b66239b436
When operating with binary files on Windows it is necessary to explicitly open them in binary mode to prevent unwanted effects. git-svn-id: svn://svn.rockbox.org/rockbox/trunk@29699 a1c6a512-1295-4272-9138-f99709370657
1358 lines
38 KiB
C
1358 lines
38 KiB
C
/***************************************************************************
|
|
* __________ __ ___.
|
|
* Open \______ \ ____ ____ | | _\_ |__ _______ ___
|
|
* Source | _// _ \_/ ___\| |/ /| __ \ / _ \ \/ /
|
|
* Jukebox | | ( <_> ) \___| < | \_\ ( <_> > < <
|
|
* Firmware |____|_ /\____/ \___ >__|_ \|___ /\____/__/\_ \
|
|
* \/ \/ \/ \/ \/
|
|
* $Id$
|
|
*
|
|
* Copyright (C) 2007 Jonas Hurrelmann
|
|
*
|
|
* A command-line tool to convert ttf file to bitmap fonts
|
|
*
|
|
* All files in this archive are subject to the GNU General Public License.
|
|
* See the file COPYING in the source tree root for full license agreement.
|
|
*
|
|
* This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
|
|
* KIND, either express or implied.
|
|
*
|
|
****************************************************************************/
|
|
#include <ft2build.h>
|
|
#include FT_FREETYPE_H
|
|
#include FT_GLYPH_H
|
|
|
|
#include <stdio.h>
|
|
#ifdef WIN32
|
|
#ifdef _MSC_VER
|
|
#define snprintf _snprintf
|
|
#endif
|
|
#else
|
|
#include <stdlib.h>
|
|
#include <unistd.h>
|
|
#endif
|
|
#include FT_SFNT_NAMES_H
|
|
#include FT_TRUETYPE_TABLES_H
|
|
|
|
#include <string.h>
|
|
/*
|
|
* Set the default values used to generate a BDF font.
|
|
*/
|
|
#ifndef DEFAULT_PLATFORM_ID
|
|
#define DEFAULT_PLATFORM_ID 3
|
|
#endif
|
|
|
|
#ifndef DEFAULT_ENCODING_ID
|
|
#define DEFAULT_ENCODING_ID 1
|
|
#endif
|
|
|
|
#define ABS(x) (((x) < 0) ? -(x) : (x))
|
|
|
|
#define VERSION "RB12"
|
|
/*
|
|
* nameID macros for getting strings from the OT font.
|
|
*/
|
|
enum {
|
|
BDFOTF_COPYRIGHT_STRING = 0,
|
|
BDFOTF_FAMILY_STRING,
|
|
BDFOTF_SUBFAMILY_STRING,
|
|
BDFOTF_UNIQUEID_STRING,
|
|
BDFOTF_FULLNAME_STRING,
|
|
BDFOTF_VENDOR_STRING,
|
|
BDFOTF_POSTSCRIPT_STRING,
|
|
BDFOTF_TRADEMARK_STRING,
|
|
};
|
|
/*
|
|
* String names for the string indexes. Used for error messages.
|
|
*/
|
|
static char *string_names[] = {
|
|
"\"Copyright\"",
|
|
"\"Family\"",
|
|
"\"SubFamily\"",
|
|
"\"Unique ID\"",
|
|
"\"Full Name\"",
|
|
"\"Vendor\"",
|
|
"\"Postscript Name\"",
|
|
"\"Trademark\""
|
|
};
|
|
|
|
|
|
/*
|
|
* The default platform and encoding ID's.
|
|
*/
|
|
static int pid = DEFAULT_PLATFORM_ID;
|
|
static int eid = DEFAULT_ENCODING_ID;
|
|
|
|
|
|
/*
|
|
* A flag indicating if a CMap was found or not.
|
|
*/
|
|
static FT_UShort nocmap;
|
|
|
|
int pct = 0; /* display ttc table if it is not zero. */
|
|
FT_Long max_char = 65535;
|
|
int pixel_size = 15;
|
|
FT_Long start_char = 0;
|
|
FT_Long limit_char;
|
|
FT_Long firstchar = 0;
|
|
FT_Long lastchar;
|
|
FT_Long ttc_index = -1;
|
|
int flg_all_ttc = 0;
|
|
short antialias = 1; /* smooth fonts with gray levels */
|
|
int oflag = 0;
|
|
char outfile[1024];
|
|
float between_chr = 0.0f;
|
|
float between_row = 0.0f;
|
|
int hv_resolution = 60;
|
|
int dump_glyphs = 0;
|
|
int digits_equally_wide = 1; /* Try to make digits equally wide */
|
|
int trimming = 0;
|
|
int trim_dp = 0; /* trim descent percent */
|
|
int trim_da = 0; /* trim descnet actual */
|
|
int trim_ap = 0; /* trim ascent precent */
|
|
int trim_aa = 0; /* trim ascent actual */
|
|
|
|
struct font_header_struct {
|
|
char header[4]; /* magic number and version bytes */
|
|
unsigned short maxwidth; /* max width in pixels */
|
|
unsigned short height; /* height in pixels */
|
|
unsigned short ascent; /* ascent (baseline) height */
|
|
unsigned short depth; /* depth 0=1-bit, 1=4-bit */
|
|
unsigned long firstchar; /* first character in font */
|
|
unsigned long defaultchar; /* default character in font */
|
|
unsigned long size; /* # characters in font */
|
|
unsigned long nbits; /* # bytes imagebits data in file */ /* = bits_size */
|
|
|
|
FT_Long noffset; /* # longs offset data in file */
|
|
FT_Long nwidth; /* # bytes width data in file */
|
|
};
|
|
|
|
struct font_struct {
|
|
struct font_header_struct header;
|
|
unsigned char *chars_data;
|
|
unsigned short *offset;
|
