mcell/utils/data_model_to_pymcell/generator_utils.cpp

/******************************************************************************
 *
 * Copyright (C) 2020 by
 * The Salk Institute for Biological Studies
 *
 * Use of this source code is governed by an MIT-style
 * license that can be found in the LICENSE file or at
 * https://opensource.org/licenses/MIT.
 *
******************************************************************************/

#include <fstream>
#include <algorithm>
#include <ctype.h>

#include "generator_utils.h"
#include "api/python_export_constants.h"
#include "api/compartment_utils.h"

#include "generator_structs.h"

using namespace MCell::API;

namespace MCell {


void check_not_empty(const Value& parent, const char* key, const std::string& msg_location) {
  string s = parent[key].asString();
  bool white_space = std::all_of(s.begin(), s.end(), ::isspace);
  if (s == "" || white_space) {
    ERROR("Unexpected empty string as value of " + msg_location + " - " + key + ".");
  }
}


static std::string replace_id_with_function(const std::string& id, const bool use_python_functions) {
  const auto& func_it = mdl_functions_to_py_bngl_map.find(id);
  if (func_it != mdl_functions_to_py_bngl_map.end()) {
    // replace
    if (use_python_functions) {
      return func_it->second.first;
    }
    else {
      // BNGL variant
      return func_it->second.second;
    }
  }
  else {
    // other id, keep it as it is
    return id;
  }
}

// when use_python_functions is true, function calls are replaced with Python function names
// when false, they are replaced with BNGL function names
std::string replace_function_calls_in_expr(const std::string& data_model_expr, const bool use_python_functions) {
  std::string res;

  // all function names are represented as ids, i.e. [a-zA-Z_][a-zA-Z0-9_]*
  // practically the same automaton as in get_used_ids
  enum state_t {
    START,
    IN_ID,
    IN_NUM
  };
  state_t state = START;
  string curr_id;
  for (size_t i = 0; i < data_model_expr.size(); i++) {
    char c = data_model_expr[i];
    switch (state) {
      case START:
        if (isalpha(c) || c == '_') {
          state = IN_ID;
          curr_id += c;
        }
        else if (isdigit(c)) {
          state = IN_NUM;
          res += c;
        }
        else {
          res += c;
        }
        break;
      case IN_ID:
        if (isalnum(c) || c == '_') {
          curr_id += c;
        }
        else {
          res += replace_id_with_function(curr_id, use_python_functions);
          res += c;
          curr_id = "";
          state = START;
        }
        break;
      case IN_NUM:
        if (isdigit(c) || c == '.' || c == 'e' || c == '+' || c == '-') {
          // ok - but must not be exp
          res += c;
        }
        else {
          if (isalpha(c) || c == '_') {
            state = IN_ID;
            curr_id += c;
          }
          else {
            state = START;
            res += c;
          }
        }
        break;
    }
  }
  if (state == IN_ID) {
    res += replace_id_with_function(curr_id, use_python_functions);
  }

  return res;
}


std::string get_module_name_w_prefix(const std::string& output_files_prefix, const std::string file_suffix) {

  if (output_files_prefix == "" || output_files_prefix.back() == '/' || output_files_prefix.back() == '\\') {
    return file_suffix;
  }
  else {
    size_t pos = output_files_prefix.find_last_of("/\\");
    if (pos == string::npos) {
      return output_files_prefix + "_" + file_suffix;
    }
    else {
      return output_files_prefix.substr(pos) +  "_" + file_suffix;
    }
  }
}


void parse_rxn_rule_side(
    Json::Value& substances_node,
    std::vector<std::string>& substances,
    std::vector<std::string>& orientations) {

  // cannot use BNGL parser directly because reactions may contain orientations

  substances.clear();
  orientations.clear();

  string str = substances_node.asString();

