sile 0.15.2

/* This is dvipdfmx, an eXtended version of dvipdfm by Mark A. Wicks.

    Copyright (C) 2008-2014 by Jin-Hwan Cho, Matthias Franz, and Shunsaku Hirata,
    the dvipdfmx project team.

    Copyright (C) 1998, 1999 by Mark A. Wicks <mwicks@kettering.edu>

    This program 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 2 of the License, or
    (at your option) any later version.

    This program 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 this program; if not, write to the Free Software
    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
*/

#include "libtexpdf.h"

#define FONT_FLAG_FIXEDPITCH (1 << 0)  /* Fixed-width font */
#define FONT_FLAG_SERIF      (1 << 1)  /* Serif font */
#define FONT_FLAG_SYMBOLIC   (1 << 2)  /* Symbolic font */
#define FONT_FLAG_SCRIPT     (1 << 3)  /* Script font */
#define FONT_FLAG_STANDARD   (1 << 5)  /* Adobe Standard Character Set */
#define FONT_FLAG_ITALIC     (1 << 6)  /* Italic */
#define FONT_FLAG_ALLCAP     (1 << 16) /* All-cap font */
#define FONT_FLAG_SMALLCAP   (1 << 17) /* Small-cap font */
#define FONT_FLAG_FORCEBOLD  (1 << 18) /* Force bold at small text sizes */

static int
is_basefont (const char *name)
{
  static const char *basefonts[] = {
    "Courier",                  "Courier-Bold",          "Courier-Oblique",
    "Courier-BoldOblique",      "Helvetica",             "Helvetica-Bold",
    "Helvetica-Oblique",        "Helvetica-BoldOblique", "Symbol",
    "Times-Roman",              "Times-Bold",            "Times-Italic",
    "Times-BoldItalic",         "ZapfDingbats"
  };
  int i;

  for (i = 0; i < 14; i++) {
    if (!strcmp(name, basefonts[i]))
      return 1;
  }

  return 0;
}

int
pdf_font_open_type1 (pdf_font *font)
{
  char    *ident;
  FILE    *fp;
  char     fontname[PDF_NAME_LEN_MAX+1];

  ASSERT(font);

  ident = pdf_font_get_ident(font);

  if (is_basefont(ident)) {
    pdf_font_set_fontname(font, ident);
    pdf_font_set_subtype (font, PDF_FONT_FONTTYPE_TYPE1);
    pdf_font_set_flags   (font,
			  (PDF_FONT_FLAG_NOEMBED|PDF_FONT_FLAG_BASEFONT));
  } else {
    fp = DPXFOPEN(ident, DPX_RES_TYPE_T1FONT);
    if (!fp)
      return -1;

    memset(fontname, 0, PDF_NAME_LEN_MAX+1);
    if (!is_pfb(fp) || t1_get_fontname(fp, fontname) < 0) {
      ERROR("Failed to read Type 1 font \"%s\".", ident);
    }
    DPXFCLOSE(fp);

    pdf_font_set_fontname(font, fontname);
    pdf_font_set_subtype (font, PDF_FONT_FONTTYPE_TYPE1);
  }

  return 0;
}

static void
get_font_attr (pdf_font *font, cff_font *cffont)
{
  char    *fontname;
  pdf_obj *descriptor;
  double   capheight, ascent, descent;
  double   italicangle, stemv;
  double   defaultwidth, nominalwidth;
  long     flags = 0, gid, i;
  static const char *L_c[] = {
    "H", "P", "Pi", "Rho", NULL
  };
  static const char *L_d[] = {
    "p", "q", "mu", "eta", NULL
  };
  static const char *L_a[] = {
    "b", "h", "lambda", NULL
  };
  t1_ginfo gm;

  defaultwidth = 500.0;
  nominalwidth = 0.0;

