1 %{ // -*- mode: c++; c-file-style: "linux"; indent-tabs-mode: t -*-
3 This file is part of LilyPond, the GNU music typesetter.
5 Copyright (C) 1996--2012 Han-Wen Nienhuys <hanwen@xs4all.nl>
6 Jan Nieuwenhuizen <janneke@gnu.org>
8 LilyPond is free software: you can redistribute it and/or modify
9 it under the terms of the GNU General Public License as published by
10 the Free Software Foundation, either version 3 of the License, or
11 (at your option) any later version.
13 LilyPond is distributed in the hope that it will be useful,
14 but WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 GNU General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with LilyPond. If not, see <http://www.gnu.org/licenses/>.
22 /* Mode and indentation are at best a rough approximation based on TAB
23 * formatting (reasonable for compatibility with unspecific editor
24 * modes as Flex modes are hard to find) and need manual correction
25 * frequently. Without a reasonably dependable way of formatting a
26 * Flex file sensibly, there is little point in trying to fix the
27 * inconsistent state of indentation.
33 after making a change to the lexer rules, run
34 flex -b <this lexer file>
37 contains no backup states, but only the reminder
38 Compressed tables always back up.
39 (don-t forget to rm lex.yy.cc :-)
48 /* Flex >= 2.5.29 fix; FlexLexer.h's multiple include bracing breaks
49 when building the actual lexer. */
56 #include "context-def.hh"
57 #include "duration.hh"
58 #include "international.hh"
59 #include "interval.hh"
60 #include "lily-guile.hh"
61 #include "lily-lexer.hh"
62 #include "lily-parser.hh"
63 #include "lilypond-version.hh"
66 #include "music-function.hh"
67 #include "parse-scm.hh"
70 #include "source-file.hh"
71 #include "std-string.hh"
72 #include "string-convert.hh"
79 #define isatty HORRIBLEKLUDGE
81 void strip_trailing_white (string&);
82 void strip_leading_white (string&);
83 string lyric_fudge (string s);
84 SCM lookup_markup_command (string s);
85 SCM lookup_markup_list_command (string s);
86 bool is_valid_version (string s);
89 #define start_quote() do { \
90 yy_push_state (quote); \
94 #define start_lyric_quote() do { \
95 yy_push_state (lyric_quote); \
99 #define yylval (*lexval_)
101 #define yylloc (*lexloc_)
103 #define YY_USER_ACTION add_lexed_char (YYLeng ());
106 SCM scan_fraction (string);
107 SCM (* scm_parse_error_handler) (void *);
117 %option yyclass="Lily_lexer"
119 %option never-interactive
136 /* The strategy concerning multibyte characters is to accept them but
137 * call YYText_utf8 for patterns that might contain them, in order to
138 * get a single code path responsible for flagging non-UTF-8 input:
139 * Patterns for accepting only valid UTF-8 without backing up are
140 * really hard to do and complex, and if nice error messages are
141 * wanted, one would need patterns catching the invalid input as well.
143 * Since editors and operating environments don't necessarily behave
144 * reasonably in the presence of mixed encodings, we flag encoding
145 * errors also in identifiers, comments, and strings where it would be
146 * conceivable to just transparently work with the byte string. But
147 * the whole point of caring about UTF-8 in here at all is too avoid
148 * stranger errors later when input passes into backends or log files
149 * or console output or error messages.