|
FT_Long *offset_long;
|
|
unsigned char *width;
|
|
};
|
|
|
|
struct ttc_table{
|
|
FT_Long ttc_count;
|
|
char **ttf_name;
|
|
};
|
|
|
|
/* exit the program with given message */
|
|
static void
|
|
panic( const char* message)
|
|
{
|
|
fprintf( stderr, "%s\n", message );
|
|
exit( 1 );
|
|
}
|
|
|
|
static void
|
|
arg_panic( const char* message, const char* arg )
|
|
{
|
|
fprintf( stderr, "%s: %s\n", message, arg );
|
|
exit( 1 );
|
|
}
|
|
|
|
static int writebyte(FILE *fp, unsigned char c)
|
|
{
|
|
return putc(c, fp) != EOF;
|
|
}
|
|
|
|
static int writeshort(FILE *fp, unsigned short s)
|
|
{
|
|
putc(s, fp);
|
|
return putc(s>>8, fp) != EOF;
|
|
}
|
|
|
|
static int writeint(FILE *fp, unsigned int l)
|
|
{
|
|
putc(l, fp);
|
|
putc(l>>8, fp);
|
|
putc(l>>16, fp);
|
|
return putc(l>>24, fp) != EOF;
|
|
}
|
|
|
|
static int writestr(FILE *fp, char *str, int count)
|
|
{
|
|
return (int)fwrite(str, 1, count, fp) == count;
|
|
}
|
|
|
|
/* print usage information */
|
|
void usage(void)
|
|
{
|
|
char help[] = {
|
|
"Usage: convttf [options] [input-files]\n"
|
|
" convttf [options] [-o output-file] [single-input-file]\n\n"
|
|
" Default output-file : <font-size>-<basename>.fnt.\n"
|
|
" When '-ta' or '-tc' is specified in command line,\n "
|
|
" default output-file is: \n"
|
|
" <font-size>-<internal postscript-name of input-file>.fnt.\n"
|
|
"Options:\n"
|
|
" -s N Start output at character encodings >= N\n"
|
|
" -l N Limit output to character encodings <= N\n"
|
|
" -p N Font size N in pixel (default N=15)\n"
|
|
" -c N Character separation in pixel.Insert space between lines.\n"
|
|
" -x Trim glyphs horizontally of nearly empty space\n"
|
|
" (to improve spacing on V's W's, etc.)\n"
|
|
" -X Set the horizontal and vertical resolution (default: 60)\n"
|
|
" -TA N Trim vertical ascent (N percent)\n"
|
|
" -TD N Trim vertical descent (N percent)\n"
|
|
" -Ta N Trim vertical ascent (N pixels)\n"
|
|
" -Td N Trim vertical descent (N pixels)\n"
|
|
" -r N Row separation in pixel.Insert space between characters\n"
|
|
" -d Debug: print converted glyph images\n"
|
|
" -tt Display the True Type Collection tables available in the font\n"
|
|
" -t N Index of true type collection. It must be start from 0.(default N=0).\n"
|
|
" -ta Convert all fonts in ttc (ignores outfile option)\n"
|
|
" -w Don't try to make digits (0-9) equally wide\n"
|
|
};
|
|
fprintf(stderr, "%s", help);
|
|
exit( 1 );
|
|
}
|
|
|
|
/* remove directory prefix and file suffix from full path*/
|
|
char *basename(char *path)
|
|
{
|
|
char *p, *b;
|
|
static char base[256];
|
|
|
|
/* remove prepended path and extension*/
|
|
b = path;
|
|
for (p=path; *p; ++p) {
|
|
#ifdef WIN32
|
|
if (*p == '/' || *p == '\\')
|
|
#else
|
|
if (*p == '/')
|
|
#endif
|
|
b = p + 1;
|
|
}
|
|
strcpy(base, b);
|
|
for (p=base; *p; ++p) {
|
|
if (*p == '.') {
|
|
*p = 0;
|
|
break;
|
|
}
|
|
}
|
|
return base;
|
|
}
|
|
|
|
|
|
void setcharmap(FT_Face face)
|
|
{
|
|
FT_Long i;
|
|
|
|
/*
|
|
* Get the requested cmap.
|
|
*/
|
|
for (i = 0; i < face->num_charmaps; i++) {
|
|
if (face->charmaps[i]->platform_id == pid &&
|
|
face->charmaps[i]->encoding_id == eid)
|
|
break;
|
|
}
|
|
|
|
if (i == face->num_charmaps && pid == 3 && eid == 1) {
|
|
/*
|
|
* Make a special case when this fails with pid == 3 and eid == 1.
|
|
* Change to eid == 0 and try again. This captures the two possible
|
|
* cases for MS fonts. Some other method should be used to cycle
|
|
* through all the alternatives later.
|
|
*/
|
|
for (i = 0; i < face->num_charmaps; i++) {
|
|
if (face->charmaps[i]->platform_id == pid &&
|
|
face->charmaps[i]->encoding_id == 0)
|
|
break;
|
|
}
|
|
if (i < face->num_charmaps) {
|
|
pid = 3;
|
|
eid = 1;
|
|
FT_Set_Charmap(face, face->charmaps[i]);
|
|
} else {
|
|
/*
|
|
* No CMAP was found.
|
|
*/
|
|
nocmap = 1;
|
|
pid = eid = -1;
|
|
}
|
|
} else {
|
|
FT_Set_Charmap(face, face->charmaps[i]);
|
|
nocmap = 0;
|
|
}
|
|
|
|
}
|
|
|
|
/*
|
|
* quote in otf2bdf.
|
|
* A generic routine to get a name from the OT name table. This routine
|
|
* always looks for English language names and checks three possibilities:
|
|
* 1. English names with the MS Unicode encoding ID.
