X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=lily%2Flexer.ll;h=caea441b0a2d88a1f61ea99c749c534bd0b3e1d6;hb=5bbfc22fce036b9b69df5e420de93e11da23c05e;hp=c2d8b6baeff1273520aa7a68eb32492ad286ec3e;hpb=ac45f622a1b01a089c056d021741e92933161f82;p=lilypond.git diff --git a/lily/lexer.ll b/lily/lexer.ll index c2d8b6baef..caea441b0a 100644 --- a/lily/lexer.ll +++ b/lily/lexer.ll @@ -1,8 +1,8 @@ -%{ // -*- mode: c++; c-file-style: "linux" -*- +%{ // -*- mode: c++; c-file-style: "linux"; indent-tabs-mode: t -*- /* This file is part of LilyPond, the GNU music typesetter. - Copyright (C) 1996--2012 Han-Wen Nienhuys + Copyright (C) 1996--2014 Han-Wen Nienhuys Jan Nieuwenhuizen LilyPond is free software: you can redistribute it and/or modify @@ -86,13 +86,19 @@ SCM lookup_markup_list_command (string s); bool is_valid_version (string s); -#define start_quote() \ - yy_push_state (quote);\ - yylval.string = new string +#define start_quote() do { \ + yy_push_state (quote); \ + yylval = SCM_EOL; \ + } while (0) -#define start_lyric_quote() \ - yy_push_state (lyric_quote);\ - yylval.string = new string +/* + The inside of \"violin1" is marked by commandquote mode +*/ + +#define start_command_quote() do { \ + yy_push_state (commandquote); \ + yylval = SCM_EOL; \ + } while (0) #define yylval (*lexval_) @@ -117,16 +123,16 @@ SCM (* scm_parse_error_handler) (void *); %option never-interactive %option warn -%x extratoken %x chords %x figures %x incl %x lyrics -%x lyric_quote %x longcomment +%x maininput %x markup %x notes %x quote +%x commandquote %x sourcefileline %x sourcefilename %x version @@ -150,32 +156,24 @@ SCM (* scm_parse_error_handler) (void *); A [a-zA-Z\200-\377] AA {A}|_ N [0-9] -AN {AA}|{N} ANY_CHAR (.|\n) -PUNCT [?!:'`] -ACCENT \\[`'"^] -SPECIAL_CHAR [&@] -NATIONAL [\001-\006\021-\027\031\036] -TEX {AA}|-|{PUNCT}|{ACCENT}|{NATIONAL}|{SPECIAL_CHAR} -DASHED_WORD {A}({AN}|-)* -DASHED_KEY_WORD \\{DASHED_WORD} - - - -ALPHAWORD {A}+ +WORD {A}([-_]{A}|{A})* +COMMAND \\{WORD} +/* SPECIAL category is for every letter that needs to get passed to + * the parser rather than being redefinable by the user */ +SPECIAL [-+*/=<>{}!?_^'',.:] +SHORTHAND (.|\\.) UNSIGNED {N}+ E_UNSIGNED \\{N}+ FRACTION {N}+\/{N}+ INT -?{UNSIGNED} REAL ({INT}\.{N}*)|(-?\.{N}+) +STRICTREAL {UNSIGNED}\.{UNSIGNED} WHITE [ \n\t\f\r] HORIZONTALWHITE [ \t] BLACK [^ \n\t\f\r] RESTNAME [rs] -NOTECOMMAND \\{A}+ -MARKUPCOMMAND \\({A}|[-_])+ -LYRICS ({AA}|{TEX})[^0-9 \t\n\r\f]* -ESCAPED [nt\\'"] +ESCAPED [nt\\''""] EXTENDER __ HYPHEN -- BOM_UTF8 \357\273\277 @@ -187,36 +185,6 @@ BOM_UTF8 \357\273\277 // swallow and ignore carriage returns } -{ANY_CHAR} { - /* Generate a token without swallowing anything */ - - /* First unswallow the eaten character */ - add_lexed_char (-YYLeng ()); - yyless (0); - - /* produce requested token */ - int type = scm_to_int (scm_caar (extra_tokens_)); - yylval.scm = scm_cdar (extra_tokens_); - extra_tokens_ = scm_cdr (extra_tokens_); - if (scm_is_null (extra_tokens_)) - yy_pop_state (); - - return type; -} - -<> { - /* Generate a token without swallowing anything */ - - /* produce requested token */ - int type = scm_to_int (scm_caar (extra_tokens_)); - yylval.scm = scm_cdar (extra_tokens_); - extra_tokens_ = scm_cdr (extra_tokens_); - if (scm_is_null (extra_tokens_)) - yy_pop_state (); - - return type; -} - /* Use the trailing context feature. Otherwise, the BOM will not be found if the file starts with an identifier definition. */ {BOM_UTF8}/.