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 yylval (*lexval_)
96 #define yylloc (*lexloc_)
98 #define YY_USER_ACTION add_lexed_char (YYLeng ());
101 SCM scan_fraction (string);
102 SCM (* scm_parse_error_handler) (void *);
112 %option yyclass="Lily_lexer"
114 %option never-interactive
131 /* The strategy concerning multibyte characters is to accept them but
132 * call YYText_utf8 for patterns that might contain them, in order to
133 * get a single code path responsible for flagging non-UTF-8 input:
134 * Patterns for accepting only valid UTF-8 without backing up are
135 * really hard to do and complex, and if nice error messages are
136 * wanted, one would need patterns catching the invalid input as well.
138 * Since editors and operating environments don't necessarily behave
139 * reasonably in the presence of mixed encodings, we flag encoding
140 * errors also in identifiers, comments, and strings where it would be
141 * conceivable to just transparently work with the byte string. But
142 * the whole point of caring about UTF-8 in here at all is too avoid
143 * stranger errors later when input passes into backends or log files
144 * or console output or error messages.
151 WORD {A}([-_]{A}|{A})*
158 REAL ({INT}\.{N}*)|(-?\.{N}+)
160 HORIZONTALWHITE [ \t]
166 BOM_UTF8 \357\273\277
172 // swallow and ignore carriage returns
175 <extratoken>{ANY_CHAR} {
176 /* Generate a token without swallowing anything */
178 /* First unswallow the eaten character */
179 add_lexed_char (-YYLeng ());
182 /* produce requested token */
183 int type = scm_to_int (scm_caar (extra_tokens_));
184 yylval = scm_cdar (extra_tokens_);
185 extra_tokens_ = scm_cdr (extra_tokens_);
186 if (scm_is_null (extra_tokens_))
192 <extratoken><<EOF>> {
193 /* Generate a token without swallowing anything */
195 /* produce requested token */
196 int type = scm_to_int (scm_caar (extra_tokens_));
197 yylval = scm_cdar (extra_tokens_);
198 extra_tokens_ = scm_cdr (extra_tokens_);
199 if (scm_is_null (extra_tokens_))
205 /* Use the trailing context feature. Otherwise, the BOM will not be
206 found if the file starts with an identifier definition. */
207 <INITIAL,chords,lyrics,figures,notes>{BOM_UTF8}/.* {
208 if (this->lexloc_->line_number () != 1 || this->lexloc_->column_number () != 0)
210 LexerWarning (_ ("stray UTF-8 BOM encountered").c_str ());
213 debug_output (_ ("Skipping UTF-8 BOM"));
216 <INITIAL,chords,figures,incl,lyrics,markup,notes>{
218 yy_push_state (longcomment);
220 %[^{\n\r][^\n\r]*[\n\r]? {
221 (void) YYText_utf8 ();
230 <INITIAL,notes,figures,chords,markup>{
236 <INITIAL,chords,lyrics,notes,figures>\\version{WHITE}* {
237 yy_push_state (version);
239 <INITIAL,chords,lyrics,notes,figures>\\sourcefilename{WHITE}* {
240 yy_push_state (sourcefilename);
242 <INITIAL,chords,lyrics,notes,figures>\\sourcefileline{WHITE}* {
243 yy_push_state (sourcefileline);
245 <version>\"[^""]*\" { /* got the version number */
246 string s (YYText_utf8 () + 1);
247 s = s.substr (0, s.rfind ('\"'));
251 SCM top_scope = scm_car (scm_last_pair (scopes_));
252 scm_module_define (top_scope, ly_symbol2scm ("version-seen"), SCM_BOOL_T);
254 if (!is_valid_version (s)) {
255 yylval = SCM_UNSPECIFIED;
259 <sourcefilename>\"[^""]*\" {
260 string s (YYText_utf8 () + 1);
261 s = s.substr (0, s.rfind ('\"'));
264 this->here_input().get_source_file ()->name_ = s;
265 message (_f ("Renaming input to: `%s'", s.c_str ()));
266 progress_indication ("\n");
