rosser theorem, equivalents

3963 of the Axiom of Choice;

3967 New in Isabelle94-2

3968 -------------------

3970 * Significantly faster resolution;

3972 * the different sections in a .thy file can now be mixed and repeated

3973 freely;

3975 * Database of theorems for FOL, HOL and ZF. New

3976 commands including qed, qed_goal and bind_thm store theorems in the database.

3978 * Simple database queries: return a named theorem (get_thm) or all theorems of

3979 a given theory (thms_of), or find out what theory a theorem was proved in

3980 (theory_of_thm);

3982 * Bugs fixed in the inductive definition and datatype packages;

3984 * The classical reasoner provides deepen_tac and depth_tac, making FOL_dup_cs

3985 and HOL_dup_cs obsolete;

3987 * Syntactic ambiguities caused by the new treatment of syntax in Isabelle94-1

3988 have been removed;

3990 * Simpler definition of function space in ZF;

3992 * new results about cardinal and ordinal arithmetic in ZF;

3994 * 'subtype' facility in HOL for introducing new types as subsets of existing

3995 types;

3998 $Id$