  /*
   * CapHeight, Ascent, and Descent is meaningfull only for Latin/Greek/Cyrillic.
   * The BlueValues and OtherBlues also have those information.
   */
  if (cff_dict_known(cffont->topdict, "FontBBox")) {
    /* Default values */
    capheight = ascent = cff_dict_get(cffont->topdict, "FontBBox", 3);
    descent = cff_dict_get(cffont->topdict, "FontBBox", 1);
  } else {
    capheight =  680.0;
    ascent    =  690.0;
    descent   = -190.0;
  }
  if (cff_dict_known(cffont->private[0], "StdVW")) {
    stemv = cff_dict_get(cffont->private[0], "StdVW", 0);
  } else {
    /*
     * We may use the following values for StemV:
     *  Thin - ExtraLight: <= 50
     *  Light: 71
     *  Regular(Normal): 88
     *  Medium: 109
     *  SemiBold(DemiBold): 135
     *  Bold - Heavy: >= 166
     */
    stemv = 88.0;
  }
  if (cff_dict_known(cffont->topdict, "ItalicAngle")) {
    italicangle = cff_dict_get(cffont->topdict, "ItalicAngle", 0);
    if (italicangle != 0.0)
      flags |= FONT_FLAG_ITALIC;
  } else {
    italicangle = 0.0;
  }

  /*
   * Use "space", "H", "p", and "b" for various values.
   * Those characters should not "seac". (no accent)
   */
  gid = cff_glyph_lookup(cffont, "space");
  if (gid >= 0 && gid < cffont->cstrings->count) {
    t1char_get_metrics(cffont->cstrings->data + cffont->cstrings->offset[gid] - 1,
		       cffont->cstrings->offset[gid+1] - cffont->cstrings->offset[gid],
		       cffont->subrs[0], &gm);
    defaultwidth = gm.wx;
  }

  for (i = 0; L_c[i] != NULL; i++) {
    gid = cff_glyph_lookup(cffont, L_c[i]);
    if (gid >= 0 && gid < cffont->cstrings->count) {
      t1char_get_metrics(cffont->cstrings->data + cffont->cstrings->offset[gid] - 1,
		      cffont->cstrings->offset[gid+1] - cffont->cstrings->offset[gid],
		      cffont->subrs[0], &gm);
      capheight = gm.bbox.ury;
      break;
    }
  }

  for (i = 0; L_d[i] != NULL; i++) {
    gid = cff_glyph_lookup(cffont, L_d[i]);
    if (gid >= 0 && gid < cffont->cstrings->count) {
      t1char_get_metrics(cffont->cstrings->data + cffont->cstrings->offset[gid] - 1,
			 cffont->cstrings->offset[gid+1] - cffont->cstrings->offset[gid],
			 cffont->subrs[0], &gm);
      descent = gm.bbox.lly;
      break;
    }
  }

  for (i = 0; L_a[i] != NULL; i++) {
    gid = cff_glyph_lookup(cffont, L_a[i]);
    if (gid >= 0 && gid < cffont->cstrings->count) {
      t1char_get_metrics(cffont->cstrings->data + cffont->cstrings->offset[gid] - 1,
			 cffont->cstrings->offset[gid+1] - cffont->cstrings->offset[gid],
			 cffont->subrs[0], &gm);
      ascent = gm.bbox.ury;
      break;
    }
  }

  if (defaultwidth != 0.0) {
    cff_dict_add(cffont->private[0], "defaultWidthX", 1);
    cff_dict_set(cffont->private[0], "defaultWidthX", 0, defaultwidth);
  }
  if (nominalwidth != 0.0) {
    cff_dict_add(cffont->private[0], "nominalWidthX", 1);
    cff_dict_set(cffont->private[0], "nominalWidthX", 0, nominalwidth);
  }
  if (cff_dict_known(cffont->private[0], "ForceBold") &&
      cff_dict_get(cffont->private[0], "ForceBold", 0)) {
    flags |= FONT_FLAG_FORCEBOLD;
  }
  if (cff_dict_known(cffont->private[0], "IsFixedPitch") &&
      cff_dict_get(cffont->private[0], "IsFixedPitch", 0)) {
    flags |= FONT_FLAG_FIXEDPITCH;
  }

  fontname   = pdf_font_get_fontname  (font);
  descriptor = pdf_font_get_descriptor(font);

  if (fontname && !strstr(fontname, "Sans")) {
    flags |= FONT_FLAG_SERIF;
  }
  if (fontname &&  strstr(fontname, "Caps")) {
    flags |= FONT_FLAG_SMALLCAP;
  }
  flags |= FONT_FLAG_SYMBOLIC; /* FIXME */