  // finite automata to parse the reaction side string, e.g. "a + b"
  enum state_t {
    START,
    CPLX,
    IN_PAREN,
    AFTER_MOL,
    AFTER_ORIENT,
    AFTER_PLUS
  };

  state_t state = START;
  string current_id;
  for (size_t i = 0; i < str.size(); i++) {
    char c = str[i];
    switch (state) {
      case START:
        if (isalnum(c) || c == '_' || c == '.' || c == '@') {
          state = CPLX;
          current_id = c;
        }
        else if (isblank(c)) {
          // ok
        }
        else {
          ERROR("Could not parse reaction side " + str + " (START).");
        }
        break;

      case CPLX:
        if (isalnum(c) || c == '_' || c == '.' || c == '@' || c == ':') {
          current_id += c;
        }
        else if (c == '(') {
          state = IN_PAREN;
          current_id += '(';
        }
        else if (isblank(c) || c == '+' || c == '\'' || c == ',' || c == ';') {
          substances.push_back(current_id);
          orientations.push_back("");
          if (c == '\'' || c == ',' || c == ';') {
            orientations.back() = c;
          }
          current_id = "";
          if (c == '+') {
            state = AFTER_PLUS;
          }
          else {
            state = AFTER_MOL;
          }
        }
        else {
          ERROR("Could not parse reaction side " + str + " (MOL_ID).");
        }
        break;

      case IN_PAREN:
        if (isalnum(c) || c == '_' || c == ',' || c == '~' || c == '!' || c == '+' || c == '?') {
          current_id += c;
        }
        else if (c == ')') {
          state = CPLX;
          current_id += ')';
        }
        else {
          ERROR("Could not parse reaction side " + str + " (IN_PAREN).");
        }
        break;

      case AFTER_MOL:
        if (c == '+') {
          state = AFTER_PLUS;
        }
        else if (c == '\'') {
          state = AFTER_ORIENT;
          orientations.back() = c;
        }
        else if (c == ',') {
          state = AFTER_ORIENT;
          orientations.back() = c;
        }
        else if (c == ';') {
          state = AFTER_ORIENT;
          orientations.back() = c;
        }
        else if (isblank(c)) {
          // ok
        }
        else {
          ERROR("Could not parse reaction side " + str + " (AFTER_MOL).");
        }
        break;

      case AFTER_ORIENT:
        if (c == '+') {
          state = AFTER_PLUS;
        }
        else if (isblank(c)) {
          // ok
        }
        else {
          ERROR("Could not parse reaction side " + str + " (AFTER_ORIENT).");
        }
        break;

      case AFTER_PLUS:
        if (isalnum(c) || c == '_' || c == '@') {
          state = CPLX;
          current_id = c;
        }
        else if (isblank(c)) {
          // ok
        }
        else {
          ERROR("Could not parse reaction side " + str + " (AFTER_PLUS).");
        }
        break;
      default:
        assert(false);
    }
  }
  if (current_id != "") {
    substances.push_back(current_id);
    orientations.push_back("");
  }
}


string remove_compartments(const std::string& species_name) {
  size_t i = 0;
  string res;
  bool in_compartment = false;
  while (i < species_name.size()) {
    char c = species_name[i];
    if (c == '@') {
      assert(!in_compartment);
      in_compartment = true;
    }
    else if (in_compartment && (!isalnum(c) && c != '_')) {
      in_compartment = false;
      if (c != ':') {
        res += c;
      }
    }
    else if (!in_compartment) {
      res += c;
    }
    i++;
  }
  return res;
}


