/* Copyright (C) 2001-2021 Artifex Software, Inc. All Rights Reserved. This software is provided AS-IS with no warranty, either express or implied. This software is distributed under license and may not be copied, modified or distributed except as expressly authorized under the terms of the license contained in the file LICENSE in this distribution. Refer to licensing information at http://www.artifex.com or contact Artifex Software, Inc., 1305 Grant Avenue - Suite 200, Novato, CA 94945, U.S.A., +1(415)492-9861, for further information. */ /* PostScript/PDF font writing utilities */ #include "memory_.h" #include /* for qsort */ #include "gx.h" #include "gserrors.h" #include "gsmatrix.h" /* for gxfont.h */ #include "gxfont.h" #include "gdevpsf.h" /* Begin enumerating the glyphs in a font or a font subset. */ static int enumerate_font_next(psf_glyph_enum_t *ppge, gs_glyph *pglyph) { gs_font *font = ppge->font; int index = (int)ppge->index; int code = font->procs.enumerate_glyph(font, &index, ppge->glyph_space, pglyph); ppge->index = index; return (index == 0 ? 1 : code < 0 ? code : 0); } static int enumerate_glyphs_next(psf_glyph_enum_t *ppge, gs_glyph *pglyph) { if (ppge->index >= ppge->subset.size) return 1; *pglyph = ppge->subset.selected.list[ppge->index++]; return 0; } static int enumerate_range_next(psf_glyph_enum_t *ppge, gs_glyph *pglyph) { if (ppge->index >= ppge->subset.size) return 1; *pglyph = (gs_glyph)(ppge->index++ + GS_MIN_CID_GLYPH); return 0; } void psf_enumerate_list_begin(psf_glyph_enum_t *ppge, gs_font *font, const gs_glyph *subset_list, uint subset_size, gs_glyph_space_t glyph_space) { ppge->font = font; ppge->subset.selected.list = subset_list; ppge->subset.size = subset_size; ppge->glyph_space = glyph_space; ppge->enumerate_next = (subset_list ? enumerate_glyphs_next : subset_size ? enumerate_range_next : enumerate_font_next); psf_enumerate_glyphs_reset(ppge); } /* Begin enumerating CID or TT glyphs in a subset given by a bit vector. */ static int enumerate_bits_next(psf_glyph_enum_t *ppge, gs_glyph *pglyph) { for (; ppge->index < ppge->subset.size; ppge->index++) if (ppge->subset.selected.bits[ppge->index >> 3] & (0x80 >> (ppge->index & 7))) { *pglyph = (gs_glyph)(ppge->index++ + GS_MIN_CID_GLYPH); return 0; } return 1; } void psf_enumerate_bits_begin(psf_glyph_enum_t *ppge, gs_font *font, const byte *subset_bits, uint subset_size, gs_glyph_space_t glyph_space) { ppge->font = font; ppge->subset.selected.bits = subset_bits; ppge->subset.size = subset_size; ppge->glyph_space = glyph_space; ppge->enumerate_next = (subset_bits ? enumerate_bits_next : subset_size ? enumerate_range_next : enumerate_font_next); psf_enumerate_glyphs_reset(ppge); } /* Reset a glyph enumeration. */ void psf_enumerate_glyphs_reset(psf_glyph_enum_t *ppge) { ppge->index = 0; } /* Enumerate the next glyph in a font or a font subset. */ /* Return 0 if more glyphs, 1 if done, <0 if error. */ int psf_enumerate_glyphs_next(psf_glyph_enum_t *ppge, gs_glyph *pglyph) { return ppge->enumerate_next(ppge, pglyph); } /* * Add composite glyph pieces to a list of glyphs. Does not sort or * remove duplicates. max_pieces is the maximum number of pieces that a * single glyph can have: if this value is not known, the caller should * use max_count. */ int psf_add_subset_pieces(gs_glyph *glyphs, uint *pcount, uint max_count, uint max_pieces, gs_font *font) { uint i; uint count = *pcount; for (i = 0; i < count; ++i) { gs_glyph_info_t info; int code; if (count + max_pieces > max_count) { /* Check first to make sure there is enough room. */ code = font->procs.glyph_info(font, glyphs[i], NULL, GLYPH_INFO_NUM_PIECES, &info); if (code < 0) continue; if (count + info.num_pieces > max_count) return_error(gs_error_rangecheck); } info.pieces = &glyphs[count]; code = font->procs.glyph_info(font, glyphs[i], NULL, GLYPH_INFO_NUM_PIECES | GLYPH_INFO_PIECES, &info); if (code >= 0) count += info.num_pieces; } *pcount = count; return 0; } /* * Sort a list of glyphs and remove duplicates. Return the number of glyphs * in the result. */ static int compare_glyphs(const void *pg1, const void *pg2) { gs_glyph g1 = *(const gs_glyph *)pg1, g2 = *(const gs_glyph *)pg2; return (g1 < g2 ? -1 : g1 > g2 ? 