/*
     This file is part of libextractor.
     Copyright (C) 2002, 2003, 2004, 2005, 2006, 2009, 2012 Vidyut Samanta and Christian Grothoff

     libextractor 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.

     libextractor 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 libextractor; see the file COPYING.  If not, write to the
     Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
     Boston, MA 02110-1301, USA.
*/
/**
 * @file main/extract.c
 * @brief command-line tool to run GNU libextractor
 * @author Christian Grothoff
 */
#include "platform.h"
#include "extractor.h"
#include "getopt.h"
#include <signal.h>

#define YES 1
#define NO 0


/**
 * Which keyword types should we print?
 */
static int *print;

/**
 * How verbose are we supposed to be?
 */
static int verbose;

/**
 * Run plugins in-process.
 */
static int in_process;

/**
 * Read file contents into memory, then feed them to extractor.
 */
static int from_memory;

#ifndef WINDOWS
/**
 * Install a signal handler to ignore SIGPIPE.
 */
static void
ignore_sigpipe ()
{
  struct sigaction oldsig;
  struct sigaction sig;

  memset (&sig, 0, sizeof (struct sigaction));
  sig.sa_handler = SIG_IGN;
  sigemptyset (&sig.sa_mask);
#ifdef SA_INTERRUPT
  sig.sa_flags = SA_INTERRUPT;  /* SunOS */
#else
  sig.sa_flags = SA_RESTART;
#endif
  if (0 != sigaction (SIGPIPE, &sig, &oldsig))
    FPRINTF (stderr,
             "Failed to install SIGPIPE handler: %s\n", strerror (errno));
}


#endif


/**
 * Information about command-line options.
 */
struct Help
{
  /**
   * Single-character option name, '\0' for none.
   */
  char shortArg;

  /**
   * Long name of the option.
   */
  const char *longArg;

  /**
   * Name of the mandatory argument, NULL for no argument.
   */
  const char *mandatoryArg;

  /**
   * Help text for the option.
   */
  const char *description;
};


/**
 * Indentation for descriptions.
 */
#define BORDER 29


/**
 * Display help text (--help).
 *
 * @param general binary name
 * @param description program description
 * @param opt program options (NULL-terminated array)
 */
static void
format_help (const char *general,
             const char *description,
             const struct Help *opt)
{
  size_t slen;
  unsigned int i;
  ssize_t j;
  size_t ml;
  size_t p;
  char scp[80];
  const char *trans;

  printf (_ ("Usage: %s\n%s\n\n"),
          gettext (general),
          gettext (description));
  printf (_ (
            "Arguments mandatory for long options are also mandatory for short options.\n"));
  slen = 0;
  i = 0;
  while (NULL != opt[i].description)
  {
    if (0 == opt[i].shortArg)
      printf ("      ");
    else
      printf ("  -%c, ",
              opt[i].shortArg);
    printf ("--%s",
            opt[i].longArg);
    slen = 8 + strlen (opt[i].longArg);
    if (NULL != opt[i].mandatoryArg)
    {
      printf ("=%s",
              opt[i].mandatoryArg);
      slen += 1 + strlen (opt[i].mandatoryArg);
    }
    if (slen > BORDER)
    {
      printf ("\n%*s", BORDER, "");
      slen = BORDER;
    }
    if (slen < BORDER)
    {
      printf ("%*s", (int) (BORDER - slen), "");
      slen = BORDER;
    }
    trans = gettext (opt[i].description);
    ml = strlen (trans);
    p = 0;
OUTER:
    while (ml - p > 78 - slen)
    {
      for (j = p + 78 - slen; j>p; j--)
      {
        if (isspace ( (unsigned char) trans[j]))
        {
          memcpy (scp,
                  &trans[p],
                  j - p);
          scp[j - p] = '\0';
          printf ("%s\n%*s",
                  scp,
                  BORDER + 2,
                  "");
          p = j + 1;
          slen = BORDER + 2;
          goto OUTER;
        }
      }
      /* could not find space to break line */
      memcpy (scp,
              &trans[p],
              78 - slen);
      scp[78 - slen] = '\0';
      printf ("%s\n%*s",
              scp,
              BORDER + 2,
              "");
      slen = BORDER + 2;
      p = p + 78 - slen;
    }
    /* print rest */
    if (p < ml)
      printf ("%s\n",
              &trans[p]);
    i++;
  }
}