// returns "" if there are multiple compartments and sets *has_multiple_compartments to
// true if it is not nullptr
string get_single_compartment(const std::string& name, bool* has_multiple_compartments) {
  std::vector<std::string> compartments;
  API::get_compartment_names(name, compartments);

  if (has_multiple_compartments != nullptr) {
    *has_multiple_compartments = false;
  }

  if (compartments.empty()) {
    return "";
  }
  else {
    string res = compartments[0];
    for (size_t i = 1; i < compartments.size(); i++) {
      if (res != compartments[i]) {
        // multiple compartments
        if (has_multiple_compartments != nullptr) {
          *has_multiple_compartments = true;
        }
        return "";
      }
    }
    return res;
  }
}


static string make_cplx(const string bngl_str, const string orient = "", const string compartment = "") {
  string res = S(MDOT) + API::NAME_CLASS_COMPLEX + "('" + fix_dots_in_simple_species(bngl_str) + "'";
  if (orient != "") {
    res += S(", ") + API::NAME_ORIENTATION + " = " + MDOT + API::NAME_ENUM_ORIENTATION + "." + orient;
  }
  if (compartment != "") {
    res += S(", ") + API::NAME_COMPARTMENT_NAME + " = '" + compartment + "'";
  }
  res += ")";
  return res;
}


bool static is_mdot_superclass(const std::string& name) {
  return
      name == S(MDOT) + API::NAME_CV_AllMolecules ||
      name == S(MDOT) + API::NAME_CV_AllVolumeMolecules ||
      name == S(MDOT) + API::NAME_CV_AllSurfaceMolecules;
}

string make_species_or_cplx(
    const SharedGenData& data,
    const std::string& name,
    const std::string& orient,
    const std::string& compartment) {

  bool is_superclass = is_mdot_superclass(name);

  if (is_superclass && orient == "" && compartment == "") {
    return name;
  }

  if (is_superclass || !data.bng_mode) {
    stringstream ss;
    const SpeciesOrMolType* species_info = data.find_species_or_mol_type_info(name);

    if (is_superclass || (species_info != nullptr && species_info->is_species)) {
      // substance was declared as species, we can use its id directly
      ss << make_id(name) << "." << API::NAME_INST << "(";

      if (orient != "") {
        assert(compartment == "");
        ss <<
            API::NAME_ORIENTATION << " = " <<
            MDOT << API::NAME_ENUM_ORIENTATION << "." << orient;
      }

      if (compartment != "") {
        assert(orient == "");
        ss <<
            API::NAME_COMPARTMENT_NAME << " = '" << compartment << "'";
      }

      ss << ")";
      return ss.str();
    }
  }

  // otherwise we will generate a BNGL string
  return make_cplx(name, orient, compartment);
}


string reaction_name_to_id(const string& json_name) {
  string res_name = json_name;
  replace(res_name.begin(), res_name.end(), ' ', '_');
  replace(res_name.begin(), res_name.end(), '.', '_');
  replace(res_name.begin(), res_name.end(), ')', '_');
  replace(res_name.begin(), res_name.end(), '(', '_');
  replace(res_name.begin(), res_name.end(), '!', '_');
  replace(res_name.begin(), res_name.end(), '~', '_');
  replace(res_name.begin(), res_name.end(), ':', '_');

  res_name = regex_replace(res_name, regex("<->"), "revto");
  res_name = regex_replace(res_name, regex("->"), "to");
  res_name = regex_replace(res_name, regex("\\+"), "plus");
  res_name = regex_replace(res_name, regex("\\?"), "any_bond");
  res_name = regex_replace(res_name, regex("'"), "_up");
  res_name = regex_replace(res_name, regex(","), "_down");
  res_name = regex_replace(res_name, regex(";"), "_any");
  res_name = regex_replace(res_name, regex("@"), "_at_");

  return res_name;
}


string get_rxn_id(Json::Value& reaction_list_item, uint& unnamed_rxn_counter) {
  string name = reaction_list_item[KEY_RXN_NAME].asString();
  if (name == "") {
    name = UNNAMED_REACTION_RULE_PREFIX + to_string(unnamed_rxn_counter);
    unnamed_rxn_counter++;
  }
  else {
    name = reaction_name_to_id(name);
  }
  return name;
}


string create_count_name(
    const string& what_to_count, const string& where_to_count,
    const bool molecules_not_species) {