156 WORD {A}([-_]{A}|{A})*
163 REAL ({INT}\.{N}*)|(-?\.{N}+)
165 HORIZONTALWHITE [ \t]
171 BOM_UTF8 \357\273\277
177 // swallow and ignore carriage returns
180 <extratoken>{ANY_CHAR} {
181 /* Generate a token without swallowing anything */
183 /* First unswallow the eaten character */
184 add_lexed_char (-YYLeng ());
187 /* produce requested token */
188 int type = scm_to_int (scm_caar (extra_tokens_));
189 yylval = scm_cdar (extra_tokens_);
190 extra_tokens_ = scm_cdr (extra_tokens_);
191 if (scm_is_null (extra_tokens_))
197 <extratoken><<EOF>> {
198 /* Generate a token without swallowing anything */
200 /* produce requested token */
201 int type = scm_to_int (scm_caar (extra_tokens_));
202 yylval = scm_cdar (extra_tokens_);
203 extra_tokens_ = scm_cdr (extra_tokens_);
204 if (scm_is_null (extra_tokens_))
210 /* Use the trailing context feature. Otherwise, the BOM will not be
211 found if the file starts with an identifier definition. */
212 <INITIAL,chords,lyrics,figures,notes>{BOM_UTF8}/.* {
213 if (this->lexloc_->line_number () != 1 || this->lexloc_->column_number () != 0)
215 LexerWarning (_ ("stray UTF-8 BOM encountered").c_str ());
218 debug_output (_ ("Skipping UTF-8 BOM"));
221 <INITIAL,chords,figures,incl,lyrics,markup,notes>{
223 yy_push_state (longcomment);
225 %[^{\n\r][^\n\r]*[\n\r] {
226 (void) YYText_utf8 ();
228 %[^{\n\r] { // backup rule
229 (void) YYText_utf8 ();
234 (void) YYText_utf8 ();
241 <INITIAL,notes,figures,chords,markup>{
247 <INITIAL,chords,lyrics,notes,figures>\\version{WHITE}* {
248 yy_push_state (version);
250 <INITIAL,chords,lyrics,notes,figures>\\sourcefilename{WHITE}* {
251 yy_push_state (sourcefilename);
253 <INITIAL,chords,lyrics,notes,figures>\\sourcefileline{WHITE}* {
254 yy_push_state (sourcefileline);
256 <version>\"[^"]*\" { /* got the version number */
257 string s (YYText_utf8 () + 1);
258 s = s.substr (0, s.rfind ('\"'));
262 SCM top_scope = scm_car (scm_last_pair (scopes_));
263 scm_module_define (top_scope, ly_symbol2scm ("version-seen"), SCM_BOOL_T);
265 if (!is_valid_version (s)) {
266 yylval = SCM_UNSPECIFIED;
270 <sourcefilename>\"[^""]*\" {
271 string s (YYText_utf8 () + 1);
272 s = s.substr (0, s.rfind ('\"'));
275 this->here_input().get_source_file ()->name_ = s;
276 message (_f ("Renaming input to: `%s'", s.c_str ()));
277 progress_indication ("\n");
278 scm_module_define (scm_car (scopes_),
279 ly_symbol2scm ("input-file-name"),
284 <sourcefileline>{INT} {
286 sscanf (YYText (), "%d", &i);
289 this->here_input ().get_source_file ()->set_line (here_input ().start (), i);
292 <version>{ANY_CHAR} {
293 LexerError (_ ("quoted string expected after \\version").c_str ());
296 <sourcefilename>{ANY_CHAR} {
297 LexerError (_ ("quoted string expected after \\sourcefilename").c_str ());
300 <sourcefileline>{ANY_CHAR} {
301 LexerError (_ ("integer expected after \\sourcefileline").c_str ());
306 (void) YYText_utf8 ();
309 (void) YYText_utf8 ();
317 <INITIAL,chords,lyrics,notes,figures>\\maininput {
321 is_main_input_ = true;
324 error (_ ("\\maininput not allowed outside init files"));
327 <INITIAL,chords,lyrics,figures,notes>\\include {
328 yy_push_state (incl);
330 <incl>\"[^""]*\" { /* got the include file name */
331 string s (YYText_utf8 ()+1);
332 s = s.substr (0, s.rfind ('"'));
334 new_input (s, sources_);