/**
 * Run --help.
 */
static void
print_help ()
{
  static struct Help help[] = {
    { 'b', "bibtex", NULL,
      gettext_noop ("print output in bibtex format") },
    { 'g', "grep-friendly", NULL,
      gettext_noop (
        "produce grep-friendly output (all results on one line per file)") },
    { 'h', "help", NULL,
      gettext_noop ("print this help") },
    { 'i', "in-process", NULL,
      gettext_noop ("run plugins in-process (simplifies debugging)") },
    { 'm', "from-memory", NULL,
      gettext_noop (
        "read data from file into memory and extract from memory") },
    { 'l', "library", "LIBRARY",
      gettext_noop ("load an extractor plugin named LIBRARY") },
    { 'L', "list", NULL,
      gettext_noop ("list all keyword types") },
    { 'n', "nodefault", NULL,
      gettext_noop ("do not use the default set of extractor plugins") },
    { 'p', "print", "TYPE",
      gettext_noop (
        "print only keywords of the given TYPE (use -L to get a list)") },
    { 'v', "version", NULL,
      gettext_noop ("print the version number") },
    { 'V', "verbose", NULL,
      gettext_noop ("be verbose") },
    { 'x', "exclude", "TYPE",
      gettext_noop ("do not print keywords of the given TYPE") },
    { 0, NULL, NULL, NULL },
  };
  format_help (_ ("extract [OPTIONS] [FILENAME]*"),
               _ ("Extract metadata from files."),
               help);

}


#if HAVE_ICONV
#include "iconv.c"
#endif

/**
 * Print a keyword list to a file.
 *
 * @param cls closure, not used
 * @param plugin_name name of the plugin that produced this value;
 *        special values can be used (i.e. '<zlib>' for zlib being
 *        used in the main libextractor library and yielding
 *        meta data).
 * @param type libextractor-type describing the meta data
 * @param format basic format information about data
 * @param data_mime_type mime-type of data (not of the original file);
 *        can be NULL (if mime-type is not known)
 * @param data actual meta-data found
 * @param data_len number of bytes in data
 * @return 0 to continue extracting, 1 to abort
 */
static int
print_selected_keywords (void *cls,
                         const char *plugin_name,
                         enum EXTRACTOR_MetaType type,
                         enum EXTRACTOR_MetaFormat format,
                         const char *data_mime_type,
                         const char *data,
                         size_t data_len)
{
  char *keyword;
#if HAVE_ICONV
  iconv_t cd;
#endif
  const char *stype;
  const char *mt;

  if (YES != print[type])
    return 0;
  if (verbose > 3)
    FPRINTF (stdout,
             _ ("Found by `%s' plugin:\n"),
             plugin_name);
  mt = EXTRACTOR_metatype_to_string (type);
  stype = (NULL == mt) ? _ ("unknown") : gettext (mt);
  switch (format)
  {
  case EXTRACTOR_METAFORMAT_UNKNOWN:
    FPRINTF (stdout,
             _ ("%s - (unknown, %u bytes)\n"),
             stype,
             (unsigned int) data_len);
    break;
  case EXTRACTOR_METAFORMAT_UTF8:
    if (0 == data_len)
      break;
#if HAVE_ICONV
    cd = iconv_open (nl_langinfo (CODESET), "UTF-8");
    if (((iconv_t) -1) != cd)
      keyword = iconv_helper (cd,
                              data,
                              data_len);
    else
#endif
    keyword = strdup (data);
    if (NULL != keyword)
    {
      FPRINTF (stdout,
               "%s - %s\n",
               stype,
               keyword);
      free (keyword);
    }
#if HAVE_ICONV
    if (((iconv_t) -1) != cd)
      iconv_close (cd);
#endif
    break;
  case EXTRACTOR_METAFORMAT_BINARY:
    FPRINTF (stdout,
             _ ("%s - (binary, %u bytes)\n"),
             stype,
             (unsigned int) data_len);
    break;
  case EXTRACTOR_METAFORMAT_C_STRING:
    FPRINTF (stdout,
             "%s - %.*s\n",
             stype,
             (int) data_len,
             data);
    break;
  default:
    break;
  }
  return 0;
}