267 scm_module_define (scm_car (scopes_),
268 ly_symbol2scm ("input-file-name"),
273 <sourcefileline>{INT} {
275 sscanf (YYText (), "%d", &i);
278 this->here_input ().get_source_file ()->set_line (here_input ().start (), i);
281 <version>{ANY_CHAR} {
282 LexerError (_ ("quoted string expected after \\version").c_str ());
285 <sourcefilename>{ANY_CHAR} {
286 LexerError (_ ("quoted string expected after \\sourcefilename").c_str ());
289 <sourcefileline>{ANY_CHAR} {
290 LexerError (_ ("integer expected after \\sourcefileline").c_str ());
295 (void) YYText_utf8 ();
298 (void) YYText_utf8 ();
306 <INITIAL,chords,lyrics,notes,figures>\\maininput {
310 main_input_level_ = include_stack_.size ();
311 is_main_input_ = true;
313 yy_push_state (maininput);
314 yy_push_state (state);
317 LexerError (_ ("\\maininput not allowed outside init files").c_str ());
320 <INITIAL,chords,lyrics,figures,notes>\\include {
321 yy_push_state (incl);
323 <incl>\"[^""]*\" { /* got the include file name */
324 string s (YYText_utf8 ()+1);
325 s = s.substr (0, s.rfind ('"'));
327 new_input (s, sources_);
330 <incl>\\{BLACK}*{WHITE}? { /* got the include identifier */
331 string s = YYText_utf8 () + 1;
332 strip_trailing_white (s);
333 if (s.length () && (s[s.length () - 1] == ';'))
334 s = s.substr (0, s.length () - 1);
336 SCM sid = lookup_identifier (s);
337 if (scm_is_string (sid)) {
338 new_input (ly_scm2string (sid), sources_);
341 string msg (_f ("wrong or undefined identifier: `%s'", s ));
343 LexerError (msg.c_str ());
344 SCM err = scm_current_error_port ();
345 scm_puts ("This value was found in the table: ", err);
346 scm_display (sid, err);
349 <incl>(\$|#) { // scm for the filename
351 Input hi = here_input();
353 SCM sval = ly_parse_scm (hi.start (), &n, hi,
354 be_safe_global && is_main_input_, parser_);
355 sval = eval_scm (sval);
357 for (int i = 0; i < n; i++)
361 char_count_stack_.back () += n;
363 if (scm_is_string (sval)) {
364 new_input (ly_scm2string (sval), sources_);
367 LexerError (_ ("string expected after \\include").c_str ());
368 if (sval != SCM_UNDEFINED) {
369 SCM err = scm_current_error_port ();
370 scm_puts ("This value was found instead: ", err);
371 scm_display (sval, err);
376 <incl,version,sourcefilename>\"[^""]* { // backup rule
377 LexerError (_ ("end quote missing").c_str ());
381 /* Flex picks the longest matching pattern including trailing
382 * contexts. Without the backup pattern, r-. does not trigger the
383 * {RESTNAME} rule but rather the {WORD}/[-_] rule coming later,
384 * needed for avoiding backup states.
387 <chords,notes,figures>{RESTNAME}/[-_] | // pseudo backup rule
388 <chords,notes,figures>{RESTNAME} {
389 char const *s = YYText ();
390 yylval = scm_from_locale_string (s);
393 <chords,notes,figures>q/[-_] | // pseudo backup rule
394 <chords,notes,figures>q {
395 yylval = SCM_UNSPECIFIED;
396 return CHORD_REPETITION;
399 <chords,notes,figures>R/[-_] | // pseudo backup rule
400 <chords,notes,figures>R {
401 yylval = SCM_UNSPECIFIED;
402 return MULTI_MEASURE_REST;
404 <INITIAL,chords,figures,lyrics,markup,notes># { //embedded scm
406 Input hi = here_input();
408 SCM sval = ly_parse_scm (hi.start (), &n, hi,
409 be_safe_global && is_main_input_, parser_);
411 if (sval == SCM_UNDEFINED)
414 for (int i = 0; i < n; i++)
418 char_count_stack_.back () += n;
424 <INITIAL,chords,figures,lyrics,markup,notes>\$ { //immediate scm
426 Input hi = here_input();