* { @@ -232,16 +200,10 @@ BOM_UTF8 \357\273\277 "%{" { yy_push_state (longcomment); } - %[^{\n\r][^\n\r]*[\n\r] { + %[^{\n\r][^\n\r]*[\n\r]? { (void) YYText_utf8 (); } - %[^{\n\r] { // backup rule - (void) YYText_utf8 (); - } - %[\n\r] { - } - %[^{\n\r][^\n\r]* { - (void) YYText_utf8 (); + %[\n\r]? { } {WHITE}+ { @@ -263,7 +225,7 @@ BOM_UTF8 \357\273\277 \\sourcefileline{WHITE}* { yy_push_state (sourcefileline); } -\"[^"]*\" { /* got the version number */ +\"[^""]*\" { /* got the version number */ string s (YYText_utf8 () + 1); s = s.substr (0, s.rfind ('\"')); @@ -272,12 +234,12 @@ BOM_UTF8 \357\273\277 SCM top_scope = scm_car (scm_last_pair (scopes_)); scm_module_define (top_scope, ly_symbol2scm ("version-seen"), SCM_BOOL_T); - if (!is_valid_version (s)) + if (!is_valid_version (s)) { + yylval = SCM_UNSPECIFIED; return INVALID; - - + } } -\"[^"]*\" { +\"[^""]*\" { string s (YYText_utf8 () + 1); s = s.substr (0, s.rfind ('\"')); @@ -328,10 +290,14 @@ BOM_UTF8 \357\273\277 if (!is_main_input_) { start_main_input (); + main_input_level_ = include_stack_.size (); is_main_input_ = true; + int state = YYSTATE; + yy_push_state (maininput); + yy_push_state (state); } else - error (_ ("\\maininput not allowed outside init files")); + LexerError (_ ("\\maininput not allowed outside init files").c_str ()); } \\include { @@ -364,12 +330,11 @@ BOM_UTF8 \357\273\277 } } (\$|#) { // scm for the filename - int n = 0; Input hi = here_input(); hi.step_forward (); - SCM sval = ly_parse_scm (hi.start (), &n, hi, - be_safe_global && is_main_input_, parser_); - sval = eval_scm (sval); + SCM sval = ly_parse_scm (hi, be_safe_global && is_main_input_, parser_); + sval = eval_scm (sval, hi); + int n = hi.end () - hi.start (); for (int i = 0; i < n; i++) { @@ -391,47 +356,58 @@ BOM_UTF8 \357\273\277 } \"[^""]* { // backup rule - error (_ ("end quote missing")); - exit (1); + LexerError (_ ("end quote missing").c_str ()); + yy_pop_state (); } + + /* Flex picks the longest matching pattern including trailing + * contexts. Without the backup pattern, r-. does not trigger the + * {RESTNAME} rule but rather the {WORD}/[-_] rule coming later, + * needed for avoiding backup states. + */ + +{RESTNAME}/[-_] | // pseudo backup rule {RESTNAME} { char const *s = YYText (); - yylval.scm = scm_from_locale_string (s); + yylval = scm_from_locale_string (s); return RESTNAME; } +q/[-_] | // pseudo backup rule q { + yylval = SCM_UNSPECIFIED; return CHORD_REPETITION; } +R/[-_] | // pseudo backup rule R { + yylval = SCM_UNSPECIFIED; return MULTI_MEASURE_REST; } # { //embedded scm - int n = 0; Input