* tree.c (array_at_struct_end_p): Look through MEM_REF.

[official-gcc.git] / gcc / fortran / scanner.c

/* Character scanner.
   Copyright (C) 2000-2016 Free Software Foundation, Inc.
   Contributed by Andy Vaught

This file is part of GCC.

GCC is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free
Software Foundation; either version 3, or (at your option) any later
version.

GCC is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
for more details.

You should have received a copy of the GNU General Public License
along with GCC; see the file COPYING3.  If not see
<http://www.gnu.org/licenses/>.  */

/* Set of subroutines to (ultimately) return the next character to the
   various matching subroutines.  This file's job is to read files and
   build up lines that are parsed by the parser.  This means that we
   handle continuation lines and "include" lines.

   The first thing the scanner does is to load an entire file into
   memory.  We load the entire file into memory for a couple reasons.
   The first is that we want to be able to deal with nonseekable input
   (pipes, stdin) and there is a lot of backing up involved during
   parsing.

   The second is that we want to be able to print the locus of errors,
   and an error on line 999999 could conflict with something on line
   one.  Given nonseekable input, we've got to store the whole thing.

   One thing that helps are the column truncation limits that give us
   an upper bound on the size of individual lines.  We don't store the
   truncated stuff.

   From the scanner's viewpoint, the higher level subroutines ask for
   new characters and do a lot of jumping backwards.  */

#include "config.h"
#include "system.h"
#include "coretypes.h"
#include "gfortran.h"
#include "toplev.h"     /* For set_src_pwd.  */
#include "debug.h"
#include "options.h"
#include "cpp.h"
#include "scanner.h"

/* List of include file search directories.  */
gfc_directorylist *include_dirs, *intrinsic_modules_dirs;

static gfc_file *file_head, *current_file;

static int continue_flag, end_flag, gcc_attribute_flag;
/* If !$omp/!$acc occurred in current comment line.  */
static int openmp_flag, openacc_flag;
static int continue_count, continue_line;
static locus openmp_locus;
static locus openacc_locus;
static locus gcc_attribute_locus;

gfc_source_form gfc_current_form;
static gfc_linebuf *line_head, *line_tail;
       
locus gfc_current_locus;
const char *gfc_source_file;
static FILE *gfc_src_file;
static gfc_char_t *gfc_src_preprocessor_lines[2];

static struct gfc_file_change
{
  const char *filename;
  gfc_linebuf *lb;
  int line;
} *file_changes;
size_t file_changes_cur, file_changes_count;
size_t file_changes_allocated;


/* Functions dealing with our wide characters (gfc_char_t) and
   sequences of such characters.  */

int
gfc_wide_fits_in_byte (gfc_char_t c)
{
  return (c <= UCHAR_MAX);
}

static inline int
wide_is_ascii (gfc_char_t c)
{
  return (gfc_wide_fits_in_byte (c) && ((unsigned char) c & ~0x7f) == 0);
}

int
gfc_wide_is_printable (gfc_char_t c)
{
  return (gfc_wide_fits_in_byte (c) && ISPRINT ((unsigned char) c));
}

gfc_char_t
gfc_wide_tolower (gfc_char_t c)
{
  return (wide_is_ascii (c) ? (gfc_char_t) TOLOWER((unsigned char) c) : c);
}

gfc_char_t
gfc_wide_toupper (gfc_char_t c)
{
  return (wide_is_ascii (c) ? (gfc_char_t) TOUPPER((unsigned char) c) : c);
}

int
gfc_wide_is_digit (gfc_char_t c)
{
  return (c >= '0' && c <= '9');
}

static inline int
wide_atoi (gfc_char_t *c)
{
#define MAX_DIGITS 20
  char buf[MAX_DIGITS+1];
  int i = 0;

  while (gfc_wide_is_digit(*c) && i < MAX_DIGITS)
    buf[i++] = *c++;
  buf[i] = '\0';
  return atoi (buf);
}

size_t
gfc_wide_strlen (const gfc_char_t *str)
{
  size_t i;

  for (i = 0; str[i]; i++)
    ;

  return i;
}

gfc_char_t *
gfc_wide_memset (gfc_char_t *b, gfc_char_t c, size_t len)
{
  size_t i;

  for (i = 0; i < len; i++)
    b[i] = c;

  return b;
}

static gfc_char_t *
wide_strcpy (gfc_char_t *dest, const gfc_char_t *src)
{
  gfc_char_t *d;

  for (d = dest; (*d = *src) != '\0'; ++src, ++d)
    ;

  return dest;
}

static gfc_char_t *
wide_strchr (const gfc_char_t *s, gfc_char_t c)
{
  do {
    if (*s == c)
      {
        return CONST_CAST(gfc_char_t *, s);
      }
  } while (*s++);
  return 0;
}

char *
gfc_widechar_to_char (const gfc_char_t *s, int length)
{
  size_t len, i;
  char *res;

  if (s == NULL)
    return NULL;

  /* Passing a negative length is used to indicate that length should be
     calculated using gfc_wide_strlen().  */
  len = (length >= 0 ? (size_t) length : gfc_wide_strlen (s));
  res = XNEWVEC (char, len + 1);

  for (i = 0; i < len; i++)
    {
      gcc_assert (gfc_wide_fits_in_byte (s[i]));
      res[i] = (unsigned char) s[i];
    }

  res[len] = '\0';
  return res;
}

gfc_char_t *
gfc_char_to_widechar (const char *s)
{
  size_t len, i;
  gfc_char_t *res;

  if (s == NULL)
    return NULL;

  len = strlen (s);
  res = gfc_get_wide_string (len + 1);

  for (i = 0; i < len; i++)
    res[i] = (unsigned char) s[i];

  res[len] = '\0';
  return res;
}

static int
wide_strncmp (const gfc_char_t *s1, const char *s2, size_t n)
{
  gfc_char_t c1, c2;

  while (n-- > 0)
    {
      c1 = *s1++;
      c2 = *s2++;
      if (c1 != c2)
        return (c1 > c2 ? 1 : -1);
      if (c1 == '\0')
        return 0;
    }
  return 0;
}

int
gfc_wide_strncasecmp (const gfc_char_t *s1, const char *s2, size_t n)
{
  gfc_char_t c1, c2;

  while (n-- > 0)
    {
      c1 = gfc_wide_tolower (*s1++);
      c2 = TOLOWER (*s2++);
      if (c1 != c2)
        return (c1 > c2 ? 1 : -1);
      if (c1 == '\0')
        return 0;
    }
  return 0;
}


/* Main scanner initialization.  */

void
gfc_scanner_init_1 (void)
{
  file_head = NULL;
  line_head = NULL;
  line_tail = NULL;

  continue_count = 0;
  continue_line = 0;

  end_flag = 0;
}


/* Main scanner destructor.  */

void
gfc_scanner_done_1 (void)
{
  gfc_linebuf *lb;
  gfc_file *f;

  while(line_head != NULL) 
    {
      lb = line_head->next;
      free (line_head);
      line_head = lb;
    }
     
  while(file_head != NULL) 
    {
      f = file_head->next;
      free (file_head->filename);
      free (file_head);
      file_head = f;    
    }
}


/* Adds path to the list pointed to by list.  */

static void
add_path_to_list (gfc_directorylist **list, const char *path,
                  bool use_for_modules, bool head, bool warn)
{
  gfc_directorylist *dir;
  const char *p;
  char *q;
  struct stat st;
  size_t len;
  int i;
  
  p = path;
  while (*p == ' ' || *p == '\t')  /* someone might do "-I include" */
    if (*p++ == '\0')
      return;