428 SCM sval = ly_parse_scm (hi.start (), &n, hi,
429 be_safe_global && is_main_input_, parser_);
431 for (int i = 0; i < n; i++)
435 char_count_stack_.back () += n;
437 sval = eval_scm (sval, '$');
439 int token = scan_scm_id (sval);
440 if (!scm_is_eq (yylval, SCM_UNSPECIFIED))
444 <INITIAL,notes,lyrics>{
446 yylval = SCM_UNSPECIFIED;
447 return DOUBLE_ANGLE_OPEN;
450 yylval = SCM_UNSPECIFIED;
451 return DOUBLE_ANGLE_CLOSE;
457 yylval = SCM_UNSPECIFIED;
461 yylval = SCM_UNSPECIFIED;
468 yylval = SCM_UNSPECIFIED;
472 yylval = SCM_UNSPECIFIED;
476 yylval = SCM_UNSPECIFIED;
482 {WORD}/[-_] | // backup rule
484 return scan_bare_word (YYText_utf8 ());
487 {COMMAND}/[-_] | // backup rule
489 return scan_escaped_word (YYText_utf8 () + 1);
492 yylval = scan_fraction (YYText ());
495 {UNSIGNED}/\/ | // backup rule
497 yylval = scm_c_read_string (YYText ());
501 yylval = scm_c_read_string (YYText () + 1);
508 char c = escaped_char (YYText ()[1]);
509 yylval = scm_cons (scm_from_locale_stringn (&c, 1),
513 yylval = scm_cons (scm_from_locale_string (YYText_utf8 ()),
520 /* yylval is union. Must remember STRING before setting SCM*/
522 yylval = scm_string_concatenate_reverse (yylval,
529 yylval = scm_cons (scm_from_locale_string (YYText ()),
539 yylval = scan_fraction (YYText ());
542 {UNSIGNED}/\/ | // backup rule
544 yylval = scm_c_read_string (YYText ());
547 {COMMAND}/[-_] | // backup rule
549 return scan_escaped_word (YYText_utf8 () + 1);
551 /* Characters needed to express durations, assignments, barchecks */
553 yylval = SCM_UNSPECIFIED;
556 [^$#{}\"\\ \t\n\r\f0-9]+ {
558 string s (YYText_utf8 ());
559 yylval = SCM_UNSPECIFIED;
565 yylval = ly_string2scm (s);
569 /* This should really just cover {} */
571 yylval = SCM_UNSPECIFIED;
572 return YYText ()[0]; // above catches all multibytes.
576 {WORD}/[-_] | // backup rule
578 return scan_bare_word (YYText_utf8 ());
580 {COMMAND}/[-_] | // backup rule
582 return scan_escaped_word (YYText_utf8 () + 1);
585 yylval = scan_fraction (YYText ());
588 {UNSIGNED}/\/ | // backup rule
590 yylval = scm_c_read_string (YYText ());
594 yylval = SCM_UNSPECIFIED;
598 yylval = SCM_UNSPECIFIED;
602 yylval = SCM_UNSPECIFIED;
606 yylval = SCM_UNSPECIFIED;
610 yylval = SCM_UNSPECIFIED;
614 yylval = SCM_UNSPECIFIED;
615 return YYText ()[0]; // WORD catches all multibyte.
622 yylval = SCM_UNSPECIFIED;
625 {COMMAND}/[-_] | // backup rule
627 string str (YYText_utf8 () + 1);
629 int token_type = MARKUP_FUNCTION;
630 SCM s = lookup_markup_command (str);
632 // lookup-markup-command returns a pair with the car
633 // being the function to call, and the cdr being the
634 // call signature specified to define-markup-command,
635 // a list of predicates.
637 if (!scm_is_pair (s)) {
638 // If lookup-markup-command was not successful, we
639 // try lookup-markup-list-command instead.
640 // If this fails as well, we just scan and return
642 s = lookup_markup_list_command (str);
644 token_type = MARKUP_LIST_FUNCTION;
646 return scan_escaped_word (str);
649 // If the list of predicates is, say,
650 // (number? number? markup?), then tokens
651 // EXPECT_MARKUP EXPECT_SCM EXPECT_SCM EXPECT_NO_MORE_ARGS
652 // will be generated. Note that we have to push them
653 // in reverse order, so the first token pushed in the
654 // loop will be EXPECT_NO_MORE_ARGS.
658 // yylval now contains the function to call as token
659 // value (for token type MARKUP_FUNCTION or
660 // MARKUP_LIST_FUNCTION).
