Пример #1
0
static void
pdf_load_builtin_font(fz_context *ctx, pdf_font_desc *fontdesc, char *fontname)
{
	unsigned char *data;
	unsigned int len;

#ifdef _WIN32
	/* SumatraPDF: prefer system fonts unless a base font is explicitly requested */
	char *clean_name = clean_font_name(fontname);
	if (!pdf_lookup_builtin_font(fontname, &len) &&
		/* cf. http://code.google.com/p/sumatrapdf/issues/detail?id=2173 */
		(clean_name == fontname || strncmp(clean_name, "Times-", 6) != 0))
	{
void pdf_load_windows_font(fz_context *ctx, pdf_font_desc *font, char *fontname);
		/* TODO: the metrics for Times-Roman and Courier don't match
		   those of Windows' Times New Roman and Courier New; for
		   some reason, Poppler doesn't seem to have this problem */
		fz_try(ctx)
		{
			pdf_load_windows_font(ctx, fontdesc, fontname);
		}
		fz_catch(ctx) { }
printf("    fontdesc->font=%p\n",fontdesc->font);
		if (fontdesc->font)
			return;
	}
Пример #2
0
static void
pdf_load_builtin_font(fz_context *ctx, pdf_font_desc *fontdesc, char *fontname, int has_descriptor)
{
	FT_Face face;
	const char *clean_name = clean_font_name(fontname);

	fontdesc->font = fz_load_system_font(ctx, fontname, 0, 0, !has_descriptor);
	if (!fontdesc->font)
	{
		unsigned char *data;
		unsigned int len;

		data = pdf_lookup_builtin_font(clean_name, &len);
		if (!data)
			fz_throw(ctx, FZ_ERROR_GENERIC, "cannot find builtin font: '%s'", fontname);

		fontdesc->font = fz_new_font_from_memory(ctx, fontname, data, len, 0, 1);
	}

	if (!strcmp(clean_name, "Symbol") || !strcmp(clean_name, "ZapfDingbats"))
		fontdesc->flags |= PDF_FD_SYMBOLIC;

	face = fontdesc->font->ft_face;
	fontdesc->ascent = 1000.0f * face->ascender / face->units_per_EM;
	fontdesc->descent = 1000.0f * face->descender / face->units_per_EM;
}
Пример #3
0
static void
pdf_load_font_descriptor(pdf_font_desc *fontdesc, pdf_document *doc, pdf_obj *dict, char *collection, char *basefont, int iscidfont, int has_encoding)
{
	pdf_obj *obj1, *obj2, *obj3, *obj;
	char *fontname, *origname;
	FT_Face face;
	fz_context *ctx = doc->ctx;

	/* Prefer BaseFont; don't bother with FontName */
	origname = basefont;

	/* SumatraPDF: handle /BaseFont /Arial,Bold+000041 /FontName /Arial,Bold */
	if (strchr(basefont, '+') && pdf_is_name(pdf_dict_gets(dict, "FontName")))
		origname = pdf_to_name(pdf_dict_gets(dict, "FontName"));

	/* cf. http://code.google.com/p/sumatrapdf/issues/detail?id=1616 */
	if (strlen(origname) > 7 && origname[6] == '+')
		origname += 7;

	/* Look through list of alternate names for built in fonts */
	fontname = clean_font_name(origname);

	fontdesc->flags = pdf_to_int(pdf_dict_gets(dict, "Flags"));
	fontdesc->italic_angle = pdf_to_real(pdf_dict_gets(dict, "ItalicAngle"));
	fontdesc->ascent = pdf_to_real(pdf_dict_gets(dict, "Ascent"));
	fontdesc->descent = pdf_to_real(pdf_dict_gets(dict, "Descent"));
	fontdesc->cap_height = pdf_to_real(pdf_dict_gets(dict, "CapHeight"));
	fontdesc->x_height = pdf_to_real(pdf_dict_gets(dict, "XHeight"));
	fontdesc->missing_width = pdf_to_real(pdf_dict_gets(dict, "MissingWidth"));

	obj1 = pdf_dict_gets(dict, "FontFile");
	obj2 = pdf_dict_gets(dict, "FontFile2");
	obj3 = pdf_dict_gets(dict, "FontFile3");
	obj = obj1 ? obj1 : obj2 ? obj2 : obj3;

