irep.cpp

Go to the documentation of this file.

/*******************************************************************\
 
Module: Internal Representation
 
Author: Daniel Kroening, kroening@kroening.com
 
\*******************************************************************/
 
 
#include "irep.h"
 
#include "string2int.h"
#include "string_hash.h"
#include "irep_hash.h"
 
irept nil_rep_storage;
 
const irept &get_nil_irep()
{
  if(nil_rep_storage.id().empty()) // initialized?
    nil_rep_storage.id(ID_nil);
  return nil_rep_storage;
}
 
void irept::move_to_named_sub(const irep_idt &name, irept &irep)
{
  #ifdef SHARING
  detach();
  #endif
  add(name).swap(irep);
  irep.clear();
}
 
void irept::move_to_sub(irept &irep)
{
  #ifdef SHARING
  detach();
  #endif
  get_sub().push_back(get_nil_irep());
  get_sub().back().swap(irep);
}
 
const irep_idt &irept::get(const irep_idt &name) const
{
  const named_subt &s = get_named_sub();
  named_subt::const_iterator it=s.find(name);
 
  if(it==s.end())
  {
    static const irep_idt empty;
    return empty;
  }
  return it->second.id();
}
 
bool irept::get_bool(const irep_idt &name) const
{
  return get(name)==ID_1;
}
 
int irept::get_int(const irep_idt &name) const
{
  return unsafe_string2int(get_string(name));
}
 
std::size_t irept::get_size_t(const irep_idt &name) const
{
  return unsafe_string2size_t(get_string(name));
}
 
long long irept::get_long_long(const irep_idt &name) const
{
  return unsafe_string2signedlonglong(get_string(name));
}
 
void irept::set(const irep_idt &name, const long long value)
{
#ifdef USE_DSTRING
  add(name).id(to_dstring(value));
#else
  add(name).id(std::to_string(value));
#endif
}
 
void irept::set_size_t(const irep_idt &name, const std::size_t value)
{
#ifdef USE_DSTRING
  add(name).id(to_dstring(value));
#else
  add(name).id(std::to_string(value));
#endif
}
 
void irept::remove(const irep_idt &name)
{
#if NAMED_SUB_IS_FORWARD_LIST
  return get_named_sub().remove(name);
#else
  named_subt &s = get_named_sub();
  s.erase(name);
#endif
}
 
const irept &irept::find(const irep_idt &name) const
{
  const named_subt &s = get_named_sub();
  auto it = s.find(name);
 
  if(it==s.end())
    return get_nil_irep();
  return it->second;
}
 
irept &irept::add(const irep_idt &name)
{
  named_subt &s = get_named_sub();
  return s[name];
}
 
irept &irept::add(const irep_idt &name, irept irep)
{
  named_subt &s = get_named_sub();
 
#if NAMED_SUB_IS_FORWARD_LIST
  return s.add(name, std::move(irep));
#else
  std::pair<named_subt::iterator, bool> entry = s.emplace(
    std::piecewise_construct,
    std::forward_as_tuple(name),
    std::forward_as_tuple(irep));
 
  if(!entry.second)
    entry.first->second = std::move(irep);
 
  return entry.first->second;
#endif
}
 
#ifdef IREP_HASH_STATS
unsigned long long irep_cmp_cnt=0;
unsigned long long irep_cmp_ne_cnt=0;
#endif
 
bool irept::operator==(const irept &other) const
{
  #ifdef IREP_HASH_STATS
  ++irep_cmp_cnt;
  #endif
  #ifdef SHARING
  if(data==other.data)
    return true;
  #endif
 
  if(id() != other.id() || get_sub() != other.get_sub()) // recursive call
  {
    #ifdef IREP_HASH_STATS
    ++irep_cmp_ne_cnt;
    #endif
    return false;
  }
 
  const auto &this_named_sub = get_named_sub();
  const auto &other_named_sub = other.get_named_sub();
 