hi = here_input(); hi.step_forward (); - SCM sval = ly_parse_scm (hi.start (), &n, hi, - be_safe_global && is_main_input_, parser_); + SCM sval = ly_parse_scm (hi, be_safe_global && is_main_input_, parser_); if (sval == SCM_UNDEFINED) error_level_ = 1; + int n = hi.end () - hi.start (); for (int i = 0; i < n; i++) { yyinput (); } char_count_stack_.back () += n; - yylval.scm = sval; + yylval = sval; return SCM_TOKEN; } \$ { //immediate scm - int n = 0; Input hi = here_input(); hi.step_forward (); - SCM sval = ly_parse_scm (hi.start (), &n, hi, - be_safe_global && is_main_input_, parser_); + SCM sval = ly_parse_scm (hi, be_safe_global && is_main_input_, parser_); + + int n = hi.end () - hi.start (); for (int i = 0; i < n; i++) { @@ -439,176 +415,240 @@ BOM_UTF8 \357\273\277 } char_count_stack_.back () += n; - sval = eval_scm (sval); - + sval = eval_scm (sval, hi, '$'); + int token = scan_scm_id (sval); - if (!scm_is_eq (yylval.scm, SCM_UNSPECIFIED)) - return token; + if (!scm_is_eq (yylval, SCM_UNSPECIFIED)) + return token; } { \<\< { + yylval = SCM_UNSPECIFIED; return DOUBLE_ANGLE_OPEN; } \>\> { + yylval = SCM_UNSPECIFIED; return DOUBLE_ANGLE_CLOSE; } } { \< { + yylval = SCM_UNSPECIFIED; return ANGLE_OPEN; } \> { + yylval = SCM_UNSPECIFIED; return ANGLE_CLOSE; } } { _ { + yylval = SCM_UNSPECIFIED; return FIGURE_SPACE; } \> { + yylval = SCM_UNSPECIFIED; return FIGURE_CLOSE; } \< { + yylval = SCM_UNSPECIFIED; return FIGURE_OPEN; } + \\\+ { + yylval = SCM_UNSPECIFIED; + return E_PLUS; + } + \\! { + yylval = SCM_UNSPECIFIED; + return E_EXCLAMATION; + } + \\\\ { + yylval = SCM_UNSPECIFIED; + return E_BACKSLASH; + } + [][] { + yylval = SCM_UNSPECIFIED; + return YYText ()[0]; + } } { - {ALPHAWORD} { + {WORD}/[-_] | // backup rule + {WORD} { return scan_bare_word (YYText_utf8 ()); } - - {NOTECOMMAND} { + \\\" { + start_command_quote (); + } + {COMMAND}/[-_] | // backup rule + {COMMAND} { return scan_escaped_word (YYText_utf8 () + 1); } {FRACTION} { - yylval.scm = scan_fraction (YYText ()); + yylval = scan_fraction (YYText ()); return FRACTION; } - {UNSIGNED}/\/ | // backup rule - {UNSIGNED} { - yylval.scm = scm_c_read_string (YYText ()); + {STRICTREAL} { + yylval = scm_c_read_string (YYText ()); + return REAL; + } + {UNSIGNED}/[/.] | // backup rule + {UNSIGNED} { + yylval = scm_c_read_string (YYText ()); return UNSIGNED; } {E_UNSIGNED} { - yylval.i = String_convert::dec2int (string (YYText () +1)); + yylval = scm_c_read_string (YYText () + 1); return E_UNSIGNED; } } -{ +{ \\{ESCAPED} { - *yylval.string += to_string (escaped_char (YYText ()[1])); + char c = escaped_char (YYText ()[1]); + yylval = scm_cons (scm_from_locale_stringn (&c, 1), + yylval); } [^\\""]+ { - *yylval.string += YYText_utf8 (); + yylval = scm_cons (scm_from_locale_string (YYText_utf8 ()), + yylval); } \" { + /* yylval is union. Must remember STRING before setting SCM*/ + + yylval = scm_string_concatenate_reverse (yylval, + SCM_UNDEFINED, + SCM_UNDEFINED); + + if (get_state () == commandquote) { + yy_pop_state (); + return