From 69f155d4fc11f2f0e1dd1bfcd804192303ba1627 Mon Sep 17 00:00:00 2001 From: Ulrich Drepper Date: Thu, 30 Apr 1998 16:57:48 +0000 Subject: Update. 1998-04-30 16:45 Ulrich Drepper * inet/ether_aton.c: Including netinet/if_ether.h is not necessary. * locale/Makefile (distribute): Add programs/repertoire.h. (localedef-modules): Add repertoire. (CPPFLAGS): Define REPERTOIREMAP_PATH. * locale/programs/repertoire.c: New file. * locale/programs/repertoire.h: New file. * locale/programs/charmap.c: Starting fixing character set handling to handle multi-byte encodings. * locale/programs/charset.c: Likewise. * locale/programs/charset.h: Likewise. * locale/programs/ld-collate.c: Likewise. * locale/programs/ld-ctype.c: Likewise. * locale/programs/linereader.c: Likewise. * locale/programs/localedef.c: Likewise. * locale/programs/locfile-kw.gperf: Likewise. * locale/programs/locfile-kw.h: Likewise. * locale/programs/locfile-token.h: Likewise. * locale/programs/locfile.h: Likewise. * locale/programs/stringtrans.c: Likewise. 1998-04-18 Philip Blundell * sysdeps/arm/memset.S: Fix off by one error. * sysdeps/unix/sysv/linux/arm/sysdep.h (PSEUDO): On error, call __syscall_error rather than syscall_error directly. 1998-04-17 Philip Blundell * sysdeps/unix/sysv/linux/arm/mmap.S: New file; implementation of mmap() syscall for ARM. * sysdeps/unix/arm/start.c: New file; startup code for ARM a.out binaries. --- locale/Makefile | 8 +- locale/programs/charmap.c | 50 ++++-- locale/programs/charset.c | 59 +++---- locale/programs/charset.h | 17 ++- locale/programs/ld-collate.c | 16 +- locale/programs/ld-ctype.c | 38 ++--- locale/programs/linereader.c | 2 +- locale/programs/localedef.c | 17 +-- locale/programs/locfile-kw.gperf | 3 +- locale/programs/locfile-kw.h | 160 ++++++++++--------- locale/programs/locfile-token.h | 5 +- locale/programs/locfile.h | 3 +- locale/programs/repertoire.c | 323 +++++++++++++++++++++++++++++++++++++++ locale/programs/repertoire.h | 38 +++++ locale/programs/stringtrans.c | 5 +- 15 files changed, 570 insertions(+), 174 deletions(-) create mode 100644 locale/programs/repertoire.c create mode 100644 locale/programs/repertoire.h (limited to 'locale') diff --git a/locale/Makefile b/locale/Makefile index 57a5bc5715..b93298310b 100644 --- a/locale/Makefile +++ b/locale/Makefile @@ -1,4 +1,4 @@ -# Copyright (C) 1991, 1992, 1995, 1996, 1997 Free Software Foundation, Inc. +# Copyright (C) 1991, 92, 95, 96, 97, 98 Free Software Foundation, Inc. # This file is part of the GNU C Library. # The GNU C Library is free software; you can redistribute it and/or @@ -30,7 +30,8 @@ distribute = localeinfo.h categories.def iso-4217.def weight.h \ $(lib-modules:=.c) config.h simple-hash.h \ charmap-kw.gperf charmap-kw.h locfile-token.h \ locfile-kw.gperf locfile-kw.h linereader.h \ - locales.h locfile.h stringtrans.h charset.h) + locales.h locfile.h stringtrans.h charset.h \ + repertoire.h) routines = setlocale findlocale loadlocale localeconv nl_langinfo \ mb_cur_max codeset_name \ newlocale duplocale freelocale @@ -53,7 +54,7 @@ vpath %.h programs vpath %.gperf programs localedef-modules := $(categories:%=ld-%) charmap charset linereader \ - locfile stringtrans + locfile stringtrans repertoire locale-modules := locale-spec lib-modules := simple-hash xmalloc xstrdup @@ -77,6 +78,7 @@ CPPFLAGS := -DLOCALE_PATH='$(localepath)' \ -DLOCALEDIR='"$(localedir)"' \ -DLOCALE_ALIAS_PATH='"$(localedir):$(i18ndir)"' \ -DCHARMAP_PATH='"$(i18ndir)/charmaps"' \ + -DREPERTOIREMAP_PATH='"$(i18ndir)/repertoiremaps"' \ -DLOCSRCDIR='"$(i18ndir)/locales"' -DHAVE_CONFIG_H \ -Iprograms $(CPPFLAGS) diff --git a/locale/programs/charmap.c b/locale/programs/charmap.c index 0cd62fbfd4..7114a237a0 100644 --- a/locale/programs/charmap.c +++ b/locale/programs/charmap.c @@ -33,6 +33,8 @@ #include "error.h" #include "linereader.h" #include "charset.h" +#include "locfile.h" +#include "repertoire.h" /* Uncomment following line for production version. */ @@ -209,6 +211,8 @@ parse_charmap (const char *filename) memset (result, '\0', sizeof (struct charset_t)); /* The default DEFAULT_WIDTH is 1. */ result->width_default = 1; + /* Let the user overwrite the repertoire map we use. */ + result->repertoiremap = repertoiremap; #define obstack_chunk_alloc malloc #define obstack_chunk_free free @@ -265,6 +269,17 @@ parse_charmap (const char *filename) lr_ignore_rest (cmfile, 1); + /* Read the repertoire map now. */ + if (result->repertoiremap == NULL) + /* This is fatal. */ + error (4, 0, _("no repertoire map specified: cannot proceed")); + + result->repertoire = repertoire_read (result->repertoiremap); + if (result->repertoire == NULL) + /* This is also fatal. */ + error (4, errno, _("cannot read repertoire map `%s'"), + result->repertoiremap); + state = 2; continue; } @@ -273,7 +288,7 @@ parse_charmap (const char *filename) && nowtok != tok_mb_cur_min && nowtok != tok_escape_char && nowtok != tok_comment_char && nowtok != tok_g0esc && nowtok != tok_g1esc && nowtok != tok_g2esc - && nowtok != tok_g3esc) + && nowtok != tok_g3esc && nowtok != tok_repertoiremap) { lr_error (cmfile, _("syntax error in prolog: %s"), _("illegal definition")); @@ -305,6 +320,18 @@ parse_charmap (const char *filename) lr_ignore_rest (cmfile, 1); continue; + case tok_repertoiremap: + if (arg->tok != tok_ident) + goto badarg; + + if (result->repertoiremap == NULL) + result->repertoiremap = obstack_copy0 (&result->mem_pool, + arg->val.str.start, + arg->val.str.len); + + lr_ignore_rest (cmfile, 1); + continue; + case tok_mb_cur_max: case tok_mb_cur_min: if (arg->tok != tok_number) @@ -437,14 +464,14 @@ argument to <%s> must be a single character"), continue; } - if (nowtok == tok_charcode) - /* Write char value in table. */ - charset_new_char (cmfile, result, now->val.charcode.nbytes, - now->val.charcode.val, from_name, to_name); + if (now->val.charcode.nbytes < result->mb_cur_min) + lr_error (cmfile, _("too few bytes in character encoding")); + else if (now->val.charcode.nbytes > result->mb_cur_max) + lr_error (cmfile, _("too many bytes in character encoding")); else - /* Determine ISO 10646 value and write into table. */ - charset_new_unicode (cmfile, result, now->val.charcode.nbytes, - now->val.charcode.val, from_name, to_name); + charset_new_char (cmfile, &result->char_table, + now->val.charcode.nbytes, + now->val.charcode.val, from_name, to_name); /* Ignore trailing comment silently. */ lr_ignore_rest (cmfile, 0); @@ -466,8 +493,7 @@ argument to <%s> must be a single character"), continue; } - /* If the previous line was not completely correct free the - used memory. */ + /* Copy the to-name in a safe place. */ to_name = (char *) obstack_copy0 (&result->mem_pool, cmfile->token.val.str.start, cmfile->token.val.str.len); @@ -694,7 +720,7 @@ new_width (struct linereader *cmfile, struct charset_t *result, { unsigned int from_val, to_val; - from_val = charset_find_value (result, from, strlen (from)); + from_val = charset_find_value (&result->char_table, from, strlen (from)); if ((wchar_t) from_val == ILLEGAL_CHAR_VALUE) { lr_error (cmfile, _("unknown character `%s'"), from); @@ -705,7 +731,7 @@ new_width (struct linereader *cmfile, struct charset_t *result, to_val = from_val; else { - to_val = charset_find_value (result, to, strlen (to)); + to_val = charset_find_value (&result->char_table, to, strlen (to)); if ((wchar_t) to_val == ILLEGAL_CHAR_VALUE) { lr_error (cmfile, _("unknown character `%s'"), to); diff --git a/locale/programs/charset.c b/locale/programs/charset.c index fdacf25328..767fafb93a 100644 --- a/locale/programs/charset.c +++ b/locale/programs/charset.c @@ -1,4 +1,4 @@ -/* Copyright (C) 1996, 1997 Free Software Foundation, Inc. +/* Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. This file is part of the GNU C Library. Contributed by Ulrich Drepper , 1996. @@ -23,6 +23,8 @@ #include #include +#include +#include #include #include #include @@ -31,49 +33,24 @@ #include "charset.h" -static void -insert_char (struct linereader *lr, struct charset_t *cs, int bytes, - unsigned int value, const char *from, const char *to); - - -void -charset_new_char (struct linereader *lr, struct charset_t *cs, int bytes, - unsigned int value, const char *from, const char *to) -{ - if (bytes < cs->mb_cur_min) - lr_error (lr, _("too few bytes in character encoding")); - else if (bytes > cs->mb_cur_max) - lr_error (lr, _("too many bytes in character encoding")); - else - insert_char (lr, cs, bytes, value, from, to); -} - - -void -charset_new_unicode (struct linereader *lr, struct charset_t *cs, int bytes, - unsigned int value, const char *from, const char *to) -{ - /* For now: perhaps support will be removed again... */ - insert_char (lr, cs, bytes, value, from, to); -} - - unsigned int -charset_find_value (const struct charset_t *cs, const char *name, size_t len) +charset_find_value (const hash_table *ht, const char *name, size_t len) { void *result; - if (find_entry ((hash_table *) &cs->char_table, name, len, &result) < 0) + if (find_entry ((hash_table *) ht, name, len, &result) < 0) return ILLEGAL_CHAR_VALUE; return (unsigned int) ((unsigned long int) result); } -static void -insert_char (struct linereader *lr, struct charset_t *cs, int bytes, - unsigned int value, const char *from, const char *to) +void +charset_new_char (struct linereader *lr, hash_table *ht, int bytes, + unsigned int value, const char *from, const char *to) { + char *from_end; + char *to_end; const char *cp; char *buf; int prefix_len, len1, len2; @@ -81,7 +58,7 @@ insert_char (struct linereader *lr, struct charset_t *cs, int bytes, if (to == NULL) { - if (insert_entry (&cs->char_table, from, strlen (from), + if (insert_entry (ht, from, strlen (from), (void *) (unsigned long int) value) < 0) lr_error (lr, _("duplicate character name `%s'"), from); @@ -111,8 +88,16 @@ insert_char (struct linereader *lr, struct charset_t *cs, int bytes, if (cp == &from[len1 - 1] || strncmp (from, to, prefix_len) != 0) goto illegal_range; - from_nr = strtoul (&from[prefix_len], NULL, 10); - to_nr = strtoul (&to[prefix_len], NULL, 10); + errno = 0; + from_nr = strtoul (&from[prefix_len], &from_end, 10); + if (*from_end != '\0' || (from_nr == ULONG_MAX && errno == ERANGE) + || ((to_nr = strtoul (&to[prefix_len], &to_end, 10)) == ULONG_MAX + && errno == ERANGE) + || *to_end != '\0') + { + lr_error (lr, _("<%s> and <%s> are illegal names for range")); + return; + } if (from_nr > to_nr) { @@ -127,7 +112,7 @@ insert_char (struct linereader *lr, struct charset_t *cs, int bytes, { sprintf (&buf[prefix_len], "%0*d", len1 - prefix_len, cnt); - if (insert_entry (&cs->char_table, buf, len1, + if (insert_entry (ht, buf, len1, (void *) (unsigned long int) (value + (cnt - from_nr))) < 0) lr_error (lr, _("duplicate character name `%s'"), buf); diff --git a/locale/programs/charset.h b/locale/programs/charset.h index 82c4ef0c57..db93f16306 100644 --- a/locale/programs/charset.h +++ b/locale/programs/charset.h @@ -1,4 +1,4 @@ -/* Copyright (C) 1996, 1997 Free Software Foundation, Inc. +/* Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. This file is part of the GNU C Library. Contributed by Ulrich Drepper , 1996. @@ -22,6 +22,7 @@ #include +#include "repertoire.h" #include "simple-hash.h" #include "linereader.h" @@ -36,6 +37,9 @@ struct width_rule struct charset_t { + const char *repertoiremap; + struct repertoire_t *repertoire; + const char *code_set_name; int mb_cur_min; int mb_cur_max; @@ -63,14 +67,11 @@ extern int be_quiet; struct charset_t *charmap_read (const char *filename); /* Prototypes for function to insert new character. */ -void charset_new_char (struct linereader *lr, struct charset_t *cs, int bytes, +void charset_new_char (struct linereader *lr, hash_table *ht, int bytes, unsigned int value, const char *from, const char *to); -void charset_new_unicode (struct linereader *lr, struct charset_t *cs, - int bytes, unsigned int value, const char *from, - const char *to); - -unsigned int charset_find_value (const struct charset_t *__cs, - const char *__name, size_t __len); +/* Return the value stored under the given key in the hashing table. */ +unsigned int charset_find_value (const hash_table *ht, + const char *name, size_t len); #endif /* charset.h */ diff --git a/locale/programs/ld-collate.c b/locale/programs/ld-collate.c index 57b97767af..a92ff1154a 100644 --- a/locale/programs/ld-collate.c +++ b/locale/programs/ld-collate.c @@ -212,7 +212,7 @@ collate_finish (struct localedef_t *locale, struct charset_t *charset) void *ptmp; unsigned int value = 0; - wch = charset_find_value (charset, patch->token, toklen); + wch = charset_find_value (&charset->char_table, patch->token, toklen); if (wch != ILLEGAL_CHAR_VALUE) { element_t *runp; @@ -1054,7 +1054,8 @@ collate_element_to (struct linereader *lr, struct localedef_t *locale, collate->combine_token = NULL; } - value = charset_find_value (charset, code->val.str.start, code->val.str.len); + value = charset_find_value (&charset->char_table, code->val.str.start, + code->val.str.len); if ((wchar_t) value != ILLEGAL_CHAR_VALUE) { lr_error (lr, _("symbol for multicharacter collating element " @@ -1181,7 +1182,8 @@ collate_symbol (struct linereader *lr, struct localedef_t *locale, wchar_t value; void *not_used; - value = charset_find_value (charset, code->val.str.start, code->val.str.len); + value = charset_find_value (&charset->char_table, code->val.str.start, + code->val.str.len); if (value != ILLEGAL_CHAR_VALUE) { lr_error (lr, _("symbol for multicharacter collating element " @@ -1268,7 +1270,7 @@ collate_order_elem (struct linereader *lr, struct localedef_t *locale, { case tok_bsymbol: /* We have a string to find in one of the three hashing tables. */ - value = charset_find_value (charset, code->val.str.start, + value = charset_find_value (&charset->char_table, code->val.str.start, code->val.str.len); if (value != ILLEGAL_CHAR_VALUE) { @@ -1533,7 +1535,8 @@ collate_weight_bsymbol (struct linereader *lr, struct localedef_t *locale, assert (code->tok == tok_bsymbol); - value = charset_find_value (charset, code->val.str.start, code->val.str.len); + value = charset_find_value (&charset->char_table, code->val.str.start, + code->val.str.len); if (value != ILLEGAL_CHAR_VALUE) { element_t *runp; @@ -1706,7 +1709,8 @@ collate_simple_weight (struct linereader *lr, struct localedef_t *locale, return -1; } - wch = charset_find_value (charset, startp, putp - startp); + wch = charset_find_value (&charset->char_table, startp, + putp - startp); if (wch != ILLEGAL_CHAR_VALUE) { element_t *pelem; diff --git a/locale/programs/ld-ctype.c b/locale/programs/ld-ctype.c index f2f32cc214..2f9a9a2caf 100644 --- a/locale/programs/ld-ctype.c +++ b/locale/programs/ld-ctype.c @@ -308,7 +308,7 @@ character %s'%s' in class `%s' must not be in class `%s'"), } /* ... and now test as a special case. */ - space_value = charset_find_value (charset, "SP", 2); + space_value = charset_find_value (&charset->char_table, "SP", 2); if ((wchar_t) space_value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -634,7 +634,8 @@ ctype_class_from (struct linereader *lr, struct localedef_t *locale, struct locale_ctype_t *ctype = locale->categories[LC_CTYPE].ctype; unsigned int value; - value = charset_find_value (charset, code->val.str.start, code->val.str.len); + value = charset_find_value (&charset->char_table, code->val.str.start, + code->val.str.len); ctype->last_class_char = value; @@ -656,7 +657,8 @@ ctype_class_to (struct linereader *lr, struct localedef_t *locale, struct locale_ctype_t *ctype = locale->categories[LC_CTYPE].ctype; unsigned int value, cnt; - value = charset_find_value (charset, code->val.str.start, code->val.str.len); + value = charset_find_value (&charset->char_table, code->val.str.start, + code->val.str.len); /* In the LC_CTYPE category it is no error when a character is not found. This has to be ignored silently. */ @@ -750,7 +752,8 @@ ctype_map_from (struct linereader *lr, struct localedef_t *locale, struct locale_ctype_t *ctype = locale->categories[LC_CTYPE].ctype; unsigned int value; - value = charset_find_value (charset, code->val.str.start, code->val.str.len); + value = charset_find_value (&charset->char_table, code->val.str.start, + code->val.str.len); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) /* In the LC_CTYPE category it is no error when a character is @@ -770,7 +773,8 @@ ctype_map_to (struct linereader *lr, struct localedef_t *locale, struct locale_ctype_t *ctype = locale->categories[LC_CTYPE].ctype; unsigned int value; - value = charset_find_value (charset, code->val.str.start, code->val.str.len); + value = charset_find_value (&charset->char_table, code->val.str.start, + code->val.str.len); if ((wchar_t) ctype->from_map_char == ILLEGAL_CHAR_VALUE || (wchar_t) value == ILLEGAL_CHAR_VALUE) @@ -948,7 +952,7 @@ set_class_defaults (struct locale_ctype_t *ctype, struct charset_t *charset) unsigned int value; tmp[0] = ch; - value = charset_find_value (charset, tmp, 1); + value = charset_find_value (&charset->char_table, tmp, 1); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1013,7 +1017,7 @@ character `%s' not defined while needed as default value"), { unsigned int value; - value = charset_find_value (charset, "space", 5); + value = charset_find_value (&charset->char_table, "space", 5); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1024,7 +1028,7 @@ character `%s' not defined while needed as default value"), else ELEM (ctype, class_collection, , value) |= BIT (tok_space); - value = charset_find_value (charset, "form-feed", 9); + value = charset_find_value (&charset->char_table, "form-feed", 9); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1035,7 +1039,7 @@ character `%s' not defined while needed as default value"), else ELEM (ctype, class_collection, , value) |= BIT (tok_space); - value = charset_find_value (charset, "newline", 7); + value = charset_find_value (&charset->char_table, "newline", 7); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1046,7 +1050,7 @@ character `%s' not defined while needed as default value"), else ELEM (ctype, class_collection, , value) |= BIT (tok_space); - value = charset_find_value (charset, "carriage-return", 15); + value = charset_find_value (&charset->char_table, "carriage-return", 15); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1057,7 +1061,7 @@ character `%s' not defined while needed as default value"), else ELEM (ctype, class_collection, , value) |= BIT (tok_space); - value = charset_find_value (charset, "tab", 3); + value = charset_find_value (&charset->char_table, "tab", 3); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1068,7 +1072,7 @@ character `%s' not defined while needed as default value"), else ELEM (ctype, class_collection, , value) |= BIT (tok_space); - value = charset_find_value (charset, "vertical-tab", 12); + value = charset_find_value (&charset->char_table, "vertical-tab", 12); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1097,7 +1101,7 @@ character `%s' not defined while needed as default value"), { unsigned int value; - value = charset_find_value (charset, "space", 5); + value = charset_find_value (&charset->char_table, "space", 5); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1108,7 +1112,7 @@ character `%s' not defined while needed as default value"), else ELEM (ctype, class_collection, , value) |= BIT (tok_blank); - value = charset_find_value (charset, "tab", 3); + value = charset_find_value (&charset->char_table, "tab", 3); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1149,7 +1153,7 @@ character `%s' not defined while needed as default value"), if ((ctype->class_collection[cnt] & mask) != 0) ctype->class_collection[cnt] |= BIT (tok_print); - space = charset_find_value (charset, "space", 5); + space = charset_find_value (&charset->char_table, "space", 5); if (space == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1178,7 +1182,7 @@ character `%s' not defined while needed as default value"), tmp[1] = (char) ch; - value_from = charset_find_value (charset, &tmp[1], 1); + value_from = charset_find_value (&charset->char_table, &tmp[1], 1); if ((wchar_t) value_from == ILLEGAL_CHAR_VALUE) { if (!be_quiet) @@ -1190,7 +1194,7 @@ character `%s' not defined while needed as default value"), /* This conversion is implementation defined. */ tmp[1] = (char) (ch + ('A' - 'a')); - value_to = charset_find_value (charset, &tmp[1], 1); + value_to = charset_find_value (&charset->char_table, &tmp[1], 1); if ((wchar_t) value_to == ILLEGAL_CHAR_VALUE) { if (!be_quiet) diff --git a/locale/programs/linereader.c b/locale/programs/linereader.c index 6692164b60..4406e1ab72 100644 --- a/locale/programs/linereader.c +++ b/locale/programs/linereader.c @@ -524,7 +524,7 @@ get_string (struct linereader *lr, const struct charset_t *charset) if (lr->translate_strings) { - value = charset_find_value (charset, &buf[startidx], + value = charset_find_value (&charset->char_table, &buf[startidx], bufact - startidx); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) illegal_string = 1; diff --git a/locale/programs/localedef.c b/locale/programs/localedef.c index 32437c12f5..831c36e429 100644 --- a/locale/programs/localedef.c +++ b/locale/programs/localedef.c @@ -81,8 +81,8 @@ static const char *charmap_file; /* Name of the locale definition file. */ static const char *input_file; -/* Name of the UCS file. */ -static const char *ucs_csn; +/* Name of the repertoire map file. */ +const char *repertoiremap; /* Name and version of program. */ @@ -99,9 +99,7 @@ static const struct argp_option options[] = { "charmap", 'f', "FILE", 0, N_("Symbolic character names defined in FILE") }, { "inputfile", 'i', "FILE", 0, N_("Source definitions are found in FILE") }, - { "code-set-name", 'u', "NAME", OPTION_HIDDEN, - N_("Specify code set for mapping ISO 10646 elements") }, - { "repertoire-map", 'u', NAME, 0, + { "repertoire-map", 'u', "FILE", 0, N_("file containing mapping from symbolic names to UCS4 values") }, { NULL, 0, NULL, 0, N_("Output control:") }, @@ -355,7 +353,7 @@ parse_opt (int key, char *arg, struct argp_state *state) input_file = arg; break; case 'u': - ucs_csn = arg; + repertoiremap = arg; break; case 'v': verbose = 1; @@ -377,10 +375,11 @@ more_help (int key, const char *text, void *input) case ARGP_KEY_HELP_EXTRA: /* We print some extra information. */ asprintf (&cp, gettext ("\ -System's directory for character maps: %s\n\ - locale files : %s\n\ +System's directory for character maps : %s\n\ + repertoire maps: %s\n\ + locale path : %s\n\ %s"), - CHARMAP_PATH, LOCALE_PATH, gettext ("\ + CHARMAP_PATH, REPERTOIREMAP_PATH, LOCALE_PATH, gettext ("\ Report bugs using the `glibcbug' script to .\n")); return cp; default: diff --git a/locale/programs/locfile-kw.gperf b/locale/programs/locfile-kw.gperf index 38150ad46b..991e9dd6f9 100644 --- a/locale/programs/locfile-kw.gperf +++ b/locale/programs/locfile-kw.gperf @@ -1,5 +1,5 @@ %{ -/* Copyright (C) 1996, 1997 Free Software Foundation, Inc. +/* Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. This file is part of the GNU C Library. Contributed by Ulrich Drepper , 1996. @@ -26,6 +26,7 @@ struct keyword_t ; %% escape_char, tok_escape_char, 0 comment_char, tok_comment_char, 0 +repertoiremap, tok_repertoiremap, 0 LC_CTYPE, tok_lc_ctype, 0 END, tok_end, 0 copy, tok_copy, 0 diff --git a/locale/programs/locfile-kw.h b/locale/programs/locfile-kw.h index 0fb0b5ca5f..bd80618e77 100644 --- a/locale/programs/locfile-kw.h +++ b/locale/programs/locfile-kw.h @@ -1,6 +1,6 @@ /* C code produced by gperf version 2.5 (GNU C++ version) */ /* Command-line: gperf -acCgopt -k1,2,5,$ -N locfile_hash programs/locfile-kw.gperf */ -/* Copyright (C) 1996, 1997 Free Software Foundation, Inc. +/* Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. This file is part of the GNU C Library. Contributed by Ulrich Drepper , 1996. @@ -24,12 +24,12 @@ #include "locfile-token.h" struct keyword_t ; -#define TOTAL_KEYWORDS 73 +#define TOTAL_KEYWORDS 74 #define MIN_WORD_LENGTH 3 #define MAX_WORD_LENGTH 17 #define MIN_HASH_VALUE 3 -#define MAX_HASH_VALUE 132 -/* maximum key range = 130, duplicates = 0 */ +#define MAX_HASH_VALUE 178 +/* maximum key range = 176, duplicates = 0 */ #ifdef __GNUC__ inline @@ -39,19 +39,19 @@ hash (register const char *str, register int len) { static const unsigned char asso_values[] = { - 133, 133, 133, 133, 133, 133, 133, 133, 133, 133, - 133, 133, 133, 133, 133, 133, 133, 133, 133, 133, - 133, 133, 133, 133, 133, 133, 133, 133, 133, 133, - 133, 133, 133, 133, 133, 133, 133, 133, 133, 133, - 133, 133, 133, 133, 133, 133, 133, 133, 133, 133, - 133, 133, 133, 133, 133, 133, 133, 133, 133, 133, - 133, 133, 133, 133, 133, 133, 133, 0, 0, 0, - 0, 0, 133, 0, 133, 133, 0, 133, 0, 20, - 133, 133, 0, 0, 0, 5, 133, 133, 133, 5, - 133, 133, 133, 133, 133, 5, 133, 0, 60, 0, - 15, 10, 20, 40, 5, 20, 133, 0, 45, 40, - 10, 0, 0, 133, 15, 50, 0, 30, 0, 10, - 15, 15, 133, 133, 133, 133, 133, 133, + 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, + 