/* Copyright 1990, 1998 The Open Group Permission to use, copy, modify, distribute, and sell this software and its documentation for any purpose is hereby granted without fee, provided that the above copyright notice appear in all copies and that both that copyright notice and this permission notice appear in supporting documentation. The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE OPEN GROUP BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. Except as contained in this notice, the name of The Open Group shall not be used in advertising or otherwise to promote the sale, use or other dealings in this Software without prior written authorization from The Open Group. */ /* * Author: Keith Packard, MIT X Consortium */ #ifdef HAVE_CONFIG_H #include #endif #include "libxfontint.h" #include #include #include #include #include /* for MAXFONTNAMELEN */ #include #include #include #include #include #include /* for sprintf() */ #include "src/util/replace.h" static char * GetInt(char *ptr, int *val) { if (*ptr == '*') { *val = -1; ptr++; } else for (*val = 0; *ptr >= '0' && *ptr <= '9';) *val = *val * 10 + *ptr++ - '0'; if (*ptr == '-') return ptr; return (char *) 0; } #define minchar(p) ((p).min_char_low + ((p).min_char_high << 8)) #define maxchar(p) ((p).max_char_low + ((p).max_char_high << 8)) static struct lconv *locale = 0; static const char *radix = ".", *plus = "+", *minus = "-"; static char * readreal(char *ptr, double *result) { char buffer[80], *p1, *p2; /* Figure out what symbols apply in this locale */ if (!locale) { locale = localeconv(); if (locale->decimal_point && *locale->decimal_point) radix = locale->decimal_point; if (locale->positive_sign && *locale->positive_sign) plus = locale->positive_sign; if (locale->negative_sign && *locale->negative_sign) minus = locale->negative_sign; } /* Copy the first 80 chars of ptr into our local buffer, changing symbols as needed. */ for (p1 = ptr, p2 = buffer; *p1 && (p2 - buffer) < sizeof(buffer) - 1; p1++, p2++) { switch(*p1) { case '~': *p2 = *minus; break; case '+': *p2 = *plus; break; case '.': *p2 = *radix; break; default: *p2 = *p1; } } *p2 = 0; /* Now we have something that strtod() can interpret... do it. */ *result = strtod(buffer, &p1); /* Return NULL if failure, pointer past number if success */ return (p1 == buffer) ? (char *)0 : (ptr + (p1 - buffer)); } #define XLFD_DOUBLE_TO_TEXT_BUF_SIZE 80 static char * xlfd_double_to_text(double value, char *buffer, int space_required) { register char *p1; int ndigits, exponent; const size_t buflen = XLFD_DOUBLE_TO_TEXT_BUF_SIZE; if (!locale) { locale = localeconv(); if (locale->decimal_point && *locale->decimal_point) radix = locale->decimal_point; if (locale->positive_sign && *locale->positive_sign) plus = locale->positive_sign; if (locale->negative_sign && *locale->negative_sign) minus = locale->negative_sign; } if (space_required) *buffer++ = ' '; /* Render the number using printf's idea of formatting */ snprintf(buffer, buflen, "%.