scan_escaped_word (ly_scm2string (yylval)); + } + yy_pop_state (); - /* yylval is union. Must remember STRING before setting SCM*/ - string *sp = yylval.string; - yylval.scm = ly_string2scm (*sp); - delete sp; - return is_lyric_state () ? LYRICS_STRING : STRING; + return STRING; } \\ { - *yylval.string += YYText (); + yylval = scm_cons (scm_from_locale_string (YYText ()), + yylval); } } { \" { - start_lyric_quote (); + start_quote (); } {FRACTION} { - yylval.scm = scan_fraction (YYText ()); + yylval = scan_fraction (YYText ()); return FRACTION; } - {UNSIGNED}/\/[^0-9] { // backup rule - yylval.scm = scm_c_read_string (YYText ()); - return UNSIGNED; + {STRICTREAL} { + yylval = scm_c_read_string (YYText ()); + return REAL; } - {UNSIGNED}/\/ | // backup rule + {UNSIGNED}/[/.] | // backup rule {UNSIGNED} { - yylval.scm = scm_c_read_string (YYText ()); + yylval = scm_c_read_string (YYText ()); return UNSIGNED; } - {NOTECOMMAND} { + \\\" { + start_command_quote (); + } + {COMMAND}/[-_] | // backup rule + {COMMAND} { return scan_escaped_word (YYText_utf8 () + 1); } - {LYRICS} { + \\.|\| { + // UTF-8 already covered by COMMAND + return scan_shorthand (YYText ()); + } + /* Characters needed to express durations, assignments */ + [*.=] { + yylval = SCM_UNSPECIFIED; + return YYText ()[0]; + } + [^|*.=$#{}\"\\ \t\n\r\f0-9][^$#{}\"\\ \t\n\r\f0-9]* { /* ugr. This sux. */ - string s (YYText_utf8 ()); + string s (YYText_utf8 ()); + yylval = SCM_UNSPECIFIED; if (s == "__") - return yylval.i = EXTENDER; + return EXTENDER; if (s == "--") - return yylval.i = HYPHEN; + return HYPHEN; s = lyric_fudge (s); + yylval = ly_string2scm (s); - char c = s[s.length () - 1]; - if (c == '{' || c == '}') // brace open is for not confusing dumb tools. - here_input ().warning ( - _ ("Brace found at end of lyric. Did you forget a space?")); - yylval.scm = ly_string2scm (s); - - - return LYRICS_STRING; + return STRING; } - . { - return YYText ()[0]; // LYRICS already catches all multibytes. + /* This should really just cover {} */ + [{}] { + yylval = SCM_UNSPECIFIED; + return YYText ()[0]; } } { - {ALPHAWORD} { + {WORD}/[-_] | // backup rule + {WORD} { return scan_bare_word (YYText_utf8 ()); } - {NOTECOMMAND} { + \\\" { + start_command_quote (); + } + {COMMAND}/[-_] | // backup rule + {COMMAND} { return scan_escaped_word (YYText_utf8 () + 1); } {FRACTION} { - yylval.scm = scan_fraction (YYText ()); + yylval = scan_fraction (YYText ()); return FRACTION; } - {UNSIGNED}/\/[^0-9] { // backup rule - yylval.scm = scm_c_read_string (YYText ()); - return UNSIGNED; - } {UNSIGNED}/\/ | // backup rule {UNSIGNED} { - yylval.scm = scm_c_read_string (YYText ()); + yylval = scm_c_read_string (YYText ()); return UNSIGNED; } - { + yylval = SCM_UNSPECIFIED; return CHORD_MINUS; } : { + yylval = SCM_UNSPECIFIED; return CHORD_COLON; } \/\+ { + yylval = SCM_UNSPECIFIED; return CHORD_BASS; } \/ { + yylval = SCM_UNSPECIFIED; return CHORD_SLASH; } \^ { + yylval = SCM_UNSPECIFIED; return CHORD_CARET; } - . { - return YYText ()[0]; // ALPHAWORD catches all