  texpdf_add_dict(descriptor,
	       texpdf_new_name("CapHeight"), texpdf_new_number(capheight));
  texpdf_add_dict(descriptor,
	       texpdf_new_name("Ascent"), texpdf_new_number(ascent));
  texpdf_add_dict(descriptor,
	       texpdf_new_name("Descent"), texpdf_new_number(descent));
  texpdf_add_dict(descriptor,
	       texpdf_new_name("ItalicAngle"), texpdf_new_number(italicangle));
  texpdf_add_dict(descriptor,
	       texpdf_new_name("StemV"), texpdf_new_number(stemv));
  texpdf_add_dict(descriptor,
	       texpdf_new_name("Flags"), texpdf_new_number(flags));
}

static void
add_metrics (pdf_font *font, cff_font *cffont, char **enc_vec, double *widths, long num_glyphs)
{
  pdf_obj *fontdict, *descriptor;
  pdf_obj *tmp_array;
  int      code, firstchar, lastchar;
  double   val;
  int      i;
  int tfm_id;
  char    *usedchars;
  double   scaling;

  fontdict   = pdf_font_get_resource  (font);
  descriptor = pdf_font_get_descriptor(font);
  usedchars  = pdf_font_get_usedchars (font);

  /*
   * The original FontBBox of the font is preserved, instead
   * of replacing it with tight bounding box calculated from
   * charstrings, to prevent Acrobat 4 from greeking text as
   * much as possible.
   */
  if (!cff_dict_known(cffont->topdict, "FontBBox")) {
    ERROR("No FontBBox?");
  }

  /* The widhts array in the font dictionary must be given relative
   * to the default scaling of 1000:1, not relative to the scaling
   * given by the font matrix.
   */
  if (cff_dict_known(cffont->topdict, "FontMatrix"))
    scaling = 1000*cff_dict_get(cffont->topdict, "FontMatrix", 0);
  else
    scaling = 1;

  tmp_array = texpdf_new_array();
  for (i = 0; i < 4; i++) {
    val = cff_dict_get(cffont->topdict, "FontBBox", i);
    texpdf_add_array(tmp_array, texpdf_new_number(ROUND(val, 1.0)));
  }
  texpdf_add_dict(descriptor, texpdf_new_name("FontBBox"), tmp_array);

  tmp_array = texpdf_new_array();
  if (num_glyphs <= 1) { /* This must be an error. */
    firstchar = lastchar = 0;
    texpdf_add_array(tmp_array, texpdf_new_number(0.0));
  } else {
    for (firstchar = 255, lastchar = 0, code = 0; code < 256; code++) {
      if (usedchars[code]) {
	if (code < firstchar) firstchar = code;
	if (code > lastchar)  lastchar  = code;
      }
    }
    if (firstchar > lastchar) {
      WARN("No glyphs actually used???");
      texpdf_release_obj(tmp_array);
      return;
    }
    tfm_id = texpdf_tfm_open(pdf_font_get_tfm_path(font), pdf_font_get_mapname(font), 0);
    for (code = firstchar; code <= lastchar; code++) {
      if (usedchars[code]) {
        double width;
        if (tfm_id < 0) /* tfm is not found */
	  width = scaling * widths[cff_glyph_lookup(cffont, enc_vec[code])];
        else
          width = 1000. * texpdf_tfm_get_width(tfm_id, code);
	texpdf_add_array(tmp_array,
		      texpdf_new_number(ROUND(width, 0.1)));
      } else {
	texpdf_add_array(tmp_array, texpdf_new_number(0.0));
      }
    }
  }

  if (texpdf_array_length(tmp_array) > 0) {
    texpdf_add_dict(fontdict,
		 texpdf_new_name("Widths"),  texpdf_ref_obj(tmp_array));
  }
  texpdf_release_obj(tmp_array);

  texpdf_add_dict(fontdict,
	       texpdf_new_name("FirstChar"), texpdf_new_number(firstchar));
  texpdf_add_dict(fontdict,
	       texpdf_new_name("LastChar"),  texpdf_new_number(lastchar));