	if (pdf_is_indirect(obj))
	{
		fz_try(ctx)
		{
			pdf_load_embedded_font(doc, fontdesc, fontname, obj);
		}
		fz_catch(ctx)
		{
			fz_rethrow_if(ctx, FZ_ERROR_TRYLATER);
			fz_warn(ctx, "ignored error when loading embedded font; attempting to load system font");
			if (origname != fontname && !iscidfont)
				pdf_load_builtin_font(ctx, fontdesc, fontname);
			else
				pdf_load_system_font(ctx, fontdesc, fontname, collection, has_encoding);
		}
	}
	else
	{
		if (origname != fontname && !iscidfont)
Пример #4
0
static void
pdf_load_font_descriptor(pdf_font_desc *fontdesc, pdf_document *xref, pdf_obj *dict, char *collection, char *basefont)
{
	pdf_obj *obj1, *obj2, *obj3, *obj;
	char *fontname;
	char *origname;
	FT_Face face;
	fz_context *ctx = xref->ctx;

	if (!strchr(basefont, ',') || strchr(basefont, '+'))
		origname = pdf_to_name(pdf_dict_gets(dict, "FontName"));
	else
		origname = basefont;
	fontname = clean_font_name(origname);

	fontdesc->flags = pdf_to_int(pdf_dict_gets(dict, "Flags"));
	fontdesc->italic_angle = pdf_to_real(pdf_dict_gets(dict, "ItalicAngle"));
	fontdesc->ascent = pdf_to_real(pdf_dict_gets(dict, "Ascent"));
	fontdesc->descent = pdf_to_real(pdf_dict_gets(dict, "Descent"));
	fontdesc->cap_height = pdf_to_real(pdf_dict_gets(dict, "CapHeight"));
	fontdesc->x_height = pdf_to_real(pdf_dict_gets(dict, "XHeight"));
	fontdesc->missing_width = pdf_to_real(pdf_dict_gets(dict, "MissingWidth"));

	obj1 = pdf_dict_gets(dict, "FontFile");
	obj2 = pdf_dict_gets(dict, "FontFile2");
	obj3 = pdf_dict_gets(dict, "FontFile3");
	obj = obj1 ? obj1 : obj2 ? obj2 : obj3;

	if (pdf_is_indirect(obj))
	{
		fz_try(ctx)
		{
			pdf_load_embedded_font(fontdesc, xref, obj);
		}
		fz_catch(ctx)
		{
			fz_warn(ctx, "ignored error when loading embedded font; attempting to load system font");
			if (origname != fontname)
				pdf_load_builtin_font(ctx, fontdesc, fontname);
			else
				pdf_load_system_font(ctx, fontdesc, fontname, collection);
			/* RJW: "cannot load font descriptor (%d %d R)", pdf_to_num(dict), pdf_to_gen(dict) */
		}
	}
	else
	{
		if (origname != fontname)
Пример #5
0
static void
pdf_load_builtin_font(fz_context *ctx, pdf_font_desc *fontdesc, char *fontname)
{
	unsigned char *data;
	unsigned int len;

	fontname = clean_font_name(fontname);

	data = pdf_lookup_builtin_font(fontname, &len);
	if (!data)
		fz_throw(ctx, "cannot find builtin font: '%s'", fontname);

	fontdesc->font = fz_new_font_from_memory(ctx, fontname, data, len, 0, 1);

	if (!strcmp(fontname, "Symbol") || !strcmp(fontname, "ZapfDingbats"))
		fontdesc->flags |= PDF_FD_SYMBOLIC;
}
Пример #6
0
static void
pdf_load_font_descriptor(pdf_font_desc *fontdesc, pdf_document *doc, pdf_obj *dict, char *collection, char *basefont, int iscidfont)
{
	pdf_obj *obj1, *obj2, *obj3, *obj;
	char *fontname, *origname;
	FT_Face face;
	fz_context *ctx = doc->ctx;

	/* Prefer BaseFont; don't bother with FontName */
	origname = basefont;