  // first remove all cterm_refixes
  regex pattern_cterm(COUNT_TERM_PREFIX);
  string what_to_count_no_cterm = regex_replace(what_to_count, pattern_cterm, "");
  string res = COUNT_PREFIX + fix_id(what_to_count_no_cterm);

  if (!molecules_not_species) {
    res += "_species";
  }

  if (where_to_count != WORLD && where_to_count != "") {
    res += "_" + where_to_count;
  }

  return res;
}


uint get_num_counts_in_mdl_string(const string& mdl_string) {
  uint res = 0;
  size_t pos = 0;
  while ((pos = mdl_string.find(COUNT, pos)) != string::npos) {
    res++;
    pos += strlen(COUNT);
  }
  return res;
}


string remove_c_comment(const string& str) {
  std::regex e ("/\\*.*\\*/");
  return std::regex_replace(str, e, "");
}


string remove_whitespace(const string& str) {
  std::regex e ("[ \t]");
  return std::regex_replace(str, e, "");
}


size_t find_end_brace_pos(const string& str, const size_t start) {
  int braces_count = 1;
  assert(str[start] == '[');
  size_t i = start + 1;
  while (braces_count > 0 && i < str.size()) {
    if (str[i] == '[') {
      braces_count++;
    }
    else if (str[i] == ']') {
      braces_count--;
    }
    i++;
  }
  return i - 1;
}


void process_single_count_term(
    const SharedGenData& data,
    const string& mdl_string,
    bool& rxn_not_mol,
    bool& molecules_not_species,
    string& what_to_count,
    string& where_to_count,
    string& orientation) {

  // mdl_string is always in the form COUNT[what,where]
  size_t start_brace = mdl_string.find('[');
  size_t comma = mdl_string.rfind(',');
  size_t end_brace = find_end_brace_pos(mdl_string, start_brace);

  if (mdl_string.find(COUNT) == string::npos) {
    ERROR("String 'COUNT' was not found in mdl_string '" + mdl_string + "'.");
  }
  if (start_brace == string::npos || comma == string::npos || end_brace == string::npos ||
      start_brace > comma || start_brace > end_brace || comma > end_brace
  ) {
    ERROR("Malformed mdl_string '" + mdl_string + "'.");
  }

  what_to_count = mdl_string.substr(start_brace + 1, comma - start_brace - 1);
  what_to_count = trim(what_to_count);

  // default is 'molecules_pattern', for now we are storing the
  // as a comment because the counting type belongs to the count term,
  // not to the whole 'reaction_output_list'
  // TODO: this should resolved in a better way
  molecules_not_species = true;
  if (what_to_count.find(MARKER_SPECIES_COMMENT) != string::npos) {
    molecules_not_species = false;
  }

  what_to_count = remove_c_comment(what_to_count);

  // process orientation
  orientation = "";
  assert(what_to_count != "");
  char last_c = what_to_count.back();
  if (last_c == '\'' || last_c == ',' || last_c == ';') {
    string s;
    s = last_c;
    orientation = convert_orientation(s);

    what_to_count = what_to_count.substr(0, what_to_count.size() - 1);
  }

  if (data.find_reaction_rule_info(what_to_count) != nullptr) {
    rxn_not_mol = true;
  }
  else {
    // if we did not find the name to be a reaction, we assume it is a BNGL pattern
    rxn_not_mol = false;
  }


  string where_tmp = mdl_string.substr(comma + 1, end_brace - comma - 1);

  size_t dot_pos = where_tmp.find('.');
  if (dot_pos != string::npos) {
    // no completely sure when a '.' can appear
    where_tmp = where_tmp.substr(dot_pos + 1);
  }
  where_tmp = trim(where_tmp);
  if (where_tmp == WORLD) {
    where_tmp = "";
  }

