path: root/0002-pymupdf.patch

diff -ru mupdf-1.17.0-source-orig/include/mupdf/fitz/config.h mupdf-1.17.0-source/include/mupdf/fitz/config.h
--- mupdf-1.17.0-source-orig/include/mupdf/fitz/config.h	2020-05-05 19:29:09.000000000 +0800
+++ mupdf-1.17.0-source/include/mupdf/fitz/config.h	2020-08-09 23:10:55.615700000 +0800
@@ -2,13 +2,13 @@
 
 #define FZ_CONFIG_H
 
-/**
+/*
 	Enable the following for spot (and hence overprint/overprint
 	simulation) capable rendering. This forces FZ_PLOTTERS_N on.
 */
 /* #define FZ_ENABLE_SPOT_RENDERING 1 */
 
-/**
+/*
 	Choose which plotters we need.
 	By default we build all the plotters in. To avoid building
 	plotters in that aren't needed, define the unwanted
@@ -19,7 +19,7 @@
 /* #define FZ_PLOTTERS_CMYK 1 */
 /* #define FZ_PLOTTERS_N 1 */
 
-/**
+/*
 	Choose which document agents to include.
 	By default all are enabled. To avoid building unwanted
 	ones, define FZ_ENABLE_... to 0.
@@ -32,12 +32,12 @@
 /* #define FZ_ENABLE_HTML 1 */
 /* #define FZ_ENABLE_EPUB 1 */
 
-/**
+/*
 	Choose whether to enable ICC color profiles.
 */
-/* #define FZ_ENABLE_ICC 1 */
+#define FZ_ENABLE_ICC 1
 
-/**
+/*
 	Choose whether to enable JPEG2000 decoding.
 	By default, it is enabled, but due to frequent security
 	issues with the third party libraries we support disabling
@@ -45,14 +45,13 @@
 */
 /* #define FZ_ENABLE_JPX 1 */
 
-/**
+/*
 	Choose whether to enable JavaScript.
-	By default JavaScript is enabled both for mutool and PDF
-	interactivity.
+	By default JavaScript is enabled both for mutool and PDF interactivity.
 */
 /* #define FZ_ENABLE_JS 1 */
 
-/**
+/*
 	Choose which fonts to include.
 	By default we include the base 14 PDF fonts,
 	DroidSansFallback from Android for CJK, and
@@ -61,35 +60,32 @@
 	unwanted fonts.
 */
 /* To avoid all noto fonts except CJK, enable: */
-/* #define TOFU */
+#define TOFU // <=== PyMuPDF: not the complete NOTO!
 
-/* To skip the CJK font, enable: (this implicitly enables TOFU_CJK_EXT
- * and TOFU_CJK_LANG) */
+/* To skip the CJK font, enable: (this implicitly enables TOFU_CJK_EXT and TOFU_CJK_LANG) */
 /* #define TOFU_CJK */
 
-/* To skip CJK Extension A, enable: (this implicitly enables
- * TOFU_CJK_LANG) */
-/* #define TOFU_CJK_EXT */
+/* To skip CJK Extension A, enable: (this implicitly enables TOFU_CJK_LANG) */
+#define TOFU_CJK_EXT // <=== PyMuPDF: causes 1.5 MB extra size
 
 /* To skip CJK language specific fonts, enable: */
-/* #define TOFU_CJK_LANG */
+#define TOFU_CJK_LANG // <=== PyMuPDF
 
 /* To skip the Emoji font, enable: */
-/* #define TOFU_EMOJI */
+#define TOFU_EMOJI // <=== PyMuPDF
 
 /* To skip the ancient/historic scripts, enable: */
-/* #define TOFU_HISTORIC */
+#define TOFU_HISTORIC // <=== PyMuPDF
 
 /* To skip the symbol font, enable: */
-/* #define TOFU_SYMBOL */
+#define TOFU_SYMBOL // <=== PyMuPDF
 
 /* To skip the SIL fonts, enable: */
-/* #define TOFU_SIL */
+// #define TOFU_SIL // <=== PyMuPDF: ok to have it, only small fonts
 