	/* Look through list of alternate names for built in fonts */
	fontname = clean_font_name(origname);

	fontdesc->flags = pdf_to_int(pdf_dict_gets(dict, "Flags"));
	fontdesc->italic_angle = pdf_to_real(pdf_dict_gets(dict, "ItalicAngle"));
	fontdesc->ascent = pdf_to_real(pdf_dict_gets(dict, "Ascent"));
	fontdesc->descent = pdf_to_real(pdf_dict_gets(dict, "Descent"));
	fontdesc->cap_height = pdf_to_real(pdf_dict_gets(dict, "CapHeight"));
	fontdesc->x_height = pdf_to_real(pdf_dict_gets(dict, "XHeight"));
	fontdesc->missing_width = pdf_to_real(pdf_dict_gets(dict, "MissingWidth"));

	obj1 = pdf_dict_gets(dict, "FontFile");
	obj2 = pdf_dict_gets(dict, "FontFile2");
	obj3 = pdf_dict_gets(dict, "FontFile3");
	obj = obj1 ? obj1 : obj2 ? obj2 : obj3;

	if (pdf_is_indirect(obj))
	{
		fz_try(ctx)
		{
			pdf_load_embedded_font(doc, fontdesc, fontname, obj);
		}
		fz_catch(ctx)
		{
			fz_rethrow_if(ctx, FZ_ERROR_TRYLATER);
			fz_warn(ctx, "ignored error when loading embedded font; attempting to load system font");
			if (origname != fontname && !iscidfont)
				pdf_load_builtin_font(ctx, fontdesc, fontname);
			else
				pdf_load_system_font(ctx, fontdesc, fontname, collection);
		}
	}
	else
	{
		if (origname != fontname && !iscidfont)
Пример #7
0
	{
void pdf_load_windows_font(fz_context *ctx, pdf_font_desc *font, char *fontname);
		/* TODO: the metrics for Times-Roman and Courier don't match
		   those of Windows' Times New Roman and Courier New; for
		   some reason, Poppler doesn't seem to have this problem */
		fz_try(ctx)
		{
			pdf_load_windows_font(ctx, fontdesc, fontname);
		}
		fz_catch(ctx) { }
printf("    fontdesc->font=%p\n",fontdesc->font);
		if (fontdesc->font)
			return;
	}
#endif
	fontname = clean_font_name(fontname);

	data = pdf_lookup_builtin_font(fontname, &len);
	if (!data)
		fz_throw(ctx, FZ_ERROR_GENERIC, "cannot find builtin font: '%s'", fontname);

	fontdesc->font = fz_new_font_from_memory(ctx, fontname, data, len, 0, 1);

	if (!strcmp(fontname, "Symbol") || !strcmp(fontname, "ZapfDingbats"))
		fontdesc->flags |= PDF_FD_SYMBOLIC;
}

static void
pdf_load_substitute_font(fz_context *ctx, pdf_font_desc *fontdesc, char *fontname, int mono, int serif, int bold, int italic)
{
	unsigned char *data;
Пример #8
0
static pdf_font_desc *
pdf_load_simple_font(pdf_document *xref, pdf_obj *dict)
{
	pdf_obj *descriptor;
	pdf_obj *encoding;
	pdf_obj *widths;
	unsigned short *etable = NULL;
	pdf_font_desc *fontdesc = NULL;
	FT_Face face;
	FT_CharMap cmap;
	int symbolic;
	int kind;

	char *basefont;
	char *fontname;
	char *estrings[256];
	char ebuffer[256][32];
	int i, k, n;
	int fterr;
	fz_context *ctx = xref->ctx;

	fz_var(fontdesc);
	fz_var(etable);

	basefont = pdf_to_name(pdf_dict_gets(dict, "BaseFont"));
	fontname = clean_font_name(basefont);

	/* Load font file */
	fz_try(ctx)
	{
		fontdesc = pdf_new_font_desc(ctx);

		descriptor = pdf_dict_gets(dict, "FontDescriptor");
		if (descriptor)
			pdf_load_font_descriptor(fontdesc, xref, descriptor, NULL, basefont);
		else
			pdf_load_builtin_font(ctx, fontdesc, fontname);