|
|
* 2. English names with the MS unknown encoding ID.
|
|
* 3. English names with the Apple Unicode encoding ID.
|
|
*
|
|
* The particular name ID mut be provided (e.g. nameID = 0 for copyright
|
|
* string, nameID = 6 for Postscript name, nameID = 1 for typeface name.
|
|
*
|
|
* If the `dash_to_space' flag is non-zero, all dashes (-) in the name will be
|
|
* replaced with the character passed.
|
|
*
|
|
* Returns the number of bytes added.
|
|
*/
|
|
static int
|
|
otf_get_english_string(FT_Face face, int nameID, int dash_to_space,
|
|
char *name, int name_size)
|
|
{
|
|
|
|
int j, encid;
|
|
FT_UInt i, nrec;
|
|
FT_SfntName sfntName;
|
|
unsigned char *s = NULL;
|
|
unsigned short slen = 0;
|
|
|
|
nrec = FT_Get_Sfnt_Name_Count(face);
|
|
|
|
for (encid = 1, j = 0; j < 2; j++, encid--) {
|
|
/*
|
|
* Locate one of the MS English font names.
|
|
*/
|
|
for (i = 0; i < nrec; i++) {
|
|
FT_Get_Sfnt_Name(face, i, &sfntName);
|
|
if (sfntName.platform_id == 3 &&
|
|
sfntName.encoding_id == encid &&
|
|
sfntName.name_id == nameID &&
|
|
(sfntName.language_id == 0x0409 ||
|
|
sfntName.language_id == 0x0809 ||
|
|
sfntName.language_id == 0x0c09 ||
|
|
sfntName.language_id == 0x1009 ||
|
|
sfntName.language_id == 0x1409 ||
|
|
sfntName.language_id == 0x1809)) {
|
|
s = sfntName.string;
|
|
slen = sfntName.string_len;
|
|
break;
|
|
}
|
|
}
|
|
|
|
if (i < nrec) {
|
|
if (slen >> 1 >= name_size) {
|
|
fprintf(stderr, "warning: %s string longer than buffer."
|
|
"Truncating to %d bytes.\n", string_names[nameID], name_size);
|
|
slen = name_size << 1;
|
|
}
|
|
|
|
/*
|
|
* Found one of the MS English font names. The name is by
|
|
* definition encoded in Unicode, so copy every second byte into
|
|
* the `name' parameter, assuming there is enough space.
|
|
*/
|
|
for (i = 1; i < slen; i += 2) {
|
|
if (dash_to_space)
|
|
*name++ = (s[i] != '-') ? s[i] : ' ';
|
|
else if (s[i] == '\r' || s[i] == '\n') {
|
|
if (s[i] == '\r' && i + 2 < slen && s[i + 2] == '\n')
|
|
i += 2;
|
|
*name++ = ' ';
|
|
*name++ = ' ';
|
|
} else
|
|
*name++ = s[i];
|
|
}
|
|
*name = 0;
|
|
return (slen >> 1);
|
|
}
|
|
}
|
|
|
|
/*
|
|
* No MS English name found, attempt to find an Apple Unicode English
|
|
* name.
|
|
*/
|
|
for (i = 0; i < nrec; i++) {
|
|
FT_Get_Sfnt_Name(face, i, &sfntName);
|
|
if (sfntName.platform_id == 0 && sfntName.language_id == 0 &&
|
|
sfntName.name_id == nameID) {
|
|
s = sfntName.string;
|
|
slen = sfntName.string_len;
|
|
break;
|
|
}
|
|
}
|
|
|
|
if (i < nrec) {
|
|
if (slen >> 1 >= name_size) {
|
|
fprintf(stderr, "warning: %s string longer than buffer."
|
|
"Truncating to %d bytes.\n", string_names[nameID], name_size);
|
|
slen = name_size << 1;
|
|
}
|
|
|
|
/*
|
|
* Found the Apple Unicode English name. The name is by definition
|
|
* encoded in Unicode, so copy every second byte into the `name'
|
|
* parameter, assuming there is enough space.
