doc-src/IsarImplementation/Thy/unused.thy
author wenzelm
Mon Sep 04 16:28:27 2006 +0200 (2006-09-04 ago)
changeset 20470 c839b38a1f32
parent 20460 351c63bb2704
child 20474 af069653f1d7
permissions -rw-r--r--
more on variables;
tuned;
wenzelm@18537
     1
wenzelm@18537
     2
text {*
wenzelm@20460
     3
wenzelm@20470
     4
  
wenzelm@20470
     5
wenzelm@20470
     6
  A \emph{fixed variable} acts like a local constant in the current
wenzelm@20470
     7
  context, representing some simple type @{text "\<alpha>"}, or some value
wenzelm@20470
     8
  @{text "x: \<tau>"} (for a fixed type expression @{text "\<tau>"}).  A
wenzelm@20470
     9
  \emph{schematic variable} acts like a placeholder for arbitrary
wenzelm@20470
    10
  elements, similar to outermost quantification.  The division between
wenzelm@20470
    11
  fixed and schematic variables tells which abstract entities are
wenzelm@20470
    12
  inside and outside the current context.
wenzelm@20470
    13
wenzelm@20470
    14
wenzelm@20460
    15
  @{index_ML Variable.trade: "Proof.context -> (thm list -> thm list) -> thm list -> thm list"} \\
wenzelm@20460
    16
wenzelm@20460
    17
wenzelm@20460
    18
wenzelm@20460
    19
  \item @{ML Variable.trade} composes @{ML Variable.import} and @{ML
wenzelm@20460
    20
  Variable.export}, i.e.\ it provides a view on facts with all
wenzelm@20460
    21
  variables being fixed in the current context.
wenzelm@20460
    22
wenzelm@20460
    23
wenzelm@20429
    24
  In practice, super-contexts emerge either by merging existing ones,
wenzelm@20429
    25
  or by adding explicit declarations.  For example, new theories are
wenzelm@20429
    26
  usually derived by importing existing theories from the library
wenzelm@20429
    27
  @{text "\<Theta> = \<Theta>\<^sub>1 + \<dots> + \<Theta>\<^isub>n"}, or 
wenzelm@20429
    28
wenzelm@20429
    29
wenzelm@20429
    30
wenzelm@18537
    31
  The Isar toplevel works differently for interactive developments
wenzelm@18537
    32
  vs.\ batch processing of theory sources.  For example, diagnostic
wenzelm@18537
    33
  commands produce a warning batch mode, because they are considered
wenzelm@18537
    34
  alien to the final theory document being produced eventually.
wenzelm@18537
    35
  Moreover, full @{text undo} with intermediate checkpoints to protect
wenzelm@18537
    36
  against destroying theories accidentally are limited to interactive
wenzelm@18537
    37
  mode.  In batch mode there is only a single strictly linear stream
wenzelm@18537
    38
  of potentially desctructive theory transformations.
wenzelm@18537
    39
wenzelm@18537
    40
  \item @{ML Toplevel.empty} is an empty transition; the Isar command
wenzelm@18537
    41
  dispatcher internally applies @{ML Toplevel.name} (for the command)
wenzelm@18537
    42
  name and @{ML Toplevel.position} for the source position.
wenzelm@18537
    43
wenzelm@20429
    44
*}
wenzelm@20429
    45