Isabelle_DOF/src/ontologies/scholarly_paper/scholarly_paper.thy

279 lines
10 KiB
Plaintext

(*************************************************************************
* Copyright (C)
* 2019 The University of Exeter
* 2018-2019 The University of Paris-Saclay
* 2018 The University of Sheffield
*
* License:
* This program can be redistributed and/or modified under the terms
* of the 2-clause BSD-style license.
*
* SPDX-License-Identifier: BSD-2-Clause
*************************************************************************)
section\<open>An example ontology for a scholarly paper\<close>
theory scholarly_paper
imports "../../DOF/Isa_COL"
begin
doc_class title =
short_title :: "string option" <= "None"
doc_class subtitle =
abbrev :: "string option" <= "None"
(* adding a contribution list and checking that it is cited as well in tech as in conclusion. ? *)
doc_class author =
email :: "string" <= "''''"
http_site :: "string" <= "''''"
orcid :: "string" <= "''''"
affiliation :: "string"
doc_class abstract =
keywordlist :: "string list" <= "[]"
principal_theorems :: "thm list"
doc_class text_section = text_element +
main_author :: "author option" <= None
fixme_list :: "string list" <= "[]"
level :: "int option" <= "None"
(* this attribute enables doc-notation support section* etc.
we follow LaTeX terminology on levels
part = Some -1
chapter = Some 0
section = Some 1
subsection = Some 2
subsubsection = Some 3
... *)
(* for scholarly paper: invariant level > 0 *)
doc_class introduction = text_section +
comment :: string
claims :: "thm list"
text\<open>Technical text-elements posses a status: they can be either an \<^emph>\<open>informal explanation\<close> /
description or a kind of introductory text to definition etc. or a \<^emph>\<open>formal statement\<close> similar
to :
\<^bold>\<open>Definition\<close> 3.1: Security.
As Security of the system we define etc...
A formal statement can, but must not have a reference to true formal Isabelle/Isar definition.
\<close>
datatype status = semiformal | description
text\<open>The class \<^verbatim>\<open>technical\<close> regroups a number of text-elements that contain typical
"technical content" in mathematical or engineering papers: definitions, theorems, lemmas, examples. \<close>
(* OPEN PROBLEM: connection between referentiable and status. This should be explicit
and computable. *)
(* TODO : RENAME "tag" by "short_name" *)
doc_class technical = text_section +
definition_list :: "string list" <= "[]"
status :: status <= "description"
formal_results :: "thm list"
type_synonym tc = technical
text\<open>A rough structuring is modeled as follows:\<close>
doc_class "definition" = technical +
referentiable :: bool <= "True"
tag :: "string" <= "''''"
doc_class "theorem" = technical +
referentiable :: bool <= "True"
tag :: "string" <= "''''"
text\<open>Note that the following two text-elements are currently set to no-keyword in LNCS style.\<close>
doc_class "lemma" = technical +
referentiable :: bool <= "True"
tag :: "string" <= "''''"
doc_class "corollary" = technical +
referentiable :: bool <= "True"
tag :: "string" <= "''''"
text\<open> \<^verbatim>\<open>examples\<close> are currently considered \<^verbatim>\<open>technical\<close>. Another alternative would be to group the
following classes into an own class: "evaluation" or "explanation" or ... \<close>
doc_class example = technical +
referentiable :: bool <= "True"
tag :: "string" <= "''''"
doc_class assertion = technical +
referentiable :: bool <= "True"
tag :: "string" <= "''''"
(* TODO :*)
(* attention : no LaTeX support yet >>> *)
doc_class "code" = technical +
checked :: bool <= "False"
tag :: "string" <= "''''"
doc_class "experiment" = technical +
tag :: "string" <= "''''"
doc_class "evaluation" = technical +
tag :: "string" <= "''''"
doc_class "data" = technical +
tag :: "string" <= "''''"
text\<open>The @{doc_class "code"} is a general stub for free-form and type-checked code-fragments
such as:
\<^enum> SML code
\<^enum> bash code
\<^enum> isar code (although this might be an unwanted concurrence to the Isabelle standard cartouche)
\<^enum> C code.
it is intended that later refinements of this "stub" as done in \<^verbatim>\<open>Isabelle_C\<close> which come with their
own content checking and, of course, presentation styles.
\<close>
doc_class "conclusion" = text_section +
main_author :: "author option" <= None
doc_class related_work = "conclusion" +
main_author :: "author option" <= None
doc_class bibliography = text_section +
style :: "string option" <= "Some ''LNCS''"
doc_class annex = "text_section" +
main_author :: "author option" <= None
text\<open> Besides subtyping, there is another relation between
doc\_classes: a class can be a \<^emph>\<open>monitor\<close> to other ones,
which is expressed by occurrence in the where clause.