662 push_extra_token(EXPECT_NO_MORE_ARGS);
664 for (; scm_is_pair(s); s = scm_cdr(s)) {
665 SCM predicate = scm_car(s);
667 if (predicate == ly_lily_module_constant ("markup-list?"))
668 push_extra_token(EXPECT_MARKUP_LIST);
669 else if (predicate == ly_lily_module_constant ("markup?"))
670 push_extra_token(EXPECT_MARKUP);
672 push_extra_token(EXPECT_SCM, predicate);
676 [^$#{}\"\\ \t\n\r\f]+ {
677 string s (YYText_utf8 ());
679 yylval = ly_string2scm (s);
683 yylval = SCM_UNSPECIFIED;
684 return YYText ()[0]; // Above is catchall for multibyte
688 <longcomment><<EOF>> {
689 LexerError (_ ("EOF found inside a comment").c_str ());
694 LexerError (_ ("EOF found inside string").c_str ());
699 yylval = SCM_UNSPECIFIED;
702 is_main_input_ = include_stack_.size () > main_input_level_;
705 main_input_level_ = 0;
707 if (YYSTATE != maininput)
709 LexerError (_ ("Unfinished main input").c_str ());
712 } while (YYSTATE != maininput);
716 if (!close_input () || !is_main_input_)
717 /* Returns YY_NULL */
720 else if (!close_input ())
721 /* Returns YY_NULL */
726 while (include_stack_.size () > main_input_level_
733 {WORD}/[-_] | // backup rule
735 return scan_bare_word (YYText_utf8 ());
737 {COMMAND}/[-_] | // backup rule
739 return scan_escaped_word (YYText_utf8 () + 1);
744 yylval = scan_fraction (YYText ());
748 -{UNSIGNED} | // backup rule
750 yylval = scm_c_read_string (YYText ());
754 {UNSIGNED}/\/ | // backup rule
756 yylval = scm_c_read_string (YYText ());
762 yylval = SCM_UNSPECIFIED;
766 -/\. | // backup rule
768 yylval = SCM_UNSPECIFIED;
772 <INITIAL,notes,figures>. {
773 yylval = SCM_UNSPECIFIED;
777 <INITIAL,lyrics,notes,figures>\\. {
778 yylval = SCM_UNSPECIFIED;
779 char c = YYText ()[1];
783 return E_ANGLE_CLOSE;
787 return E_EXCLAMATION;
793 return E_BRACKET_OPEN;
797 return E_BRACKET_CLOSE;
809 string msg = _f ("invalid character: `%s'", YYText_utf8 ());
810 LexerError (msg.c_str ());
811 yylval = SCM_UNSPECIFIED;
812 return '%'; // Better not return half a utf8 character.
817 /* Make the lexer generate a token of the given type as the next token.
818 TODO: make it possible to define a value for the token as well */
820 Lily_lexer::push_extra_token (int token_type, SCM scm)
822 if (scm_is_null (extra_tokens_))
824 if (YY_START != extratoken)
825 hidden_state_ = YY_START;
826 yy_push_state (extratoken);
828 extra_tokens_ = scm_acons (scm_from_int (token_type), scm, extra_tokens_);
832 Lily_lexer::push_chord_state (SCM alist)
834 SCM p = scm_assq (alist, pitchname_tab_stack_);
836 if (scm_is_false (p))
837 p = scm_cons (alist, alist_to_hashq (alist));
838 pitchname_tab_stack_ = scm_cons (p, pitchname_tab_stack_);
839 yy_push_state (chords);
843 Lily_lexer::push_figuredbass_state ()
845 yy_push_state (figures);
849 Lily_lexer::push_initial_state ()
851 yy_push_state (INITIAL);
855 Lily_lexer::push_lyric_state ()
857 yy_push_state (lyrics);
861 Lily_lexer::push_markup_state ()
863 yy_push_state (markup);
867 Lily_lexer::push_note_state (SCM alist)
869 bool extra = (YYSTATE == extratoken);
871 SCM p = scm_assq (alist, pitchname_tab_stack_);
876 if (scm_is_false (p))
877 p = scm_cons (alist, alist_to_hashq (alist));
878 pitchname_tab_stack_ = scm_cons (p, pitchname_tab_stack_);
879 yy_push_state (notes);
882 hidden_state_ = YYSTATE;
883 yy_push_state (extratoken);
888 Lily_lexer::pop_state ()
890 bool extra = (YYSTATE == extratoken);
895 if (YYSTATE == notes || YYSTATE == chords)
896 pitchname_tab_stack_ = scm_cdr (pitchname_tab_stack_);
901 hidden_state_ = YYSTATE;
902 yy_push_state (extratoken);
907 Lily_lexer::identifier_type (SCM sid)
909 int k = try_special_identifiers (&yylval , sid);
910 return k >= 0 ? k : SCM_IDENTIFIER;
915 Lily_lexer::scan_escaped_word (string str)
917 // use more SCM for this.