/**
 * Print a keyword list to a file without new lines.
 *
 * @param cls closure, not used
 * @param plugin_name name of the plugin that produced this value;
 *        special values can be used (i.e. '<zlib>' for zlib being
 *        used in the main libextractor library and yielding
 *        meta data).
 * @param type libextractor-type describing the meta data
 * @param format basic format information about data
 * @param data_mime_type mime-type of data (not of the original file);
 *        can be NULL (if mime-type is not known)
 * @param data actual meta-data found
 * @param data_len number of bytes in data
 * @return 0 to continue extracting, 1 to abort
 */
static int
print_selected_keywords_grep_friendly (void *cls,
                                       const char *plugin_name,
                                       enum EXTRACTOR_MetaType type,
                                       enum EXTRACTOR_MetaFormat format,
                                       const char *data_mime_type,
                                       const char *data,
                                       size_t data_len)
{
  char *keyword;
#if HAVE_ICONV
  iconv_t cd;
#endif
  const char *mt;

  if (YES != print[type])
    return 0;
  mt = EXTRACTOR_metatype_to_string (type);
  if (NULL == mt)
    mt = gettext_noop ("unknown");
  switch (format)
  {
  case EXTRACTOR_METAFORMAT_UNKNOWN:
    break;
  case EXTRACTOR_METAFORMAT_UTF8:
    if (0 == data_len)
      return 0;
    if (verbose > 1)
      FPRINTF (stdout,
               "%s: ",
               gettext (mt));
#if HAVE_ICONV
    cd = iconv_open (nl_langinfo (CODESET), "UTF-8");
    if (((iconv_t) -1) != cd)
      keyword = iconv_helper (cd,
                              data,
                              data_len);
    else
#endif
    keyword = strdup (data);
    if (NULL != keyword)
    {
      FPRINTF (stdout,
               "`%s' ",
               keyword);
      free (keyword);
    }
#if HAVE_ICONV
    if (((iconv_t) -1) != cd)
      iconv_close (cd);
#endif
    break;
  case EXTRACTOR_METAFORMAT_BINARY:
    break;
  case EXTRACTOR_METAFORMAT_C_STRING:
    if (verbose > 1)
      FPRINTF (stdout,
               "%s ",
               gettext (mt));
    FPRINTF (stdout,
             "`%s'",
             data);
    break;
  default:
    break;
  }
  return 0;
}


/**
 * Entry in the map we construct for each file.
 */
struct BibTexMap
{
  /**
   * Name in bibTeX
   */
  const char *bibTexName;

  /**
   * Meta type for the value.
   */
  enum EXTRACTOR_MetaType le_type;

  /**
   * The value itself.
   */
  char *value;
};


/**
 * Type of the entry for bibtex.
 */
static char *entry_type;

/**
 * Mapping between bibTeX strings, libextractor
 * meta data types and values for the current document.
 */
static struct BibTexMap btm[] = {
  { "title", EXTRACTOR_METATYPE_TITLE, NULL},
  { "year", EXTRACTOR_METATYPE_PUBLICATION_YEAR, NULL },
  { "author", EXTRACTOR_METATYPE_AUTHOR_NAME, NULL },
  { "book", EXTRACTOR_METATYPE_BOOK_TITLE, NULL},
  { "edition", EXTRACTOR_METATYPE_BOOK_EDITION, NULL},
  { "chapter", EXTRACTOR_METATYPE_BOOK_CHAPTER_NUMBER, NULL},
  { "journal", EXTRACTOR_METATYPE_JOURNAL_NAME, NULL},
  { "volume", EXTRACTOR_METATYPE_JOURNAL_VOLUME, NULL},
  { "number", EXTRACTOR_METATYPE_JOURNAL_NUMBER, NULL},
  { "pages", EXTRACTOR_METATYPE_PAGE_COUNT, NULL },
  { "pages", EXTRACTOR_METATYPE_PAGE_RANGE, NULL },
  { "school", EXTRACTOR_METATYPE_AUTHOR_INSTITUTION, NULL},
  { "publisher", EXTRACTOR_METATYPE_PUBLISHER, NULL },
  { "address", EXTRACTOR_METATYPE_PUBLISHER_ADDRESS, NULL },
  { "institution", EXTRACTOR_METATYPE_PUBLISHER_INSTITUTION, NULL },
  { "series", EXTRACTOR_METATYPE_PUBLISHER_SERIES, NULL},
  { "month", EXTRACTOR_METATYPE_PUBLICATION_MONTH, NULL },
  { "url", EXTRACTOR_METATYPE_URL, NULL},
  { "note", EXTRACTOR_METATYPE_COMMENT, NULL},
  { "eprint", EXTRACTOR_METATYPE_BIBTEX_EPRINT, NULL },
  { "type", EXTRACTOR_METATYPE_PUBLICATION_TYPE, NULL },
  { NULL, 0, NULL }
};