  /* Strip trailing directory separators from the path, as this
     will confuse Windows systems.  */
  len = strlen (p);
  q = (char *) alloca (len + 1);
  memcpy (q, p, len + 1);
  i = len - 1;
  while (i >=0 && IS_DIR_SEPARATOR (q[i]))
    q[i--] = '\0';

  if (stat (q, &st))
    {
      if (errno != ENOENT)
        gfc_warning_now (0, "Include directory %qs: %s", path,
                         xstrerror(errno));
      else if (warn)
        gfc_warning_now (OPT_Wmissing_include_dirs,
                         "Nonexistent include directory %qs", path);
      return;
    }
  else if (!S_ISDIR (st.st_mode))
    {
      gfc_fatal_error ("%qs is not a directory", path);
      return;
    }

  if (head || *list == NULL)
    {
      dir = XCNEW (gfc_directorylist);
      if (!head)
        *list = dir;
    }
  else
    {
      dir = *list;
      while (dir->next)
        dir = dir->next;

      dir->next = XCNEW (gfc_directorylist);
      dir = dir->next;
    }

  dir->next = head ? *list : NULL;
  if (head)
    *list = dir;
  dir->use_for_modules = use_for_modules;
  dir->path = XCNEWVEC (char, strlen (p) + 2);
  strcpy (dir->path, p);
  strcat (dir->path, "/");      /* make '/' last character */
}


void
gfc_add_include_path (const char *path, bool use_for_modules, bool file_dir,
                      bool warn)
{
  add_path_to_list (&include_dirs, path, use_for_modules, file_dir, warn);

  /* For '#include "..."' these directories are automatically searched.  */
  if (!file_dir)
    gfc_cpp_add_include_path (xstrdup(path), true);
}


void
gfc_add_intrinsic_modules_path (const char *path)
{
  add_path_to_list (&intrinsic_modules_dirs, path, true, false, false);
}


/* Release resources allocated for options.  */

void
gfc_release_include_path (void)
{
  gfc_directorylist *p;

  while (include_dirs != NULL)
    {
      p = include_dirs;
      include_dirs = include_dirs->next;
      free (p->path);
      free (p);
    }

  while (intrinsic_modules_dirs != NULL)
    {
      p = intrinsic_modules_dirs;
      intrinsic_modules_dirs = intrinsic_modules_dirs->next;
      free (p->path);
      free (p);
    }

  free (gfc_option.module_dir);
}


static FILE *
open_included_file (const char *name, gfc_directorylist *list,
                    bool module, bool system)
{
  char *fullname;
  gfc_directorylist *p;
  FILE *f;

  for (p = list; p; p = p->next)
    {
      if (module && !p->use_for_modules)
        continue;

      fullname = (char *) alloca(strlen (p->path) + strlen (name) + 1);
      strcpy (fullname, p->path);
      strcat (fullname, name);

      f = gfc_open_file (fullname);
      if (f != NULL)
        {
          if (gfc_cpp_makedep ())
            gfc_cpp_add_dep (fullname, system);

          return f;
        }
    }

  return NULL;
}


/* Opens file for reading, searching through the include directories
   given if necessary.  If the include_cwd argument is true, we try
   to open the file in the current directory first.  */

FILE *
gfc_open_included_file (const char *name, bool include_cwd, bool module)
{
  FILE *f = NULL;

  if (IS_ABSOLUTE_PATH (name) || include_cwd)
    {
      f = gfc_open_file (name);
      if (f && gfc_cpp_makedep ())
        gfc_cpp_add_dep (name, false);
    }

  if (!f)
    f = open_included_file (name, include_dirs, module, false);

  return f;
}


/* Test to see if we're at the end of the main source file.  */

int
gfc_at_end (void)
{
  return end_flag;
}


/* Test to see if we're at the end of the current file.  */

int
gfc_at_eof (void)
{
  if (gfc_at_end ())
    return 1;

  if (line_head == NULL)
    return 1;                   /* Null file */

  if (gfc_current_locus.lb == NULL)
    return 1;

  return 0;
}


/* Test to see if we're at the beginning of a new line.  */

int
gfc_at_bol (void)
{
  if (gfc_at_eof ())
    return 1;

  return (gfc_current_locus.nextc == gfc_current_locus.lb->line);
}


/* Test to see if we're at the end of a line.  */

int
gfc_at_eol (void)
{
  if (gfc_at_eof ())
    return 1;

  return (*gfc_current_locus.nextc == '\0');
}

static void
add_file_change (const char *filename, int line)
{
  if (file_changes_count == file_changes_allocated)
    {
      if (file_changes_allocated)
        file_changes_allocated *= 2;
      else
        file_changes_allocated = 16;
      file_changes = XRESIZEVEC (struct gfc_file_change, file_changes,
                                 file_changes_allocated);
    }
  file_changes[file_changes_count].filename = filename;
  file_changes[file_changes_count].lb = NULL;
  file_changes[file_changes_count++].line = line;
}

static void
report_file_change (gfc_linebuf *lb)
{
  size_t c = file_changes_cur;
  while (c < file_changes_count
         && file_changes[c].lb == lb)
    {
      if (file_changes[c].filename)
        (*debug_hooks->start_source_file) (file_changes[c].line,
                                           file_changes[c].filename);
      else
        (*debug_hooks->end_source_file) (file_changes[c].line);
      ++c;
    }
  file_changes_cur = c;
}

void
gfc_start_source_files (void)
{
  /* If the debugger wants the name of the main source file,
     we give it.  */
  if (debug_hooks->start_end_main_source_file)
    (*debug_hooks->start_source_file) (0, gfc_source_file);

  file_changes_cur = 0;
  report_file_change (gfc_current_locus.lb);
}

void
gfc_end_source_files (void)
{
  report_file_change (NULL);

  if (debug_hooks->start_end_main_source_file)
    (*debug_hooks->end_source_file) (0);
}

/* Advance the current line pointer to the next line.  */

void
gfc_advance_line (void)
{
  if (gfc_at_end ())
    return;

  if (gfc_current_locus.lb == NULL) 
    {
      end_flag = 1;
      return;
    } 

  if (gfc_current_locus.lb->next
      && !gfc_current_locus.lb->next->dbg_emitted)
    {
      report_file_change (gfc_current_locus.lb->next);
      gfc_current_locus.lb->next->dbg_emitted = true;
    }

  gfc_current_locus.lb = gfc_current_locus.lb->next;

  if (gfc_current_locus.lb != NULL)      
    gfc_current_locus.nextc = gfc_current_locus.lb->line;
  else 
    {
      gfc_current_locus.nextc = NULL;
      end_flag = 1;
    }       
}


/* Get the next character from the input, advancing gfc_current_file's
   locus.  When we hit the end of the line or the end of the file, we
   start returning a '\n' in order to complete the current statement.
   No Fortran line conventions are implemented here.

   Requiring explicit advances to the next line prevents the parse
   pointer from being on the wrong line if the current statement ends
   prematurely.  */

static gfc_char_t
next_char (void)
{
  gfc_char_t c;
  
  if (gfc_current_locus.nextc == NULL)
    return '\n';

  c = *gfc_current_locus.nextc++;
  if (c == '\0')
    {
      gfc_current_locus.nextc--; /* Remain on this line.  */
      c = '\n';
    }

  return c;
}


/* Skip a comment.  When we come here the parse pointer is positioned
   immediately after the comment character.  If we ever implement
   compiler directives within comments, here is where we parse the
   directive.  */

static void
skip_comment_line (void)
{
  gfc_char_t c;

  do
    {
      c = next_char ();
    }
  while (c != '\n');

  gfc_advance_line ();
}


int
gfc_define_undef_line (void)
{
  char *tmp;

  /* All lines beginning with '#' are either #define or #undef.  */
  if (debug_info_level != DINFO_LEVEL_VERBOSE || gfc_peek_ascii_char () != '#')
    return 0;

  if (wide_strncmp (gfc_current_locus.nextc, "#define ", 8) == 0)
    {
      tmp = gfc_widechar_to_char (&gfc_current_locus.nextc[8], -1);
      (*debug_hooks->define) (gfc_linebuf_linenum (gfc_current_locus.lb),
                              tmp);
      free (tmp);
    }

  if (wide_strncmp (gfc_current_locus.nextc, "#undef ", 7) == 0)
    {
      tmp = gfc_widechar_to_char (&gfc_current_locus.nextc[7], -1);
      (*debug_hooks->undef) (gfc_linebuf_linenum (gfc_current_locus.lb),
                             tmp);
      free (tmp);
    }