*le", XLFD_NDIGITS, value); /* Find and read the exponent value */ for (p1 = buffer + strlen(buffer); *p1-- != 'e' && p1[1] != 'E';); exponent = atoi(p1 + 2); if (value == 0.0) exponent = 0; /* Figure out how many digits are significant */ while (p1 >= buffer && (!isdigit((unsigned char)*p1) || *p1 == '0')) p1--; ndigits = 0; while (p1 >= buffer) if (isdigit((unsigned char)*p1--)) ndigits++; /* Figure out notation to use */ if (exponent >= XLFD_NDIGITS || ndigits - exponent > XLFD_NDIGITS + 1) { /* Scientific */ snprintf(buffer, buflen, "%.*le", ndigits - 1, value); } else { /* Fixed */ ndigits -= exponent + 1; if (ndigits < 0) ndigits = 0; snprintf(buffer, buflen, "%.*lf", ndigits, value); if (exponent < 0) { p1 = buffer; while (*p1 && *p1 != '0') p1++; while (*p1++) p1[-1] = *p1; } } /* Last step, convert the locale-specific sign and radix characters to our own. */ for (p1 = buffer; *p1; p1++) { if (*p1 == *minus) *p1 = '~'; else if (*p1 == *plus) *p1 = '+'; else if (*p1 == *radix) *p1 = '.'; } return buffer - space_required; } double xlfd_round_double(double x) { /* Utility for XLFD users to round numbers to XLFD_NDIGITS significant digits. How do you round to n significant digits on a binary machine? */ #ifdef HAVE_FLOAT_H #include /* if we have IEEE 754 fp, we can round to binary digits... */ #if (FLT_RADIX == 2) && (DBL_DIG == 15) && (DBL_MANT_DIG == 53) #ifndef M_LN2 #define M_LN2 0.69314718055994530942 #endif #ifndef M_LN10 #define M_LN10 2.30258509299404568402 #endif /* convert # of decimal digits to # of binary digits */ #define XLFD_NDIGITS_2 ((int)(XLFD_NDIGITS * M_LN10 / M_LN2 + 0.5)) union conv_d { double d; unsigned char b[8]; } d; int i,j,k,d_exp; if (x == 0) return x; /* do minor sanity check for IEEE 754 fp and correct byte order */ d.d = 1.0; if (sizeof(double) == 8 && d.b[7] == 0x3f && d.b[6] == 0xf0) { /* * this code will round IEEE 754 double to XLFD_NDIGITS_2 binary digits */ d.d = x; d_exp = (d.b[7] << 4) | (d.b[6] >> 4); i = (DBL_MANT_DIG-XLFD_NDIGITS_2) >> 3; j = 1 << ((DBL_MANT_DIG-XLFD_NDIGITS_2) & 0x07); for (; i<7; i++) { k = d.b[i] + j; d.b[i] = k; if (k & 0x100) j = 1; else break; } if ((i==7) && ((d.b[6] & 0xf0) != ((d_exp<<4) & 0xf0))) { /* mantissa overflow: increment exponent */ d_exp = (d_exp & 0x800 ) | ((d_exp & 0x7ff) + 1); d.b[7] = d_exp >> 4; d.b[6] = (d.b[6] & 0x0f) | (d_exp << 4); } i = (DBL_MANT_DIG-XLFD_NDIGITS_2) >> 3; j = 1 << ((DBL_MANT_DIG-XLFD_NDIGITS_2) & 0x07); d.b[i] &= ~(j-1); for (;--i>=0;) d.b[i] = 0; return d.d; } else #endif #endif /* HAVE_FLOAT_H */ { /* * If not IEEE 754: Let printf() do it for you. */ char buffer[40]; snprintf(buffer, sizeof(buffer), "%.*lg", XLFD_NDIGITS, x); return atof(buffer); } } static char * GetMatrix(char *ptr, FontScalablePtr vals, int which) { double *matrix; if (which == PIXELSIZE_MASK) matrix = vals->pixel_matrix; else if (which == POINTSIZE_MASK) matrix = vals->point_matrix; else return (char *)0; while (isspace((unsigned char)*ptr)) ptr++; if (*ptr == '[') { /* This is a matrix containing real numbers. It would be nice to use strtod() or sscanf() to read the numbers, but those don't handle '~' for minus and we