/**
 * Clean up the bibtex processor in preparation for the next round.
 */
static void
cleanup_bibtex ()
{
  unsigned int i;

  for (i = 0; NULL != btm[i].bibTexName; i++)
  {
    free (btm[i].value);
    btm[i].value = NULL;
  }
  free (entry_type);
  entry_type = NULL;
}


/**
 * Callback function for printing meta data in bibtex format.
 *
 * @param cls closure, not used
 * @param plugin_name name of the plugin that produced this value;
 *        special values can be used (i.e. '<zlib>' for zlib being
 *        used in the main libextractor library and yielding
 *        meta data).
 * @param type libextractor-type describing the meta data
 * @param format basic format information about data
 * @param data_mime_type mime-type of data (not of the original file);
 *        can be NULL (if mime-type is not known)
 * @param data actual meta-data found
 * @param data_len number of bytes in data
 * @return 0 to continue extracting (always)
 */
static int
print_bibtex (void *cls,
              const char *plugin_name,
              enum EXTRACTOR_MetaType type,
              enum EXTRACTOR_MetaFormat format,
              const char *data_mime_type,
              const char *data,
              size_t data_len)
{
  unsigned int i;

  if (0 == data_len)
    return 0;
  if (YES != print[type])
    return 0;
  if (EXTRACTOR_METAFORMAT_UTF8 != format)
    return 0;
  if (EXTRACTOR_METATYPE_BIBTEX_ENTRY_TYPE == type)
  {
    entry_type = strdup (data);
    return 0;
  }
  for (i = 0; NULL != btm[i].bibTexName; i++)
    if ( (NULL == btm[i].value) &&
         (btm[i].le_type == type) )
      btm[i].value = strdup (data);
  return 0;
}


/**
 * Print the computed bibTeX entry.
 *
 * @param fn file for which the entry was created.
 */
static void
finish_bibtex (const char *fn)
{
  unsigned int i;
  ssize_t n;
  const char *et;
  char temp[20];

  if (NULL != entry_type)
    et = entry_type;
  else
    et = "misc";
  if ( (NULL == btm[0].value) ||
       (NULL == btm[1].value) ||
       (NULL == btm[2].value) )
    FPRINTF (stdout,
             "@%s %s { ",
             et,
             fn);
  else
  {
    snprintf (temp,
              sizeof (temp),
              "%.5s%.5s%.5s",
              btm[2].value,
              btm[1].value,
              btm[0].value);
    for (n = strlen (temp) - 1; n>=0; n--)
      if (! isalnum ( (unsigned char) temp[n]) )
        temp[n] = '_';
      else
        temp[n] = tolower ( (unsigned char) temp[n]);
    FPRINTF (stdout,
             "@%s %s { ",
             et,
             temp);
  }
  for (i = 0; NULL != btm[i].bibTexName; i++)
    if (NULL != btm[i].value)
      FPRINTF (stdout,
               "\t%s = {%s},\n",
               btm[i].bibTexName,
               btm[i].value);
  FPRINTF (stdout, "%s", "}\n\n");
}


