ltlmix: add support for the I/O variants

* bin/ltlmix.cc: Add options --ins, --outs, as well as the
two-argument form of -A/-P.
* bin/common_ioap.hh, bin/common_ioap.cc (is_output): New function.
* spot/tl/apcollect.cc, spot/tl/apcollect.hh (create_atomic_prop_set):
Allow the prefix string to be changed.
* spot/tl/randomltl.cc, spot/tl/randomltl.hh: Add support for an I/O
version with two set of atomic proposition, and a predicate to decide
if the original proposition was input or output.
* tests/core/ltlmix.test: More tests.
This commit is contained in:
Alexandre Duret-Lutz 2024-08-26 11:42:09 +02:00
parent 6fa42c90b8
commit 844fb887d9
8 changed files with 286 additions and 103 deletions

View file

@ -88,6 +88,72 @@ list_aps_in_formula(spot::formula f)
return aps;
}
bool
is_output(const std::string& a, const char* filename, int linenum)
{
if (auto it = identifier_map.find(a); it != identifier_map.end())
return it->second;
bool found_in = false;
for (const std::regex& r: regex_in)
if (std::regex_search(a, r))
{
found_in = true;
break;
}
bool found_out = false;
for (const std::regex& r: regex_out)
if (std::regex_search(a, r))
{
found_out = true;
break;
}
if (all_input_aps.has_value() == all_output_aps.has_value())
{
if (!all_input_aps.has_value())
{
// If the atomic proposition hasn't been classified
// because neither --ins nor --out were specified,
// attempt to classify automatically using the first
// letter.
int fl = a[0];
if (fl == 'i' || fl == 'I')
found_in = true;
else if (fl == 'o' || fl == 'O')
found_out = true;
}
if (found_in && found_out)
error_at_line(2, 0, filename, linenum,
"'%s' matches both --ins and --outs",
a.c_str());
if (!found_in && !found_out)
{
if (all_input_aps.has_value() || all_output_aps.has_value())
error_at_line(2, 0, filename, linenum,
"one of --ins or --outs should match '%s'",
a.c_str());
else
error_at_line(2, 0, filename, linenum,
"since '%s' does not start with 'i' or 'o', "
"it is unclear if it is an input or "
"an output;\n use --ins or --outs",
a.c_str());
}
}
else
{
// if we had only --ins or only --outs, anything not
// matching that was given is assumed to belong to the
// other one.
if (!all_input_aps.has_value() && !found_out)
found_in = true;
else if (!all_output_aps.has_value() && !found_in)
found_out = true;
}
return found_out;
}
// Takes a set of the atomic propositions appearing in the formula,
// and separate them into two vectors: input APs and output APs.
std::pair<std::vector<std::string>, std::vector<std::string>>
@ -97,71 +163,7 @@ filter_list_of_aps(spot::formula f, const char* filename, int linenum)
// now iterate over the list of atomic propositions to filter them
std::vector<std::string> matched[2]; // 0 = input, 1 = output
for (const std::string& a: aps)
{
if (auto it = identifier_map.find(a); it != identifier_map.end())
{
matched[it->second].push_back(a);
continue;
}
bool found_in = false;
for (const std::regex& r: regex_in)
if (std::regex_search(a, r))
{
found_in = true;
break;
}
bool found_out = false;
for (const std::regex& r: regex_out)
if (std::regex_search(a, r))
{
found_out = true;
break;
}
if (all_input_aps.has_value() == all_output_aps.has_value())
{
if (!all_input_aps.has_value())
{
// If the atomic proposition hasn't been classified
// because neither --ins nor --out were specified,
// attempt to classify automatically using the first
// letter.
int fl = a[0];
if (fl == 'i' || fl == 'I')
found_in = true;
else if (fl == 'o' || fl == 'O')
found_out = true;
}
if (found_in && found_out)
error_at_line(2, 0, filename, linenum,
"'%s' matches both --ins and --outs",
a.c_str());
if (!found_in && !found_out)
{
if (all_input_aps.has_value() || all_output_aps.has_value())
error_at_line(2, 0, filename, linenum,
"one of --ins or --outs should match '%s'",
a.c_str());
else
error_at_line(2, 0, filename, linenum,
"since '%s' does not start with 'i' or 'o', "
"it is unclear if it is an input or "
"an output;\n use --ins or --outs",
a.c_str());
}
}
else
{
// if we had only --ins or only --outs, anything not
// matching that was given is assumed to belong to the
// other one.
if (!all_input_aps.has_value() && !found_out)
found_in = true;
else if (!all_output_aps.has_value() && !found_in)
found_out = true;
}
matched[found_out].push_back(a);
}
matched[is_output(a, filename, linenum)].push_back(a);
return {matched[0], matched[1]};
}