  /* Skip the rest of the line.  */
  skip_comment_line ();

  return 1;
}


/* Return true if GCC$ was matched.  */
static bool
skip_gcc_attribute (locus start)
{
  bool r = false;
  char c;
  locus old_loc = gfc_current_locus;

  if ((c = next_char ()) == 'g' || c == 'G')
    if ((c = next_char ()) == 'c' || c == 'C')
      if ((c = next_char ()) == 'c' || c == 'C')
        if ((c = next_char ()) == '$')
          r = true;

  if (r == false)
    gfc_current_locus = old_loc;
  else
   {
      gcc_attribute_flag = 1;
      gcc_attribute_locus = old_loc;
      gfc_current_locus = start;
   }

  return r;
}

/* Return true if CC was matched.  */
static bool
skip_free_oacc_sentinel (locus start, locus old_loc)
{
  bool r = false;
  char c;

  if ((c = next_char ()) == 'c' || c == 'C')
    if ((c = next_char ()) == 'c' || c == 'C')
      r = true;

  if (r)
   {
      if ((c = next_char ()) == ' ' || c == '\t'
          || continue_flag)
        {
          while (gfc_is_whitespace (c))
            c = next_char ();
          if (c != '\n' && c != '!')
            {
              openacc_flag = 1;
              openacc_locus = old_loc;
              gfc_current_locus = start;
            }
          else 
            r = false;
        }
      else
        {
          gfc_warning_now (0, "!$ACC at %C starts a commented "
                           "line as it neither is followed "
                           "by a space nor is a "
                           "continuation line");
          r = false;
        }
   }

  return r;
}

/* Return true if MP was matched.  */
static bool
skip_free_omp_sentinel (locus start, locus old_loc)
{
  bool r = false;
  char c;

  if ((c = next_char ()) == 'm' || c == 'M')
    if ((c = next_char ()) == 'p' || c == 'P')
      r = true;

  if (r)
   {
      if ((c = next_char ()) == ' ' || c == '\t'
          || continue_flag)
        {
          while (gfc_is_whitespace (c))
            c = next_char ();
          if (c != '\n' && c != '!')
            {
              openmp_flag = 1;
              openmp_locus = old_loc;
              gfc_current_locus = start;
            }
          else 
            r = false;
        }
      else
        {
          gfc_warning_now (0, "!$OMP at %C starts a commented "
                           "line as it neither is followed "
                           "by a space nor is a "
                           "continuation line");
          r = false;
        }
   }

  return r;
}

/* Comment lines are null lines, lines containing only blanks or lines
   on which the first nonblank line is a '!'.
   Return true if !$ openmp or openacc conditional compilation sentinel was
   seen.  */

static bool
skip_free_comments (void)
{
  locus start;
  gfc_char_t c;
  int at_bol;

  for (;;)
    {
      at_bol = gfc_at_bol ();
      start = gfc_current_locus;
      if (gfc_at_eof ())
        break;

      do
        c = next_char ();
      while (gfc_is_whitespace (c));

      if (c == '\n')
        {
          gfc_advance_line ();
          continue;
        }

      if (c == '!')
        {
          /* Keep the !GCC$ line.  */
          if (at_bol && skip_gcc_attribute (start))
            return false;

          /* If -fopenmp/-fopenacc, we need to handle here 2 things:
             1) don't treat !$omp/!$acc as comments, but directives
             2) handle OpenMP/OpenACC conditional compilation, where
                !$ should be treated as 2 spaces (for initial lines
                only if followed by space).  */
          if (at_bol)
          {
            if ((flag_openmp || flag_openmp_simd)
                && flag_openacc)
              {
                locus old_loc = gfc_current_locus;
                if (next_char () == '$')
                  {
                    c = next_char ();
                    if (c == 'o' || c == 'O')
                      {
                        if (skip_free_omp_sentinel (start, old_loc))
                          return false;
                        gfc_current_locus = old_loc;
                        next_char ();
                        c = next_char ();
                      }
                    else if (c == 'a' || c == 'A')
                      {
                        if (skip_free_oacc_sentinel (start, old_loc))
                          return false;
                        gfc_current_locus = old_loc;
                        next_char ();
                        c = next_char ();
                      }
                    if (continue_flag || c == ' ' || c == '\t')
                      {
                        gfc_current_locus = old_loc;
                        next_char ();
                        openmp_flag = openacc_flag = 0;
                        return true;
                      }
                  }
                gfc_current_locus = old_loc;
              }
            else if ((flag_openmp || flag_openmp_simd)
                     && !flag_openacc)
              {
                locus old_loc = gfc_current_locus;
                if (next_char () == '$')
                  {
                    c = next_char ();
                    if (c == 'o' || c == 'O')
                      {
                        if (skip_free_omp_sentinel (start, old_loc))
                          return false;
                        gfc_current_locus = old_loc;
                        next_char ();
                        c = next_char ();
                      }
                    if (continue_flag || c == ' ' || c == '\t')
                      {
                        gfc_current_locus = old_loc;
                        next_char ();
                        openmp_flag = 0;
                        return true;
                      }
                  }
                gfc_current_locus = old_loc;
              }
            else if (flag_openacc
                     && !(flag_openmp || flag_openmp_simd))
              {
                locus old_loc = gfc_current_locus;
                if (next_char () == '$')
                  {
                    c = next_char ();
                      if (c == 'a' || c == 'A')
                        {
                          if (skip_free_oacc_sentinel (start, old_loc))
                            return false;
                          gfc_current_locus = old_loc;
                          next_char();
                          c = next_char();
                        }
                      if (continue_flag || c == ' ' || c == '\t')
                        {
                          gfc_current_locus = old_loc;
                          next_char();
                          openacc_flag = 0;
                          return true;
                        }
                  }
                gfc_current_locus = old_loc;
              }
          }
          skip_comment_line ();
          continue;
        }

      break;
    }

  if (openmp_flag && at_bol)
    openmp_flag = 0;

  if (openacc_flag && at_bol)
    openacc_flag = 0;

  gcc_attribute_flag = 0;
  gfc_current_locus = start;
  return false;
}

/* Return true if MP was matched in fixed form.  */
static bool
skip_fixed_omp_sentinel (locus *start)
{
  gfc_char_t c;
  if (((c = next_char ()) == 'm' || c == 'M')
      && ((c = next_char ()) == 'p' || c == 'P'))
    {
      c = next_char ();
      if (c != '\n'
          && (continue_flag
              || c == ' ' || c == '\t' || c == '0'))
        {
          do
            c = next_char ();
          while (gfc_is_whitespace (c));
          if (c != '\n' && c != '!')
            {
              /* Canonicalize to *$omp.  */
              *start->nextc = '*';
              openmp_flag = 1;
              gfc_current_locus = *start;
              return true;
            }
        }
    }
  return false;
}

/* Return true if CC was matched in fixed form.  */
static bool
skip_fixed_oacc_sentinel (locus *start)
{
  gfc_char_t c;
  if (((c = next_char ()) == 'c' || c == 'C')
      && ((c = next_char ()) == 'c' || c == 'C'))
    {
      c = next_char ();
      if (c != '\n'
          && (continue_flag
              || c == ' ' || c == '\t' || c == '0'))
        {
          do
            c = next_char ();
          while (gfc_is_whitespace (c));
          if (c != '\n' && c != '!')
            {
              /* Canonicalize to *$acc.  */
              *start->nextc = '*';
              openacc_flag = 1;
              gfc_current_locus = *start;
              return true;
            }
        }
    }
  return false;
}