337 <incl>\\{BLACK}*{WHITE}? { /* got the include identifier */
338 string s = YYText_utf8 () + 1;
339 strip_trailing_white (s);
340 if (s.length () && (s[s.length () - 1] == ';'))
341 s = s.substr (0, s.length () - 1);
343 SCM sid = lookup_identifier (s);
344 if (scm_is_string (sid)) {
345 new_input (ly_scm2string (sid), sources_);
348 string msg (_f ("wrong or undefined identifier: `%s'", s ));
350 LexerError (msg.c_str ());
351 SCM err = scm_current_error_port ();
352 scm_puts ("This value was found in the table: ", err);
353 scm_display (sid, err);
356 <incl>(\$|#) { // scm for the filename
358 Input hi = here_input();
360 SCM sval = ly_parse_scm (hi.start (), &n, hi,
361 be_safe_global && is_main_input_, parser_);
362 sval = eval_scm (sval);
364 for (int i = 0; i < n; i++)
368 char_count_stack_.back () += n;
370 if (scm_is_string (sval)) {
371 new_input (ly_scm2string (sval), sources_);
374 LexerError (_ ("string expected after \\include").c_str ());
375 if (sval != SCM_UNDEFINED) {
376 SCM err = scm_current_error_port ();
377 scm_puts ("This value was found instead: ", err);
378 scm_display (sval, err);
383 <incl,version,sourcefilename>\"[^""]* { // backup rule
384 error (_ ("end quote missing"));
388 /* Flex picks the longest matching pattern including trailing
389 * contexts. Without the backup pattern, r-. does not trigger the
390 * {RESTNAME} rule but rather the {WORD}/[-_] rule coming later,
391 * needed for avoiding backup states.
394 <chords,notes,figures>{RESTNAME}/[-_] | // pseudo backup rule
395 <chords,notes,figures>{RESTNAME} {
396 char const *s = YYText ();
397 yylval = scm_from_locale_string (s);
400 <chords,notes,figures>q/[-_] | // pseudo backup rule
401 <chords,notes,figures>q {
402 yylval = SCM_UNSPECIFIED;
403 return CHORD_REPETITION;
406 <chords,notes,figures>R/[-_] | // pseudo backup rule
407 <chords,notes,figures>R {
408 yylval = SCM_UNSPECIFIED;
409 return MULTI_MEASURE_REST;
411 <INITIAL,chords,figures,lyrics,markup,notes># { //embedded scm
413 Input hi = here_input();
415 SCM sval = ly_parse_scm (hi.start (), &n, hi,
416 be_safe_global && is_main_input_, parser_);
418 if (sval == SCM_UNDEFINED)
421 for (int i = 0; i < n; i++)
425 char_count_stack_.back () += n;
431 <INITIAL,chords,figures,lyrics,markup,notes>\$ { //immediate scm
433 Input hi = here_input();
435 SCM sval = ly_parse_scm (hi.start (), &n, hi,
436 be_safe_global && is_main_input_, parser_);
438 for (int i = 0; i < n; i++)
442 char_count_stack_.back () += n;
444 sval = eval_scm (sval, '$');
446 int token = scan_scm_id (sval);
447 if (!scm_is_eq (yylval, SCM_UNSPECIFIED))
451 <INITIAL,notes,lyrics>{
453 yylval = SCM_UNSPECIFIED;
454 return DOUBLE_ANGLE_OPEN;
457 yylval = SCM_UNSPECIFIED;
458 return DOUBLE_ANGLE_CLOSE;
464 yylval = SCM_UNSPECIFIED;
468 yylval = SCM_UNSPECIFIED;
475 yylval = SCM_UNSPECIFIED;
479 yylval = SCM_UNSPECIFIED;
483 yylval = SCM_UNSPECIFIED;
489 {WORD}/[-_] | // backup rule
491 return scan_bare_word (YYText_utf8 ());
494 {COMMAND}/[-_] | // backup rule
496 return scan_escaped_word (YYText_utf8 () + 1);
499 yylval = scan_fraction (YYText ());
502 {UNSIGNED}/\/ | // backup rule
504 yylval = scm_c_read_string (YYText ());
508 yylval = scm_c_read_string (YYText () + 1);
515 char c = escaped_char (YYText ()[1]);
