FIXME-TODO
changeset 768 e9e205b904e2
parent 760 c1989de100b4
child 778 54f186bb5e3e
equal deleted inserted replaced
767:37285ec4387d 768:e9e205b904e2
     6 
     6 
     7 - have FSet.thy to have a simple infrastructure for 
     7 - have FSet.thy to have a simple infrastructure for 
     8   finite sets (syntax should be \<lbrace> \<rbrace>,
     8   finite sets (syntax should be \<lbrace> \<rbrace>,
     9   look at Set.thy how syntax is been introduced)
     9   look at Set.thy how syntax is been introduced)
    10 
    10 
    11 - think about what happens if things go wrong (like
    11 - Handle theorems that include Ball/Bex. Would it help 
    12   theorem cannot be lifted) / proper diagnostic 
    12   if we introduced separate Bex and Ball constants for 
    13   messages for the user
    13   quotienting?
    14 
       
    15 - Handle theorems that include Ball/Bex
       
    16 
    14 
    17 - user should be able to give quotient_respects and 
    15 - user should be able to give quotient_respects and 
    18   preserves theorems in a more natural form.
    16   preserves theorems in a more natural form.
    19 
    17 
    20 - the test in the (_, Const _) needs to be fixed
    18 - the test in the (_, Const _) needs to be fixed
    21 
    19 
    22 Lower Priority
    20 Lower Priority
    23 ==============
    21 ==============
    24 
    22 
       
    23 - Maybe a quotient_definition should already require
       
    24   a proof of the respectfulness (in this way one
       
    25   already excludes non-sensical definitions)
       
    26 
    25 - accept partial equvalence relations
    27 - accept partial equvalence relations
    26 
    28 
    27 - what happens if the original theorem contains bounded
    29 - think about what happens if things go wrong (like
    28   quantifiers? can such theorems be lifted? If not, would 
    30   theorem cannot be lifted) / proper diagnostic 
    29   it help if we introduced separate Bex and Ball constants
    31   messages for the user
    30   for quotienting?
       
    31 
    32 
    32 - inductions from the datatype package have a strange
    33 - inductions from the datatype package have a strange
    33   order of quantifiers in assumptions.
    34   order of quantifiers in assumptions.
    34 
    35 
    35 - wrapper that translates an an original theorem given
    36 - wrapper that translates an an original theorem given
    54 - Maybe quotient and equiv theorems like the ones for
    55 - Maybe quotient and equiv theorems like the ones for
    55   [QuotList, QuotOption, QuotPair...] could be automatically
    56   [QuotList, QuotOption, QuotPair...] could be automatically
    56   proven?
    57   proven?
    57 
    58 
    58 - Examples: Finite multiset.
    59 - Examples: Finite multiset.
       
    60 
       
    61 - The current syntax of the quotient_definition is
       
    62 
       
    63       "qconst :: qty"
       
    64       as "rconst"
       
    65 
       
    66   Is it possible to have the more Isabelle-like
       
    67   syntax
       
    68    
       
    69       qconst :: "qty"
       
    70       as "rconst"
       
    71 
       
    72   That means "qconst :: qty" is not read as a term, but
       
    73   as two entities.