/* Skip comment lines in fixed source mode.  We have the same rules as
   in skip_free_comment(), except that we can have a 'c', 'C' or '*'
   in column 1, and a '!' cannot be in column 6.  Also, we deal with
   lines with 'd' or 'D' in column 1, if the user requested this.  */

static void
skip_fixed_comments (void)
{
  locus start;
  int col;
  gfc_char_t c;

  if (! gfc_at_bol ())
    {
      start = gfc_current_locus;
      if (! gfc_at_eof ())
        {
          do
            c = next_char ();
          while (gfc_is_whitespace (c));

          if (c == '\n')
            gfc_advance_line ();
          else if (c == '!')
            skip_comment_line ();
        }

      if (! gfc_at_bol ())
        {
          gfc_current_locus = start;
          return;
        }
    }

  for (;;)
    {
      start = gfc_current_locus;
      if (gfc_at_eof ())
        break;

      c = next_char ();
      if (c == '\n')
        {
          gfc_advance_line ();
          continue;
        }

      if (c == '!' || c == 'c' || c == 'C' || c == '*')
        {
          if (skip_gcc_attribute (start))
            {
              /* Canonicalize to *$omp.  */
              *start.nextc = '*';
              return;
            }

          /* If -fopenmp/-fopenacc, we need to handle here 2 things:
             1) don't treat !$omp/!$acc|c$omp/c$acc|*$omp / *$acc as comments, 
                but directives
             2) handle OpenMP/OpenACC conditional compilation, where
                !$|c$|*$ should be treated as 2 spaces if the characters
                in columns 3 to 6 are valid fixed form label columns
                characters.  */
          if (gfc_current_locus.lb != NULL
              && continue_line < gfc_linebuf_linenum (gfc_current_locus.lb))
            continue_line = gfc_linebuf_linenum (gfc_current_locus.lb);

          if ((flag_openmp || flag_openmp_simd) && !flag_openacc)
            {
              if (next_char () == '$')
                {
                  c = next_char ();
                  if (c == 'o' || c == 'O')
                    {
                      if (skip_fixed_omp_sentinel (&start))
                        return;
                    }
                  else
                    goto check_for_digits;
                }
              gfc_current_locus = start;
            }

          if (flag_openacc && !(flag_openmp || flag_openmp_simd))
            {
              if (next_char () == '$')
                {
                  c = next_char ();
                  if (c == 'a' || c == 'A')
                    {
                      if (skip_fixed_oacc_sentinel (&start))
                        return;
                    }
                  else
                    goto check_for_digits;
                }
              gfc_current_locus = start;
            }

          if (flag_openacc || flag_openmp || flag_openmp_simd)
            {
              if (next_char () == '$')
                {
                  c = next_char ();
                  if (c == 'a' || c == 'A')
                    {
                      if (skip_fixed_oacc_sentinel (&start))
                        return;
                    }
                  else if (c == 'o' || c == 'O')
                    {
                      if (skip_fixed_omp_sentinel (&start))
                        return;
                    }
                  else
                    goto check_for_digits;
                }
              gfc_current_locus = start;
            }

          skip_comment_line ();
          continue;

          gcc_unreachable ();
check_for_digits:
          {
            int digit_seen = 0;

            for (col = 3; col < 6; col++, c = next_char ())
              if (c == ' ')
                continue;
              else if (c == '\t')
                {
                  col = 6;
                  break;
                }
              else if (c < '0' || c > '9')
                break;
              else
                digit_seen = 1;

            if (col == 6 && c != '\n'
                && ((continue_flag && !digit_seen)
                    || c == ' ' || c == '\t' || c == '0'))
              {
                gfc_current_locus = start;
                start.nextc[0] = ' ';
                start.nextc[1] = ' ';
                continue;
              }
            }
          skip_comment_line ();
          continue;
        }

      if (gfc_option.flag_d_lines != -1 && (c == 'd' || c == 'D'))
        {
          if (gfc_option.flag_d_lines == 0)
            {
              skip_comment_line ();
              continue;
            }
          else
            *start.nextc = c = ' ';
        }

      col = 1;

      while (gfc_is_whitespace (c))
        {
          c = next_char ();
          col++;
        }

      if (c == '\n')
        {
          gfc_advance_line ();
          continue;
        }

      if (col != 6 && c == '!')
        {
          if (gfc_current_locus.lb != NULL
              && continue_line < gfc_linebuf_linenum (gfc_current_locus.lb))
            continue_line = gfc_linebuf_linenum (gfc_current_locus.lb);
          skip_comment_line ();
          continue;
        }

      break;
    }

  openmp_flag = 0;
  openacc_flag = 0;
  gcc_attribute_flag = 0;
  gfc_current_locus = start;
}


/* Skips the current line if it is a comment.  */

void
gfc_skip_comments (void)
{
  if (gfc_current_form == FORM_FREE)
    skip_free_comments ();
  else
    skip_fixed_comments ();
}


/* Get the next character from the input, taking continuation lines
   and end-of-line comments into account.  This implies that comment
   lines between continued lines must be eaten here.  For higher-level
   subroutines, this flattens continued lines into a single logical
   line.  The in_string flag denotes whether we're inside a character
   context or not.  */

gfc_char_t
gfc_next_char_literal (gfc_instring in_string)
{
  locus old_loc;
  int i, prev_openmp_flag, prev_openacc_flag;
  gfc_char_t c;

  continue_flag = 0;
  prev_openacc_flag = prev_openmp_flag = 0;

restart:
  c = next_char ();
  if (gfc_at_end ())
    {
      continue_count = 0;
      return c;
    }

  if (gfc_current_form == FORM_FREE)
    {
      bool openmp_cond_flag;

      if (!in_string && c == '!')
        {
          if (gcc_attribute_flag
              && memcmp (&gfc_current_locus, &gcc_attribute_locus,
                 sizeof (gfc_current_locus)) == 0)
            goto done;

          if (openmp_flag
              && memcmp (&gfc_current_locus, &openmp_locus,
                 sizeof (gfc_current_locus)) == 0)
            goto done;

          if (openacc_flag
              && memcmp (&gfc_current_locus, &openacc_locus,
                 sizeof (gfc_current_locus)) == 0)
            goto done;

          /* This line can't be continued */
          do
            {
              c = next_char ();
            }
          while (c != '\n');

          /* Avoid truncation warnings for comment ending lines.  */
          gfc_current_locus.lb->truncated = 0;

          goto done;
        }

      /* Check to see if the continuation line was truncated.  */
      if (warn_line_truncation && gfc_current_locus.lb != NULL
          && gfc_current_locus.lb->truncated)
        {
          int maxlen = flag_free_line_length;
          gfc_char_t *current_nextc = gfc_current_locus.nextc;

          gfc_current_locus.lb->truncated = 0;
          gfc_current_locus.nextc =  gfc_current_locus.lb->line + maxlen;
          gfc_warning_now (OPT_Wline_truncation,
                           "Line truncated at %L", &gfc_current_locus);
          gfc_current_locus.nextc = current_nextc;
        }

      if (c != '&')
        goto done;

      /* If the next nonblank character is a ! or \n, we've got a
         continuation line.  */
      old_loc = gfc_current_locus;

      c = next_char ();
      while (gfc_is_whitespace (c))
        c = next_char ();

      /* Character constants to be continued cannot have commentary
         after the '&'. However, there are cases where we may think we
         are still in a string and we are looking for a possible
         doubled quote and we end up here. See PR64506.  */

      if (in_string && c != '\n')
        {
          gfc_current_locus = old_loc;
          c = '&';
          goto done;
        }

      if (c != '!' && c != '\n')
        {
          gfc_current_locus = old_loc;
          c = '&';
          goto done;
        }

      if (flag_openmp)
        prev_openmp_flag = openmp_flag;
      if (flag_openacc)
        prev_openacc_flag = openacc_flag;

      continue_flag = 1;
      if (c == '!')
        skip_comment_line ();
      else
        gfc_advance_line ();
      
      if (gfc_at_eof ())
        goto not_continuation;

      /* We've got a continuation line.  If we are on the very next line after
         the last continuation, increment the continuation line count and
         check whether the limit has been exceeded.  */
      if (gfc_linebuf_linenum (gfc_current_locus.lb) == continue_line + 1)
        {
          if (++continue_count == gfc_option.max_continue_free)
            {
              if (gfc_notification_std (GFC_STD_GNU) || pedantic)
                gfc_warning (0, "Limit of %d continuations exceeded in "
                             "statement at %C", gfc_option.max_continue_free);
            }
        }