516 yylval = scm_cons (scm_from_locale_stringn (&c, 1),
520 yylval = scm_cons (scm_from_locale_string (YYText_utf8 ()),
527 /* yylval is union. Must remember STRING before setting SCM*/
529 yylval = scm_string_concatenate_reverse (yylval,
533 return is_lyric_state () ? LYRICS_STRING : STRING;
536 yylval = scm_cons (scm_from_locale_string (YYText ()),
543 start_lyric_quote ();
546 yylval = scan_fraction (YYText ());
549 {UNSIGNED}/\/ | // backup rule
551 yylval = scm_c_read_string (YYText ());
554 {COMMAND}/[-_] | // backup rule
556 return scan_escaped_word (YYText_utf8 () + 1);
558 /* Characters needed to express durations, assignments, barchecks */
560 yylval = SCM_UNSPECIFIED;
563 [^$#{}\"\\ \t\n\r\f0-9]+ {
565 string s (YYText_utf8 ());
566 yylval = SCM_UNSPECIFIED;
572 yylval = ly_string2scm (s);
574 return LYRICS_STRING;
576 /* This should really just cover {} */
578 yylval = SCM_UNSPECIFIED;
579 return YYText ()[0]; // above catches all multibytes.
583 {WORD}/[-_] | // backup rule
585 return scan_bare_word (YYText_utf8 ());
587 {COMMAND}/[-_] | // backup rule
589 return scan_escaped_word (YYText_utf8 () + 1);
592 yylval = scan_fraction (YYText ());
595 {UNSIGNED}/\/ | // backup rule
597 yylval = scm_c_read_string (YYText ());
601 yylval = SCM_UNSPECIFIED;
605 yylval = SCM_UNSPECIFIED;
609 yylval = SCM_UNSPECIFIED;
613 yylval = SCM_UNSPECIFIED;
617 yylval = SCM_UNSPECIFIED;
621 yylval = SCM_UNSPECIFIED;
622 return YYText ()[0]; // WORD catches all multibyte.
629 yylval = SCM_UNSPECIFIED;
632 {COMMAND}/[-_] | // backup rule
634 string str (YYText_utf8 () + 1);
636 int token_type = MARKUP_FUNCTION;
637 SCM s = lookup_markup_command (str);
639 // lookup-markup-command returns a pair with the car
640 // being the function to call, and the cdr being the
641 // call signature specified to define-markup-command,
642 // a list of predicates.
644 if (!scm_is_pair (s)) {
645 // If lookup-markup-command was not successful, we
646 // try lookup-markup-list-command instead.
647 // If this fails as well, we just scan and return
649 s = lookup_markup_list_command (str);
651 token_type = MARKUP_LIST_FUNCTION;
653 return scan_escaped_word (str);
656 // If the list of predicates is, say,
657 // (number? number? markup?), then tokens
658 // EXPECT_MARKUP EXPECT_SCM EXPECT_SCM EXPECT_NO_MORE_ARGS
659 // will be generated. Note that we have to push them
660 // in reverse order, so the first token pushed in the
661 // loop will be EXPECT_NO_MORE_ARGS.
665 // yylval now contains the function to call as token
666 // value (for token type MARKUP_FUNCTION or
667 // MARKUP_LIST_FUNCTION).
669 push_extra_token(EXPECT_NO_MORE_ARGS);
671 for (; scm_is_pair(s); s = scm_cdr(s)) {
672 SCM predicate = scm_car(s);
674 if (predicate == ly_lily_module_constant ("markup-list?"))
675 push_extra_token(EXPECT_MARKUP_LIST);
676 else if (predicate == ly_lily_module_constant ("markup?"))
677 push_extra_token(EXPECT_MARKUP);
679 push_extra_token(EXPECT_SCM, predicate);
683 [^$#{}\"\\ \t\n\r\f]+ {
684 string s (YYText_utf8 ());
686 yylval = ly_string2scm (s);
690 yylval = SCM_UNSPECIFIED;
691 return YYText ()[0]; // Above is catchall for multibyte
695 <longcomment><<EOF>> {
696 LexerError (_ ("EOF found inside a comment").c_str ());
697 is_main_input_ = false; // should be safe , can't have \include in --safe.