  return;
}


static long
write_fontfile (pdf_font *font, cff_font *cffont)
{
  pdf_obj   *descriptor;
  pdf_obj   *fontfile, *stream_dict;
  cff_index *topdict;
  long       private_size, stream_data_len, charstring_len;
  long       topdict_offset, offset;
#define  WBUF_SIZE 1024
  card8     *stream_data_ptr, wbuf[WBUF_SIZE];

  descriptor = pdf_font_get_descriptor(font);

  topdict = cff_new_index(1);
  /*
   * Force existence of Encoding.
   */
  if (!cff_dict_known(cffont->topdict, "CharStrings"))
    cff_dict_add(cffont->topdict, "CharStrings", 1);
  if (!cff_dict_known(cffont->topdict, "charset"))
    cff_dict_add(cffont->topdict, "charset", 1);
  if (!cff_dict_known(cffont->topdict, "Encoding"))
    cff_dict_add(cffont->topdict, "Encoding", 1);
  private_size = cff_dict_pack((cffont->private)[0], wbuf, WBUF_SIZE);
  /* Private dict is required (but may have size 0) */
  if (!cff_dict_known(cffont->topdict, "Private"))
    cff_dict_add(cffont->topdict, "Private", 2);
  topdict->offset[1] = cff_dict_pack(cffont->topdict, wbuf, WBUF_SIZE) + 1;

  /*
   * Estimate total size of fontfile.
   */
  charstring_len = cff_index_size(cffont->cstrings);

  stream_data_len = 4; /* header size */
  stream_data_len += cff_index_size(cffont->name);
  stream_data_len += cff_index_size(topdict);
  stream_data_len += cff_index_size(cffont->string);
  stream_data_len += cff_index_size(cffont->gsubr);
  /* We are using format 1 for Encoding and format 0 for charset.
   * TODO: Should implement cff_xxx_size().
   */
  stream_data_len += 2 + (cffont->encoding->num_entries)*2 + 1 + (cffont->encoding->num_supps)*3;
  stream_data_len += 1 + (cffont->charsets->num_entries)*2;
  stream_data_len += charstring_len;
  stream_data_len += private_size;

  /*
   * Now we create FontFile data.
   */
  stream_data_ptr = NEW(stream_data_len, card8);
  /*
   * Data Layout order as described in CFF spec., sec 2 "Data Layout".
   */
  offset = 0;
  /* Header */
  offset += cff_put_header(cffont,
			   stream_data_ptr + offset, stream_data_len - offset);
  /* Name */
  offset += cff_pack_index(cffont->name,
			   stream_data_ptr + offset, stream_data_len - offset);
  /* Top DICT */
  topdict_offset = offset;
  offset += cff_index_size(topdict);
  /* Strings */
  offset += cff_pack_index(cffont->string,
			   stream_data_ptr + offset, stream_data_len - offset);
  /* Global Subrs */
  offset += cff_pack_index(cffont->gsubr,
			   stream_data_ptr + offset, stream_data_len - offset);
  /* Encoding */
  /* TODO: don't write Encoding entry if the font is always used
   * with PDF Encoding information. Applies to type1c.c as well.
   */
  cff_dict_set(cffont->topdict, "Encoding", 0, offset);
  offset += cff_pack_encoding(cffont,
			      stream_data_ptr + offset, stream_data_len - offset);
  /* charset */
  cff_dict_set(cffont->topdict, "charset", 0, offset);
  offset += cff_pack_charsets(cffont,
			      stream_data_ptr + offset, stream_data_len - offset);
  /* CharStrings */
  cff_dict_set(cffont->topdict, "CharStrings", 0, offset);
  offset += cff_pack_index(cffont->cstrings,
			   stream_data_ptr + offset, charstring_len);
  /* Private */
  if ((cffont->private)[0] && private_size > 0) {
    private_size = cff_dict_pack(cffont->private[0],
				 stream_data_ptr + offset, private_size);
    cff_dict_set(cffont->topdict, "Private", 1, offset);
    cff_dict_set(cffont->topdict, "Private", 0, private_size);
  }
  offset += private_size;