  // remove all [ALL] and replace box1[box1_sr1] -> box1_box1_sr1
  where_to_count = "";
  size_t i = 0;
  while (i < where_tmp.size()) {
    char c = where_tmp[i];
    if (c == '[') {
      // followed by ALL]?
      if (i + 4 < where_tmp.size() && where_tmp.substr(i + 1, 4) == "ALL]") {
        // skip
        i += 5;
      }
      else {
        // replace
        where_to_count += '_';
        i++;
      }
    }
    else if (c == ']') {
      // ignore
      i++;
    }
    else {
      // keep character
      where_to_count += c;
      i++;
    }
  }
}

// sets val if the name_or_value is a floating point value,
// if not, tries to find the parameter and reads its value
// returns true on success
// parameters are not evaluated and only one level is tried,
// returns false if value was not obtained
bool get_parameter_value(Json::Value& mcell, const string& name_or_value, double& val) {
  try {
    val = stod(name_or_value);
    return true;
  }
  catch (const std::invalid_argument&) {
    // not a float, try to get parameter value
    if (mcell.isMember(KEY_PARAMETER_SYSTEM) && mcell.isMember(KEY_MODEL_PARAMETERS)) {
      Json::Value& params = mcell[KEY_PARAMETER_SYSTEM][KEY_MODEL_PARAMETERS];
      for (Value::ArrayIndex i = 0; i < params.size(); i++) {
        if (params[i][KEY_NAME].asString() == name_or_value) {
          try {
            if (params[i].isMember(KEY__EXTRAS)) {
              val = stod(params[i][KEY__EXTRAS][KEY_PAR_VALUE].asString());
            }
            else {
              val = stod(params[i][KEY_PAR_EXPRESSION].asString());
            }
            return true;
          }
          catch (const std::invalid_argument&) {
            return false;
          }
        }
      }
    }
  }
  return false;
}


bool is_volume_mol_type(Json::Value& mcell, const std::string& mol_type_name) {

  string mol_type_name_no_comp = remove_compartments(mol_type_name);

  Value& define_molecules = get_node(mcell, KEY_DEFINE_MOLECULES);
  check_version(KEY_DEFINE_MOLECULES, define_molecules, VER_DM_2014_10_24_1638);

  Value& molecule_list = get_node(define_molecules, KEY_MOLECULE_LIST);
  for (Value::ArrayIndex i = 0; i < molecule_list.size(); i++) {
    Value& molecule_list_item = molecule_list[i];
    check_version(KEY_MOLECULE_LIST, molecule_list_item, VER_DM_2018_10_16_1632);

    string name = molecule_list_item[KEY_MOL_NAME].asString();
    if (name != mol_type_name_no_comp) {
      continue;
    }

    string mol_type = molecule_list_item[KEY_MOL_TYPE].asString();
    CHECK_PROPERTY(mol_type == VALUE_MOL_TYPE_2D || mol_type == VALUE_MOL_TYPE_3D);
    return mol_type == VALUE_MOL_TYPE_3D;
  }

  ERROR("Could not find species or molecule type '" + mol_type_name_no_comp + "'.");
}


static void get_mol_types_in_species(const std::string& species_name, vector<string>& mol_types) {
  mol_types.clear();

  size_t i = 0;
  string current_name;
  bool in_name = true;
  while (i < species_name.size()) {
    char c = species_name[i];
    if (c == '(') {
      in_name = false;
      assert(current_name != "");
      mol_types.push_back(current_name);
      current_name = "";
    }
    else if (c == '.') {
      in_name = true;
    }
    else if (in_name && !isspace(c)) {
      current_name += c;
    }
    i++;
  }

  if (current_name != "") {
    mol_types.push_back(current_name);
  }
}


bool is_volume_species(Json::Value& mcell, const std::string& species_name) {
  if (is_simple_species(species_name)) {
    return is_volume_mol_type(mcell, species_name);
  }
  else {
    vector<string> mol_types;
    get_mol_types_in_species( species_name, mol_types);
    for (string& mt: mol_types) {
      if (!is_volume_mol_type(mcell, mt)) {
        // surface
        return false;
      }
    }
    return true;
  }
}

} // namespace MCell