698 yylval = SCM_UNSPECIFIED;
700 yyterminate (); // can't move this, since it actually rets a YY_NULL
704 yylval = SCM_UNSPECIFIED;
707 /* 2 = init.ly + current file.
708 > because we're before closing, but is_main_input_ should
711 is_main_input_ = include_stack_.size () > 2;
712 if (!close_input () || !is_main_input_)
713 /* Returns YY_NULL */
716 else if (!close_input ())
717 /* Returns YY_NULL */
722 {WORD}/[-_] | // backup rule
724 return scan_bare_word (YYText_utf8 ());
726 {COMMAND}/[-_] | // backup rule
728 return scan_escaped_word (YYText_utf8 () + 1);
733 yylval = scan_fraction (YYText ());
737 -{UNSIGNED} | // backup rule
739 yylval = scm_c_read_string (YYText ());
743 {UNSIGNED}/\/ | // backup rule
745 yylval = scm_c_read_string (YYText ());
751 yylval = SCM_UNSPECIFIED;
755 -/\. | // backup rule
757 yylval = SCM_UNSPECIFIED;
761 <INITIAL,notes,figures>. {
762 yylval = SCM_UNSPECIFIED;
766 <INITIAL,lyrics,notes,figures>\\. {
767 yylval = SCM_UNSPECIFIED;
768 char c = YYText ()[1];
772 return E_ANGLE_CLOSE;
776 return E_EXCLAMATION;
782 return E_BRACKET_OPEN;
786 return E_BRACKET_CLOSE;
798 string msg = _f ("invalid character: `%s'", YYText_utf8 ());
799 LexerError (msg.c_str ());
800 yylval = SCM_UNSPECIFIED;
801 return '%'; // Better not return half a utf8 character.
806 /* Make the lexer generate a token of the given type as the next token.
807 TODO: make it possible to define a value for the token as well */
809 Lily_lexer::push_extra_token (int token_type, SCM scm)
811 if (scm_is_null (extra_tokens_))
813 if (YY_START != extratoken)
814 hidden_state_ = YY_START;
815 yy_push_state (extratoken);
817 extra_tokens_ = scm_acons (scm_from_int (token_type), scm, extra_tokens_);
821 Lily_lexer::push_chord_state (SCM alist)
823 SCM p = scm_assq (alist, pitchname_tab_stack_);
825 if (scm_is_false (p))
826 p = scm_cons (alist, alist_to_hashq (alist));
827 pitchname_tab_stack_ = scm_cons (p, pitchname_tab_stack_);
828 yy_push_state (chords);
832 Lily_lexer::push_figuredbass_state ()
834 yy_push_state (figures);
838 Lily_lexer::push_initial_state ()
840 yy_push_state (INITIAL);
844 Lily_lexer::push_lyric_state ()
846 yy_push_state (lyrics);
850 Lily_lexer::push_markup_state ()
852 yy_push_state (markup);
856 Lily_lexer::push_note_state (SCM alist)
858 bool extra = (YYSTATE == extratoken);
860 SCM p = scm_assq (alist, pitchname_tab_stack_);
865 if (scm_is_false (p))
866 p = scm_cons (alist, alist_to_hashq (alist));
867 pitchname_tab_stack_ = scm_cons (p, pitchname_tab_stack_);
868 yy_push_state (notes);
871 hidden_state_ = YYSTATE;
872 yy_push_state (extratoken);
877 Lily_lexer::pop_state ()
879 bool extra = (YYSTATE == extratoken);
884 if (YYSTATE == notes || YYSTATE == chords)
885 pitchname_tab_stack_ = scm_cdr (pitchname_tab_stack_);
890 hidden_state_ = YYSTATE;
891 yy_push_state (extratoken);
896 Lily_lexer::identifier_type (SCM sid)
898 int k = try_special_identifiers (&yylval , sid);
899 return k >= 0 ? k : SCM_IDENTIFIER;
904 Lily_lexer::scan_escaped_word (string str)
906 // use more SCM for this.