1 : 0); } int psf_sort_glyphs(gs_glyph *glyphs, int count) { int i, n; qsort(glyphs, count, sizeof(*glyphs), compare_glyphs); for (i = n = 0; i < count; ++i) if (i == 0 || glyphs[i] != glyphs[i - 1]) glyphs[n++] = glyphs[i]; return n; } /* * Return the index of a given glyph in a sorted list of glyphs, or -1 * if the glyph is not present. */ int psf_sorted_glyphs_index_of(const gs_glyph *glyphs, int count, gs_glyph glyph) { int lo = 0, hi = count - 1; if (hi < 0) return -1; if (glyph < glyphs[0] || glyph > glyphs[hi]) return -1; /* * Loop invariants: hi > lo; * glyphs[lo] <= glyph <= glyphs[hi]. */ while (hi - lo > 1) { int mid = (lo + hi) >> 1; if (glyph >= glyphs[mid]) lo = mid; else hi = mid; } return (glyph == glyphs[lo] ? lo : glyph == glyphs[hi] ? hi : -1); } /* Determine whether a sorted list of glyphs includes a given glyph. */ bool psf_sorted_glyphs_include(const gs_glyph *glyphs, int count, gs_glyph glyph) { return psf_sorted_glyphs_index_of(glyphs, count, glyph) >= 0; } /* Check that all selected glyphs can be written. */ int psf_check_outline_glyphs(gs_font_base *pfont, psf_glyph_enum_t *ppge, glyph_data_proc_t glyph_data) { uint members = GLYPH_INFO_WIDTH0 << pfont->WMode; gs_glyph glyph; int code, good_glyphs = 0; while ((code = psf_enumerate_glyphs_next(ppge, &glyph)) != 1) { gs_glyph_data_t gdata; gs_font_type1 *ignore_font; gs_glyph_info_t info; if (code < 0) return code; gdata.memory = pfont->memory; code = glyph_data(pfont, glyph, &gdata, &ignore_font); /* * If the glyph isn't defined by a CharString, glyph_data will * return a typecheck error. But if there's merely a glyph in * in the Encoding that isn't defined, glyph_data will return an * undefined error, which is OK. */ if (code < 0) { if (code == gs_error_undefined) continue; return code; } gs_glyph_data_free(&gdata, "psf_check_outline_glyphs"); /* * If the font has a CDevProc or calls a non-standard OtherSubr, * glyph_info will return a rangecheck error. */ code = pfont->procs.glyph_info((gs_font *)pfont, glyph, NULL, members, &info); /* It may be that a single glyph is bad (eg no (h)sbw), we'll ignore it */ /* here, the glyph may not be included in any subset, or not used at all */ /* (ie the /.notdef). If an invalid glyph is actually used then the text */ /* processing will still signal an error causing the document to fail. */ if(code == gs_error_invalidfont || code == gs_error_rangecheck) continue; if (code < 0) return code; good_glyphs++; } if(good_glyphs) return 0; else return_error(gs_error_invalidfont); } /* Gather glyph information for a Type 1 or Type 2 font. */ int psf_get_outline_glyphs(psf_outline_glyphs_t *pglyphs, gs_font_base *pfont, gs_glyph *orig_subset_glyphs, uint orig_subset_size, glyph_data_proc_t glyph_data) { gs_glyph notdef = GS_NO_GLYPH; gs_glyph *subset_glyphs = orig_subset_glyphs; uint subset_size = orig_subset_size; /* Currently its impossible to hit this code (subset_glyphs is always NULL) and if * we ever did, there's a problem with countof(pglyphs->subset_data), the count * will always be incorrect it seems. Since we never use the code we could just * leave it in place, but Coverity complains. We could remove it, but it might * actually be useful one day (if fixed) so for now, ifdef it out. */ #if 0 if (subset_glyphs) { if (subset_size > countof(pglyphs->subset_data)) return_error(gs_error_limitcheck); memcpy(pglyphs->subset_data, orig_subset_glyphs, sizeof(gs_glyph) * subset_size); subset_glyphs = pglyphs->subset_data; } #endif { /* * Make sure that this font can be written out. Specifically, it * must have no CharStrings defined by PostScript procedures, no * non-standard OtherSubrs, and no CDevProc. */ psf_glyph_enum_t genum; int code; psf_enumerate_glyphs_begin(&genum, (gs_font *)pfont, subset_glyphs, (subset_glyphs ? subset_size : 0), GLYPH_SPACE_NAME); code = psf_check_outline_glyphs(pfont, &genum, glyph_data); if (code < 0) return code; } { /* * Detect the .notdef glyph, needed for subset fonts and to * eliminate unnecessary Encoding assignments. */ psf_glyph_enum_t genum; gs_glyph glyph; int code; psf_enumerate_glyphs_begin(&genum, (gs_font *)pfont, NULL, 0, GLYPH_SPACE_NAME); while ((code = psf_enumerate_glyphs_next(&genum, &glyph)) != 1) { if (gs_font_glyph_is_notdef(pfont, glyph)) { notdef = glyph; break; } } } #if 0 if (subset_glyphs) { /* * For subset fonts, we must ensure that characters referenced * by seac are also included. Note that seac creates at most * 2 pieces. */ int code = psf_add_subset_pieces(subset_glyphs, &subset_size, countof(pglyphs->subset_data) - 1, 2, (gs_font *)pfont); uint keep_size, i; if (code < 0) return code; /* Subset fonts require .notdef. */ if (notdef == GS_NO_GLYPH) return_error(gs_error_rangecheck); /* Remove undefined glyphs. */ for (i = 0, keep_size = 0; i < subset_size; ++i) { gs_glyph_info_t info; gs_glyph glyph = subset_glyphs[i]; /* * The documentation for the glyph_info procedure says that * using members = 0 is an inexpensive way to find out * whether a given glyph exists, but the implementations * don't actually do this. Request an inexpensive value. */ if (pfont->procs.glyph_info((gs_font *)pfont, glyph, NULL, GLYPH_INFO_NUM_PIECES, &info) >= 0) subset_glyphs[keep_size++] = glyph; } subset_size = keep_size; /* Sort the glyphs. Make sure .notdef is included. */ subset_glyphs[subset_size++] = notdef; subset_size = psf_sort_glyphs(subset_glyphs, subset_size); } #endif pglyphs->notdef = notdef; pglyphs->subset_glyphs = subset_glyphs; pglyphs->subset_size = subset_size; return 0; }