While sub-classing refers to data-inheritance of attributes,
a monitor captures structural constraints -- the order --
in which instances of monitored classes may occur.
The control of monitors is done by the commands:
\<^item> \<^verbatim>\<open> monitor <oid::class_type, <attributes-defs> > \<close>
\<^item> \<^verbatim>\<open> close_monitor <oid[::class_type],<attributes-updates>> \<close>
where the automaton of the monitor class is expected
to be in a final state.
Monitors can be nested.
Classes neither directly or indirectly (via inheritance)
mentioned in the monitor clause are \<^emph>\<open>independent\<close> from
the monitor and may occur freely, \ie{} in arbitrary order.n \<close>
text \<open>underlying idea: a monitor class automatically receives a
\<^verbatim>\<open>trace\<close> attribute in which a list of observed class-ids is maintained.
The \<^verbatim>\<open>trace\<close> is a \<^emph>\<open>`predefined id`\<close> like \<^verbatim>\<open>main\<close> in C. It can be accessed
like any other attribute of a class instance, \ie{} a document item.\<close>
doc_class article =
style_id :: string <= "''LNCS''"
version :: "(int \<times> int \<times> int)" <= "(0,0,0)"
accepts "(title ~~
\<lbrakk>subtitle\<rbrakk> ~~
\<lbrace>author\<rbrace>\<^sup>+ ~~
abstract ~~
\<lbrace>introduction\<rbrace>\<^sup>+ ~~
\<lbrace>technical\<rbrace>\<^sup>+ ~~
\<lbrace>conclusion\<rbrace>\<^sup>+ ~~
bibliography ~~
\<lbrace>annex\<rbrace>\<^sup>* )"
ML\<open>
structure Scholarly_paper_trace_invariant =
struct
local
fun group f g cidS [] = []
|group f g cidS (a::S) = case find_first (f a) cidS of
NONE => [a] :: group f g cidS S
| SOME cid => let val (pref,suff) = chop_prefix (g cid) S
in (a::pref)::(group f g cidS suff) end;
fun partition ctxt cidS trace =
let fun find_lead (x,_) = DOF_core.is_subclass ctxt x;
fun find_cont cid (cid',_) = DOF_core.is_subclass ctxt cid' cid
in group find_lead find_cont cidS trace end;
fun dest_option _ (Const (@{const_name "None"}, _)) = NONE
| dest_option f (Const (@{const_name "Some"}, _) $ t) = SOME (f t)
in
fun check ctxt cidS mon_id pos =
let val trace = AttributeAccess.compute_trace_ML ctxt mon_id pos @{here}
val groups = partition (Context.proof_of ctxt) cidS trace
fun get_level_raw oid = AttributeAccess.compute_attr_access ctxt "level" oid @{here} @{here};
fun get_level oid = dest_option (snd o HOLogic.dest_number) (get_level_raw (oid));
fun check_level_hd a = case (get_level (snd a)) of
NONE => error("Invariant violation: leading section" ^ snd a ^
" must have lowest level")
| SOME X => X
fun check_group_elem level_hd a = case (get_level (snd a)) of
NONE => true
| SOME y => if level_hd <= y then true
(* or < ? But this is too strong ... *)
else error("Invariant violation: "^
"subsequent section " ^ snd a ^
" must have higher level.");
fun check_group [] = true
|check_group [_] = true
|check_group (a::S) = forall (check_group_elem (check_level_hd a)) (S)
in if forall check_group groups then ()
else error"Invariant violation: leading section must have lowest level"
end
end
end
\<close>
setup\<open> let val cidS = ["scholarly_paper.introduction","scholarly_paper.technical",
"scholarly_paper.example", "scholarly_paper.conclusion"];
fun body moni_oid _ ctxt = (Scholarly_paper_trace_invariant.check
ctxt cidS moni_oid @{here};
true)
in DOF_core.update_class_invariant "scholarly_paper.article" body end\<close>
(* some test code *)
ML\<open>
(*
val trace = AttributeAccess.compute_trace_ML (Context.Proof @{context}) "this" @{here} @{here}
val groups = partition ( @{context}) cidS trace
val _::_::_::_:: _ ::_ ::_ ::a::_ = groups;
check;
fun get_level_raw oid = AttributeAccess.compute_attr_access (Context.Proof @{context}) "level" oid @{here} @{here};
fun get_level oid = dest_option (snd o HOLogic.dest_number) (get_level_raw (oid));
fun check_level_hd a = case (get_level (snd a)) of
NONE => error("Invariant violation: leading section" ^ snd a ^
" must have lowest level")
| SOME X => X
fun check_group_elem level_hd a = case (get_level (snd a)) of
NONE => true
| SOME y => if y > level_hd then true
else error("Invariant violation: subsequent section " ^ snd a ^
" must have higher level.");
fun check_group a = map (check_group_elem (check_level_hd (hd a))) (tl a) ;
*)
\<close>
end