      /* Now find where it continues. First eat any comment lines.  */
      openmp_cond_flag = skip_free_comments ();

      if (gfc_current_locus.lb != NULL
          && continue_line < gfc_linebuf_linenum (gfc_current_locus.lb))
        continue_line = gfc_linebuf_linenum (gfc_current_locus.lb);

      if (flag_openmp)
        if (prev_openmp_flag != openmp_flag && !openacc_flag)
          {
            gfc_current_locus = old_loc;
            openmp_flag = prev_openmp_flag;
            c = '&';
            goto done;
          }

      if (flag_openacc)
        if (prev_openacc_flag != openacc_flag && !openmp_flag)
          {
            gfc_current_locus = old_loc;
            openacc_flag = prev_openacc_flag;
            c = '&';
            goto done;
          }

      /* Now that we have a non-comment line, probe ahead for the
         first non-whitespace character.  If it is another '&', then
         reading starts at the next character, otherwise we must back
         up to where the whitespace started and resume from there.  */

      old_loc = gfc_current_locus;

      c = next_char ();
      while (gfc_is_whitespace (c))
        c = next_char ();

      if (openmp_flag && !openacc_flag)
        {
          for (i = 0; i < 5; i++, c = next_char ())
            {
              gcc_assert (gfc_wide_tolower (c) == (unsigned char) "!$omp"[i]);
              if (i == 4)
                old_loc = gfc_current_locus;
            }
          while (gfc_is_whitespace (c))
            c = next_char ();
        }
      if (openacc_flag && !openmp_flag)
        {
          for (i = 0; i < 5; i++, c = next_char ())
            {
              gcc_assert (gfc_wide_tolower (c) == (unsigned char) "!$acc"[i]);
              if (i == 4)
                old_loc = gfc_current_locus;
            }
          while (gfc_is_whitespace (c))
            c = next_char ();
        }

      /* In case we have an OpenMP directive continued by OpenACC
         sentinel, or vice versa, we get both openmp_flag and
         openacc_flag on.  */

      if (openacc_flag && openmp_flag)
        {
          int is_openmp = 0;
          for (i = 0; i < 5; i++, c = next_char ())
            {
              if (gfc_wide_tolower (c) != (unsigned char) "!$acc"[i])
                is_openmp = 1;
              if (i == 4)
                old_loc = gfc_current_locus;
            }
          gfc_error (is_openmp ? "Wrong OpenACC continuation at %C: "
                     "expected !$ACC, got !$OMP"
                     : "Wrong OpenMP continuation at %C: "
                     "expected !$OMP, got !$ACC");
        }

      if (c != '&')
        {
          if (in_string)
            {
              gfc_current_locus.nextc--;
              if (warn_ampersand && in_string == INSTRING_WARN)
                gfc_warning (OPT_Wampersand, 
                             "Missing %<&%> in continued character "
                             "constant at %C");
            }
          else if (!in_string && (c == '\'' || c == '"'))
              goto done;
          /* Both !$omp and !$ -fopenmp continuation lines have & on the
             continuation line only optionally.  */
          else if (openmp_flag || openacc_flag || openmp_cond_flag)
            gfc_current_locus.nextc--;
          else
            {
              c = ' ';
              gfc_current_locus = old_loc;
              goto done;
            }
        }
    }
  else /* Fixed form.  */
    {
      /* Fixed form continuation.  */
      if (in_string != INSTRING_WARN && c == '!')
        {
          /* Skip comment at end of line.  */
          do
            {
              c = next_char ();
            }
          while (c != '\n');

          /* Avoid truncation warnings for comment ending lines.  */
          gfc_current_locus.lb->truncated = 0;
        }

      if (c != '\n')
        goto done;

      /* Check to see if the continuation line was truncated.  */
      if (warn_line_truncation && gfc_current_locus.lb != NULL
          && gfc_current_locus.lb->truncated)
        {
          gfc_current_locus.lb->truncated = 0;
          gfc_warning_now (OPT_Wline_truncation,
                           "Line truncated at %L", &gfc_current_locus);
        }

      if (flag_openmp)
        prev_openmp_flag = openmp_flag;
      if (flag_openacc)
        prev_openacc_flag = openacc_flag;

      continue_flag = 1;
      old_loc = gfc_current_locus;

      gfc_advance_line ();
      skip_fixed_comments ();

      /* See if this line is a continuation line.  */
      if (flag_openmp && openmp_flag != prev_openmp_flag && !openacc_flag)
        {
          openmp_flag = prev_openmp_flag;
          goto not_continuation;
        }
      if (flag_openacc && openacc_flag != prev_openacc_flag && !openmp_flag)
        {
          openacc_flag = prev_openacc_flag;
          goto not_continuation;
        }

      /* In case we have an OpenMP directive continued by OpenACC
         sentinel, or vice versa, we get both openmp_flag and
         openacc_flag on.  */
      if (openacc_flag && openmp_flag)
        {
          int is_openmp = 0;
          for (i = 0; i < 5; i++)
            {
              c = next_char ();
              if (gfc_wide_tolower (c) != (unsigned char) "*$acc"[i])
                is_openmp = 1;
            }
          gfc_error (is_openmp ? "Wrong OpenACC continuation at %C: "
                     "expected !$ACC, got !$OMP"
                     : "Wrong OpenMP continuation at %C: "
                     "expected !$OMP, got !$ACC");
        }
      else if (!openmp_flag && !openacc_flag)
        for (i = 0; i < 5; i++)
          {
            c = next_char ();
            if (c != ' ')
              goto not_continuation;
          }
      else if (openmp_flag)
        for (i = 0; i < 5; i++)
          {
            c = next_char ();
            if (gfc_wide_tolower (c) != (unsigned char) "*$omp"[i])
              goto not_continuation;
          }
      else if (openacc_flag)
        for (i = 0; i < 5; i++)
          {
            c = next_char ();
            if (gfc_wide_tolower (c) != (unsigned char) "*$acc"[i])
              goto not_continuation;
          }

      c = next_char ();
      if (c == '0' || c == ' ' || c == '\n')
        goto not_continuation;

      /* We've got a continuation line.  If we are on the very next line after
         the last continuation, increment the continuation line count and
         check whether the limit has been exceeded.  */
      if (gfc_linebuf_linenum (gfc_current_locus.lb) == continue_line + 1)
        {
          if (++continue_count == gfc_option.max_continue_fixed)
            {
              if (gfc_notification_std (GFC_STD_GNU) || pedantic)
                gfc_warning (0, "Limit of %d continuations exceeded in "
                             "statement at %C",
                             gfc_option.max_continue_fixed);
            }
        }

      if (gfc_current_locus.lb != NULL
          && continue_line < gfc_linebuf_linenum (gfc_current_locus.lb))
        continue_line = gfc_linebuf_linenum (gfc_current_locus.lb);
    }

  /* Ready to read first character of continuation line, which might
     be another continuation line!  */
  goto restart;

not_continuation:
  c = '\n';
  gfc_current_locus = old_loc;

done:
  if (c == '\n')
    continue_count = 0;
  continue_flag = 0;
  return c;
}


/* Get the next character of input, folded to lowercase.  In fixed
   form mode, we also ignore spaces.  When matcher subroutines are
   parsing character literals, they have to call
   gfc_next_char_literal().  */

gfc_char_t
gfc_next_char (void)
{
  gfc_char_t c;

  do
    {
      c = gfc_next_char_literal (NONSTRING);
    }
  while (gfc_current_form == FORM_FIXED && gfc_is_whitespace (c));

  return gfc_wide_tolower (c);
}

char
gfc_next_ascii_char (void)
{
  gfc_char_t c = gfc_next_char ();

  return (gfc_wide_fits_in_byte (c) ? (unsigned char) c
                                    : (unsigned char) UCHAR_MAX);
}


gfc_char_t
gfc_peek_char (void)
{
  locus old_loc;
  gfc_char_t c;

  old_loc = gfc_current_locus;
  c = gfc_next_char ();
  gfc_current_locus = old_loc;

  return c;
}


char
gfc_peek_ascii_char (void)
{
  gfc_char_t c = gfc_peek_char ();

  return (gfc_wide_fits_in_byte (c) ? (unsigned char) c
                                    : (unsigned char) UCHAR_MAX);
}