908 // SCM sym = ly_symbol2scm (str.c_str ());
910 yylval = SCM_UNSPECIFIED;
911 int i = lookup_keyword (str);
912 if (i == MARKUP && is_lyric_state ())
917 SCM sid = lookup_identifier (str);
918 if (sid != SCM_UNDEFINED)
919 return scan_scm_id (sid);
921 string msg (_f ("unknown escaped string: `\\%s'", str));
922 LexerError (msg.c_str ());
924 yylval = ly_string2scm (str);
930 Lily_lexer::scan_scm_id (SCM sid)
932 if (is_music_function (sid))
934 int funtype = SCM_FUNCTION;
938 SCM s = get_music_function_signature (sid);
939 SCM cs = scm_car (s);
941 if (scm_is_pair (cs))
946 if (scm_is_eq (cs, ly_lily_module_constant ("ly:music?")))
947 funtype = MUSIC_FUNCTION;
948 else if (scm_is_eq (cs, ly_lily_module_constant ("ly:event?")))
949 funtype = EVENT_FUNCTION;
950 else if (ly_is_procedure (cs))
951 funtype = SCM_FUNCTION;
952 else programming_error ("Bad syntax function predicate");
954 push_extra_token (EXPECT_NO_MORE_ARGS);
955 for (s = scm_cdr (s); scm_is_pair (s); s = scm_cdr (s))
957 SCM optional = SCM_UNDEFINED;
960 if (scm_is_pair (cs))
962 optional = SCM_CDR (cs);
966 if (cs == Pitch_type_p_proc)
967 push_extra_token (EXPECT_PITCH);
968 else if (cs == Duration_type_p_proc)
969 push_extra_token (EXPECT_DURATION);
970 else if (ly_is_procedure (cs))
971 push_extra_token (EXPECT_SCM, cs);
974 programming_error ("Function parameter without type-checking predicate");
977 if (!scm_is_eq (optional, SCM_UNDEFINED))
978 push_extra_token (EXPECT_OPTIONAL, optional);
983 return identifier_type (sid);
987 Lily_lexer::scan_bare_word (string str)
989 SCM sym = ly_symbol2scm (str.c_str ());
990 if ((YYSTATE == notes) || (YYSTATE == chords)) {
991 SCM handle = SCM_BOOL_F;
992 if (scm_is_pair (pitchname_tab_stack_))
993 handle = scm_hashq_get_handle (scm_cdar (pitchname_tab_stack_), sym);
995 if (scm_is_pair (handle)) {
996 yylval = scm_cdr (handle);
997 if (unsmob_pitch (yylval))
998 return (YYSTATE == notes) ? NOTENAME_PITCH : TONICNAME_PITCH;
999 else if (scm_is_symbol (yylval))
1002 else if ((YYSTATE == chords)
1003 && (handle = scm_hashq_get_handle (chordmodifier_tab_, sym))!= SCM_BOOL_F)
1005 yylval = scm_cdr (handle);
1006 return CHORD_MODIFIER;
1009 yylval = ly_string2scm (str);
1014 Lily_lexer::get_state () const
1016 if (YY_START == extratoken)
1017 return hidden_state_;
1023 Lily_lexer::is_note_state () const
1025 return get_state () == notes;
1029 Lily_lexer::is_chord_state () const
1031 return get_state () == chords;
1035 Lily_lexer::is_lyric_state () const
1037 return get_state () == lyrics;
1041 Lily_lexer::is_figure_state () const
1043 return get_state () == figures;
1046 // The extra_token parameter specifies how to convert multiple values
1047 // into additional tokens. For '#', additional values get pushed as
1048 // SCM_IDENTIFIER. For '$', they get checked for their type and get
1049 // pushed as a corresponding *_IDENTIFIER token. Since the latter
1050 // tampers with yylval, it can only be done from the lexer itself, so
1051 // this function is private.