#ifdef WINDOWS
static int
_wchar_to_str (const wchar_t *wstr, char **retstr, UINT cp)
{
  char *str;
  int len, lenc;
  BOOL lossy = FALSE;
  DWORD error;

  SetLastError (0);
  len = WideCharToMultiByte (cp, 0, wstr, -1, NULL, 0, NULL, (cp == CP_UTF8 ||
                                                              cp == CP_UTF7) ?
                             NULL : &lossy);
  error = GetLastError ();
  if (len <= 0)
    return -1;

  str = malloc (sizeof (char) * len);

  SetLastError (0);
  lenc = WideCharToMultiByte (cp, 0, wstr, -1, str, len, NULL, (cp == CP_UTF8 ||
                                                                cp == CP_UTF7) ?
                              NULL : &lossy);
  error = GetLastError ();
  if (lenc != len)
  {
    free (str);
    return -3;
  }
  *retstr = str;
  if (lossy)
    return 1;
  return 0;
}


#endif


/**
 * Makes a copy of argv that consists of a single memory chunk that can be
 * freed with a single call to free ();
 */
static char **
_make_continuous_arg_copy (int argc, char *const *argv)
{
  size_t argvsize = 0;
  int i;
  char **new_argv;
  char *p;
  for (i = 0; i < argc; i++)
    argvsize += strlen (argv[i]) + 1 + sizeof (char *);
  new_argv = malloc (argvsize + sizeof (char *));
  if (NULL == new_argv)
    return NULL;
  p = (char *) &new_argv[argc + 1];
  for (i = 0; i < argc; i++)
  {
    new_argv[i] = p;
    strcpy (p, argv[i]);
    p += strlen (argv[i]) + 1;
  }
  new_argv[argc] = NULL;
  return (char **) new_argv;
}


/**
 * Returns utf-8 encoded arguments.
 * Returned argv has u8argv[u8argc] == NULL.
 * Returned argv is a single memory block, and can be freed with a single
 *   free () call.
 *
 * @param argc argc (as given by main())
 * @param argv argv (as given by main())
 * @param u8argc a location to store new argc in (though it's th same as argc)
 * @param u8argv a location to store new argv in
 * @return 0 on success, -1 on failure
 */
static int
_get_utf8_args (int argc, char *const *argv, int *u8argc, char ***u8argv)
{
#ifdef WINDOWS
  wchar_t *wcmd;
  wchar_t **wargv;
  int wargc;
  int i;
  char **split_u8argv;

  wcmd = GetCommandLineW ();
  if (NULL == wcmd)
    return -1;
  wargv = CommandLineToArgvW (wcmd, &wargc);
  if (NULL == wargv)
    return -1;

  split_u8argv = malloc (wargc * sizeof (char *));

  for (i = 0; i < wargc; i++)
  {
    if (_wchar_to_str (wargv[i], &split_u8argv[i], CP_UTF8) != 0)
    {
      int j;
      int e = errno;
      for (j = 0; j < i; j++)
        free (split_u8argv[j]);
      free (split_u8argv);
      LocalFree (wargv);
      errno = e;
      return -1;
    }
  }

  *u8argv = _make_continuous_arg_copy (wargc, split_u8argv);
  if (NULL == *u8argv)
  {
    free (split_u8argv);
    return -1;
  }
  *u8argc = wargc;

  for (i = 0; i < wargc; i++)
    free (split_u8argv[i]);
  free (split_u8argv);
#else
  *u8argv = _make_continuous_arg_copy (argc, argv);
  if (NULL == *u8argv)
    return -1;
  *u8argc = argc;
#endif
  return 0;
}


/**
 * Main function for the 'extract' tool.  Invoke with a list of
 * filenames to extract keywords from.
 *
 * @param argc number of arguments in argv
 * @param argv command line options and filename to run on
 * @return 0 on success
 */
int
main (int argc, char *argv[])
{
  unsigned int i;
  struct EXTRACTOR_PluginList *plugins;
  int option_index;
  int c;
  char *libraries = NULL;
  int nodefault = NO;
  int defaultAll = YES;
  int bibtex = NO;
  int grepfriendly = NO;
  int ret = 0;
  EXTRACTOR_MetaDataProcessor processor = NULL;
  char **utf8_argv;
  int utf8_argc;