/* Recover from an error.  We try to get past the current statement
   and get lined up for the next.  The next statement follows a '\n'
   or a ';'.  We also assume that we are not within a character
   constant, and deal with finding a '\'' or '"'.  */

void
gfc_error_recovery (void)
{
  gfc_char_t c, delim;

  if (gfc_at_eof ())
    return;

  for (;;)
    {
      c = gfc_next_char ();
      if (c == '\n' || c == ';')
        break;

      if (c != '\'' && c != '"')
        {
          if (gfc_at_eof ())
            break;
          continue;
        }
      delim = c;

      for (;;)
        {
          c = next_char ();

          if (c == delim)
            break;
          if (c == '\n')
            return;
          if (c == '\\')
            {
              c = next_char ();
              if (c == '\n')
                return;
            }
        }
      if (gfc_at_eof ())
        break;
    }
}


/* Read ahead until the next character to be read is not whitespace.  */

void
gfc_gobble_whitespace (void)
{
  static int linenum = 0;
  locus old_loc;
  gfc_char_t c;

  do
    {
      old_loc = gfc_current_locus;
      c = gfc_next_char_literal (NONSTRING);
      /* Issue a warning for nonconforming tabs.  We keep track of the line
         number because the Fortran matchers will often back up and the same
         line will be scanned multiple times.  */
      if (warn_tabs && c == '\t')
        {
          int cur_linenum = LOCATION_LINE (gfc_current_locus.lb->location);
          if (cur_linenum != linenum)
            {
              linenum = cur_linenum;
              gfc_warning_now (OPT_Wtabs, "Nonconforming tab character at %C");
            }
        }
    }
  while (gfc_is_whitespace (c));

  gfc_current_locus = old_loc;
}


/* Load a single line into pbuf.

   If pbuf points to a NULL pointer, it is allocated.
   We truncate lines that are too long, unless we're dealing with
   preprocessor lines or if the option -ffixed-line-length-none is set,
   in which case we reallocate the buffer to fit the entire line, if
   need be.
   In fixed mode, we expand a tab that occurs within the statement
   label region to expand to spaces that leave the next character in
   the source region.

   If first_char is not NULL, it's a pointer to a single char value holding
   the first character of the line, which has already been read by the
   caller.  This avoids the use of ungetc().

   load_line returns whether the line was truncated.

   NOTE: The error machinery isn't available at this point, so we can't
         easily report line and column numbers consistent with other 
         parts of gfortran.  */

static int
load_line (FILE *input, gfc_char_t **pbuf, int *pbuflen, const int *first_char)
{
  static int linenum = 0, current_line = 1;
  int c, maxlen, i, preprocessor_flag, buflen = *pbuflen;
  int trunc_flag = 0, seen_comment = 0;
  int seen_printable = 0, seen_ampersand = 0, quoted = ' ';
  gfc_char_t *buffer;
  bool found_tab = false;

  /* Determine the maximum allowed line length.  */
  if (gfc_current_form == FORM_FREE)
    maxlen = flag_free_line_length;
  else if (gfc_current_form == FORM_FIXED)
    maxlen = flag_fixed_line_length;
  else
    maxlen = 72;

  if (*pbuf == NULL)
    {
      /* Allocate the line buffer, storing its length into buflen.
         Note that if maxlen==0, indicating that arbitrary-length lines
         are allowed, the buffer will be reallocated if this length is
         insufficient; since 132 characters is the length of a standard
         free-form line, we use that as a starting guess.  */
      if (maxlen > 0)
        buflen = maxlen;
      else
        buflen = 132;

      *pbuf = gfc_get_wide_string (buflen + 1);
    }

  i = 0;
  buffer = *pbuf;

  if (first_char)
    c = *first_char;
  else
    c = getc (input);

  /* In order to not truncate preprocessor lines, we have to
     remember that this is one.  */
  preprocessor_flag = (c == '#' ? 1 : 0);

  for (;;)
    {
      if (c == EOF)
        break;

      if (c == '\n')
        {
          /* Check for illegal use of ampersand. See F95 Standard 3.3.1.3.  */
          if (gfc_current_form == FORM_FREE 
              && !seen_printable && seen_ampersand)
            {
              if (pedantic)
                gfc_error_now ("%<&%> not allowed by itself in line %d",
                               current_line);
              else
                gfc_warning_now (0, "%<&%> not allowed by itself in line %d",
                                 current_line);
            }
          break;
        }

      if (c == '\r' || c == '\0')
        goto next_char;                 /* Gobble characters.  */

      if (c == '&')
        {
          if (seen_ampersand)
            {
              seen_ampersand = 0;
              seen_printable = 1;
            }
          else
            seen_ampersand = 1;
        }

      if ((c != '&' && c != '!' && c != ' ') || (c == '!' && !seen_ampersand))
        seen_printable = 1;

      /* Is this a fixed-form comment?  */
      if (gfc_current_form == FORM_FIXED && i == 0
          && (c == '*' || c == 'c' || c == 'd'))
        seen_comment = 1;

      if (quoted == ' ')
        {
          if (c == '\'' || c == '"')
            quoted = c;
        }
      else if (c == quoted)
        quoted = ' ';

      /* Is this a free-form comment?  */
      if (c == '!' && quoted == ' ')
        seen_comment = 1;

      /* Vendor extension: "<tab>1" marks a continuation line.  */
      if (found_tab)
        {
          found_tab = false;
          if (c >= '1' && c <= '9')
            {
              *(buffer-1) = c;
              goto next_char;
            }
        }

      if (gfc_current_form == FORM_FIXED && c == '\t' && i < 6)
        {
          found_tab = true;

          if (warn_tabs && seen_comment == 0 && current_line != linenum)
            {
              linenum = current_line;
              gfc_warning_now (OPT_Wtabs,
                               "Nonconforming tab character in column %d "
                               "of line %d", i+1, linenum);
            }

          while (i < 6)
            {
              *buffer++ = ' ';
              i++;
            }

          goto next_char;
        }

      *buffer++ = c;
      i++;

      if (maxlen == 0 || preprocessor_flag)
        {
          if (i >= buflen)
            {
              /* Reallocate line buffer to double size to hold the
                overlong line.  */
              buflen = buflen * 2;
              *pbuf = XRESIZEVEC (gfc_char_t, *pbuf, (buflen + 1));
              buffer = (*pbuf) + i;
            }
        }
      else if (i >= maxlen)
        {
          bool trunc_warn = true;

          /* Enhancement, if the very next non-space character is an ampersand
             or comment that we would otherwise warn about, don't mark as
             truncated.  */

          /* Truncate the rest of the line.  */
          for (;;)
            {
              c = getc (input);
              if (c == '\r' || c == ' ')
                continue;

              if (c == '\n' || c == EOF)
                break;

              if (!trunc_warn && c != '!')
                trunc_warn = true;

              if (trunc_warn && ((gfc_current_form == FORM_FIXED && c == '&')
                  || c == '!'))
                trunc_warn = false;

              if (c == '!')
                seen_comment = 1;

              if (trunc_warn && !seen_comment)
                trunc_flag = 1;
            }

          c = '\n';
          continue;
        }

next_char:
      c = getc (input);
    }

  /* Pad lines to the selected line length in fixed form.  */
  if (gfc_current_form == FORM_FIXED
      && flag_fixed_line_length != 0
      && !preprocessor_flag
      && c != EOF)
    {
      while (i++ < maxlen)
        *buffer++ = ' ';
    }