cannot force them to use a "." for the radix. We'll have to do the hard work ourselves (in readreal()). */ if ((ptr = readreal(++ptr, matrix + 0)) && (ptr = readreal(ptr, matrix + 1)) && (ptr = readreal(ptr, matrix + 2)) && (ptr = readreal(ptr, matrix + 3))) { while (isspace((unsigned char)*ptr)) ptr++; if (*ptr != ']') ptr = (char *)0; else { ptr++; while (isspace((unsigned char)*ptr)) ptr++; if (*ptr == '-') { if (which == POINTSIZE_MASK) vals->values_supplied |= POINTSIZE_ARRAY; else vals->values_supplied |= PIXELSIZE_ARRAY; } else ptr = (char *)0; } } } else { int value; if ((ptr = GetInt(ptr, &value))) { vals->values_supplied &= ~which; if (value > 0) { matrix[3] = (double)value; if (which == POINTSIZE_MASK) { matrix[3] /= 10.0; vals->values_supplied |= POINTSIZE_SCALAR; } else vals->values_supplied |= PIXELSIZE_SCALAR; /* If we're concocting the pixelsize array from a scalar, we will need to normalize element 0 for the pixel shape. This is done in FontFileCompleteXLFD(). */ matrix[0] = matrix[3]; matrix[1] = matrix[2] = 0.0; } else if (value < 0) { if (which == POINTSIZE_MASK) vals->values_supplied |= POINTSIZE_WILDCARD; else vals->values_supplied |= PIXELSIZE_WILDCARD; } } } return ptr; } static void append_ranges(char *fname, size_t fnamelen, int nranges, fsRange *ranges) { if (nranges) { int i; strlcat(fname, "[", fnamelen); for (i = 0; i < nranges && strlen(fname) < 1010; i++) { size_t curlen; if (i) strlcat(fname, " ", fnamelen); curlen = strlen(fname); snprintf(fname + curlen, fnamelen - curlen, "%d", minchar(ranges[i])); if (ranges[i].min_char_low == ranges[i].max_char_low && ranges[i].min_char_high == ranges[i].max_char_high) continue; snprintf(fname + curlen, fnamelen - curlen, "_%d", maxchar(ranges[i])); } strlcat(fname, "]", fnamelen); } } Bool FontParseXLFDName(char *fname, FontScalablePtr vals, int subst) { register char *ptr; register char *ptr1, *ptr2, *ptr3, *ptr4; register char *ptr5; FontScalableRec tmpvals; char replaceChar = '0'; char tmpBuf[1024]; size_t tlen; size_t fnamelen = MAXFONTNAMELEN; /* assumed for now */ int spacingLen; int l; char *p; bzero(&tmpvals, sizeof(tmpvals)); if (subst != FONT_XLFD_REPLACE_VALUE) *vals = tmpvals; if (!(*(ptr = fname) == '-' || (*ptr++ == '*' && *ptr == '-')) || /* fndry */ !(ptr = strchr(ptr + 1, '-')) || /* family_name */ !(ptr1 = ptr = strchr(ptr + 1, '-')) || /* weight_name */ !(ptr = strchr(ptr + 1, '-')) || /* slant */ !(ptr = strchr(ptr + 1, '-')) || /* setwidth_name */ !(ptr = strchr(ptr + 1, '-')) || /* add_style_name */ !(ptr = strchr(ptr + 1, '-')) || /* pixel_size */ !(ptr = GetMatrix(ptr + 1, &tmpvals, PIXELSIZE_MASK)) || !(ptr2 = ptr = GetMatrix(ptr + 1, &tmpvals, POINTSIZE_MASK)) || !(ptr = GetInt(ptr + 1, &tmpvals.x)) || /* resolution_x */ !(ptr3 = ptr = GetInt(ptr + 1, &tmpvals.y)) || /* resolution_y */ !(ptr4 = ptr = strchr(ptr + 1, '-')) || /* spacing */ !(ptr5 = ptr = GetInt(ptr + 1, &tmpvals.width)) || /* average_width */ !