919 // SCM sym = ly_symbol2scm (str.c_str ());
921 yylval = SCM_UNSPECIFIED;
922 int i = lookup_keyword (str);
927 SCM sid = lookup_identifier (str);
928 if (Music *m = unsmob_music (sid))
930 m->set_spot (override_input (last_input_));
933 if (sid != SCM_UNDEFINED)
934 return scan_scm_id (sid);
936 string msg (_f ("unknown escaped string: `\\%s'", str));
937 LexerError (msg.c_str ());
939 yylval = ly_string2scm (str);
945 Lily_lexer::scan_scm_id (SCM sid)
947 if (is_music_function (sid))
949 int funtype = SCM_FUNCTION;
953 SCM s = get_music_function_signature (sid);
954 SCM cs = scm_car (s);
956 if (scm_is_pair (cs))
961 if (scm_is_eq (cs, ly_lily_module_constant ("ly:music?")))
962 funtype = MUSIC_FUNCTION;
963 else if (scm_is_eq (cs, ly_lily_module_constant ("ly:event?")))
964 funtype = EVENT_FUNCTION;
965 else if (ly_is_procedure (cs))
966 funtype = SCM_FUNCTION;
967 else programming_error ("Bad syntax function predicate");
969 push_extra_token (EXPECT_NO_MORE_ARGS);
970 for (s = scm_cdr (s); scm_is_pair (s); s = scm_cdr (s))
972 SCM optional = SCM_UNDEFINED;
975 if (scm_is_pair (cs))
977 optional = SCM_CDR (cs);
981 if (cs == Pitch_type_p_proc)
982 push_extra_token (EXPECT_PITCH);
983 else if (cs == Duration_type_p_proc)
984 push_extra_token (EXPECT_DURATION);
985 else if (ly_is_procedure (cs))
986 push_extra_token (EXPECT_SCM, cs);
989 programming_error ("Function parameter without type-checking predicate");
992 if (!scm_is_eq (optional, SCM_UNDEFINED))
993 push_extra_token (EXPECT_OPTIONAL, optional);
998 return identifier_type (sid);
1002 Lily_lexer::scan_bare_word (string str)
1004 SCM sym = ly_symbol2scm (str.c_str ());
1005 if ((YYSTATE == notes) || (YYSTATE == chords)) {
1006 SCM handle = SCM_BOOL_F;
1007 if (scm_is_pair (pitchname_tab_stack_))
1008 handle = scm_hashq_get_handle (scm_cdar (pitchname_tab_stack_), sym);
1010 if (scm_is_pair (handle)) {
1011 yylval = scm_cdr (handle);
1012 if (unsmob_pitch (yylval))
1013 return (YYSTATE == notes) ? NOTENAME_PITCH : TONICNAME_PITCH;
1014 else if (scm_is_symbol (yylval))
1017 else if ((YYSTATE == chords)
1018 && (handle = scm_hashq_get_handle (chordmodifier_tab_, sym))!= SCM_BOOL_F)
1020 yylval = scm_cdr (handle);
1021 return CHORD_MODIFIER;
1024 yylval = ly_string2scm (str);
1029 Lily_lexer::get_state () const
1031 if (YY_START == extratoken)
1032 return hidden_state_;
1038 Lily_lexer::is_note_state () const
1040 return get_state () == notes;
1044 Lily_lexer::is_chord_state () const
1046 return get_state () == chords;
1050 Lily_lexer::is_lyric_state () const
1052 return get_state () == lyrics;
1056 Lily_lexer::is_figure_state () const
1058 return get_state () == figures;
1061 // The extra_token parameter specifies how to convert multiple values
1062 // into additional tokens. For '#', additional values get pushed as
1063 // SCM_IDENTIFIER. For '$', they get checked for their type and get
1064 // pushed as a corresponding *_IDENTIFIER token. Since the latter
1065 // tampers with yylval, it can only be done from the lexer itself, so
1066 // this function is private.