  *buffer = '\0';
  *pbuflen = buflen;
  current_line++;

  return trunc_flag;
}


/* Get a gfc_file structure, initialize it and add it to
   the file stack.  */

static gfc_file *
get_file (const char *name, enum lc_reason reason ATTRIBUTE_UNUSED)
{
  gfc_file *f;

  f = XCNEW (gfc_file);

  f->filename = xstrdup (name);

  f->next = file_head;
  file_head = f;

  f->up = current_file;
  if (current_file != NULL)
    f->inclusion_line = current_file->line;

  linemap_add (line_table, reason, false, f->filename, 1);

  return f;
}


/* Deal with a line from the C preprocessor. The
   initial octothorp has already been seen.  */

static void
preprocessor_line (gfc_char_t *c)
{
  bool flag[5];
  int i, line;
  gfc_char_t *wide_filename;
  gfc_file *f;
  int escaped, unescape;
  char *filename;

  c++;
  while (*c == ' ' || *c == '\t')
    c++;

  if (*c < '0' || *c > '9')
    goto bad_cpp_line;

  line = wide_atoi (c);

  c = wide_strchr (c, ' ');
  if (c == NULL)
    {
      /* No file name given.  Set new line number.  */
      current_file->line = line;
      return;
    }

  /* Skip spaces.  */
  while (*c == ' ' || *c == '\t')
    c++;

  /* Skip quote.  */
  if (*c != '"')
    goto bad_cpp_line;
  ++c;

  wide_filename = c;

  /* Make filename end at quote.  */
  unescape = 0;
  escaped = false;
  while (*c && ! (!escaped && *c == '"'))
    {
      if (escaped)
        escaped = false;
      else if (*c == '\\')
        {
          escaped = true;
          unescape++;
        }
      ++c;
    }

  if (! *c)
    /* Preprocessor line has no closing quote.  */
    goto bad_cpp_line;

  *c++ = '\0';

  /* Undo effects of cpp_quote_string.  */
  if (unescape)
    {
      gfc_char_t *s = wide_filename;
      gfc_char_t *d = gfc_get_wide_string (c - wide_filename - unescape);

      wide_filename = d;
      while (*s)
        {
          if (*s == '\\')
            *d++ = *++s;
          else
            *d++ = *s;
          s++;
        }
      *d = '\0';
    }

  /* Get flags.  */

  flag[1] = flag[2] = flag[3] = flag[4] = false;

  for (;;)
    {
      c = wide_strchr (c, ' ');
      if (c == NULL)
        break;

      c++;
      i = wide_atoi (c);

      if (1 <= i && i <= 4)
        flag[i] = true;
    }

  /* Convert the filename in wide characters into a filename in narrow
     characters.  */
  filename = gfc_widechar_to_char (wide_filename, -1);

  /* Interpret flags.  */

  if (flag[1]) /* Starting new file.  */
    {
      f = get_file (filename, LC_RENAME);
      add_file_change (f->filename, f->inclusion_line);
      current_file = f;
    }

  if (flag[2]) /* Ending current file.  */
    {
      if (!current_file->up
          || filename_cmp (current_file->up->filename, filename) != 0)
        {
          linemap_line_start (line_table, current_file->line, 80);
          /* ??? One could compute the exact column where the filename
             starts and compute the exact location here.  */
          gfc_warning_now_at (linemap_position_for_column (line_table, 1),
                              0, "file %qs left but not entered",
                              filename);
          current_file->line++;
          if (unescape)
            free (wide_filename);
          free (filename);
          return;
        }

      add_file_change (NULL, line);
      current_file = current_file->up;
      linemap_add (line_table, LC_RENAME, false, current_file->filename,
                   current_file->line);
    }

  /* The name of the file can be a temporary file produced by
     cpp. Replace the name if it is different.  */

  if (filename_cmp (current_file->filename, filename) != 0)
    {
       /* FIXME: we leak the old filename because a pointer to it may be stored
          in the linemap.  Alternative could be using GC or updating linemap to
          point to the new name, but there is no API for that currently.  */
      current_file->filename = xstrdup (filename);
    }

  /* Set new line number.  */
  current_file->line = line;
  if (unescape)
    free (wide_filename);
  free (filename);
  return;

 bad_cpp_line:
  linemap_line_start (line_table, current_file->line, 80);
  /* ??? One could compute the exact column where the directive
     starts and compute the exact location here.  */
  gfc_warning_now_at (linemap_position_for_column (line_table, 2), 0,
                      "Illegal preprocessor directive");
  current_file->line++;
}


static bool load_file (const char *, const char *, bool);

/* include_line()-- Checks a line buffer to see if it is an include
   line.  If so, we call load_file() recursively to load the included
   file.  We never return a syntax error because a statement like
   "include = 5" is perfectly legal.  We return false if no include was
   processed or true if we matched an include.  */

static bool
include_line (gfc_char_t *line)
{
  gfc_char_t quote, *c, *begin, *stop;
  char *filename;

  c = line;

  if (flag_openmp || flag_openmp_simd)
    {
      if (gfc_current_form == FORM_FREE)
        {
          while (*c == ' ' || *c == '\t')
            c++;
          if (*c == '!' && c[1] == '$' && (c[2] == ' ' || c[2] == '\t'))
            c += 3;
        }
      else
        {
          if ((*c == '!' || *c == 'c' || *c == 'C' || *c == '*')
              && c[1] == '$' && (c[2] == ' ' || c[2] == '\t'))
            c += 3;
        }
    }

  while (*c == ' ' || *c == '\t')
    c++;

  if (gfc_wide_strncasecmp (c, "include", 7))
    return false;

  c += 7;
  while (*c == ' ' || *c == '\t')
    c++;

  /* Find filename between quotes.  */
  
  quote = *c++;
  if (quote != '"' && quote != '\'')
    return false;

  begin = c;

  while (*c != quote && *c != '\0')
    c++;

  if (*c == '\0')
    return false;

  stop = c++;
  
  while (*c == ' ' || *c == '\t')
    c++;

  if (*c != '\0' && *c != '!')
    return false;

  /* We have an include line at this point.  */

  *stop = '\0'; /* It's ok to trash the buffer, as this line won't be
                   read by anything else.  */

  filename = gfc_widechar_to_char (begin, -1);
  if (!load_file (filename, NULL, false))
    exit (FATAL_EXIT_CODE);

  free (filename);
  return true;
}


/* Load a file into memory by calling load_line until the file ends.  */

static bool
load_file (const char *realfilename, const char *displayedname, bool initial)
{
  gfc_char_t *line;
  gfc_linebuf *b;
  gfc_file *f;
  FILE *input;
  int len, line_len;
  bool first_line;
  struct stat st;
  int stat_result;
  const char *filename;
  /* If realfilename and displayedname are different and non-null then
     surely realfilename is the preprocessed form of
     displayedname.  */
  bool preprocessed_p = (realfilename && displayedname
                         && strcmp (realfilename, displayedname));

  filename = displayedname ? displayedname : realfilename;

  for (f = current_file; f; f = f->up)
    if (filename_cmp (filename, f->filename) == 0)
      {
        fprintf (stderr, "%s:%d: Error: File '%s' is being included "
                 "recursively\n", current_file->filename, current_file->line,
                 filename);
        return false;
      }

  if (initial)
    {
      if (gfc_src_file)
        {
          input = gfc_src_file;
          gfc_src_file = NULL;
        }
      else
        input = gfc_open_file (realfilename);

      if (input == NULL)
        {
          gfc_error_now ("Can't open file %qs", filename);
          return false;
        }
    }
  else
    {
      input = gfc_open_included_file (realfilename, false, false);
      if (input == NULL)
        {
          fprintf (stderr, "%s:%d: Error: Can't open included file '%s'\n",
                   current_file->filename, current_file->line, filename);
          return false;
        }
      stat_result = stat (realfilename, &st);
      if (stat_result == 0 && !S_ISREG(st.st_mode))
        {
          fprintf (stderr, "%s:%d: Error: Included path '%s'"
                   " is not a regular file\n",
                   current_file->filename, current_file->line, filename);
          fclose (input);
          return false;
        }
    }

  /* Load the file.

     A "non-initial" file means a file that is being included.  In
     that case we are creating an LC_ENTER map.