|
|
*/
|
|
for (i = 1; i < slen; i += 2) {
|
|
if (dash_to_space)
|
|
*name++ = (s[i] != '-') ? s[i] : ' ';
|
|
else if (s[i] == '\r' || s[i] == '\n') {
|
|
if (s[i] == '\r' && i + 2 < slen && s[i + 2] == '\n')
|
|
i += 2;
|
|
*name++ = ' ';
|
|
*name++ = ' ';
|
|
} else
|
|
*name++ = s[i];
|
|
}
|
|
*name = 0;
|
|
return (slen >> 1);
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
|
|
int get_ttc_table(char *path, struct ttc_table *ttcname )
|
|
{
|
|
|
|
|
|
FT_Error err;
|
|
FT_Library library;
|
|
FT_Face face;
|
|
FT_Long i;
|
|
char xlfd[BUFSIZ];
|
|
|
|
/* init number of ttf in ttc */
|
|
ttcname->ttc_count = 0;
|
|
|
|
/* Initialize engine */
|
|
if ( ( err = FT_Init_FreeType( &library ) ) != 0 )
|
|
{
|
|
panic( "Error while initializing engine" );
|
|
return err;
|
|
}
|
|
|
|
|
|
/* Load face */
|
|
err = FT_New_Face( library, path, (FT_Long) 0, &face );
|
|
if ( err )
|
|
{
|
|
arg_panic( "Could not find/open font", path );
|
|
return err;
|
|
}
|
|
|
|
ttcname->ttc_count = face->num_faces;
|
|
ttcname->ttf_name = malloc( sizeof(char*) * ttcname->ttc_count);
|
|
|
|
for(i = 0; i < ttcname->ttc_count; i++)
|
|
{
|
|
err = FT_New_Face( library, path, i, &face );
|
|
if ( err == FT_Err_Cannot_Open_Stream )
|
|
arg_panic( "Could not find/open font", path );
|
|
otf_get_english_string(face, BDFOTF_POSTSCRIPT_STRING, 0, xlfd,
|
|
sizeof(xlfd));
|
|
ttcname->ttf_name[i] = malloc(sizeof(char) * (strlen(xlfd) + 1 ));
|
|
strcpy(ttcname->ttf_name[i], xlfd);
|
|
}
|
|
return 0;
|
|
}
|
|
|
|
void print_ttc_table(char* path)
|
|
{
|
|
struct ttc_table ttcname;
|
|
FT_Long i;
|
|
|
|
get_ttc_table(path, &ttcname);
|
|
printf("ttc header count = %ld \n\n", ttcname.ttc_count);
|
|
printf("Encoding tables available in the true type collection\n\n");
|
|
printf("INDEX\tPOSTSCRIPT NAME\n");
|
|
printf("-----------------------------------------------------\n");
|
|
for(i = 0; i < ttcname.ttc_count; i++)
|
|
{
|
|
printf("%ld\t%s\n", i, ttcname.ttf_name[i]);
|
|
}
|
|
for(i = 0; i < ttcname.ttc_count; i++)
|
|
{
|
|
free(ttcname.ttf_name[i]);
|
|
}
|
|
printf("\n\n");
|
|
free(ttcname.ttf_name);
|
|
|
|
return;
|
|
}
|
|
|
|
FT_Long getcharindex(FT_Face face, FT_Long code)
|
|
{
|
|
FT_Long idx;
|
|
if (nocmap) {
|
|
if (code >= face->num_glyphs)
|
|
idx = 0;
|
|
else
|
|
idx = code;
|
|
} else
|
|
idx = FT_Get_Char_Index( face, code);
|
|
|
|
if ( idx <= 0 || idx > face->num_glyphs)
|
|
return 0;
|
|
else
|
|
return idx;
|
|
}
|
|
|
|
void print_raw_glyph( FT_Face face)
|
|
{
|
|
int pixel,row,col,width;
|
|
width = face->glyph->metrics.width >> 6;
|
|
|
|
printf("\n---Raw-Glyph---\n");
|
|
for(row=0; row < face->glyph->metrics.height >> 6; row++)
|
|
{
|
|
printf("_");
|
|
for(col=0; col < width; col++)
|
|
{
|
|
pixel = *(face->glyph->bitmap.buffer+width*row+col)/26;
|
|
if ( pixel ) printf("%d",pixel); else printf(" ");
|
|
}
|
|
printf("_\n");
|
|
}
|
|
printf("----End-----\n");
|
|
}
|
|
|
|
int glyph_width( FT_Face face, FT_Long code, FT_Long digit_width )
|
|
{
|
|
int width;
|
|
|
|
if (code >= '0' && code <= '9' && digit_width)
|
|
{
|
|
width = digit_width;
|
|
}
|
|
else
|
|
{
|
|
int pitch, h_adv;
|
|
unsigned spacing = (unsigned)(between_chr * (1<<6));/* convert to fixed point */
|
|
|
|
pitch = ABS(face->glyph->bitmap.pitch);
|
|
h_adv = face->glyph->metrics.horiAdvance >> 6;
|
|
width = (face->glyph->metrics.width + spacing) >> 6;
|
|
|
|
if(pitch == 0) pitch = h_adv;
|
|
if(width < pitch) width = pitch;
|
|
}
|
|
|
|
return width;
|
|
}
|
|
|
|
FT_Long check_digit_width( FT_Face face )
|
|
{
|
|
FT_Long code;
|
|
FT_Long last_advance = -1;
|
|
|
|
for (code='0'; code <= '9'; ++code)
|
|
{
|
|
FT_Glyph_Metrics* metrics;
|
|
|
|
FT_Load_Char(face, code, FT_LOAD_RENDER | FT_LOAD_NO_BITMAP);
|
|
metrics = &face->glyph->metrics;
|
|
|
|
if ((last_advance != -1 && last_advance != metrics->horiAdvance) ||
|
|
metrics->horiBearingX < 0)
|
|
{
|
|
last_advance = 0;
|
|
break;
|
|
}
|
|
|
|
last_advance = metrics->horiAdvance;
|
|
}
|
|
|
|
return last_advance >> 6;
|
|
}
|
|
|
|
void trim_glyph( FT_GlyphSlot glyph, int *empty_first_col,
|
|
int *empty_last_col, int *width )
|
|
{
|
|
int row;
|
|
int stride = glyph->bitmap.pitch;
|
|
int end = stride-1;
|
|
int trim_left = 2, trim_right = 2;
|
|
|
|
const unsigned char limit = 64u;