#if ENABLE_NLS
  setlocale (LC_ALL, "");
  textdomain (PACKAGE);
#endif
#ifndef WINDOWS
  ignore_sigpipe ();
#endif
  if (NULL == (print = malloc (sizeof (int) * EXTRACTOR_metatype_get_max ())))
  {
    FPRINTF (stderr,
             "malloc failed: %s\n",
             strerror (errno));
    return 1;
  }
  for (i = 0; i < EXTRACTOR_metatype_get_max (); i++)
    print[i] = YES;   /* default: print everything */

  if (0 != _get_utf8_args (argc, argv, &utf8_argc, &utf8_argv))
  {
    FPRINTF (stderr, "Failed to get arguments: %s\n", strerror (errno));
    return 1;
  }

  while (1)
  {
    static struct option long_options[] = {
      {"bibtex", 0, 0, 'b'},
      {"grep-friendly", 0, 0, 'g'},
      {"help", 0, 0, 'h'},
      {"in-process", 0, 0, 'i'},
      {"from-memory", 0, 0, 'm'},
      {"list", 0, 0, 'L'},
      {"library", 1, 0, 'l'},
      {"nodefault", 0, 0, 'n'},
      {"print", 1, 0, 'p'},
      {"verbose", 0, 0, 'V'},
      {"version", 0, 0, 'v'},
      {"exclude", 1, 0, 'x'},
      {0, 0, 0, 0}
    };
    option_index = 0;
    c = getopt_long (utf8_argc,
                     utf8_argv,
                     "abghiml:Lnp:vVx:",
                     long_options,
                     &option_index);

    if (c == -1)
      break;  /* No more flags to process */
    switch (c)
    {
    case 'b':
      bibtex = YES;
      if (NULL != processor)
      {
        FPRINTF (stderr,
                 "%s",
                 _ (
                   "Illegal combination of options, cannot combine multiple styles of printing.\n"));
        free (utf8_argv);
        return 0;
      }
      processor = &print_bibtex;
      break;
    case 'g':
      grepfriendly = YES;
      if (NULL != processor)
      {
        FPRINTF (stderr,
                 "%s",
                 _ (
                   "Illegal combination of options, cannot combine multiple styles of printing.\n"));
        free (utf8_argv);
        return 0;
      }
      processor = &print_selected_keywords_grep_friendly;
      break;
    case 'h':
      print_help ();
      free (utf8_argv);
      return 0;
    case 'i':
      in_process = YES;
      break;
    case 'm':
      from_memory = YES;
      break;
    case 'l':
      libraries = optarg;
      break;
    case 'L':
      i = 0;
      while (NULL != EXTRACTOR_metatype_to_string (i))
        printf ("%s\n",
                gettext (EXTRACTOR_metatype_to_string (i++)));
      free (utf8_argv);
      return 0;
    case 'n':
      nodefault = YES;
      break;
    case 'p':
      if (NULL == optarg)
      {
        FPRINTF (stderr,
                 _ (
                   "You must specify an argument for the `%s' option (option ignored).\n"),
                 "-p");
        break;
      }
      if (YES == defaultAll)
      {
        defaultAll = NO;
        i = 0;
        while (NULL != EXTRACTOR_metatype_to_string (i))
          print[i++] = NO;
      }
      i = 0;
      while (NULL != EXTRACTOR_metatype_to_string (i))
      {
        if ( (0 == strcmp (optarg,
                           EXTRACTOR_metatype_to_string (i))) ||
             (0 == strcmp (optarg,
                           gettext (EXTRACTOR_metatype_to_string (i)))) )