1069 Lily_lexer::eval_scm (SCM readerdata, char extra_token)
1071 SCM sval = SCM_UNDEFINED;
1073 if (!SCM_UNBNDP (readerdata))
1075 sval = ly_eval_scm (scm_car (readerdata),
1076 *unsmob_input (scm_cdr (readerdata)),
1077 be_safe_global && is_main_input_,
1081 if (SCM_UNBNDP (sval))
1084 return SCM_UNSPECIFIED;
1087 if (extra_token && SCM_VALUESP (sval))
1089 sval = scm_struct_ref (sval, SCM_INUM0);
1091 if (scm_is_pair (sval)) {
1092 for (SCM p = scm_reverse (scm_cdr (sval));
1096 SCM v = scm_car (p);
1097 if (Music *m = unsmob_music (v))
1099 if (!unsmob_input (m->get_property ("origin")))
1100 m->set_spot (override_input (last_input_));
1104 switch (extra_token) {
1106 token = scan_scm_id (v);
1107 if (!scm_is_eq (yylval, SCM_UNSPECIFIED))
1108 push_extra_token (token, yylval);
1111 push_extra_token (SCM_IDENTIFIER, v);
1115 sval = scm_car (sval);
1117 sval = SCM_UNSPECIFIED;
1120 if (Music *m = unsmob_music (sval))
1122 if (!unsmob_input (m->get_property ("origin")))
1123 m->set_spot (override_input (last_input_));
1129 /* Check for valid UTF-8 that has no overlong or surrogate codes and
1130 is in the range 0-0x10ffff */
1133 Lily_lexer::YYText_utf8 ()
1135 const char * const p = YYText ();
1136 for (int i=0; p[i];) {
1137 if ((p[i] & 0xff) < 0x80) {
1141 int oldi = i; // start of character
1142 int more = 0; // # of followup bytes, 0 if bad
1143 switch (p[i++] & 0xff) {
1144 // 0xc0 and 0xc1 are overlong prefixes for
1145 // 0x00-0x3f and 0x40-0x7f respectively, bad.
1146 case 0xc2: // 0x80-0xbf
1147 case 0xc3: // 0xc0-0xff
1148 case 0xc4: // 0x100-0x13f
1149 case 0xc5: // 0x140-0x17f
1150 case 0xc6: // 0x180-0x1bf
1151 case 0xc7: // 0x1c0-0x1ff
1152 case 0xc8: // 0x200-0x23f
1153 case 0xc9: // 0x240-0x27f
1154 case 0xca: // 0x280-0x2bf
1155 case 0xcb: // 0x2c0-0x2ff
1156 case 0xcc: // 0x300-0x33f
1157 case 0xcd: // 0x340-0x37f
1158 case 0xce: // 0x380-0x3bf
1159 case 0xcf: // 0x3c0-0x3ff
1160 case 0xd0: // 0x400-0x43f
1161 case 0xd1: // 0x440-0x47f
1162 case 0xd2: // 0x480-0x4bf
1163 case 0xd3: // 0x4c0-0x4ff
1164 case 0xd4: // 0x500-0x53f
1165 case 0xd5: // 0x540-0x57f
1166 case 0xd6: // 0x580-0x5bf
1167 case 0xd7: // 0x5c0-0x5ff
1168 case 0xd8: // 0x600-0x63f
1169 case 0xd9: // 0x640-0x67f
1170 case 0xda: // 0x680-0x6bf
1171 case 0xdb: // 0x6c0-0x6ff
1172 case 0xdc: // 0x700-0x73f
1173 case 0xdd: // 0x740-0x77f
1174 case 0xde: // 0x780-0x7bf
1175 case 0xdf: // 0x7c0-0x7ff
1176 more = 1; // 2-byte sequences, 0x80-0x7ff
1179 // don't allow overlong sequences for 0-0x7ff
1180 if ((p[i] & 0xff) < 0xa0)
1182 case 0xe1: // 0x1000-0x1fff
1183 case 0xe2: // 0x2000-0x2fff
1184 case 0xe3: // 0x3000-0x3fff
1185 case 0xe4: // 0x4000-0x4fff
1186 case 0xe5: // 0x5000-0x5fff
1187 case 0xe6: // 0x6000-0x6fff
1188 case 0xe7: // 0x7000-0x7fff
1189 case 0xe8: // 0x8000-0x8fff
1190 case 0xe9: // 0x9000-0x9fff