  /* Finally Top DICT */
  topdict->data = NEW(topdict->offset[1] - 1, card8);
  cff_dict_pack (cffont->topdict, topdict->data, topdict->offset[1] - 1);
  cff_pack_index(topdict,
		 stream_data_ptr + topdict_offset, cff_index_size(topdict));
  cff_release_index(topdict);

  /* Copyright and Trademark Notice ommited. */

  /* Flush Font File */
  fontfile    = texpdf_new_stream(STREAM_COMPRESS);
  stream_dict = texpdf_stream_dict(fontfile);
  texpdf_add_dict(descriptor,
	       texpdf_new_name("FontFile3"), texpdf_ref_obj (fontfile));
  texpdf_add_dict(stream_dict,
	       texpdf_new_name("Subtype"),   texpdf_new_name("Type1C"));
  texpdf_add_stream (fontfile, (void *) stream_data_ptr,  offset);
  texpdf_release_obj(fontfile);

  RELEASE(stream_data_ptr);

  return offset;
}


int
pdf_font_load_type1 (pdf_font *font)
{
  pdf_obj      *fontdict;
  int           encoding_id;
  char         *usedchars, *ident;
  char         *fontname, *uniqueTag;
  char         *fullname; /* With pseudo unique tag */
  cff_font     *cffont;
  cff_charsets *charset;
  char        **enc_vec;
  double        defaultwidth, nominalwidth;
  double       *widths;
  card16       *GIDMap, num_glyphs = 0;
  FILE         *fp;
  long          offset;
  int           code, verbose;

  ASSERT(font);

  if (!pdf_font_is_in_use(font)) {
    return 0;
  }

  verbose     = pdf_font_get_verbose();

  encoding_id = pdf_font_get_encoding  (font);
  fontdict    = pdf_font_get_resource  (font);

                pdf_font_get_descriptor(font);
  usedchars   = pdf_font_get_usedchars (font);
  ident       = pdf_font_get_ident     (font);
  fontname    = pdf_font_get_fontname  (font);
  uniqueTag   = pdf_font_get_uniqueTag (font);
  if (!usedchars || !ident || !fontname) {
    ERROR("Type1: Unexpected error.");
  }

  fp = DPXFOPEN(ident, DPX_RES_TYPE_T1FONT);
  if (!fp) {
    ERROR("Type1: Could not open Type1 font: %s", ident);
  }

  GIDMap     = NULL;
  num_glyphs = 0;

  if (encoding_id >= 0) {
    enc_vec = NULL;
  } else {
    enc_vec = NEW(256, char *);
    for (code = 0; code <= 0xff; code++) {
      enc_vec[code] = NULL;
    }
  }

  cffont = t1_load_font(enc_vec, 0, fp);
  if (!cffont) {
    ERROR("Could not load Type 1 font: %s", ident);
  }
  DPXFCLOSE(fp);

  fullname = NEW(strlen(fontname) + 8, char);
  sprintf(fullname, "%6s+%s", uniqueTag, fontname);

  /*
   * Encoding related things.
   */
  if (encoding_id >= 0) {
    enc_vec = pdf_encoding_get_encoding(encoding_id);
  } else {
    pdf_obj *tounicode;

    /*
     * Create enc_vec and ToUnicode CMap for built-in encoding.
     */
    if (!texpdf_lookup_dict(fontdict, "ToUnicode")) {
    tounicode = pdf_create_ToUnicode_CMap(fullname,
					  enc_vec, usedchars);
    if (tounicode) {
      texpdf_add_dict(fontdict,
                     texpdf_new_name("ToUnicode"),
                     texpdf_ref_obj (tounicode));
      texpdf_release_obj(tounicode);
      }
    }
  }

  cff_set_name(cffont, fullname);
  RELEASE(fullname);

  /* defaultWidthX, CapHeight, etc. */
  get_font_attr(font, cffont);
  if (cff_dict_known(cffont->private[0], "defaultWidthX")) {
    defaultwidth = cff_dict_get(cffont->private[0], "defaultWidthX", 0);
  } else {
    defaultwidth = 0.0;
  }
  if (cff_dict_known(cffont->private[0], "nominalWidthX")) {
    nominalwidth = cff_dict_get(cffont->private[0], "nominalWidthX", 0);
  } else {
    nominalwidth = 0.0;
  }