  // walk in sync, ignoring comments, until end of both maps
  named_subt::const_iterator this_it = this_named_sub.begin();
  named_subt::const_iterator other_it = other_named_sub.begin();
 
  while(this_it != this_named_sub.end() || other_it != other_named_sub.end())
  {
    if(this_it != this_named_sub.end() && is_comment(this_it->first))
    {
      this_it++;
      continue;
    }
 
    if(other_it != other_named_sub.end() && is_comment(other_it->first))
    {
      other_it++;
      continue;
    }
 
    if(
      this_it == this_named_sub.end() ||   // reached end of 'this'
      other_it == other_named_sub.end() || // reached the end of 'other'
      this_it->first != other_it->first ||
      this_it->second != other_it->second) // recursive call
    {
#ifdef IREP_HASH_STATS
      ++irep_cmp_ne_cnt;
#endif
      return false;
    }
 
    this_it++;
    other_it++;
  }
 
  // reached the end of both
  return true;
}
 
bool irept::full_eq(const irept &other) const
{
  #ifdef SHARING
  if(data==other.data)
    return true;
  #endif
 
  if(id()!=other.id())
    return false;
 
  const irept::subt &i1_sub=get_sub();
  const irept::subt &i2_sub=other.get_sub();
 
  if(i1_sub.size() != i2_sub.size())
    return false;
 
  const irept::named_subt &i1_named_sub=get_named_sub();
  const irept::named_subt &i2_named_sub=other.get_named_sub();
 
  if(i1_named_sub.size() != i2_named_sub.size())
    return false;
 
  for(std::size_t i=0; i<i1_sub.size(); i++)
    if(!i1_sub[i].full_eq(i2_sub[i]))
      return false;
 
  {
    irept::named_subt::const_iterator i1_it=i1_named_sub.begin();
    irept::named_subt::const_iterator i2_it=i2_named_sub.begin();
 
    for(; i1_it!=i1_named_sub.end(); i1_it++, i2_it++)
      if(i1_it->first!=i2_it->first ||
         !i1_it->second.full_eq(i2_it->second))
        return false;
  }
 
  return true;
}
 
bool irept::ordering(const irept &other) const
{
  return compare(other)<0;
 
  #if 0
  if(X.data<Y.data)
    return true;
  if(Y.data<X.data)
    return false;
 
  if(X.sub.size()<Y.sub.size())
    return true;
  if(Y.sub.size()<X.sub.size())
    return false;
 
  {
    irept::subt::const_iterator it1, it2;
 
    for(it1=X.sub.begin(),
        it2=Y.sub.begin();
        it1!=X.sub.end() && it2!=Y.sub.end();
        it1++,
        it2++)
    {
      if(ordering(*it1, *it2))
        return true;
      if(ordering(*it2, *it1))
        return false;
    }
 
    INVARIANT(it1==X.sub.end() && it2==Y.sub.end(),
              "Unequal lengths will return before this");
  }
 
  if(X.named_sub.size()<Y.named_sub.size())
    return true;
  if(Y.named_sub.size()<X.named_sub.size())
    return false;
 
  {
    irept::named_subt::const_iterator it1, it2;
 
    for(it1=X.named_sub.begin(),
        it2=Y.named_sub.begin();
        it1!=X.named_sub.end() && it2!=Y.named_sub.end();
        it1++,
        it2++)
    {
      if(it1->first<it2->first)
        return true;
      if(it2->first<it1->first)
        return false;
 
      if(ordering(it1->second, it2->second))
        return true;
      if(ordering(it2->second, it1->second))
        return false;
    }
 
    INVARIANT(it1==X.named_sub.end() && it2==Y.named_sub.end(),
              "Unequal lengths will return before this");
  }
 
  return false;
  #endif
}
 
int irept::compare(const irept &i) const
{
  int r;
 
  r=id().compare(i.id());
  if(r!=0)
    return r;
 
  const subt::size_type size=get_sub().size(),
        i_size=i.get_sub().size();
 
  if(size<i_size)
    return -1;
  if(size>i_size)
    return 1;
 
  {
    irept::subt::const_iterator it1, it2;
 
    for(it1=get_sub().begin(),
        it2=i.get_sub().begin();
        it1!=get_sub().end() && it2!=i.get_sub().end();
        it1++,
        it2++)
    {
      r=it1->compare(*it2);
      if(r!=0)
        return r;
    }
 
    INVARIANT(it1==get_sub().end() && it2==i.get_sub().end(),
              "Unequal lengths will return before this");
  }
 
  const auto n_size = number_of_non_comments(get_named_sub()),
             i_n_size = number_of_non_comments(i.get_named_sub());
 
  if(n_size<i_n_size)
    return -1;
  if(n_size>i_n_size)
    return 1;
 
  {
    irept::named_subt::const_iterator it1, it2;
 