|
|
const unsigned char *image = glyph->bitmap.buffer;
|
|
|
|
if (*width < 2)
|
|
return; /* nothing to do? */
|
|
|
|
for(row=0; row< glyph->metrics.height >> 6; row++)
|
|
{
|
|
const unsigned char *column = image+row*stride;
|
|
if (*column++ < limit && trim_left)
|
|
{
|
|
if (*column >= limit/2)
|
|
trim_left = 1;
|
|
}
|
|
else
|
|
trim_left = 0;
|
|
|
|
column = image+row*stride+end;
|
|
if (*column-- < limit && trim_right)
|
|
{
|
|
if (*column >= limit/2)
|
|
trim_right = 1;
|
|
}
|
|
else
|
|
trim_right = 0;
|
|
}
|
|
|
|
|
|
(*width) -= trim_left + trim_right;
|
|
if (*width < 0) *width = 0;
|
|
|
|
*empty_first_col = trim_left;
|
|
*empty_last_col = trim_right;
|
|
}
|
|
|
|
void convttf(char* path, char* destfile, FT_Long face_index)
|
|
{
|
|
FT_Error err;
|
|
FT_Library library;
|
|
FT_Face face;
|
|
int w,h;
|
|
int row,col;
|
|
int empty_first_col, empty_last_col;
|
|
FT_Long charindex;
|
|
FT_Long idx = 0;
|
|
FT_Long code;
|
|
FT_Long digit_width = 0;
|
|
float extra_space;
|
|
FT_Long char_count;
|
|
char use_long_offset;
|
|
int done = 0;
|
|
char char_name[1024];
|
|
int converted_char_count = 0;
|
|
int failed_char_count = 0;
|
|
|
|
int depth = 2;
|
|
unsigned char bit_shift = 1u << depth;
|
|
unsigned char pixel_per_byte = CHAR_BIT / bit_shift;
|
|
struct font_struct export_font;
|
|
char pad[] = {0,0,0,0};
|
|
int skip,i;
|
|
FILE *file;
|
|
|
|
/* Initialize engine */
|
|
if ( ( err = FT_Init_FreeType( &library ) ) != 0 )
|
|
panic( "Error while initializing engine" );
|
|
|
|
/* Load face */
|
|
err = FT_New_Face( library, path, (FT_Long) face_index, &face );
|
|
if ( err == FT_Err_Cannot_Open_Stream )
|
|
arg_panic( "Could not find/open font\n", path );
|
|
else if ( err )
|
|
arg_panic( "Error while opening font\n", path );
|
|
|
|
|
|
setcharmap( face );
|
|
/* Set font header data */
|
|
|
|
export_font.header.header[0] = 'R';
|
|
export_font.header.header[1] = 'B';
|
|
export_font.header.header[2] = '1';
|
|
export_font.header.header[3] = '2';
|
|
#if 0
|
|
export_font.header.height = 0;
|
|
export_font.header.ascent = 0;
|
|
#endif
|
|
|
|
extra_space = (float)(between_row-trim_aa-trim_da);
|
|
FT_Set_Char_Size( face, 0, pixel_size << 6, hv_resolution, hv_resolution );
|
|
export_font.header.ascent =
|
|
((face->size->metrics.ascender*(100-trim_ap)/100) >> 6) - trim_aa;
|
|
|
|
export_font.header.height =
|
|
(((face->size->metrics.ascender*(100-trim_ap)/100) -
|
|
(face->size->metrics.descender*(100-trim_dp)/100)) >> 6) + extra_space;
|
|
|
|
printf("\n");
|
|
printf("Please wait, converting %s:\n", path);
|
|
|
|
/* "face->num_glyphs" is NG.; */
|
|
if ( limit_char == 0 ) limit_char = max_char;
|
|
if ( limit_char > max_char ) limit_char = max_char;
|
|
|
|
char_count = 0;
|
|
|
|
|
|
|
|
export_font.header.maxwidth = 1;
|
|
export_font.header.depth = 1;
|
|
firstchar = limit_char;
|
|
lastchar = start_char;
|
|
|
|
if (digits_equally_wide)
|
|
digit_width = check_digit_width(face);
|
|
|
|
/* calculate memory usage */
|
|
for(code = start_char; code <= limit_char ; code++ )
|
|
{
|
|
charindex = getcharindex( face, code);
|
|
if ( !(charindex) ) continue;
|
|
err = FT_Load_Glyph( face, charindex,
|
|
(FT_LOAD_RENDER | FT_LOAD_NO_BITMAP) );
|
|
if ( err ) continue;
|
|
|
|
w = glyph_width( face, code, digit_width );
|
|
if (w == 0) continue;
|
|
empty_first_col = empty_last_col = 0;
|
|
if(trimming)
|
|
trim_glyph( face->glyph, &empty_first_col, &empty_last_col, &w);
|
|
|
|
if (export_font.header.maxwidth < w)
|
|
export_font.header.maxwidth = w;
|
|
|
|
|
|
char_count++;
|
|
idx += (w*export_font.header.height + pixel_per_byte - 1)/pixel_per_byte;
|
|
|
|
if (code >= lastchar)
|
|
lastchar = code;
|
|
|
|
if (code <= firstchar)
|
|
firstchar = code;
|
|
}
|
|
export_font.header.defaultchar = firstchar;
|
|
export_font.header.firstchar = firstchar;
|
|
export_font.header.size = lastchar - firstchar + 1;
|
|
export_font.header.nbits = idx;
|
|
export_font.header.noffset = export_font.header.size;
|
|
export_font.header.nwidth = export_font.header.size;
|
|
|
|
/* check if we need to use long offsets */
|
|
use_long_offset = (export_font.header.nbits >= 0xFFDB );
|
|
|
|
/* allocate memory */
|
|
export_font.offset = NULL;
|
|
export_font.offset_long = NULL;
|
|
if (use_long_offset)
|
|