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, + 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, + 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, + 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, + 179, 179, 179, 179, 179, 179, 179, 179, 179, 179, + 179, 179, 179, 179, 179, 179, 179, 0, 0, 0, + 0, 0, 179, 0, 179, 179, 0, 179, 0, 45, + 179, 179, 0, 0, 0, 5, 179, 179, 179, 10, + 179, 179, 179, 179, 179, 5, 179, 0, 5, 0, + 15, 20, 5, 20, 40, 20, 179, 25, 15, 50, + 10, 0, 0, 179, 45, 50, 0, 30, 0, 5, + 10, 60, 179, 179, 179, 179, 179, 179, }; register int hval = len; @@ -88,102 +88,110 @@ locfile_hash (register const char *str, register int len) {"",}, {"t_fmt", tok_t_fmt, 0}, {"LC_MESSAGES", tok_lc_messages, 0}, - {"",}, - {"charconv", tok_charconv, 0}, + {"",}, {"",}, {"UNDEFINED", tok_undefined, 0}, {"LC_NUMERIC", tok_lc_numeric, 0}, {"",}, {"collating-element", tok_collating_element, 0}, {"position", tok_position, 0}, - {"copy", tok_copy, 0}, - {"print", tok_print, 0}, {"",}, - {"toupper", tok_toupper, 0}, + {"alpha", tok_alpha, 0}, + {"",}, {"",}, {"positive_sign", tok_positive_sign, 0}, {"",}, {"d_fmt", tok_d_fmt, 0}, + {"",}, + {"forward", tok_forward, 0}, {"",}, {"",}, - {"era", tok_era, 0}, - {"p_sep_by_space", tok_p_sep_by_space, 0}, - {"LC_COLLATE", tok_lc_collate, 0}, - {"noexpr", tok_noexpr, 0}, - {"tolower", tok_tolower, 0}, - {"day", tok_day, 0}, - {"era_t_fmt", tok_era_t_fmt, 0}, + {"abmon", tok_abmon, 0}, + {"collating-symbol", tok_collating_symbol, 0}, + {"d_t_fmt", tok_d_t_fmt, 0}, + {"backward", tok_backward, 0}, + {"",}, {"punct", tok_punct, 0}, - {"LC_MONETARY", tok_lc_monetary, 0}, - {"comment_char", tok_comment_char, 0}, + {"",}, {"",}, {"",}, + {"p_sep_by_space", tok_p_sep_by_space, 0}, + {"digit", tok_digit, 0}, + {"",}, {"",}, {"",}, {"",}, + {"cntrl", tok_cntrl, 0}, + {"p_sign_posn", tok_p_sign_posn, 0}, {"",}, + {"charconv", tok_charconv, 0}, {"n_sep_by_space", tok_n_sep_by_space, 0}, - {"digit", tok_digit, 0}, - {"order_start", tok_order_start, 0}, - {"forward", tok_forward, 0}, + {"print", tok_print, 0}, + {"xdigit", tok_xdigit, 0}, + {"toupper", tok_toupper, 0}, {"negative_sign", tok_negative_sign, 0}, {"",}, - {"nostr", tok_nostr, 0}, - {"yesstr", tok_yesstr, 0}, - {"d_t_fmt", tok_d_t_fmt, 0}, - {"",}, - {"era_d_fmt", tok_era_d_fmt, 0}, - {"alpha", tok_alpha, 0}, - {"era_d_t_fmt", tok_era_d_t_fmt, 0}, + {"LC_COLLATE", tok_lc_collate, 0}, + {"n_sign_posn", tok_n_sign_posn, 0}, + {"tolower", tok_tolower, 0}, + {"",}, {"",}, + {"int_curr_symbol", tok_int_curr_symbol, 0}, + {"noexpr", tok_noexpr, 0}, {"",}, {"mon", tok_mon, 0}, - {"order_end", tok_order_end, 0}, + {"copy", tok_copy, 0}, {"t_fmt_ampm", tok_t_fmt_ampm, 0}, - {"xdigit", tok_xdigit, 0}, + {"LC_MONETARY", tok_lc_monetary, 0}, {"mon_thousands_sep", tok_mon_thousands_sep, 0}, - {"",}, {"",}, {"",}, - {"collating-symbol", tok_collating_symbol, 0}, - {"yesexpr", tok_yesexpr, 0}, - {"era_year", tok_era_year, 0}, - {"charclass", tok_charclass, 0}, - {"upper", tok_upper, 0}, - {"p_sign_posn", tok_p_sign_posn, 0}, + {"era", tok_era, 0}, + {"",}, {"",}, {"",}, {"",}, + {"p_cs_precedes", tok_p_cs_precedes, 0}, + {"era_t_fmt", tok_era_t_fmt, 0}, + {"blank", tok_blank, 0}, {"",}, - {"thousands_sep", tok_thousands_sep, 0}, + {"comment_char", tok_comment_char, 0}, + {"day", tok_day, 0}, {"",}, - {"graph", tok_graph, 0}, + {"currency_symbol", tok_currency_symbol, 0}, {"",}, {"mon_decimal_point", tok_mon_decimal_point, 0}, - {"p_cs_precedes", tok_p_cs_precedes, 0}, + {"n_cs_precedes", tok_n_cs_precedes, 0}, + {"",}, {"",}, {"",}, {"",}, {"",}, + {"era_d_fmt", tok_era_d_fmt, 0}, + {"alt_digits", tok_alt_digits, 0}, + {"era_d_t_fmt", tok_era_d_t_fmt, 0}, {"",}, - {"space", tok_space, 0}, - {"n_sign_posn", tok_n_sign_posn, 0}, + {"grouping", tok_grouping, 0}, {"",}, + {"space", tok_space, 0}, + {"",}, {"",}, {"decimal_point", tok_decimal_point, 0}, + {"charclass", tok_charclass, 0}, + {"int_frac_digits", tok_int_frac_digits, 0}, + {"order_start", tok_order_start, 0}, + {"mon_grouping", tok_mon_grouping, 0}, + {"thousands_sep", tok_thousands_sep, 0}, {"from", tok_from, 0}, + {"nostr", tok_nostr, 0}, + {"",}, {"",}, {"",}, {"",}, {"lower", tok_lower, 0}, - {"",}, {"",}, - {"n_cs_precedes", tok_n_cs_precedes, 0}, + {"",}, {"",}, {"",}, + {"order_end", tok_order_end, 0}, {"",}, - {"abmon", tok_abmon, 0}, - {"escape_char", tok_escape_char, 0}, + {"frac_digits", tok_frac_digits, 0}, {"",}, {"",}, {"",}, - {"int_curr_symbol", tok_int_curr_symbol, 0}, + {"alnum", tok_alnum, 0}, {"",}, {"",}, - {"backward", tok_backward, 0}, + {"repertoiremap", tok_repertoiremap, 0}, {"",}, + {"upper", tok_upper, 0}, + {"escape_char", tok_escape_char, 0}, + {"",}, {"",}, {"",}, {"abday", tok_abday, 0}, - {"",}, {"",}, {"",}, {"",}, - {"currency_symbol", tok_currency_symbol, 0}, - {"frac_digits", tok_frac_digits, 0}, - {"",}, - {"grouping", tok_grouping, 0}, + {"yesstr", tok_yesstr, 