1054 Lily_lexer::eval_scm (SCM readerdata, char extra_token)
1056 SCM sval = SCM_UNDEFINED;
1058 if (!SCM_UNBNDP (readerdata))
1060 sval = ly_eval_scm (scm_car (readerdata),
1061 *unsmob_input (scm_cdr (readerdata)),
1062 be_safe_global && is_main_input_,
1066 if (SCM_UNBNDP (sval))
1069 return SCM_UNSPECIFIED;
1072 if (extra_token && SCM_VALUESP (sval))
1074 sval = scm_struct_ref (sval, SCM_INUM0);
1076 if (scm_is_pair (sval)) {
1077 for (SCM v = scm_reverse (scm_cdr (sval));
1082 switch (extra_token) {
1084 token = scan_scm_id (scm_car (v));
1085 if (!scm_is_eq (yylval, SCM_UNSPECIFIED))
1086 push_extra_token (token, yylval);
1089 push_extra_token (SCM_IDENTIFIER, scm_car (v));
1093 sval = scm_car (sval);
1095 sval = SCM_UNSPECIFIED;
1101 /* Check for valid UTF-8 that has no overlong or surrogate codes and
1102 is in the range 0-0x10ffff */
1105 Lily_lexer::YYText_utf8 ()
1107 const char * const p = YYText ();
1108 for (int i=0; p[i];) {
1109 if ((p[i] & 0xff) < 0x80) {
1113 int oldi = i; // start of character
1114 int more = 0; // # of followup bytes, 0 if bad
1115 switch (p[i++] & 0xff) {
1116 // 0xc0 and 0xc1 are overlong prefixes for
1117 // 0x00-0x3f and 0x40-0x7f respectively, bad.
1118 case 0xc2: // 0x80-0xbf
1119 case 0xc3: // 0xc0-0xff
1120 case 0xc4: // 0x100-0x13f
1121 case 0xc5: // 0x140-0x17f
1122 case 0xc6: // 0x180-0x1bf
1123 case 0xc7: // 0x1c0-0x1ff
1124 case 0xc8: // 0x200-0x23f
1125 case 0xc9: // 0x240-0x27f
1126 case 0xca: // 0x280-0x2bf
1127 case 0xcb: // 0x2c0-0x2ff
1128 case 0xcc: // 0x300-0x33f
1129 case 0xcd: // 0x340-0x37f
1130 case 0xce: // 0x380-0x3bf
1131 case 0xcf: // 0x3c0-0x3ff
1132 case 0xd0: // 0x400-0x43f
1133 case 0xd1: // 0x440-0x47f
1134 case 0xd2: // 0x480-0x4bf
1135 case 0xd3: // 0x4c0-0x4ff
1136 case 0xd4: // 0x500-0x53f
1137 case 0xd5: // 0x540-0x57f
1138 case 0xd6: // 0x580-0x5bf
1139 case 0xd7: // 0x5c0-0x5ff
1140 case 0xd8: // 0x600-0x63f
1141 case 0xd9: // 0x640-0x67f
1142 case 0xda: // 0x680-0x6bf
1143 case 0xdb: // 0x6c0-0x6ff
1144 case 0xdc: // 0x700-0x73f
1145 case 0xdd: // 0x740-0x77f
1146 case 0xde: // 0x780-0x7bf
1147 case 0xdf: // 0x7c0-0x7ff
1148 more = 1; // 2-byte sequences, 0x80-0x7ff
1151 // don't allow overlong sequences for 0-0x7ff
1152 if ((p[i] & 0xff) < 0xa0)
1154 case 0xe1: // 0x1000-0x1fff
1155 case 0xe2: // 0x2000-0x2fff
1156 case 0xe3: // 0x3000-0x3fff
1157 case 0xe4: // 0x4000-0x4fff
1158 case 0xe5: // 0x5000-0x5fff
1159 case 0xe6: // 0x6000-0x6fff
1160 case 0xe7: // 0x7000-0x7fff
1161 case 0xe8: // 0x8000-0x8fff
1162 case 0xe9: // 0x9000-0x9fff
1163 case 0xea: // 0xa000-0xafff
1164 case 0xeb: // 0xb000-0xbfff
1165 case 0xec: // 0xc000-0xcfff
1166 more = 2; // 3-byte sequences, 0x7ff-0xcfff