export_font.offset_long =
|
|
malloc( sizeof(FT_Long)* export_font.header.noffset );
|
|
else
|
|
export_font.offset =
|
|
malloc( sizeof(unsigned short)* export_font.header.noffset );
|
|
|
|
export_font.width =
|
|
malloc( sizeof(unsigned char) * export_font.header.nwidth );
|
|
export_font.chars_data =
|
|
malloc( sizeof(unsigned char) * export_font.header.nbits );
|
|
|
|
/* for now we use the full height for each character */
|
|
h = export_font.header.height;
|
|
|
|
idx = 0;
|
|
|
|
for( code = firstchar; code <= lastchar; code++ )
|
|
{
|
|
FT_GlyphSlot slot;
|
|
FT_Bitmap* source;
|
|
unsigned char* src;
|
|
unsigned char* tmpbuf;
|
|
int start_y;
|
|
|
|
int glyph_height;
|
|
int stride;
|
|
unsigned char* buf;
|
|
unsigned char* endbuf;
|
|
|
|
/* insert empty pixels on the left */
|
|
int col_off;
|
|
int numbits;
|
|
unsigned int field;
|
|
|
|
/* Get gylph index from the char and render it */
|
|
charindex = getcharindex( face, code);
|
|
if ( !charindex )
|
|
{
|
|
if ( use_long_offset )
|
|
export_font.offset_long[code - firstchar] = export_font.offset_long[0];
|
|
else
|
|
export_font.offset[code - firstchar] = export_font.offset[0];
|
|
export_font.width[code - firstchar] = export_font.width[0];
|
|
continue;
|
|
}
|
|
|
|
err = FT_Load_Glyph( face, charindex ,
|
|
(FT_LOAD_RENDER | FT_LOAD_NO_BITMAP) );
|
|
if ( err ) {
|
|
continue;
|
|
}
|
|
if FT_HAS_GLYPH_NAMES( face )
|
|
FT_Get_Glyph_Name( face, charindex, char_name, 16);
|
|
else
|
|
char_name[0] = '\0';
|
|
|
|
slot = face->glyph;
|
|
source = &slot->bitmap;
|
|
#if 0
|
|
print_raw_glyph( face );
|
|
#endif
|
|
w = glyph_width( face, code, digit_width );
|
|
if (w == 0) continue;
|
|
empty_first_col = empty_last_col = 0;
|
|
|
|
if(trimming)
|
|
trim_glyph( face->glyph, &empty_first_col, &empty_last_col, &w );
|
|
|
|
if ( use_long_offset )
|
|
export_font.offset_long[code - firstchar] = idx;
|
|
else
|
|
export_font.offset[code - firstchar] = idx;
|
|
|
|
export_font.width[code - firstchar] = w;
|
|
|
|
/* copy the glyph bitmap to a full sized glyph bitmap */
|
|
src = source->buffer;
|
|
tmpbuf = malloc(sizeof(unsigned char) * w * h);
|
|
memset(tmpbuf, 0xff, w*h);
|
|
start_y = export_font.header.ascent - slot->bitmap_top;
|
|
|
|
glyph_height = source->rows;
|
|
stride = source->pitch;
|
|
buf = tmpbuf;
|
|
endbuf = tmpbuf + w*h;
|
|
|
|
err = 0;
|
|
/* insert empty pixels on the left */
|
|
col_off = w - stride;
|
|
if (col_off > 1) col_off /= 2;
|
|
if (col_off < 0) col_off = 0;
|
|
|
|
for(row=0; row < glyph_height; row++)
|
|
{
|
|
if(row+start_y < 0 || row+start_y >= h)
|
|
continue;
|
|
for(col = empty_first_col; col < stride; col++)
|
|
{
|
|
unsigned char *tsrc, *dst;
|
|
dst = buf + (w*(start_y+row)) + col + col_off;
|
|
tsrc = src + stride*row + col;
|
|
if (dst < endbuf && dst >= tmpbuf)
|
|
*dst = 0xff - *tsrc;
|
|
else {
|
|
err = 1;
|
|
printf("Error! row: %3d col: %3d\n", row, col);
|
|
}
|
|
}
|
|
}
|
|
if(err) print_raw_glyph(face);
|
|
|
|
buf = tmpbuf;
|
|
field = 0;
|
|
numbits = pixel_per_byte;
|
|
|
|
for(row=0; row < h; row++)
|
|
{
|
|
for(col=0; col < w; col++)
|
|
{
|
|
unsigned int src2 = *buf++;
|
|
unsigned int cur_col = (src2 + 8) / 17;
|
|
field |= (cur_col << (bit_shift*(pixel_per_byte-numbits)));
|
|
|
|
if (--numbits == 0)
|
|
{
|
|
export_font.chars_data[idx++] = (unsigned char)field;
|
|
numbits = pixel_per_byte;
|
|
field = 0;
|
|
}
|
|
}
|
|
}
|
|
|
|
/* Pad last byte */
|
|
if (numbits != pixel_per_byte)
|
|
{
|
|
export_font.chars_data[idx++] = (unsigned char)field;
|
|
}
|
|
|
|
if( dump_glyphs )
|
|
{
|
|
/* debug: dump char */
|
|
unsigned char bit_max = (1 << bit_shift) - 1;
|
|
printf("\n---Converted Glyph Dump---\n");
|
|
|
|
if ( code > 32 && code < 255 ) {
|
|
unsigned char current_data;
|
|
unsigned char font_bits;
|
|
|
|
row = h;
|
|
if(use_long_offset)
|
|
buf = &(export_font.chars_data[export_font.offset_long[
|
|
code - firstchar]]);
|
|
else
|
|
buf = &(export_font.chars_data[export_font.offset[
|
|
code - firstchar]]);
|
|
numbits = pixel_per_byte;
|
|
current_data = *buf;
|
|
do
|
|
{
|
|
col = w;
|
|
printf("-");
|
|
do
|
|
{
|
|
font_bits = current_data & bit_max;
|
|
if (font_bits==bit_max)
|
|
printf(" ");
|
|
else
|
|
{
|
|
if(font_bits > bit_max/2)
|
|
printf(".");
|
|
else
|
|
printf("@");
|
|
}
|
|