		/* Some chinese documents mistakenly consider WinAnsiEncoding to be codepage 936 */
		if (!*fontdesc->font->name &&
			!pdf_dict_gets(dict, "ToUnicode") &&
			!strcmp(pdf_to_name(pdf_dict_gets(dict, "Encoding")), "WinAnsiEncoding") &&
			pdf_to_int(pdf_dict_gets(descriptor, "Flags")) == 4)
		{
			/* note: without the comma, pdf_load_font_descriptor would prefer /FontName over /BaseFont */
			char *cp936fonts[] = {
				"\xCB\xCE\xCC\xE5", "SimSun,Regular",
				"\xBA\xDA\xCC\xE5", "SimHei,Regular",
				"\xBF\xAC\xCC\xE5_GB2312", "SimKai,Regular",
				"\xB7\xC2\xCB\xCE_GB2312", "SimFang,Regular",
				"\xC1\xA5\xCA\xE9", "SimLi,Regular",
				NULL
			};
			for (i = 0; cp936fonts[i]; i += 2)
				if (!strcmp(basefont, cp936fonts[i]))
					break;
			if (cp936fonts[i])
			{
				fz_warn(ctx, "workaround for S22PDF lying about chinese font encodings");
				pdf_drop_font(ctx, fontdesc);
				fontdesc = pdf_new_font_desc(ctx);
				pdf_load_font_descriptor(fontdesc, xref, descriptor, "Adobe-GB1", cp936fonts[i+1]);
				fontdesc->encoding = pdf_load_system_cmap(ctx, "GBK-EUC-H");
				fontdesc->to_unicode = pdf_load_system_cmap(ctx, "Adobe-GB1-UCS2");
				fontdesc->to_ttf_cmap = pdf_load_system_cmap(ctx, "Adobe-GB1-UCS2");
				/* RJW: "cannot load font" */

				face = fontdesc->font->ft_face;
				kind = ft_kind(face);
				goto skip_encoding;
			}
		}

		face = fontdesc->font->ft_face;
		kind = ft_kind(face);

		/* Encoding */

		symbolic = fontdesc->flags & 4;

		if (face->num_charmaps > 0)
			cmap = face->charmaps[0];
		else
			cmap = NULL;

		for (i = 0; i < face->num_charmaps; i++)
		{
			FT_CharMap test = face->charmaps[i];

			if (kind == TYPE1)
			{
				if (test->platform_id == 7)
					cmap = test;
			}

			if (kind == TRUETYPE)
			{
				if (test->platform_id == 1 && test->encoding_id == 0)
					cmap = test;
				if (test->platform_id == 3 && test->encoding_id == 1)
					cmap = test;
			}
		}

		if (cmap)
		{
			fterr = FT_Set_Charmap(face, cmap);
			if (fterr)
				fz_warn(ctx, "freetype could not set cmap: %s", ft_error_string(fterr));
		}
		else
			fz_warn(ctx, "freetype could not find any cmaps");

		etable = fz_malloc_array(ctx, 256, sizeof(unsigned short));
		fontdesc->size += 256 * sizeof(unsigned short);
		for (i = 0; i < 256; i++)
		{
			estrings[i] = NULL;
			etable[i] = 0;
		}

		encoding = pdf_dict_gets(dict, "Encoding");
		if (encoding)
		{
			if (pdf_is_name(encoding))
				pdf_load_encoding(estrings, pdf_to_name(encoding));

			if (pdf_is_dict(encoding))
			{
				pdf_obj *base, *diff, *item;

				base = pdf_dict_gets(encoding, "BaseEncoding");
				if (pdf_is_name(base))
					pdf_load_encoding(estrings, pdf_to_name(base));
				else if (!fontdesc->is_embedded && !symbolic)
					pdf_load_encoding(estrings, "StandardEncoding");

				diff = pdf_dict_gets(encoding, "Differences");
				if (pdf_is_array(diff))
				{
					n = pdf_array_len(diff);
					k = 0;
					for (i = 0; i < n; i++)
					{
						item = pdf_array_get(diff, i);
						if (pdf_is_int(item))
							k = pdf_to_int(item);
						if (pdf_is_name(item))
							estrings[k++] = pdf_to_name(item);
						if (k < 0) k = 0;
						if (k > 255) k = 255;
					}
				}
			}
		}

		/* start with the builtin encoding */
		for (i = 0; i < 256; i++)
			etable[i] = ft_char_index(face, i);

		/* encode by glyph name where we can */
		fz_lock(ctx, FZ_LOCK_FREETYPE);
		if (kind == TYPE1)
		{
			for (i = 0; i < 256; i++)
			{
				if (estrings[i])
				{
					etable[i] = FT_Get_Name_Index(face, estrings[i]);
					if (etable[i] == 0)
					{
						int aglcode = pdf_lookup_agl(estrings[i]);
						const char **dupnames = pdf_lookup_agl_duplicates(aglcode);
						while (*dupnames)
						{
							etable[i] = FT_Get_Name_Index(face, (char*)*dupnames);
							if (etable[i])
								break;
							dupnames++;
						}
					}
				}
			}
		}