 /* To skip the Base14 fonts, enable: */
 /* #define TOFU_BASE14 */
-/* (You probably really don't want to do that except for measurement
- * purposes!) */
+/* (You probably really don't want to do that except for measurement purposes!) */
 
 /* ---------- DO NOT EDIT ANYTHING UNDER THIS LINE ---------- */
 
diff -ru mupdf-1.17.0-source-orig/source/fitz/encode-fax.c mupdf-1.17.0-source/source/fitz/encode-fax.c
--- mupdf-1.17.0-source-orig/source/fitz/encode-fax.c	2020-05-05 19:29:09.000000000 +0800
+++ mupdf-1.17.0-source/source/fitz/encode-fax.c	2020-08-09 23:10:55.615700000 +0800
@@ -120,7 +120,7 @@
 {
 	int a, b;
 
-	if (!line)
+	if (!line || x >= w)
 		return w;
 
 	if (x == -1)
@@ -147,7 +147,7 @@
 static inline int
 find_changing_color(const unsigned char *line, int x, int w, int color)
 {
-	if (!line)
+	if (!line || x >= w)
 		return w;
 	x = find_changing(line, x, w);
 	if (x < w && getbit(line, x) != color)
diff -ru mupdf-1.17.0-source-orig/source/pdf/pdf-font-add.c mupdf-1.17.0-source/source/pdf/pdf-font-add.c
--- mupdf-1.17.0-source-orig/source/pdf/pdf-font-add.c	2020-05-05 19:29:09.000000000 +0800
+++ mupdf-1.17.0-source/source/pdf/pdf-font-add.c	2020-08-09 23:10:55.615700000 +0800
@@ -21,10 +21,14 @@
 #else
 	const char *kind = FT_Get_X11_Font_Format(face);
 #endif
-	if (!strcmp(kind, "TrueType")) return 2;
-	if (!strcmp(kind, "Type 1")) return 1;
-	if (!strcmp(kind, "CFF")) return 3;
-	if (!strcmp(kind, "CID Type 1")) return 1;
+	if (!strcmp(kind, "TrueType"))
+		return 2;
+	if (!strcmp(kind, "Type 1"))
+		return 1;
+	if (!strcmp(kind, "CFF"))
+		return 3;
+	if (!strcmp(kind, "CID Type 1"))
+		return 1;
 	return 0;
 }
 
@@ -54,7 +58,7 @@
 	return fz_lookup_base14_font(ctx, pdf_clean_font_name(font->name), &size) == data;
 }
 
-static pdf_obj*
+static pdf_obj *
 pdf_add_font_file(fz_context *ctx, pdf_document *doc, fz_font *font)
 {
 	fz_buffer *buf = font->buffer;
@@ -136,9 +140,15 @@
 			switch (ft_font_file_kind(face))
 			{
 			default:
-			case 1: pdf_dict_put_drop(ctx, fdobj, PDF_NAME(FontFile), fileref); break;
-			case 2: pdf_dict_put_drop(ctx, fdobj, PDF_NAME(FontFile2), fileref); break;
-			case 3: pdf_dict_put_drop(ctx, fdobj, PDF_NAME(FontFile3), fileref); break;
+			case 1:
+				pdf_dict_put_drop(ctx, fdobj, PDF_NAME(FontFile), fileref);
+				break;
+			case 2:
+				pdf_dict_put_drop(ctx, fdobj, PDF_NAME(FontFile2), fileref);
+				break;
+			case 3:
+				pdf_dict_put_drop(ctx, fdobj, PDF_NAME(FontFile3), fileref);
+				break;
 			}
 		}
 
@@ -151,7 +161,7 @@
 }
 
 static void
-pdf_add_simple_font_widths(fz_context *ctx, pdf_document *doc, pdf_obj *fobj, fz_font *font, const char * const encoding[])
+pdf_add_simple_font_widths(fz_context *ctx, pdf_document *doc, pdf_obj *fobj, fz_font *font, const char *const encoding[])
 {
 	int width_table[256];
 	pdf_obj *widths;
@@ -196,7 +206,12 @@
 }
 