if (--numbits == 0)
|
|
{
|
|
current_data = *(++buf);
|
|
numbits = pixel_per_byte;
|
|
}
|
|
else
|
|
{
|
|
current_data >>= bit_shift;
|
|
}
|
|
} while (--col);
|
|
printf("-\n");
|
|
} while (--row);
|
|
}
|
|
buf = NULL;
|
|
printf("---End Glyph Dump---\n");
|
|
}
|
|
|
|
free(tmpbuf);
|
|
converted_char_count++;
|
|
done = (100*(converted_char_count))/char_count;
|
|
printf("Converted %s %d (%d%%)\e[K\r",
|
|
char_name,converted_char_count,done); fflush(stdout);
|
|
}
|
|
|
|
file = fopen(destfile, "wb");
|
|
printf("Writing %s\n", destfile);
|
|
|
|
/* font info */
|
|
writestr(file, VERSION, 4);
|
|
writeshort(file, export_font.header.maxwidth);
|
|
writeshort(file, export_font.header.height);
|
|
writeshort(file, export_font.header.ascent);
|
|
writeshort(file, export_font.header.depth);
|
|
writeint(file, export_font.header.firstchar);
|
|
writeint(file, export_font.header.defaultchar);
|
|
writeint(file, export_font.header.size);
|
|
writeint(file, export_font.header.nbits);
|
|
writeint(file, export_font.header.noffset);
|
|
writeint(file, export_font.header.nwidth);
|
|
|
|
fwrite( (char*)export_font.chars_data, 1,
|
|
export_font.header.nbits, file);
|
|
free(export_font.chars_data);
|
|
|
|
|
|
|
|
if ( use_long_offset )
|
|
{
|
|
skip = ((export_font.header.nbits + 3) & ~3) -
|
|
export_font.header.nbits;
|
|
fwrite(pad, 1, skip, file); /* pad */
|
|
for(i = 0; i < export_font.header.noffset; i++)
|
|
writeint(file, export_font.offset_long[i]);
|
|
}
|
|
else
|
|
{
|
|
skip = ((export_font.header.nbits + 1) & ~1) -
|
|
export_font.header.nbits;
|
|
fwrite(pad, 1, skip, file); /* pad */
|
|
for(i = 0; i < export_font.header.noffset; i++)
|
|
writeshort(file, export_font.offset[i]);
|
|
}
|
|
|
|
for(i = 0; i < export_font.header.nwidth; i++)
|
|
writebyte(file, export_font.width[i]);
|
|
free(export_font.width);
|
|
|
|
if ( use_long_offset )
|
|
free(export_font.offset_long);
|
|
else
|
|
free(export_font.offset);
|
|
|
|
fclose(file);
|
|
FT_Done_Face( face );
|
|
FT_Done_FreeType( library );
|
|
printf("done (converted %d glyphs, %d errors).\e[K\n\n",
|
|
converted_char_count, failed_char_count);
|
|
|
|
}
|
|
|
|
void convttc(char* path)
|
|
{
|
|
struct ttc_table ttcname;
|
|
FT_Long i;
|
|
|
|
get_ttc_table(path, &ttcname);
|
|
|
|
if (ttcname.ttc_count == 0)
|
|
{
|
|
printf("This file is a not true type font.\n");
|
|
return;
|
|
}
|
|
|
|
/* default */
|
|
if (!flg_all_ttc && ttc_index == -1)
|
|
{
|
|
if (!oflag)
|
|
{ /* generate filename */
|
|
snprintf(outfile, sizeof(outfile),
|
|
"%d-%s.fnt", pixel_size, basename(path));
|
|
}
|
|
convttf(path, outfile, (FT_Long) 0);
|
|
}
|
|
|
|
/* set face_index of ttc */
|
|
else if (!flg_all_ttc)
|
|
{
|
|
print_ttc_table(path);
|
|
if ( !oflag )
|
|
{
|
|
if (ttc_index >= 0 &&
|
|
ttc_index < ttcname.ttc_count)
|
|
{
|
|
if (strcmp(ttcname.ttf_name[ttc_index], "") != 0)
|
|
{
|
|
snprintf(outfile, sizeof(outfile), "%d-%s.fnt",
|
|
pixel_size, ttcname.ttf_name[ttc_index]);
|
|
}
|
|
else
|
|
{
|
|
snprintf(outfile, sizeof(outfile), "%d-%s-%ld.fnt",
|
|
pixel_size, basename(path), ttc_index);
|
|
}
|
|
}
|
|
else
|
|
{
|
|
printf("illegal face index of ttc.\n");
|
|
}
|
|
}
|
|
convttf(path, outfile, ttc_index);
|
|
}
|
|
else { /* convert all fonts */
|
|
print_ttc_table(path);
|
|
for(i = 0; i < ttcname.ttc_count; i++)
|
|
{
|
|
snprintf(outfile, sizeof(outfile), "%d-%s.fnt",
|
|
pixel_size, ttcname.ttf_name[i]);
|
|
convttf(path, outfile, i);
|
|
}
|
|
}
|
|
|
|
for(i = 0; i < ttcname.ttc_count; i++)
|
|
{
|
|
free(ttcname.ttf_name[i]);
|
|
}
|
|
free(ttcname.ttf_name);
|
|
}
|
|
|
|
|
|
|
|
/* parse command line options*/
|
|
void getopts(int *pac, char ***pav)
|
|
{
|
|
char *p;
|
|
char **av;
|
|
int ac;
|
|
ac = *pac;
|
|
av = *pav;
|
|
|
|
limit_char = max_char;
|
|
start_char = 0;
|
|
|
|
while (ac > 0 && av[0][0] == '-') {
|
|
p = &av[0][1];
|
|
while( *p)
|
|
switch(*p++) {
|
|
case 'h':case 'H':
|
|
usage();
|
|
break;
|
|
case ' ': /* multiple -args on av[]*/
|
|
while( *p && *p == ' ')
|
|
p++;
|
|
if( *p++ != '-') /* next option must have dash*/
|
|
p = "";
|
|
break; /* proceed to next option*/
|
|
case 'o': /* set output file*/
|
|
oflag = 1;
|
|
if (*p) {
|
|
strcpy(outfile, p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
strcpy(outfile, av[0]);
|