		/* encode by glyph name where we can */
		if (kind == TRUETYPE)
		{
			/* Unicode cmap */
			if (!symbolic && face->charmap && face->charmap->platform_id == 3)
			{
				for (i = 0; i < 256; i++)
				{
					if (estrings[i])
					{
						int aglcode = pdf_lookup_agl(estrings[i]);
						if (!aglcode)
							etable[i] = FT_Get_Name_Index(face, estrings[i]);
						else
							etable[i] = ft_char_index(face, aglcode);
					}
				}
			}

			/* MacRoman cmap */
			else if (!symbolic && face->charmap && face->charmap->platform_id == 1)
			{
				for (i = 0; i < 256; i++)
				{
					if (estrings[i])
					{
						k = lookup_mre_code(estrings[i]);
						if (k <= 0)
							etable[i] = FT_Get_Name_Index(face, estrings[i]);
						else
							etable[i] = ft_char_index(face, k);
					}
				}
			}

			/* Symbolic cmap */
			else
			{
				for (i = 0; i < 256; i++)
				{
					if (estrings[i])
					{
						etable[i] = FT_Get_Name_Index(face, estrings[i]);
						if (etable[i] == 0)
							etable[i] = ft_char_index(face, i);
					}
				}
			}
		}

		/* try to reverse the glyph names from the builtin encoding */
		for (i = 0; i < 256; i++)
		{
			if (etable[i] && !estrings[i])
			{
				if (FT_HAS_GLYPH_NAMES(face))
				{
					fterr = FT_Get_Glyph_Name(face, etable[i], ebuffer[i], 32);
					if (fterr)
						fz_warn(ctx, "freetype get glyph name (gid %d): %s", etable[i], ft_error_string(fterr));
					if (ebuffer[i][0])
						estrings[i] = ebuffer[i];
				}
				else
				{
					estrings[i] = (char*) pdf_win_ansi[i]; /* discard const */
				}
			}
		}
		fz_unlock(ctx, FZ_LOCK_FREETYPE);

		fontdesc->encoding = pdf_new_identity_cmap(ctx, 0, 1);
		fontdesc->size += pdf_cmap_size(ctx, fontdesc->encoding);
		fontdesc->cid_to_gid_len = 256;
		fontdesc->cid_to_gid = etable;

		pdf_load_to_unicode(xref, fontdesc, estrings, NULL, pdf_dict_gets(dict, "ToUnicode"));
		/* RJW: "cannot load to_unicode" */

	skip_encoding:

		/* Widths */

		pdf_set_default_hmtx(ctx, fontdesc, fontdesc->missing_width);

		widths = pdf_dict_gets(dict, "Widths");
		if (widths)
		{
			int first, last;

			first = pdf_to_int(pdf_dict_gets(dict, "FirstChar"));
			last = pdf_to_int(pdf_dict_gets(dict, "LastChar"));

			if (first < 0 || last > 255 || first > last)
				first = last = 0;

			for (i = 0; i < last - first + 1; i++)
			{
				int wid = pdf_to_int(pdf_array_get(widths, i));
				pdf_add_hmtx(ctx, fontdesc, i + first, i + first, wid);
			}
		}
		else
		{
			fz_lock(ctx, FZ_LOCK_FREETYPE);
			fterr = FT_Set_Char_Size(face, 1000, 1000, 72, 72);
			if (fterr)
				fz_warn(ctx, "freetype set character size: %s", ft_error_string(fterr));
			for (i = 0; i < 256; i++)
			{
				pdf_add_hmtx(ctx, fontdesc, i, i, ft_width(ctx, fontdesc, i));
			}
			fz_unlock(ctx, FZ_LOCK_FREETYPE);
		}

		pdf_end_hmtx(ctx, fontdesc);
	}
	fz_catch(ctx)
	{
		if (fontdesc && etable != fontdesc->cid_to_gid)
			fz_free(ctx, etable);
		pdf_drop_font(ctx, fontdesc);
		fz_throw(ctx, "cannot load simple font (%d %d R)", pdf_to_num(dict), pdf_to_gen(dict));
	}
	return fontdesc;
}