     An "initial" file means a main file; one that is not included.
     That file has already got at least one (surely more) line map(s)
     created by gfc_init.  So the subsequent map created in that case
     must have LC_RENAME reason.

     This latter case is not true for a preprocessed file.  In that
     case, although the file is "initial", the line maps created by
     gfc_init was used during the preprocessing of the file.  Now that
     the preprocessing is over and we are being fed the result of that
     preprocessing, we need to create a brand new line map for the
     preprocessed file, so the reason is going to be LC_ENTER.  */

  f = get_file (filename, (initial && !preprocessed_p) ? LC_RENAME : LC_ENTER);
  if (!initial)
    add_file_change (f->filename, f->inclusion_line);
  current_file = f;
  current_file->line = 1;
  line = NULL;
  line_len = 0;
  first_line = true;

  if (initial && gfc_src_preprocessor_lines[0])
    {
      preprocessor_line (gfc_src_preprocessor_lines[0]);
      free (gfc_src_preprocessor_lines[0]);
      gfc_src_preprocessor_lines[0] = NULL;
      if (gfc_src_preprocessor_lines[1])
        {
          preprocessor_line (gfc_src_preprocessor_lines[1]);
          free (gfc_src_preprocessor_lines[1]);
          gfc_src_preprocessor_lines[1] = NULL;
        }
    }

  for (;;)
    {
      int trunc = load_line (input, &line, &line_len, NULL);

      len = gfc_wide_strlen (line);
      if (feof (input) && len == 0)
        break;

      /* If this is the first line of the file, it can contain a byte
         order mark (BOM), which we will ignore:
           FF FE is UTF-16 little endian,
           FE FF is UTF-16 big endian,
           EF BB BF is UTF-8.  */
      if (first_line
          && ((line_len >= 2 && line[0] == (unsigned char) '\xFF'
                             && line[1] == (unsigned char) '\xFE')
              || (line_len >= 2 && line[0] == (unsigned char) '\xFE'
                                && line[1] == (unsigned char) '\xFF')
              || (line_len >= 3 && line[0] == (unsigned char) '\xEF'
                                && line[1] == (unsigned char) '\xBB'
                                && line[2] == (unsigned char) '\xBF')))
        {
          int n = line[1] == (unsigned char) '\xBB' ? 3 : 2;
          gfc_char_t *new_char = gfc_get_wide_string (line_len);

          wide_strcpy (new_char, &line[n]);
          free (line);
          line = new_char;
          len -= n;
        }

      /* There are three things this line can be: a line of Fortran
         source, an include line or a C preprocessor directive.  */

      if (line[0] == '#')
        {
          /* When -g3 is specified, it's possible that we emit #define
             and #undef lines, which we need to pass to the middle-end
             so that it can emit correct debug info.  */
          if (debug_info_level == DINFO_LEVEL_VERBOSE
              && (wide_strncmp (line, "#define ", 8) == 0
                  || wide_strncmp (line, "#undef ", 7) == 0))
            ;
          else
            {
              preprocessor_line (line);
              continue;
            }
        }

      /* Preprocessed files have preprocessor lines added before the byte
         order mark, so first_line is not about the first line of the file
         but the first line that's not a preprocessor line.  */
      first_line = false;

      if (include_line (line))
        {
          current_file->line++;
          continue;
        }

      /* Add line.  */

      b = XCNEWVAR (gfc_linebuf, gfc_linebuf_header_size
                    + (len + 1) * sizeof (gfc_char_t));


      b->location
        = linemap_line_start (line_table, current_file->line++, len);
      /* ??? We add the location for the maximum column possible here,
         because otherwise if the next call creates a new line-map, it
         will not reserve space for any offset.  */
      if (len > 0)
        linemap_position_for_column (line_table, len);

      b->file = current_file;
      b->truncated = trunc;
      wide_strcpy (b->line, line);

      if (line_head == NULL)
        line_head = b;
      else
        line_tail->next = b;

      line_tail = b;

      while (file_changes_cur < file_changes_count)
        file_changes[file_changes_cur++].lb = b;
    }

  /* Release the line buffer allocated in load_line.  */
  free (line);

  fclose (input);

  if (!initial)
    add_file_change (NULL, current_file->inclusion_line + 1);
  current_file = current_file->up;
  linemap_add (line_table, LC_LEAVE, 0, NULL, 0);
  return true;
}


/* Open a new file and start scanning from that file. Returns true
   if everything went OK, false otherwise.  If form == FORM_UNKNOWN
   it tries to determine the source form from the filename, defaulting
   to free form.  */

bool
gfc_new_file (void)
{
  bool result;

  if (gfc_cpp_enabled ())
    {
      result = gfc_cpp_preprocess (gfc_source_file);
      if (!gfc_cpp_preprocess_only ())
        result = load_file (gfc_cpp_temporary_file (), gfc_source_file, true);
    }
  else
    result = load_file (gfc_source_file, NULL, true);

  gfc_current_locus.lb = line_head;
  gfc_current_locus.nextc = (line_head == NULL) ? NULL : line_head->line;

#if 0 /* Debugging aid.  */
  for (; line_head; line_head = line_head->next)
    printf ("%s:%3d %s\n", LOCATION_FILE (line_head->location),
            LOCATION_LINE (line_head->location), line_head->line);

  exit (SUCCESS_EXIT_CODE);
#endif

  return result;
}

static char *
unescape_filename (const char *ptr)
{
  const char *p = ptr, *s;
  char *d, *ret;
  int escaped, unescape = 0;

  /* Make filename end at quote.  */
  escaped = false;
  while (*p && ! (! escaped && *p == '"'))
    {
      if (escaped)
        escaped = false;
      else if (*p == '\\')
        {
          escaped = true;
          unescape++;
        }
      ++p;
    }

  if (!*p || p[1])
    return NULL;

  /* Undo effects of cpp_quote_string.  */
  s = ptr;
  d = XCNEWVEC (char, p + 1 - ptr - unescape);
  ret = d;

  while (s != p)
    {
      if (*s == '\\')
        *d++ = *++s;
      else
        *d++ = *s;
      s++;
    }
  *d = '\0';
  return ret;
}

/* For preprocessed files, if the first tokens are of the form # NUM.
   handle the directives so we know the original file name.  */

const char *
gfc_read_orig_filename (const char *filename, const char **canon_source_file)
{
  int c, len;
  char *dirname, *tmp;

  gfc_src_file = gfc_open_file (filename);
  if (gfc_src_file == NULL)
    return NULL;

  c = getc (gfc_src_file);

  if (c != '#')
    return NULL;

  len = 0;
  load_line (gfc_src_file, &gfc_src_preprocessor_lines[0], &len, &c);

  if (wide_strncmp (gfc_src_preprocessor_lines[0], "# 1 \"", 5) != 0)
    return NULL;

  tmp = gfc_widechar_to_char (&gfc_src_preprocessor_lines[0][5], -1);
  filename = unescape_filename (tmp);
  free (tmp);
  if (filename == NULL)
    return NULL;

  c = getc (gfc_src_file);

  if (c != '#')
    return filename;

  len = 0;
  load_line (gfc_src_file, &gfc_src_preprocessor_lines[1], &len, &c);

  if (wide_strncmp (gfc_src_preprocessor_lines[1], "# 1 \"", 5) != 0)
    return filename;

  tmp = gfc_widechar_to_char (&gfc_src_preprocessor_lines[1][5], -1);
  dirname = unescape_filename (tmp);
  free (tmp);
  if (dirname == NULL)
    return filename;

  len = strlen (dirname);
  if (len < 3 || dirname[len - 1] != '/' || dirname[len - 2] != '/')
    {
      free (dirname);
      return filename;
    }
  dirname[len - 2] = '\0';
  set_src_pwd (dirname);

  if (! IS_ABSOLUTE_PATH (filename))
    {
      char *p = XCNEWVEC (char, len + strlen (filename));

      memcpy (p, dirname, len - 2);
      p[len - 2] = '/';
      strcpy (p + len - 1, filename);
      *canon_source_file = p;
    }

  free (dirname);
  return filename;
}