        {
          print[i] = YES;
          break;
        }
        i++;
      }
      if (NULL == EXTRACTOR_metatype_to_string (i))
      {
        FPRINTF (stderr,
                 "Unknown keyword type `%s', use option `%s' to get a list.\n",
                 optarg,
                 "-L");
        free (utf8_argv);
        return -1;
      }
      break;
    case 'v':
      printf ("extract v%s\n", PACKAGE_VERSION);
      free (utf8_argv);
      return 0;
    case 'V':
      verbose++;
      break;
    case 'x':
      i = 0;
      while (NULL != EXTRACTOR_metatype_to_string (i))
      {
        if ( (0 == strcmp (optarg,
                           EXTRACTOR_metatype_to_string (i))) ||
             (0 == strcmp (optarg,
                           gettext (EXTRACTOR_metatype_to_string (i)))) )
        {
          print[i] = NO;
          break;
        }
        i++;
      }
      if (NULL == EXTRACTOR_metatype_to_string (i))
      {
        FPRINTF (stderr,
                 "Unknown keyword type `%s', use option `%s' to get a list.\n",
                 optarg,
                 "-L");
        free (utf8_argv);
        return -1;
      }
      break;
    default:
      FPRINTF (stderr,
               "%s",
               _ ("Use --help to get a list of options.\n"));
      free (utf8_argv);
      return -1;
    }   /* end of parsing commandline */
  }         /* while (1) */
  if (optind < 0)
  {
    FPRINTF (stderr,
             "%s", "Unknown error parsing options\n");
    free (print);
    free (utf8_argv);
    return -1;
  }
  if (utf8_argc - optind < 1)
  {
    FPRINTF (stderr,
             "%s", "Invoke with list of filenames to extract keywords form!\n");
    free (print);
    free (utf8_argv);
    return -1;
  }

  /* build list of libraries */
  if (NO == nodefault)
    plugins = EXTRACTOR_plugin_add_defaults (in_process
                                             ? EXTRACTOR_OPTION_IN_PROCESS
                                             : EXTRACTOR_OPTION_DEFAULT_POLICY);
  else
    plugins = NULL;
  if (NULL != libraries)
    plugins = EXTRACTOR_plugin_add_config (plugins,
                                           libraries,
                                           in_process
                                           ? EXTRACTOR_OPTION_IN_PROCESS
                                           : EXTRACTOR_OPTION_DEFAULT_POLICY);
  if (NULL == processor)
    processor = &print_selected_keywords;

  /* extract keywords */
  if (YES == bibtex)
    FPRINTF (stdout,
             "%s", _ ("% BiBTeX file\n"));
  for (i = optind; i < utf8_argc; i++)
  {
    errno = 0;
    if (YES == grepfriendly)
      FPRINTF (stdout, "%s ", utf8_argv[i]);
    else if (NO == bibtex)
      FPRINTF (stdout,
               _ ("Keywords for file %s:\n"),
               utf8_argv[i]);
    else
      cleanup_bibtex ();
    if (NO == from_memory)
      EXTRACTOR_extract (plugins,
                         utf8_argv[i],
                         NULL, 0,
                         processor,
                         NULL);
    else
    {
      struct stat sb;
      unsigned char *data = NULL;
      int f = OPEN (utf8_argv[i], O_RDONLY
#if WINDOWS
                    | O_BINARY
#endif
                    );
      if ( (-1 != f) &&
           (0 == FSTAT (f, &sb)) &&
           (NULL != (data = malloc ((size_t) sb.st_size))) &&
           (sb.st_size == READ (f, data, (size_t) sb.st_size) ) )
      {
        EXTRACTOR_extract (plugins,
                           NULL,
                           data, sb.st_size,
                           processor,
                           NULL);
      }
      else
      {
        if (verbose > 0)
          FPRINTF (stderr,
                   "%s: %s: %s\n",
                   utf8_argv[0], utf8_argv[i], strerror (errno));
        ret = 1;
      }
      if (NULL != data)
        free (data);
      if (-1 != f)
        (void) CLOSE (f);
    }
    if (YES == grepfriendly)
      FPRINTF (stdout, "%s", "\n");
    continue;
  }
  if (YES == grepfriendly)
    FPRINTF (stdout, "%s", "\n");
  if (bibtex)
    finish_bibtex (utf8_argv[i]);
  if (verbose > 0)
    FPRINTF (stdout, "%s", "\n");
  free (print);
  free (utf8_argv);
  EXTRACTOR_plugin_remove_all (plugins);
  plugins = NULL;
  cleanup_bibtex (); /* actually free's stuff */
  return ret;
}


/* end of extract.c */