 /* Different states of starting, same width as last, or consecutive glyph */
-enum { FW_START, FW_SAME, FW_RUN };
+enum
+{
+	FW_START,
+	FW_SAME,
+	FW_RUN
+};
 
 /* ToDo: Ignore the default sized characters */
 static void
@@ -329,7 +344,7 @@
 }
 
 /* Descendant font construction used for CID font creation from ttf or Adobe type1 */
-static pdf_obj*
+static pdf_obj *
 pdf_add_descendant_cid_font(fz_context *ctx, pdf_document *doc, fz_font *font)
 {
 	FT_Face face = font->ft_face;
@@ -372,9 +387,9 @@
 	int n;
 	for (n = 1; k + n < size; ++n)
 	{
-		if ((k & 0xFF00) != ((k+n) & 0xFF00)) /* high byte changes */
+		if ((k & 0xFF00) != ((k + n) & 0xFF00)) /* high byte changes */
 			break;
-		if (table[k] + n != table[k+n])
+		if (table[k] + n != table[k + n])
 			break;
 	}
 	return n;
@@ -474,7 +489,7 @@
 							fz_append_printf(ctx, buf, "%d beginbfrange\n", num_seq);
 						count = 0;
 					}
-					fz_append_printf(ctx, buf, "<%04x> <%04x> <%04x>\n", k, k+n-1, table[k]);
+					fz_append_printf(ctx, buf, "<%04x> <%04x> <%04x>\n", k, k + n - 1, table[k]);
 					++count;
 				}
 			}
@@ -582,7 +597,7 @@
 		const char *glyph = glyph_names[i];
 		if (glyph)
 		{
-			if (last != i-1)
+			if (last != i - 1)
 				pdf_array_push_int(ctx, diff, i);
 			last = i;
 			pdf_array_push_name(ctx, diff, glyph);
@@ -624,9 +639,15 @@
 	switch (encoding)
 	{
 	default:
-	case PDF_SIMPLE_ENCODING_LATIN: enc = fz_glyph_name_from_windows_1252; break;
-	case PDF_SIMPLE_ENCODING_GREEK: enc = fz_glyph_name_from_iso8859_7; break;
-	case PDF_SIMPLE_ENCODING_CYRILLIC: enc = fz_glyph_name_from_koi8u; break;
+	case PDF_SIMPLE_ENCODING_LATIN:
+		enc = fz_glyph_name_from_windows_1252;
+		break;
+	case PDF_SIMPLE_ENCODING_GREEK:
+		enc = fz_glyph_name_from_iso8859_7;
+		break;
+	case PDF_SIMPLE_ENCODING_CYRILLIC:
+		enc = fz_glyph_name_from_koi8u;
+		break;
 	}
 
 	fobj = pdf_add_new_dict(ctx, doc, 10);
@@ -667,8 +688,7 @@
 	return fref;
 }
 
-int
-pdf_font_writing_supported(fz_font *font)
+int pdf_font_writing_supported(fz_font *font)
 {
 	if (font->ft_face == NULL || font->buffer == NULL || font->buffer->len < 4)
 		return 0;
@@ -686,7 +706,7 @@
 {
 	pdf_obj *fref, *font, *subfont, *fontdesc;
 	pdf_obj *dfonts;
-	fz_rect bbox = { -200, -200, 1200, 1200 };
+	fz_rect bbox = {-200, -200, 1200, 1200};
 	unsigned char digest[16];
 	int flags;
 
@@ -769,3 +789,21 @@
 
 	return fref;
 }
+
+void jm_valid_chars(fz_context *ctx, fz_font *font, void *arr)
+{
+	FT_Face face = font->ft_face;
+	FT_ULong ucs;
+	FT_UInt gid;
+	long *table = (long *)arr;
+	fz_lock(ctx, FZ_LOCK_FREETYPE);
+	ucs = FT_Get_First_Char(face, &gid);
+	while (gid > 0)
+	{
+		if (gid < (FT_ULong)face->num_glyphs && face->num_glyphs > 0)
+			table[gid] = (long)ucs;
+		ucs = FT_Get_Next_Char(face, ucs, &gid);
+	}
+	fz_unlock(ctx, FZ_LOCK_FREETYPE);
+	return;
+}