word: implement twa word parsing

* spot/twaalgos/word.hh: add parse_word method and a new constructor
* spot/twaalgos/word.cc: implement word parsing
* python/spot/__init__.py: add parse_word method binding
* tests/python/word.ipynb: add word parsing tests
This commit is contained in:
Amaury Fauchille 2016-02-02 19:07:42 +01:00 committed by Alexandre Duret-Lutz
parent 61b1f200b6
commit 1fd76ee918
4 changed files with 393 additions and 7 deletions

View file

@ -1,6 +1,6 @@
// -*- coding: utf-8 -*-
// Copyright (C) 2013, 2014, 2015 Laboratoire de Recherche et Développement
// de l'Epita (LRDE).
// Copyright (C) 2013, 2014, 2015, 2016 Laboratoire de Recherche et
// Développement de l'Epita (LRDE).
//
// This file is part of Spot, a model checking library.
//
@ -20,10 +20,13 @@
#include <spot/twaalgos/word.hh>
#include <spot/twa/bddprint.hh>
#include <spot/twa/bdddict.hh>
#include <spot/tl/parse.hh>
#include <spot/tl/simplify.hh>
#include <spot/tl/apcollect.hh>
namespace spot
{
twa_word::twa_word(const twa_run_ptr run)
twa_word::twa_word(const twa_run_ptr& run)
: dict_(run->aut->get_dict())
{
for (auto& i: run->prefix)
@ -33,6 +36,11 @@ namespace spot
dict_->register_all_variables_of(run->aut, this);
}
twa_word::twa_word(const bdd_dict_ptr& dict)
: dict_(dict)
{
}
void
twa_word::simplify()
{
@ -105,4 +113,143 @@ namespace spot
os << '}';
return os;
}
namespace
{
static void word_parse_error(const std::string& word,
size_t i, parse_error_list pel)
{
for (auto& err: pel)
{
err.first.begin.column += i;
err.first.end.column += i;
}
std::ostringstream os;
format_parse_errors(os, word, pel);
throw parse_error(os.str());
}
static void word_parse_error(const std::string& word, size_t i,
const std::string& message)
{
if (i == std::string::npos)
i = word.size();
std::ostringstream s;
s << ">>> " << word << '\n';
for (auto j = i + 4; j > 0; --j)
s << ' ';
s << '^' << '\n';
s << message << '\n';
throw parse_error(s.str());
}
static size_t skip_next_formula(const std::string& word, size_t begin)
{
bool quoted = false;
auto size = word.size();
for (auto j = begin; j < size; ++j)
{
auto c = word[j];
if (!quoted && (c == ';' || c == '}'))
return j;
if (c == '"')
quoted = !quoted;
else if (quoted && c == '\\')
++j;
}
if (quoted)
word_parse_error(word, word.size(), "Unclosed string");
return std::string::npos;
}
}
twa_word_ptr parse_word(const std::string& word, const bdd_dict_ptr& dict)
{
atomic_prop_set aps;
parse_error_list pel;
tl_simplifier tls(dict);
twa_word_ptr tw = make_twa_word(dict);
size_t i = 0;
auto ind = i;
auto extract_bdd =
[&](typename twa_word::seq_t& seq)
{
auto sub = word.substr(i, ind - i);
formula f = spot::parse_infix_boolean(sub, pel);
if (!pel.empty())
word_parse_error(word, i, pel);
atomic_prop_collect(f, &aps);
seq.push_back(tls.as_bdd(f));
if (word[ind] == '}')
return true;
// Skip blanks after semi-colon
i = word.find_first_not_of(' ', ind + 1);
return false;
};
// Parse the prefix part. Can be empty.
while (word.substr(i, 6) != std::string("cycle{"))
{
ind = skip_next_formula(word, i);
if (ind == std::string::npos)
word_parse_error(word, word.size(),
"A twa_word must contain a cycle");
if (word[ind] == '}')
word_parse_error(word, ind, "Expected ';' delimiter :"
"'}' stands for ending a cycle");
// Exract formula, convert it to bdd and add it to the prefix sequence
extract_bdd(tw->prefix);
if (i == std::string::npos)
word_parse_error(word, ind + 1, "Missing cycle in formula");
}
// Consume "cycle{"
i += 6;
while (true)
{
ind = skip_next_formula(word, i);
if (ind == std::string::npos)
word_parse_error(word, word.size(),
"Missing ';' or '}' after formula");
// Extract formula, convert it to bdd and add it to the cycle sequence
// Break if an '}' is encountered
if (extract_bdd(tw->cycle))
break;
if (i == std::string::npos)
word_parse_error(word, ind + 1,
"Missing end of cycle character : '}'");
}
if (ind != word.size() - 1)
word_parse_error(word, ind + 1, "Input should be finished after cycle");
for (auto ap: aps)
dict->register_proposition(ap, tw.get());
return tw;
}
twa_graph_ptr twa_word::as_automaton() const
{
twa_graph_ptr aut = make_twa_graph(dict_);
aut->prop_weak(true);
aut->prop_deterministic(true);
size_t i = 0;
aut->new_states(prefix.size() + cycle.size());
for (auto b: prefix)
{
aut->new_edge(i, i + 1, b);
++i;
}
size_t j = i;
auto b = cycle.begin();
auto end = --cycle.end();
for (; b != end; ++b)
{
aut->new_edge(i, i + 1, *b);
++i;
}
// Close the loop
aut->new_edge(i, j, *b);
return aut;
}
}

View file

@ -1,5 +1,5 @@
// -*- coding: utf-8 -*-
// Copyright (C) 2013, 2014, 2015 Laboratoire de Recherche et
// Copyright (C) 2013, 2014, 2015, 2016 Laboratoire de Recherche et
// Développement de l'Epita (LRDE).
//
// This file is part of Spot, a model checking library.
@ -28,7 +28,8 @@ namespace spot
/// \brief An infinite word stored as a lasso.
struct SPOT_API twa_word final
{
twa_word(const twa_run_ptr run);
twa_word(const bdd_dict_ptr& dict);
twa_word(const twa_run_ptr& run);
~twa_word()
{
dict_->unregister_all_my_variables(this);
@ -45,10 +46,26 @@ namespace spot
return dict_;
}
twa_graph_ptr as_automaton() const;
SPOT_API
friend std::ostream& operator<<(std::ostream& os, const twa_word& w);
private:
bdd_dict_ptr dict_;
};
typedef std::shared_ptr<twa_word> twa_word_ptr;
inline twa_word_ptr make_twa_word(const bdd_dict_ptr& dict)
{
return std::make_shared<twa_word>(dict);
}
inline twa_word_ptr make_twa_word(const twa_run_ptr& run)
{
return std::make_shared<twa_word>(run);
}
SPOT_API
twa_word_ptr parse_word(const std::string& word, const bdd_dict_ptr& dict);
}