  /* Create CFF encoding, charset, sort glyphs */
#define MAX_GLYPHS 1024
  GIDMap = NEW(MAX_GLYPHS, card16);
  {
    int     prev, duplicate;
    long    gid;
    char   *glyph;
    s_SID   sid;

    cffont->encoding = NEW(1, cff_encoding);
    cffont->encoding->format      = 1;
    cffont->encoding->num_entries = 0;
    cffont->encoding->data.range1 = NEW(256, cff_range1);
    cffont->encoding->num_supps   = 0;
    cffont->encoding->supp        = NEW(256, cff_map);

    charset = NEW(1, cff_charsets);
    charset->format      = 0;
    charset->num_entries = 0;
    charset->data.glyphs = NEW(MAX_GLYPHS, s_SID);

    gid = cff_glyph_lookup(cffont, ".notdef");
    if (gid < 0)
      ERROR("Type 1 font with no \".notdef\" glyph???");
    GIDMap[0] = (card16) gid;
    if (verbose > 2)
      MESG("[glyphs:/.notdef");
    num_glyphs =  1;
    for (prev = -2, code = 0; code <= 0xff; code++) {
      glyph = enc_vec[code];

      if (!usedchars[code])
	continue;
      if (glyph && !strcmp(glyph, ".notdef")) {
	WARN("Character mapped to .notdef used in font: %s",
	     fontname);
	usedchars[code] = 0;
	continue;
      }

      gid = cff_glyph_lookup(cffont, glyph);
      if (gid < 1 || gid >= cffont->cstrings->count) {
	WARN("Glyph \"%s\" missing in font \"%s\".", glyph, fontname);
	usedchars[code] = 0;
	continue;
      }

      for (duplicate = 0; duplicate < code; duplicate++) {
	if (usedchars[duplicate] &&
	    enc_vec[duplicate]   && !strcmp(enc_vec[duplicate], glyph))
	  break;
      }

      sid = cff_add_string(cffont, glyph, 1); /* FIXME */
      if (duplicate < code) { /* found duplicates */
	cffont->encoding->supp[cffont->encoding->num_supps].code  = duplicate;
	cffont->encoding->supp[cffont->encoding->num_supps].glyph = sid;
	cffont->encoding->num_supps += 1;
      } else {
	GIDMap[num_glyphs] = (card16) gid;
	charset->data.glyphs[charset->num_entries] = sid;
	charset->num_entries += 1;
	if (code != prev + 1) {
	  cffont->encoding->num_entries += 1;
	  cffont->encoding->data.range1[cffont->encoding->num_entries-1].first  = code;
	  cffont->encoding->data.range1[cffont->encoding->num_entries-1].n_left = 0;
	} else {
	  cffont->encoding->data.range1[cffont->encoding->num_entries-1].n_left += 1;
	}
	prev = code;
	num_glyphs++;

	if (verbose > 2) {
	  MESG("/%s", glyph);
	}

      }
    }
    if (cffont->encoding->num_supps > 0) {
      cffont->encoding->format |= 0x80;
    } else {
      RELEASE(cffont->encoding->supp); /* FIXME */
      cffont->encoding->supp = NULL;
    }
  }

  widths = NEW(cffont->cstrings->count, double);
  /*
   * No more string will be added.
   * The Type 1 seac operator may add another glyph but the glyph name of
   * those glyphs are contained in standard string. The String Index will
   * not be modified after here.
   * BUT: We cannot update the String Index yet because then we wouldn't be
   * able to find the GIDs of the base and accent characters (unless they
   * have been used already).
   */

  {
    cff_index *cstring;
    t1_ginfo   gm;
    card16     gid, gid_orig;
    long       dstlen_max, srclen;
    card8     *srcptr, *dstptr;

    offset  = dstlen_max = 0L;
    cstring = cff_new_index(cffont->cstrings->count);
    cstring->data      = NULL;
    cstring->offset[0] = 1;
    