multibyte. - } } { \\score { + yylval = SCM_UNSPECIFIED; return SCORE; } - {MARKUPCOMMAND} { + \\score-lines { + yylval = SCM_UNSPECIFIED; + return SCORELINES; + } + \\\" { + start_command_quote (); + } + {COMMAND}/[-_] | // backup rule + {COMMAND} { string str (YYText_utf8 () + 1); int token_type = MARKUP_FUNCTION; @@ -638,62 +678,68 @@ BOM_UTF8 \357\273\277 // in reverse order, so the first token pushed in the // loop will be EXPECT_NO_MORE_ARGS. - yylval.scm = scm_car(s); + yylval = scm_car(s); // yylval now contains the function to call as token // value (for token type MARKUP_FUNCTION or // MARKUP_LIST_FUNCTION). - push_extra_token(EXPECT_NO_MORE_ARGS); + push_extra_token (here_input (), EXPECT_NO_MORE_ARGS); s = scm_cdr(s); for (; scm_is_pair(s); s = scm_cdr(s)) { SCM predicate = scm_car(s); if (predicate == ly_lily_module_constant ("markup-list?")) - push_extra_token(EXPECT_MARKUP_LIST); + push_extra_token (here_input (), EXPECT_MARKUP_LIST); else if (predicate == ly_lily_module_constant ("markup?")) - push_extra_token(EXPECT_MARKUP); + push_extra_token (here_input (), EXPECT_MARKUP); else - push_extra_token(EXPECT_SCM, predicate); + push_extra_token (here_input (), EXPECT_SCM, predicate); } return token_type; } - [{}] { - return YYText ()[0]; - } [^$#{}\"\\ \t\n\r\f]+ { string s (YYText_utf8 ()); - char c = s[s.length () - 1]; - /* brace open is for not confusing dumb tools. */ - if (c == '{' || c == '}') - here_input ().warning ( - _ ("Brace found at end of markup. Did you forget a space?")); - yylval.scm = ly_string2scm (s); - - + yylval = ly_string2scm (s); return STRING; } - . { - return YYText()[0]; // Above is catchall for multibyte + [{}] { + yylval = SCM_UNSPECIFIED; + return YYText ()[0]; } } <> { LexerError (_ ("EOF found inside a comment").c_str ()); - is_main_input_ = false; // should be safe , can't have \include in --safe. - if (!close_input ()) - yyterminate (); // can't move this, since it actually rets a YY_NULL + yy_pop_state (); } -<> { if (is_main_input_) +<> { + LexerError (_ ("EOF found inside string").c_str ()); + yy_pop_state (); +} + +<> { + yylval = SCM_UNSPECIFIED; + if (is_main_input_) { - /* 2 = init.ly + current file. - > because we're before closing, but is_main_input_ should - reflect after. - */ - is_main_input_ = include_stack_.size () > 2; - if (!close_input ()) + is_main_input_ = include_stack_.size () > main_input_level_; + if (!is_main_input_) + { + main_input_level_ = 0; + pop_state (); + if (YYSTATE != maininput) + { + LexerError (_ ("Unfinished main input").c_str ()); + do { + yy_pop_state (); + } while (YYSTATE != maininput); + } + extra_tokens_ = SCM_EOL; + yy_pop_state (); + } + if (!close_input () || !is_main_input_) /* Returns YY_NULL */ yyterminate (); } @@ -702,78 +748,63 @@ BOM_UTF8 \357\273\277 yyterminate (); } +{ANY_CHAR} { + while (include_stack_.size () > main_input_level_ + && close_input ()) + ; + yyterminate (); +} + { - {DASHED_WORD} { + {WORD}/[-_] | // backup rule + {WORD} { return scan_bare_word (YYText_utf8 ()); } - {DASHED_KEY_WORD} { + \\\" { + start_command_quote (); + } + {COMMAND}/[-_] | // backup rule + {COMMAND} { return scan_escaped_word (YYText_utf8 () + 1); } } +{FRACTION} { + yylval = scan_fraction (YYText ()); + return FRACTION; +} + -{UNSIGNED} | // backup rule {REAL} { - yylval.scm = scm_c_read_string (YYText ()); - return REAL; -} --\. { // backup rule - yylval.scm = scm_from_double (0.0); + yylval = scm_c_read_string (YYText ()); return REAL; } +{UNSIGNED}/\/ | // backup rule {UNSIGNED} { - yylval.scm = scm_c_read_string (YYText ()); + yylval = scm_c_read_string (YYText ()); return UNSIGNED; } -[{}] { - +-/\. { // backup rule + yylval = SCM_UNSPECIFIED; return YYText ()[0]; } -[*:=] { - char c = YYText ()[0]; - - return c; -} -. { +{SPECIAL} { + yylval = SCM_UNSPECIFIED; return YYText ()[0]; } -\\. { - char c = YYText ()[1]; - - switch (c) { - case '>': - return E_ANGLE_CLOSE; - case '<': - return E_ANGLE_OPEN; - case '!': - return E_EXCLAMATION; - case '(': - return E_OPEN; - case ')': - return E_CLOSE; - case '[': - return E_BRACKET_OPEN; - case '+': - return E_PLUS; - case ']': - return E_BRACKET_CLOSE; - case '~': - return E_TILDE; - case '\\': - return E_BACKSLASH; - - default: - return E_CHAR; - } +{SHORTHAND} { + return scan_shorthand (YYText_utf8 ()); // should not be utf-8 } <*>.[\200-\277]* { string msg = _f ("invalid character: `%s'", YYText_utf8 ()); LexerError (msg.c_str ()); + yylval = SCM_UNSPECIFIED; return '%'; // Better not return half a utf8 character. } @@ -782,21 +813,35 @@ BOM_UTF8 \357\273\277 /* Make the lexer generate a token of the given type as the next token. TODO: make it possible to define a value for the token as well */ void -Lily_lexer::push_extra_token (int token_type, SCM scm) +Lily_lexer::push_extra_token (Input const &where, int token_type, SCM scm) +{ + extra_tokens_ = scm_cons (scm_cons2 (make_input (where), + scm_from_int (token_type), + scm), extra_tokens_); +} + +int +Lily_lexer::pop_extra_token () { if (scm_is_null (extra_tokens_)) - { - if (YY_START != extratoken) - hidden_state_ = YY_START; - yy_push_state (extratoken); - } - extra_tokens_ = scm_acons (scm_from_int (token_type), scm, extra_tokens_); + return -1; + + /* produce requested token */ + yylloc = *unsmob_input (scm_caar (extra_tokens_)); + int type = scm_to_int (scm_cadar (extra_tokens_)); + yylval = scm_cddar (extra_tokens_); + extra_tokens_ = scm_cdr (extra_tokens_); + return type; } void -Lily_lexer::push_chord_state (SCM tab) +Lily_lexer::push_chord_state (SCM alist) { - pitchname_tab_stack_ = scm_cons (tab, pitchname_tab_stack_); + SCM p = scm_assq (alist, pitchname_tab_stack_); + + if (scm_is_false (p)) + p = scm_cons (alist, alist_to_hashq (alist)); + pitchname_tab_stack_ = scm_cons (p, pitchname_tab_stack_); yy_push_state (chords); } @@ -825,60 +870,82 @@ Lily_lexer::push_markup_state () } void -Lily_lexer::push_note_state (SCM tab) +Lily_lexer::push_note_state (SCM alist) { - pitchname_tab_stack_ = scm_cons (tab, pitchname_tab_stack_); + SCM p = scm_assq (alist, pitchname_tab_stack_); + + if (scm_is_false (p)) + p = scm_cons (alist, alist_to_hashq (alist)); + pitchname_tab_stack_ = scm_cons (p, pitchname_tab_stack_); yy_push_state (notes); } void Lily_lexer::pop_state () { - bool extra = (YYSTATE == extratoken); - - if (extra) - yy_pop_state (); - if (YYSTATE == notes || YYSTATE == chords) pitchname_tab_stack_ = scm_cdr (pitchname_tab_stack_); - yy_pop_state (); + // don't cross the maininput threshold + if (YYSTATE != maininput) + yy_pop_state (); - if (extra) { - hidden_state_ = YYSTATE; - yy_push_state (extratoken); - } } int Lily_lexer::identifier_type (SCM sid) { - int k = try_special_identifiers (&yylval.scm , sid); + int k = try_special_identifiers (&yylval , sid); return k >= 0 ? k : SCM_IDENTIFIER; } int -Lily_lexer::scan_escaped_word (string str) +Lily_lexer::scan_escaped_word (const string &str) { // use more SCM for this. // SCM sym = ly_symbol2scm (str.c_str ()); + yylval = SCM_UNSPECIFIED; int i = lookup_keyword (str); - if (i == MARKUP && is_lyric_state ()) - return LYRIC_MARKUP; + if (i != -1) return i; SCM sid = lookup_identifier (str); + if (Music *m = unsmob_music (sid)) + { + m->set_spot (override_input (here_input ())); + } + if (sid != SCM_UNDEFINED) return scan_scm_id (sid); string msg (_f ("unknown escaped string: `\\%s'", str)); LexerError (msg.c_str ()); - yylval.scm = ly_string2scm (str); + yylval = ly_string2scm (str); + + return STRING; +} + +int +Lily_lexer::scan_shorthand (const string &str) +{ + SCM sid = lookup_identifier (str); + if (Music *m = unsmob_music (sid)) + { + m->set_spot (override_input (here_input ())); + } + + if (sid != SCM_UNDEFINED) + return scan_scm_id (sid); + + string msg (_f ("undefined character or shorthand: %s", str)); + LexerError (msg.c_str ()); + + yylval = ly_string2scm (str); return STRING; } @@ -890,9 +957,9 @@ Lily_lexer::scan_scm_id (SCM sid) { int funtype = SCM_FUNCTION; - yylval.scm = get_music_function_transform (sid); + yylval = sid; - SCM s = scm_object_property (yylval.scm, ly_symbol2scm ("music-function-signature")); + SCM s = get_music_function_signature (sid); SCM cs = scm_car (s); if (scm_is_pair (cs)) @@ -908,7 +975,7 @@ Lily_lexer::scan_scm_id (SCM sid) funtype = SCM_FUNCTION; else programming_error ("Bad syntax function predicate"); - push_extra_token (EXPECT_NO_MORE_ARGS); + push_extra_token (here_input (), EXPECT_NO_MORE_ARGS); for (s = scm_cdr (s); scm_is_pair (s); s = scm_cdr (s)) { SCM optional = SCM_UNDEFINED; @@ -920,60 +987,53 @@ Lily_lexer::scan_scm_id (SCM sid) cs = SCM_CAR (cs); } - if (cs == Pitch_type_p_proc) - push_extra_token (EXPECT_PITCH); - else if (cs == Duration_type_p_proc) - push_extra_token (EXPECT_DURATION); - else if (ly_is_procedure (cs)) - push_extra_token (EXPECT_SCM, cs); + if (ly_is_procedure (cs)) + push_extra_token (here_input (), EXPECT_SCM, cs); else { programming_error ("Function parameter without type-checking predicate"); continue; } if (!scm_is_eq (optional, SCM_UNDEFINED)) - push_extra_token (EXPECT_OPTIONAL, optional); + push_extra_token (here_input (), EXPECT_OPTIONAL, optional); } return funtype; } - yylval.scm = sid; + yylval = sid; return identifier_type (sid); } int -Lily_lexer::scan_bare_word (string str) +Lily_lexer::scan_bare_word (const string &str) { SCM sym = ly_symbol2scm (str.c_str ()); if ((YYSTATE == notes) || (YYSTATE == chords)) { SCM handle = SCM_BOOL_F; if (scm_is_pair (pitchname_tab_stack_)) - handle = scm_hashq_get_handle (scm_car (pitchname_tab_stack_), sym); + handle = scm_hashq_get_handle (scm_cdar (pitchname_tab_stack_), sym); if (scm_is_pair (handle)) { - yylval.scm = scm_cdr (handle); - if (unsmob_pitch (yylval.scm)) + yylval = scm_cdr (handle); + if (Pitch::unsmob (yylval)) return (YYSTATE == notes) ? NOTENAME_PITCH : TONICNAME_PITCH; - else if (scm_is_symbol (yylval.scm)) + else if (scm_is_symbol (yylval)) return DRUM_PITCH; } else if ((YYSTATE == chords) && (handle = scm_hashq_get_handle (chordmodifier_tab_, sym))!= SCM_BOOL_F) { - yylval.scm = scm_cdr (handle); + yylval = scm_cdr (handle); return CHORD_MODIFIER; } } - yylval.scm = ly_string2scm (str); + yylval = ly_string2scm (str); return STRING; } int Lily_lexer::get_state () const { - if (YY_START == extratoken) - return hidden_state_; - else - return YY_START; + return YY_START; } bool @@ -1000,15 +1060,22 @@ Lily_lexer::is_figure_state () const return get_state () == figures; } +// The extra_token parameter specifies how to convert multiple values +// into additional tokens. For '#', additional values get pushed as +// SCM_IDENTIFIER. For '$', they get checked for their type and get +// pushed as a corresponding *_IDENTIFIER token. Since the latter +// tampers with yylval, it can only be done from the lexer itself, so +// this function is private. + SCM -Lily_lexer::eval_scm (SCM readerdata) +Lily_lexer::eval_scm (SCM readerdata, Input hi, char extra_token) { SCM sval = SCM_UNDEFINED; if (!SCM_UNBNDP (readerdata)) { - sval = ly_eval_scm (scm_car (readerdata), - *unsmob_input (scm_cdr (readerdata)), + sval = ly_eval_scm (readerdata, + hi, be_safe_global && is_main_input_, parser_); } @@ -1018,6 +1085,48 @@ Lily_lexer::eval_scm (SCM readerdata) error_level_ = 1; return SCM_UNSPECIFIED; } + + if (extra_token && SCM_VALUESP (sval)) + { + sval = scm_struct_ref (sval, SCM_INUM0); + + if (scm_is_pair (sval)) { + for (SCM p = scm_reverse (scm_cdr (sval)); + scm_is_pair (p); + p = scm_cdr (p)) + { + SCM v = scm_car (p); + if (Music *m = unsmob_music (v)) + { + if (!unsmob_input (m->get_property ("origin"))) + m->set_spot (override_input (here_input ())); + } + + int token; + switch (extra_token) { + case '$': + token = scan_scm_id (v); + if (!scm_is_eq (yylval, SCM_UNSPECIFIED)) + push_extra_token (here_input (), + token, yylval); + break; + case '#': + push_extra_token (here_input (), + SCM_IDENTIFIER, v); + break; + } + } + sval = scm_car (sval); + } else + sval = SCM_UNSPECIFIED; + } + + if (Music *m = unsmob_music (sval)) + { + if (!unsmob_input (m->get_property ("origin"))) + m->set_spot (override_input (here_input ())); + } + return sval; }