View file

@ -45,6 +45,15 @@ extern std::vector<std::regex> regex_out;
// map identifier to input/output (false=input, true=output)
extern std::unordered_map<std::string, bool> identifier_map;
// Given an atomic proposition AP and the above
// regex_in/regex_out/identifier_map, decide if this AP is an output
// (true) or input (false.
bool
is_output(const std::string& ap,
const char* filename = nullptr, int linenum = 0);
// Separate the set of the atomic propositions appearing in f, into
// two vectors: input APs and output APs, becased on regex_in,
// regex_out, and identifier_map.

View file

@ -24,6 +24,7 @@
#include "common_setup.hh"
#include "common_finput.hh"
#include "common_output.hh"
#include "common_ioap.hh"
#include "common_conv.hh"
#include "common_cout.hh"
#include "common_range.hh"
@ -36,7 +37,9 @@ enum {
OPT_BOOLEAN_PRIORITIES = 256,
OPT_DUMP_PRIORITIES,
OPT_DUPS,
OPT_INS,
OPT_LTL_PRIORITIES,
OPT_OUTS,
OPT_SEED,
OPT_TREE_SIZE,
};
@ -70,18 +73,21 @@ static const argp_option options[] = {
{ nullptr, 0, nullptr, 0, "Generation parameters:", 2 },
{ "allow-dups", OPT_DUPS, nullptr, 0,
"allow duplicate formulas to be output", 0 },
{ "ap-count", 'A', "N", 0,
{ "ap-count", 'A', "N[,M]", 0,
"rename the atomic propositions in each selected formula by drawing "
"randomly from N atomic propositions (the rewriting is bijective "
"if N is larger than the original set)", 0 },
"if N is larger than the original set). If M is specified, two sets "
"of atomic propositions are used to represent inputs and outputs, and "
"options --ins/--outs can be used to classify the original propositions.",
0 },
{ "boolean", 'B', nullptr, 0,
"generate Boolean combinations of formulas (default)", 0 },
{ "formulas", 'n', "INT", 0,
"number of formulas to generate (default: 1);\n"
"use a negative value for unbounded generation", 0 },
{ "ltl", 'L', nullptr, 0, "generate LTL combinations of subformulas", 0 },
{ "polarized-ap", 'P', "N", 0,
"similar to -A N, but randomize the polarity of the new atomic "
{ "polarized-ap", 'P', "N[,M]", 0,
"similar to -A, but randomize the polarity of the new atomic "
"propositions", 0 },
{ "random-conjuncts", 'C', "N", 0,
"generate random-conjunctions of N conjuncts; "
@ -92,6 +98,12 @@ static const argp_option options[] = {
{ "tree-size", OPT_TREE_SIZE, "RANGE", 0,
"tree size of main pattern generated (default: 5);\n"
"input formulas count as size 1.", 0 },
{ "ins", OPT_INS, "PROPS", 0,
"comma-separated list of atomic propositions to consider as input, "
"interpreted as a regex if enclosed in slashes", 0 },
{ "outs", OPT_OUTS, "PROPS", 0,
"comma-separated list of atomic propositions to consider as putput, "
"interpreted as a regex if enclosed in slashes", 0 },
RANGE_DOC,
/**************************************************/
{ nullptr, 0, nullptr, 0, "Adjusting probabilities:", 4 },
@ -140,7 +152,9 @@ static int opt_seed = 0;
static range opt_tree_size = { 5, 5 };
static bool opt_unique = true;
static int opt_ap_count = 0;
static int opt_out_ap_count = 0;
static bool opt_literal = false;
static bool opt_io = false;
namespace
{
@ -160,8 +174,12 @@ namespace