1191 case 0xea: // 0xa000-0xafff
1192 case 0xeb: // 0xb000-0xbfff
1193 case 0xec: // 0xc000-0xcfff
1194 more = 2; // 3-byte sequences, 0x7ff-0xcfff
1196 case 0xed: // 0xd000-0xdfff
1197 // Don't allow surrogate codes 0xd800-0xdfff
1198 if ((p[i] & 0xff) >= 0xa0)
1200 case 0xee: // 0xe000-0xefff
1201 case 0xef: // 0xf000-0xffff
1202 more = 2; // 3-byte sequences,
1203 // 0xd000-0xd7ff, 0xe000-0xffff
1206 // don't allow overlong sequences for 0-0xffff
1207 if ((p[i] & 0xff) < 0x90)
1209 case 0xf1: // 0x40000-0x7ffff
1210 case 0xf2: // 0x80000-0xbffff
1211 case 0xf3: // 0xc0000-0xfffff
1212 more = 3; // 4-byte sequences, 0x10000-0xfffff
1215 // don't allow more than 0x10ffff
1216 if ((p[i] & 0xff) >= 0x90)
1218 more = 3; // 4-byte sequence, 0x100000-0x10ffff
1222 // check that all continuation bytes are valid
1224 if ((p[i++] & 0xc0) != 0x80)
1230 Input h = here_input ();
1231 h.set (h.get_source_file (), h.start () + oldi, h.start () + i);
1232 h.warning (_ ("non-UTF-8 input").c_str ());
1239 urg, belong to string (_convert)
1240 and should be generalised
1243 strip_leading_white (string&s)
1246 for (; i < s.length (); i++)
1247 if (!isspace (s[i]))
1254 strip_trailing_white (string&s)
1256 ssize i = s.length ();
1258 if (!isspace (s[i]))
1261 s = s.substr (0, i + 1);
1266 Lilypond_version oldest_version ("2.7.38");
1270 is_valid_version (string s)
1272 Lilypond_version current ( MAJOR_VERSION "." MINOR_VERSION "." PATCH_LEVEL );
1273 Lilypond_version ver (s);
1274 if (int (ver) < oldest_version)
1276 non_fatal_error (_f ("file too old: %s (oldest supported: %s)", ver.to_string (), oldest_version.to_string ()));
1277 non_fatal_error (_ ("consider updating the input with the convert-ly script"));
1283 non_fatal_error (_f ("program too old: %s (file requires: %s)", current.to_string (), ver.to_string ()));
1294 lyric_fudge (string s)
1298 while ((i = s.find ('_', i)) != string::npos)
1306 Convert "NUM/DEN" into a '(NUM . DEN) cons.
1309 scan_fraction (string frac)
1311 ssize i = frac.find ('/');
1312 string left = frac.substr (0, i);
1313 string right = frac.substr (i + 1, (frac.length () - i + 1));
1315 int n = String_convert::dec2int (left);
1316 int d = String_convert::dec2int (right);
1317 return scm_cons (scm_from_int (n), scm_from_int (d));
1321 lookup_markup_command (string s)
1323 SCM proc = ly_lily_module_constant ("lookup-markup-command");
1324 return scm_call_1 (proc, ly_string2scm (s));
1328 lookup_markup_list_command (string s)
1330 SCM proc = ly_lily_module_constant ("lookup-markup-list-command");
1331 return scm_call_1 (proc, ly_string2scm (s));
1334 /* Shut up lexer warnings. */
1354 silence_lexer_warnings ()
1356 (void) yy_start_stack_ptr;
1357 (void) yy_start_stack_depth;
1358 (void) yy_start_stack;
1359 (void) yy_push_state;
1360 (void) yy_pop_state;
1361 (void) yy_top_state;
1362 (void) silence_lexer_warnings;