    /* The num_glyphs increases if "seac" operators are used. */
    for (gid = 0; gid < num_glyphs; gid++) {
      if (offset + CS_STR_LEN_MAX >= dstlen_max) {
	dstlen_max += CS_STR_LEN_MAX * 2;
	cstring->data = RENEW(cstring->data, dstlen_max, card8);
      }
      gid_orig = GIDMap[gid];

      dstptr   = cstring->data + cstring->offset[gid] - 1;
      srcptr   = cffont->cstrings->data + cffont->cstrings->offset[gid_orig] - 1;
      srclen   = cffont->cstrings->offset[gid_orig + 1] - cffont->cstrings->offset[gid_orig];

      offset  += t1char_convert_charstring(dstptr, CS_STR_LEN_MAX,
					   srcptr, srclen,
					   cffont->subrs[0], defaultwidth, nominalwidth, &gm);
      cstring->offset[gid + 1] = offset + 1;
      if (gm.use_seac) {
	long  bchar_gid, achar_gid, i;
	const char *bchar_name, *achar_name;

	/*
	 * NOTE:
	 *  1. seac.achar and seac.bchar must be contained in the CFF standard string.
	 *  2. Those characters need not to be encoded.
	 *  3. num_glyphs == charsets->num_entries + 1.
	 */
	achar_name = t1_get_standard_glyph(gm.seac.achar);
	achar_gid  = cff_glyph_lookup(cffont, achar_name);
	bchar_name = t1_get_standard_glyph(gm.seac.bchar);
	bchar_gid  = cff_glyph_lookup(cffont, bchar_name);
	if (achar_gid < 0) {
	  WARN("Accent char \"%s\" not found. Invalid use of \"seac\" operator.",
	       achar_name);
	  continue;
	}
	if (bchar_gid < 0) {
	  WARN("Base char \"%s\" not found. Invalid use of \"seac\" operator.",
	       bchar_name);
	  continue;
	}

	for (i = 0; i < num_glyphs; i++) {
	  if (GIDMap[i] == achar_gid)
	    break;
	}
	if (i == num_glyphs) {
	  if (verbose > 2)
	    MESG("/%s", achar_name);
	  GIDMap[num_glyphs++] = achar_gid;
	  charset->data.glyphs[charset->num_entries] = cff_get_seac_sid(cffont, achar_name);
	  charset->num_entries += 1;
	}

	for (i = 0; i < num_glyphs; i++) {
	  if (GIDMap[i] == bchar_gid)
	    break;
	}
	if (i == num_glyphs) {
	  if (verbose > 2)
	    MESG("/%s", bchar_name);
	  GIDMap[num_glyphs++] = bchar_gid;
	  charset->data.glyphs[charset->num_entries] = cff_get_seac_sid(cffont, bchar_name);
	  charset->num_entries += 1;
	}
      }
      widths[gid] = gm.wx;
    }
    cstring->count = num_glyphs;

    cff_release_index(cffont->subrs[0]);
    cffont->subrs[0] = NULL;
    RELEASE(cffont->subrs);
    cffont->subrs    = NULL;

    cff_release_index(cffont->cstrings);
    cffont->cstrings = cstring;

    cff_release_charsets(cffont->charsets);
    cffont->charsets = charset;
  }
  if (verbose > 2)
    MESG("]");

  /* Now we can update the String Index */
  cff_dict_update  (cffont->topdict,    cffont);
  cff_dict_update  (cffont->private[0], cffont);
  cff_update_string(cffont);

  add_metrics(font, cffont, enc_vec, widths, num_glyphs);

  offset = write_fontfile(font, cffont);
  if (verbose > 1)
    MESG("[%u glyphs][%ld bytes]", num_glyphs, offset);

  cff_close(cffont);

  /* Cleanup */
  if (encoding_id < 0 && enc_vec) {
    for (code = 0; code < 256; code++) {
      if (enc_vec[code])
	RELEASE(enc_vec[code]);
      enc_vec[code] = NULL;
    }
    RELEASE(enc_vec);
  }
  if (widths)
    RELEASE(widths);
  if (GIDMap)
    RELEASE(GIDMap);

  /*
   * Maybe writing Charset is recommended for subsetted font.
   */

  return 0;
}