process_formula(spot::formula f, const char* filename = nullptr,
int linenum = 0) override
{
(void) filename;
(void) linenum;
if (opt_io)
// Filter the atomic propositions of each formula in order to
// report those that are not classifiable. Throw the result
// of that filtering away, as we only care about the potential
// diagnostics.
(void) filter_list_of_aps(f, filename, linenum);
opt->sub.insert(f);
return 0;
}
@ -170,6 +188,24 @@ namespace
static sub_processor subreader;
std::pair<int, int> to_int_pair(const char* arg, const char* opt)
{
const char* comma = strchr(arg, ',');
if (!comma)
{
int res = to_int(arg, opt);
return {res, 0};
}
else
{
std::string arg1(arg, comma);
int res1 = to_int(arg1.c_str(), opt);
int res2 = to_int(comma + 1, opt);
return {res1, res2};
}
}
static int
parse_opt(int key, char* arg, struct argp_state*)
{
@ -178,7 +214,8 @@ parse_opt(int key, char* arg, struct argp_state*)
switch (key)
{
case 'A':
opt_ap_count = to_int(arg, "-A/--ap-count");
std::tie(opt_ap_count, opt_out_ap_count) =
to_int_pair(arg, "-A/--ap-count");
opt_literal = false;
break;
case 'B':
@ -199,7 +236,8 @@ parse_opt(int key, char* arg, struct argp_state*)
opt_formulas = to_int(arg, "-n/--formulas");
break;
case 'P':
opt_ap_count = to_int(arg, "-P/--polarized-ap");
std::tie(opt_ap_count, opt_out_ap_count) =
to_int_pair(arg, "-P/--polarized-ap");
opt_literal = true;
break;
case OPT_BOOLEAN_PRIORITIES:
@ -214,6 +252,20 @@ parse_opt(int key, char* arg, struct argp_state*)
case OPT_DUPS:
opt_unique = false;
break;
case OPT_INS:
{
all_input_aps.emplace(std::vector<std::string>{});
split_aps(arg, *all_input_aps);
opt_io = true;
break;
}
case OPT_OUTS:
{
all_output_aps.emplace(std::vector<std::string>{});
split_aps(arg, *all_output_aps);
opt_io = true;
break;
}
case OPT_SEED:
opt_seed = to_int(arg, "--seed");
break;
@ -249,6 +301,19 @@ main(int argc, char* argv[])
check_no_formula("combine");
if (opt_io && !opt_out_ap_count)
error(2, 0,
"options --ins and --outs only make sense when the "
"two-argument version of '-A N,M' or '-P N,M' is used.");
if (opt_out_ap_count > 0)
// Do not require --ins/--outs to be used, as the input
// pattern may use atomic propositions starting with i/o
// already. Setting opt_io will cause the subreader to
// complain about unclassifible atomic propositions.
opt_io = true;
if (opt_io)
process_io_options();
if (subreader.run())
return 2;
@ -257,11 +322,16 @@ main(int argc, char* argv[])
spot::srand(opt_seed);
std::function<bool(spot::formula)> output_p = nullptr;
if (opt_out_ap_count)
output_p = [&](spot::formula f) { return is_output(f.ap_name()); };
spot::randltlgenerator rg
(opt_ap_count,
[&] (){
spot::option_map opts;
opts.set("output", output);
opts.set("out_ap_size", opt_out_ap_count);
opts.set("tree_size_min", opt_tree_size.min);
opts.set("tree_size_max", opt_tree_size.max);
opts.set("seed", opt_seed);
@ -269,7 +339,7 @@ main(int argc, char* argv[])
opts.set("unique", opt_unique);
opts.set("literals", opt_literal);
return opts;
}(), opt_pL, nullptr, opt_pB, &opt->sub);
}(), opt_pL, nullptr, opt_pB, &opt->sub, output_p);
if (opt_dump_priorities)
{