support phantom types as quotient types

merged

merged

tuned

datatype dtyp with explicit sort information; tuned messages;

tuned;

updated generated file;

tuned quickcheck's response

hiding constants and facts in the Quickcheck_Exhaustive and Quickcheck_Narrowing theory;

merged

replace more uses of 'lemmas' with explicit 'lemma'; replace uses of 'simplified' attribute with 'unfolded'; remove unused intermediate lemmas.

Add Quotient_Rat: an example of using the quotient package with partial equivalence relations, defining rational numbers.

fix spelling

fix spelling

added IMP/Live_True.thy

replace many uses of 'lemmas' with 'lemma'; remove many unused intermediate lemmas.

prove class instances without extra lemmas

finite class instance for word type; remove unused lemmas

remove unused lemmas

generalize some lemmas

merged

tidied Word.thy; put attributes directly on lemmas instead of using 'declare'; replace various 'lemmas' commands with ordinary 'lemma'.

remove redundant lemma word_diff_minus

remove some duplicate lemmas, simplify some proofs

Quotient_Info stores only relation maps

hiding definitional facts in Quickcheck; introducing catch_match more honestly

added dependencies

added an example file with lifting of constants with contravariant and co/contravariant types

merged

make ctxt the first parameter

context/theory parametres tuned

maps are taken from enriched type infrastracture, rewritten lifting of constants, now we can lift even contravariant and co/contravariant types

add induction rule for list_all2

deactivating quickcheck_narrowing if Efficient_Nat theory is loaded

tuned quickcheck's response

more systematic lemma name

adding examples for quickcheck narrowing about partial functions

removing special code generator setup for hd and last function because this causes problems with quickcheck narrowing as the Haskell Prelude functions throw errors that cannot be caught instead of PatternFail exceptions

HOLCF/ex/Letrec.thy: keep class 'domain' as default sort

more error checking for fixrec

reinstate old functions cfst and csnd as abbreviations

merged

tuned

merged

added a specific tactic and method that deal with partial equivalence relations

use same order of facts for preplay as for actual reconstruction -- Metis sometimes exhibits very different timings depending on the order of the facts

avoid multiple TFF1 declarations

updated TFF1 support

updated Metis to 20110926 version

remove unnecessary sublocale instantiations in HOL-Probability (for clarity and speedup); remove Infinite_Product_Measure.product_prob_space which was a duplicate of Probability_Measure.product_prob_space

remove mem_(c)ball_0 and centre_in_(c)ball from simpset, as rules mem_(c)ball always match instead

add cancellation simprocs for type enat

tuned

increasing quickcheck's timeout in the example theory to avoid failures on the testing infrastructure

tuned proofs

added lemmas

tuned proof

real is better supported than real_of_nat, use it in the nat => ereal coercion

merged

the note about morphisms moved in the description part

updating documentation about quiet and verbose options in quickcheck

making the default behaviour of quickcheck a little bit less verbose; adapting quickcheck examples

adding verbose configuration to quickcheck

random reporting compilation returns if counterexample is genuine or potentially spurious, and takes genuine_only option as argument

the reporting random testing also returns if the counterexample is genuine or potentially spurious

exhaustive returns if a counterexample is genuine or potentially spurious in the presence of assumptions more correctly

inverted flag potential to genuine_only in the quickcheck narrowing Haskell code

NEWS

documenting the genuine_only option in quickcheck;

renaming potential flag to genuine_only flag with an inverse semantics

quickcheck narrowing continues searching after found a potentially spurious counterexample

outputing the potentially spurious counterexample and continue search

dynamic genuine_flag in compilation of random and exhaustive

indicating where the restart should occur; making safe_if dynamic

merged

enforce parantheses around SKIP {_}

adding genuine flag to predicate_compile_quickcheck and prolog_quickcheck (cf. 5e46c225370e); adjusting smart quickcheck compilation to new signature of exhaustive generators (cf. 1f5fc44254d7);

merged

remove Library/Diagonalize.thy, because Library/Nat_Bijection.thy includes all the same functionality

missing dependency

improved var names

invariant holds before loop

caret_range based on BreakIterator, which handles combined unicode characters as well;

misc tuning;

some localization;

eliminated some legacy operations;

more antiquotations;

tuned whitespace;

eliminated some legacy operations;

removed dead code, which has never been active in recorded history;

do not open ML structures;

tuned signature;

hide quickcheck constants Abs_cfun and Rep_cfun, to avoid clash with HOLCF

hiding internal constants and facts more properly

removing catch_match' now that catch_match is polymorphic

adapting exhaustive generators in record package

outputing if counterexample is potentially spurious or not

making catch_match polymorphic

compilations return genuine flag to quickcheck framework

extending quickcheck's result by the genuine flag

reporting random compilation also catches match exceptions internally

the narrowing also indicates if counterexample is potentially spurious

the simple exhaustive compilation also indicates if counterexample is potentially spurious;

quickcheck-random compilation also indicates if the counterexample is potentially spurious or not

changing the exhaustive generator signatures; replacing the hard-wired result type by its own identifier

quickcheck's compilation returns if it is genuine counterexample or a counterexample due to a match exception

adding examples for quickcheck-random

removing exception handling now that is caught at some other point; tuned message

quickcheck random can also find potential counterexamples; moved catch_match definition; split quickcheck setup;

merged

merged IMP/Util into IMP/Vars

use list theorems in Dining Cryptographers and Koepf Duermuth Countermeasure

cardinality of sets of lists

do not import examples Probability theory

moved theorems about distribution to the definition; removed oopsed-lemma

rename finite_prob_space.setsum_distribution, it collides with prob_space.setsum_distribution

remove duplicate theorem setsum_real_distribution

clarified modules (again) -- NB: both Document and Protocol are specific to this particular prover;

updated Sledgehammer docs with new/renamed options

added "minimize" option for more control over automatic minimization