(ptr = strchr(ptr + 1, '-')) || /* charset_registry */ strchr(ptr + 1, '-'))/* charset_encoding */ return FALSE; /* Lop off HP charset subsetting enhancement. Interpreting this field requires allocating some space in which to return the results. So, to prevent memory leaks, this procedure will simply lop off and ignore charset subsetting, and initialize the relevant vals fields to zero. It's up to the caller to make its own call to FontParseRanges() if it's interested in the charset subsetting. */ if (subst != FONT_XLFD_REPLACE_NONE && (p = strchr(strrchr(fname, '-'), '['))) { tmpvals.values_supplied |= CHARSUBSET_SPECIFIED; *p = '\0'; } /* Fill in deprecated fields for the benefit of rasterizers that care about them. */ tmpvals.pixel = (tmpvals.pixel_matrix[3] >= 0) ? (int)(tmpvals.pixel_matrix[3] + .5) : (int)(tmpvals.pixel_matrix[3] - .5); tmpvals.point = (tmpvals.point_matrix[3] >= 0) ? (int)(tmpvals.point_matrix[3] * 10 + .5) : (int)(tmpvals.point_matrix[3] * 10 - .5); spacingLen = ptr4 - ptr3 + 1; switch (subst) { case FONT_XLFD_REPLACE_NONE: *vals = tmpvals; break; case FONT_XLFD_REPLACE_STAR: replaceChar = '*'; /* FALLTHROUGH */ case FONT_XLFD_REPLACE_ZERO: strlcpy(tmpBuf, ptr2, sizeof(tmpBuf)); ptr5 = tmpBuf + (ptr5 - ptr2); ptr3 = tmpBuf + (ptr3 - ptr2); ptr2 = tmpBuf; ptr = ptr1 + 1; ptr = strchr(ptr, '-') + 1; /* skip weight */ ptr = strchr(ptr, '-') + 1; /* skip slant */ ptr = strchr(ptr, '-') + 1; /* skip setwidth_name */ ptr = strchr(ptr, '-') + 1; /* skip add_style_name */ if ((ptr - fname) + spacingLen + strlen(ptr5) + 10 >= (unsigned)1024) return FALSE; *ptr++ = replaceChar; *ptr++ = '-'; *ptr++ = replaceChar; *ptr++ = '-'; *ptr++ = '*'; *ptr++ = '-'; *ptr++ = '*'; if (spacingLen > 2) { memmove(ptr, ptr3, spacingLen); ptr += spacingLen; } else { *ptr++ = '-'; *ptr++ = '*'; *ptr++ = '-'; } *ptr++ = replaceChar; strlcpy(ptr, ptr5, fnamelen - (ptr - fname)); *vals = tmpvals; break; case FONT_XLFD_REPLACE_VALUE: if (vals->values_supplied & PIXELSIZE_MASK) { tmpvals.values_supplied = (tmpvals.values_supplied & ~PIXELSIZE_MASK) | (vals->values_supplied & PIXELSIZE_MASK); tmpvals.pixel_matrix[0] = vals->pixel_matrix[0]; tmpvals.pixel_matrix[1] = vals->pixel_matrix[1]; tmpvals.pixel_matrix[2] = vals->pixel_matrix[2]; tmpvals.pixel_matrix[3] = vals->pixel_matrix[3]; } if (vals->values_supplied & POINTSIZE_MASK) { tmpvals.values_supplied = (tmpvals.values_supplied & ~POINTSIZE_MASK) | (vals->values_supplied & POINTSIZE_MASK); tmpvals.point_matrix[0] = vals->point_matrix[0]; tmpvals.point_matrix[1] = vals->point_matrix[1]; tmpvals.point_matrix[2] = vals->point_matrix[2]; tmpvals.point_matrix[3] = vals->point_matrix[3]; } if (vals->x >= 0) tmpvals.x = vals->x; if (vals->y >= 0) tmpvals.y = vals->y; if (vals->width >= 0) tmpvals.width = vals->width; else if (vals->width < -1) /* overload: -1 means wildcard */ tmpvals.width = -vals->width; p = ptr1 + 1; /* weight field */ l = strchr(p, '-') - p; snprintf(tmpBuf, sizeof(tmpBuf), "%*.*s", l, l, p); p += l + 1; /* slant field */ l = strchr(p, '-') - p; tlen = strlen(tmpBuf); snprintf(tmpBuf + tlen, sizeof(tmpBuf) - tlen, "-%*.