0}, + {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, - {"cntrl", tok_cntrl, 0}, - {"",}, {"",}, {"",}, {"",}, - {"blank", tok_blank, 0}, - {"",}, {"",}, {"",}, {"",}, - {"int_frac_digits", tok_int_frac_digits, 0}, - {"",}, {"",}, {"",}, {"",}, - {"alt_digits", tok_alt_digits, 0}, + {"yesexpr", tok_yesexpr, 0}, + {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, + {"graph", tok_graph, 0}, {"",}, {"",}, {"",}, {"",}, {"am_pm", tok_am_pm, 0}, + {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, + {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, - {"alnum", tok_alnum, 0}, - {"",}, - {"mon_grouping", tok_mon_grouping, 0}, + {"era_year", tok_era_year, 0}, }; if (len <= MAX_WORD_LENGTH && len >= MIN_WORD_LENGTH) diff --git a/locale/programs/locfile-token.h b/locale/programs/locfile-token.h index 6d1543c320..7845b4ba76 100644 --- a/locale/programs/locfile-token.h +++ b/locale/programs/locfile-token.h @@ -1,4 +1,4 @@ -/* Copyright (C) 1996, 1997 Free Software Foundation, Inc. +/* Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. This file is part of the GNU C Library. Contributed by Ulrich Drepper , 1996. @@ -49,6 +49,8 @@ enum token_t tok_g2esc, tok_g3esc, + tok_charids, + tok_code_set_name, tok_mb_cur_max, tok_mb_cur_min, @@ -56,6 +58,7 @@ enum token_t tok_width, tok_width_variable, tok_width_default, + tok_repertoiremap, tok_lc_ctype, tok_copy, diff --git a/locale/programs/locfile.h b/locale/programs/locfile.h index 604e726e20..697af64979 100644 --- a/locale/programs/locfile.h +++ b/locale/programs/locfile.h @@ -1,4 +1,4 @@ -/* Copyright (C) 1996, 1997 Free Software Foundation, Inc. +/* Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. This file is part of the GNU C Library. Contributed by Ulrich Drepper , 1996. @@ -55,6 +55,7 @@ struct localedef_t /* Declared in localedef.c. */ extern int be_quiet; +extern const char *repertoiremap; /* Found in localedef.c. */ void def_to_process (const char *name, int category); diff --git a/locale/programs/repertoire.c b/locale/programs/repertoire.c new file mode 100644 index 0000000000..1f219ec81f --- /dev/null +++ b/locale/programs/repertoire.c @@ -0,0 +1,323 @@ +/* Copyright (C) 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Ulrich Drepper , 1998. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include +#include + +#include "linereader.h" +#include "charset.h" +#include "repertoire.h" +#include "simple-hash.h" + + +extern void *xmalloc (size_t __n); + + +/* Simple keyword hashing for the repertoiremap. */ +static struct repertoire_t *parse_repertoiremap (const char *filename); +static const struct keyword_t *repertoiremap_hash (const char *str, int len); + + +struct repertoire_t * +repertoire_read (const char *filename) +{ + const char *pathnfile; + struct repertoire_t *result = NULL; + + if (euidaccess (filename, R_OK) >= 0) + pathnfile = filename; + else if (filename[0] != '/') + { + char *cp = xmalloc (strlen (filename) + sizeof CHARMAP_PATH + 1); + stpcpy (stpcpy (stpcpy (cp, CHARMAP_PATH), "/"), filename); + + pathnfile = (const char *) cp; + } + else + pathnfile = NULL; + + if (pathnfile != NULL) + { + result = parse_repertoiremap (pathnfile); + + if (result == NULL && !be_quiet) + error (0, errno, _("repertoire map file `%s' not found"), filename); + } + + return result; +} + + +static struct repertoire_t * +parse_repertoiremap (const char *filename) +{ + struct linereader *cmfile; + struct repertoire_t *result; + int state; + char *from_name = NULL; + char *to_name = NULL; + + /* Determine path. */ + cmfile = lr_open (filename, repertoiremap_hash); + if (cmfile == NULL) + { + if (strchr (filename, '/') == NULL) + { + /* Look in the systems charmap directory. */ + char *buf = xmalloc (strlen (filename) + 1 + + sizeof (REPERTOIREMAP_PATH)); + + stpcpy (stpcpy (stpcpy (buf, REPERTOIREMAP_PATH), "/"), filename); + cmfile = lr_open (buf, repertoiremap_hash); + + if (cmfile == NULL) + free (buf); + } + + if (cmfile == NULL) + return NULL; + } + + /* Allocate room for result. */ + result = (struct repertoire_t *) xmalloc (sizeof (struct repertoire_t)); + memset (result, '\0', sizeof (struct repertoire_t)); + +#define obstack_chunk_alloc malloc +#define obstack_chunk_free free + obstack_init (&result->mem_pool); + + if (init_hash (&result->char_table, 256)) + { + free (result); + return NULL; + } + + /* We use a state machine to describe the charmap description file + format. */ + state = 1; + while (1) + { + /* What's on? */ + struct token *now = lr_token (cmfile, NULL); + enum token_t nowtok = now->tok; + struct token *arg; + + if (nowtok == tok_eof) + break; + + switch (state) + { + case 1: + /* We haven't yet read any character definition. This is where + we accept escape_char and comment_char definitions. */ + if (nowtok == tok_eol) + /* Ignore empty lines. */ + continue; + + if (nowtok == tok_escape_char || nowtok == tok_comment_char) + { + /* We know that we need an argument. */ + arg = lr_token (cmfile, NULL); + + if (arg->tok != tok_ident) + { + lr_error (cmfile, _("syntax error in prolog: %s"), + _("bad argument")); + + lr_ignore_rest (cmfile, 0); + continue; + } + + if (arg->val.str.len != 1) + { + lr_error (cmfile, _("\ +argument to <%s> must be a single character"), + nowtok == tok_escape_char ? "escape_char" + : "comment_char"); + + lr_ignore_rest (cmfile, 0); + continue; + } + + if (nowtok == tok_escape_char) + cmfile->escape_char = *arg->val.str.start; + else + cmfile->comment_char = *arg->val.str.start; + + lr_ignore_rest (cmfile, 1); + continue; + } + + if (nowtok == tok_charids) + { + lr_ignore_rest (cmfile, 1); + + state = 2; + continue; + } + + /* Otherwise we start reading the character definitions. */ + state = 2; + /* FALLTHROUGH */ + + case 2: + /* We are now are in the body. Each line + must have the format "%s %s %s\n" or "%s...%s %s %s\n". */ + if (nowtok == tok_eol) + /* Ignore empty lines. */ + continue; + + if (nowtok == tok_end) + { + state = 90; + continue; + } + + if (nowtok != tok_bsymbol) + { + lr_error (cmfile, + _("syntax error in repertoire map definition: %s"), + _("no symbolic name given")); + + lr_ignore_rest (cmfile, 0); + continue; + } + + /* If the previous line was not completely correct free the + used memory. */ + if (from_name != NULL) + obstack_free (&result->mem_pool, from_name); + + from_name = (char *) obstack_copy0 (&result->mem_pool, + now->val.str.start, + now->val.str.len); + to_name = NULL; + + state = 3; + continue; + + case 3: + /* We have two possibilities: We can see an ellipsis or an + encoding value. */ + if (nowtok == tok_ellipsis) + { + state = 4; + continue; + } + /* FALLTHROUGH */ + + case 5: + /* We expect a value of the form or where + the xxx mean a hexadecimal value. */ + state = 2; + + errno = 0; + if (nowtok != tok_ucs2 && nowtok != tok_ucs4) + { + lr_error (cmfile, + _("syntax error in repertoire map definition: %s"), + _("no or value given")); + + lr_ignore_rest (cmfile, 0); + continue; + } + + /* We've found a new valid definition. */ + charset_new_char (cmfile, &result->char_table, 4, + now->val.charcode.val, from_name, to_name); + + /* Ignore the rest of the line. */ + lr_ignore_rest (cmfile, 0); + + from_name = NULL; + to_name = NULL; + + continue; + + case 4: + if (nowtok != tok_bsymbol) + { + lr_error (cmfile, + _("syntax error in repertoire map definition: %s"), + _("no symbolic name given for end of range")); + + lr_ignore_rest (cmfile, 0); + state = 2; + continue; + } + + /* Copy the to-name in a safe place. */ + to_name = (char *) obstack_copy0 (&result->mem_pool, + cmfile->token.val.str.start, + cmfile->token.val.str.len); + + state = 5; + continue; + + case 90: + if (nowtok != tok_charids) + lr_error (cmfile, _("\ +`%1$s' definition does not end with `END %1$s'"), "CHARIDS"); + + lr_ignore_rest (cmfile, nowtok == tok_charids); + break; + } + + break; + } + + if (state != 2 && state != 90 && !be_quiet) + error (0, 0, _("%s: premature end of file"), cmfile->fname); + + lr_close (cmfile); + + return result; +} + + +static const struct keyword_t * +repertoiremap_hash (const char *str, int len) +{ + static const struct keyword_t wordlist[0] = + { + {"escape_char", tok_escape_char, 1}, + {"comment_char", tok_comment_char, 1}, + {"CHARIDS", tok_charids, 0}, + {"END", tok_end, 0}, + }; + + if (len == 11 && memcmp (wordlist[0].name, str, 11) == 0) + return &wordlist[0]; + if (len == 12 && memcmp (wordlist[1].name, str, 12) == 0) + return &wordlist[1]; + if (len == 7 && memcmp (wordlist[2].name, str, 7) == 0) + return &wordlist[2]; + if (len == 3 && memcmp (wordlist[3].name, str, 3) == 0) + return &wordlist[3]; + + return NULL; +} diff --git a/locale/programs/repertoire.h b/locale/programs/repertoire.h new file mode 100644 index 0000000000..7befeb4e0d --- /dev/null +++ b/locale/programs/repertoire.h @@ -0,0 +1,38 @@ +/* Copyright (C) 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Ulrich Drepper , 1998. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef _REPERTOIREMAP_H +#define _REPERTOIREMAP_H 1 + +#include + +#include "simple-hash.h" +#include "linereader.h" + +struct repertoire_t +{ + struct obstack mem_pool; + hash_table char_table; +}; + + +/* Prototypes for repertoire map handling functions. */ +struct repertoire_t *repertoire_read (const char *filename); + +#endif /* repertoiremap.h */ diff --git a/locale/programs/stringtrans.c b/locale/programs/stringtrans.c index 34b107e295..b810129678 100644 --- a/locale/programs/stringtrans.c +++ b/locale/programs/stringtrans.c @@ -1,4 +1,4 @@ -/* Copyright (C) 1996, 1997 Free Software Foundation, Inc. +/* Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. This file is part of the GNU C Library. Contributed by Ulrich Drepper , 1996. @@ -92,7 +92,8 @@ translate_string (char *str, struct charset_t *charset) return NULL; } - value = charset_find_value (charset, str + 1, tp - (str + 1)); + value = charset_find_value (&charset->char_table, str + 1, + tp - (str + 1)); if ((wchar_t) value == ILLEGAL_CHAR_VALUE) { free (buf); -- cgit v1.2.3