|
}
|
|
break;
|
|
case 'l': /* set encoding limit*/
|
|
if (*p) {
|
|
limit_char = atoi(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
limit_char = atoi(av[0]);
|
|
}
|
|
break;
|
|
case 's': /* set encoding start*/
|
|
if (*p) {
|
|
start_char = atol(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
start_char = atol(av[0]);
|
|
}
|
|
break;
|
|
case 'p': /* set pixel size*/
|
|
if (*p) {
|
|
pixel_size = atoi(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
pixel_size = atoi(av[0]);
|
|
}
|
|
break;
|
|
case 'c': /* set spaece between characters */
|
|
{
|
|
if (*p) {
|
|
between_chr = atof(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
between_chr = atof(av[0]);
|
|
}
|
|
break;
|
|
}
|
|
case 'd':
|
|
dump_glyphs = 1;
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
break;
|
|
case 'x':
|
|
trimming = 1;
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
break;
|
|
case 'X':
|
|
if (*p) {
|
|
hv_resolution = atoi(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
hv_resolution = atoi(av[0]);
|
|
}
|
|
break;
|
|
case 'r':
|
|
if (*p) {
|
|
between_row = atof(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
between_row = atof(av[0]);
|
|
}
|
|
break;
|
|
case 'T':
|
|
if(*p == 'A') {
|
|
if(*(++p)) {
|
|
trim_ap = atoi(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
trim_ap = atoi(av[0]);
|
|
}
|
|
break;
|
|
}
|
|
if(*p == 'D') {
|
|
if(*(++p)) {
|
|
trim_dp = atoi(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
trim_dp = atoi(av[0]);
|
|
}
|
|
break;
|
|
}
|
|
if(*p == 'a') {
|
|
if(*(++p)) {
|
|
trim_aa = atoi(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
trim_aa = atoi(av[0]);
|
|
}
|
|
break;
|
|
}
|
|
if(*p == 'd') {
|
|
if(*(++p)) {
|
|
trim_da = atoi(p);
|
|
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
trim_da = atoi(av[0]);
|
|
}
|
|
break;
|
|
}
|
|
fprintf(stderr, "Unknown option ignored: %s\n", p-1);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
break;
|
|
case 't': /* display ttc table */
|
|
if (*p == 't') {
|
|
pct = 1;
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
|
|
else if (*p == 'a') {
|
|
flg_all_ttc = 1;
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
|
|
else if (*p) {
|
|
ttc_index = atoi(p);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
else {
|
|
av++; ac--;
|
|
if (ac > 0)
|
|
ttc_index = atoi(av[0]);
|
|
}
|
|
break;
|
|
case 'w': /* Don't try to make digits equally wide */
|
|
digits_equally_wide = 0;
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
break;
|
|
|
|
default:
|
|
fprintf(stderr, "Unknown option ignored: %s\n", p-1);
|
|
while (*p && *p != ' ')
|
|
p++;
|
|
}
|
|
++av; --ac;
|
|
}
|
|
*pac = ac;
|
|
*pav = av;
|
|
}
|
|
|
|
|
|
int main(int ac, char **av)
|
|
{
|
|
int ret = 0;
|
|
|
|
++av; --ac; /* skip av[0]*/
|
|
|
|
getopts(&ac, &av); /* read command line options*/
|
|
|
|
if (ac < 1)
|
|
{
|
|
usage();
|
|
}
|
|
if (oflag)
|
|
{
|
|
if (ac > 1)
|
|
{
|
|
usage();
|
|
}
|
|
}
|
|
|
|
if (limit_char < start_char)
|
|
{
|
|
usage();
|
|
exit(0);
|
|
}
|
|
|
|
while (pct && ac > 0)
|
|
{
|
|
print_ttc_table(av[0]);
|
|
++av; --ac;
|
|
exit(0);
|
|
}
|
|
|
|
while (ac > 0)
|
|
{
|
|
convttc(av[0]);
|
|
++av; --ac;
|
|
}
|
|
|
|
exit(ret);
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
* Trie node structure.
|
|
*/
|
|
typedef struct {
|
|
unsigned short key; /* Key value. */
|
|
unsigned short val; /* Data for the key. */
|
|
unsigned long sibs; /* Offset of siblings from trie beginning. */
|
|
unsigned long kids; /* Offset of children from trie beginning. */
|
|
} node_t;
|
|
|
|
/*
|
|
* The trie used for remapping codes.
|
|
*/
|
|
static node_t *nodes;
|
|
static unsigned long nodes_used = 0;
|
|
|
|
int
|
|
otf2bdf_remap(unsigned short *code)
|
|
{
|
|
unsigned long i, n, t;
|
|
unsigned short c, codes[2];
|
|
|
|
/*
|
|
* If no mapping table was loaded, then simply return the code.
|
|
*/
|
|
if (nodes_used == 0)
|
|
return 1;
|
|
|
|
c = *code;
|
|
codes[0] = (c >> 8) & 0xff;
|
|
codes[1] = c & 0xff;
|
|
|
|
for (i = n = 0; i < 2; i++) {
|
|
t = nodes[n].kids;
|
|
if (t == 0)
|
|
return 0;
|
|
for (; nodes[t].sibs && nodes[t].key != codes[i]; t = nodes[t].sibs);
|
|
if (nodes[t].key != codes[i])
|
|
return 0;
|
|
n = t;
|
|
}
|
|
|
|
*code = nodes[n].val;
|
|
return 1;
|
|
}
|