*s", l, l, p); p += l + 1; /* setwidth_name */ l = strchr(p, '-') - p; tlen = strlen(tmpBuf); snprintf(tmpBuf + tlen, sizeof(tmpBuf) - tlen, "-%*.*s", l, l, p); p += l + 1; /* add_style_name field */ l = strchr(p, '-') - p; tlen = strlen(tmpBuf); snprintf(tmpBuf + tlen, sizeof(tmpBuf) - tlen, "-%*.*s", l, l, p); strlcat(tmpBuf, "-", sizeof(tmpBuf)); if ((tmpvals.values_supplied & PIXELSIZE_MASK) == PIXELSIZE_ARRAY) { char buffer[XLFD_DOUBLE_TO_TEXT_BUF_SIZE]; strlcat(tmpBuf, "[", sizeof(tmpBuf)); strlcat(tmpBuf, xlfd_double_to_text(tmpvals.pixel_matrix[0], buffer, 0), sizeof(tmpBuf)); strlcat(tmpBuf, xlfd_double_to_text(tmpvals.pixel_matrix[1], buffer, 1), sizeof(tmpBuf)); strlcat(tmpBuf, xlfd_double_to_text(tmpvals.pixel_matrix[2], buffer, 1), sizeof(tmpBuf)); strlcat(tmpBuf, xlfd_double_to_text(tmpvals.pixel_matrix[3], buffer, 1), sizeof(tmpBuf)); strlcat(tmpBuf, "]", sizeof(tmpBuf)); } else { tlen = strlen(tmpBuf); snprintf(tmpBuf + tlen, sizeof(tmpBuf) - tlen, "%d", (int)(tmpvals.pixel_matrix[3] + .5)); } strlcat(tmpBuf, "-", sizeof(tmpBuf)); if ((tmpvals.values_supplied & POINTSIZE_MASK) == POINTSIZE_ARRAY) { char buffer[XLFD_DOUBLE_TO_TEXT_BUF_SIZE]; strlcat(tmpBuf, "[", sizeof(tmpBuf)); strlcat(tmpBuf, xlfd_double_to_text(tmpvals.point_matrix[0], buffer, 0), sizeof(tmpBuf)); strlcat(tmpBuf, xlfd_double_to_text(tmpvals.point_matrix[1], buffer, 1), sizeof(tmpBuf)); strlcat(tmpBuf, xlfd_double_to_text(tmpvals.point_matrix[2], buffer, 1), sizeof(tmpBuf)); strlcat(tmpBuf, xlfd_double_to_text(tmpvals.point_matrix[3], buffer, 1), sizeof(tmpBuf)); strlcat(tmpBuf, "]", sizeof(tmpBuf)); } else { tlen = strlen(tmpBuf); snprintf(tmpBuf + tlen, sizeof(tmpBuf) - tlen, "%d", (int)(tmpvals.point_matrix[3] * 10.0 + .5)); } tlen = strlen(tmpBuf); snprintf(tmpBuf + tlen, sizeof(tmpBuf) - tlen, "-%d-%d%*.*s%d%s", tmpvals.x, tmpvals.y, spacingLen, spacingLen, ptr3, tmpvals.width, ptr5); strlcpy(ptr1 + 1, tmpBuf, fnamelen - (ptr1 - fname)); if ((vals->values_supplied & CHARSUBSET_SPECIFIED) && !vals->nranges) strlcat(fname, "[]", fnamelen); else append_ranges(fname, fnamelen, vals->nranges, vals->ranges); break; } return TRUE; } fsRange *FontParseRanges(char *name, int *nranges) { int n; unsigned long l; char *p1, *p2; fsRange *result = (fsRange *)0; name = strchr(name, '-'); for (n = 1; name && n < 14; n++) name = strchr(name + 1, '-'); *nranges = 0; if (!name || !(p1 = strchr(name, '['))) return (fsRange *)0; p1++; while (*p1 && *p1 != ']') { fsRange thisrange; l = strtol(p1, &p2, 0); if (p2 == p1 || l > 0xffff) break; thisrange.max_char_low = thisrange.min_char_low = l & 0xff; thisrange.max_char_high = thisrange.min_char_high = l >> 8; p1 = p2; if (*p1 == ']' || *p1 == ' ') { while (*p1 == ' ') p1++; if (add_range(&thisrange, nranges, &result, TRUE) != Successful) break; } else if (*p1 == '_') { l = strtol(++p1, &p2, 0); if (p2 == p1 || l > 0xffff) break; thisrange.max_char_low = l & 0xff; thisrange.max_char_high = l >> 8; p1 = p2; if (*p1 == ']' || *p1 == ' ') { while (*p1 == ' ') p1++; if (add_range(&thisrange, nranges, &result, TRUE) != Successful) break; } } else break; } return result; }