1168 case 0xed: // 0xd000-0xdfff
1169 // Don't allow surrogate codes 0xd800-0xdfff
1170 if ((p[i] & 0xff) >= 0xa0)
1172 case 0xee: // 0xe000-0xefff
1173 case 0xef: // 0xf000-0xffff
1174 more = 2; // 3-byte sequences,
1175 // 0xd000-0xd7ff, 0xe000-0xffff
1178 // don't allow overlong sequences for 0-0xffff
1179 if ((p[i] & 0xff) < 0x90)
1181 case 0xf1: // 0x40000-0x7ffff
1182 case 0xf2: // 0x80000-0xbffff
1183 case 0xf3: // 0xc0000-0xfffff
1184 more = 3; // 4-byte sequences, 0x10000-0xfffff
1187 // don't allow more than 0x10ffff
1188 if ((p[i] & 0xff) >= 0x90)
1190 more = 3; // 4-byte sequence, 0x100000-0x10ffff
1194 // check that all continuation bytes are valid
1196 if ((p[i++] & 0xc0) != 0x80)
1202 Input h = here_input ();
1203 h.set (h.get_source_file (), h.start () + oldi, h.start () + i);
1204 h.warning (_ ("non-UTF-8 input").c_str ());
1211 urg, belong to string (_convert)
1212 and should be generalised
1215 strip_leading_white (string&s)
1218 for (; i < s.length (); i++)
1219 if (!isspace (s[i]))
1226 strip_trailing_white (string&s)
1228 ssize i = s.length ();
1230 if (!isspace (s[i]))
1233 s = s.substr (0, i + 1);
1238 Lilypond_version oldest_version ("2.7.38");
1242 is_valid_version (string s)
1244 Lilypond_version current ( MAJOR_VERSION "." MINOR_VERSION "." PATCH_LEVEL );
1245 Lilypond_version ver (s);
1246 if (int (ver) < oldest_version)
1248 non_fatal_error (_f ("file too old: %s (oldest supported: %s)", ver.to_string (), oldest_version.to_string ()));
1249 non_fatal_error (_ ("consider updating the input with the convert-ly script"));
1255 non_fatal_error (_f ("program too old: %s (file requires: %s)", current.to_string (), ver.to_string ()));
1266 lyric_fudge (string s)
1270 while ((i = s.find ('_', i)) != string::npos)
1278 Convert "NUM/DEN" into a '(NUM . DEN) cons.
1281 scan_fraction (string frac)
1283 ssize i = frac.find ('/');
1284 string left = frac.substr (0, i);
1285 string right = frac.substr (i + 1, (frac.length () - i + 1));
1287 int n = String_convert::dec2int (left);
1288 int d = String_convert::dec2int (right);
1289 return scm_cons (scm_from_int (n), scm_from_int (d));
1293 lookup_markup_command (string s)
1295 SCM proc = ly_lily_module_constant ("lookup-markup-command");
1296 return scm_call_1 (proc, ly_string2scm (s));
1300 lookup_markup_list_command (string s)
1302 SCM proc = ly_lily_module_constant ("lookup-markup-list-command");
1303 return scm_call_1 (proc, ly_string2scm (s));
1306 /* Shut up lexer warnings. */
1326 silence_lexer_warnings ()
1328 (void) yy_start_stack_ptr;
1329 (void) yy_start_stack_depth;
1330 (void) yy_start_stack;
1331 (void) yy_push_state;
1332 (void) yy_pop_state;
1333 (void) yy_top_state;
1334 (void) silence_lexer_warnings;