    // clang-format off
    for(it1 = get_named_sub().begin(),
        it2 = i.get_named_sub().begin();
        it1 != get_named_sub().end() ||
        it2 != i.get_named_sub().end();
        ) // no iterator increments
    // clang-format on
    {
      if(it1 != get_named_sub().end() && is_comment(it1->first))
      {
        it1++;
        continue;
      }
 
      if(it2 != i.get_named_sub().end() && is_comment(it2->first))
      {
        it2++;
        continue;
      }
 
      // the case that both it1 and it2 are .end() is treated
      // by the loop guard; furthermore, the number of non-comments
      // must be the same
      INVARIANT(it1 != get_named_sub().end(), "not at end of get_named_sub()");
      INVARIANT(
        it2 != i.get_named_sub().end(), "not at end of i.get_named_sub()");
 
      r=it1->first.compare(it2->first);
      if(r!=0)
        return r;
 
      r=it1->second.compare(it2->second);
      if(r!=0)
        return r;
 
      it1++;
      it2++;
    }
 
    INVARIANT(
      it1 == get_named_sub().end() && it2 == i.get_named_sub().end(),
      "Unequal number of non-comments will return before this");
  }
 
  // equal
  return 0;
}
 
bool irept::operator<(const irept &other) const
{
  return ordering(other);
}
 
#ifdef IREP_HASH_STATS
unsigned long long irep_hash_cnt=0;
#endif
 
std::size_t irept::number_of_non_comments(const named_subt &named_sub)
{
  std::size_t result = 0;
 
  for(const auto &n : named_sub)
    if(!is_comment(n.first))
      result++;
 
  return result;
}
 
std::size_t irept::hash() const
{
#if HASH_CODE
  if(read().hash_code!=0)
    return read().hash_code;
  #endif
 
  const irept::subt &sub=get_sub();
  const irept::named_subt &named_sub=get_named_sub();
 
  std::size_t result=hash_string(id());
 
  for(const auto &irep : sub)
    result = hash_combine(result, irep.hash());
 
  std::size_t number_of_named_ireps = 0;
 
  for(const auto &irep_entry : named_sub)
  {
    if(!is_comment(irep_entry.first)) // this variant ignores comments
    {
      result = hash_combine(result, hash_string(irep_entry.first));
      result = hash_combine(result, irep_entry.second.hash());
      number_of_named_ireps++;
    }
  }
 
  result = hash_finalize(result, sub.size() + number_of_named_ireps);
 
#if HASH_CODE
  read().hash_code=result;
#endif
#ifdef IREP_HASH_STATS
  ++irep_hash_cnt;
#endif
  return result;
}
 
std::size_t irept::full_hash() const
{
  const irept::subt &sub=get_sub();
  const irept::named_subt &named_sub=get_named_sub();
 
  std::size_t result=hash_string(id());
 
  for(const auto &irep : sub)
    result = hash_combine(result, irep.full_hash());
 
  // this variant includes all named_sub elements
  for(const auto &irep_entry : named_sub)
  {
    result = hash_combine(result, hash_string(irep_entry.first));
    result = hash_combine(result, irep_entry.second.full_hash());
  }
 
  const std::size_t named_sub_size = named_sub.size();
  result = hash_finalize(result, named_sub_size + sub.size());
 
  return result;
}
 
static void indent_str(std::string &s, unsigned indent)
{
  for(unsigned i=0; i<indent; i++)
    s+=' ';
}
 
std::string irept::pretty(unsigned indent, unsigned max_indent) const
{
  if(max_indent>0 && indent>max_indent)
    return "";
 
  std::string result;
 
  if(!id().empty())
  {
    result+=id_string();
    indent+=2;
  }
 
  for(const auto &irep_entry : get_named_sub())
  {
    result+="\n";
    indent_str(result, indent);
 
    result+="* ";
    result += id2string(irep_entry.first);
    result+=": ";
 
    result += irep_entry.second.pretty(indent + 2, max_indent);
  }
 
  std::size_t count=0;
 
  for(const auto &irep : get_sub())
  {
    result+="\n";
    indent_str(result, indent);
 
    result+=std::to_string(count++);
    result+=": ";
 
    result += irep.pretty(indent + 2, max_indent);
  }
 
  return result;
}
 
irep_pretty_diagnosticst::irep_pretty_diagnosticst(const irept &irep)
  : irep(irep)
{
}