X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=lily%2Flexer.ll;h=2ef59f43f5243618951de8eb6d4c0ea90b23b03f;hb=bdf8a9979fbc14018078288d00e76b5b75f2fe85;hp=71fb2479eec8596df6d390f989aa0b46f9176782;hpb=a93cdac09beaeb940a1776a5177fb823d1fa8337;p=lilypond.git diff --git a/lily/lexer.ll b/lily/lexer.ll index 71fb2479ee..2ef59f43f5 100644 --- a/lily/lexer.ll +++ b/lily/lexer.ll @@ -1,13 +1,23 @@ %{ // -*-Fundamental-*- /* - lexer.ll -- implement the Flex lexer + This file is part of LilyPond, the GNU music typesetter. - source file of the LilyPond music typesetter + Copyright (C) 1996--2011 Han-Wen Nienhuys + Jan Nieuwenhuizen - (c) 1996--2007 Han-Wen Nienhuys - Jan Nieuwenhuizen -*/ + LilyPond is free software: you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation, either version 3 of the License, or + (at your option) any later version. + + LilyPond 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 General Public License for more details. + You should have received a copy of the GNU General Public License + along with LilyPond. If not, see . +*/ /* backup rules @@ -36,17 +46,19 @@ using namespace std; #include "context-def.hh" -#include "identifier-smob.hh" +#include "duration.hh" #include "international.hh" #include "interval.hh" #include "lily-guile.hh" #include "lily-lexer.hh" +#include "lily-parser.hh" #include "lilypond-version.hh" #include "main.hh" #include "music.hh" #include "music-function.hh" #include "parse-scm.hh" #include "parser.hh" +#include "pitch.hh" #include "source-file.hh" #include "std-string.hh" #include "string-convert.hh" @@ -75,10 +87,10 @@ bool is_valid_version (string s); yylval.string = new string #define yylval \ - (*(YYSTYPE*)lexval) + (*(YYSTYPE*)lexval_) #define yylloc \ - (*(YYLTYPE*)lexloc) + (*(YYLTYPE*)lexloc_) #define YY_USER_ACTION add_lexed_char (YYLeng ()); @@ -120,22 +132,20 @@ AN {AA}|{N} ANY_CHAR (.|\n) PUNCT [?!:'`] ACCENT \\[`'"^] +SPECIAL_CHAR [&@] NATIONAL [\001-\006\021-\027\031\036] -TEX {AA}|-|{PUNCT}|{ACCENT}|{NATIONAL} -WORD {A}{AN}* +TEX {AA}|-|{PUNCT}|{ACCENT}|{NATIONAL}|{SPECIAL_CHAR} DASHED_WORD {A}({AN}|-)* DASHED_KEY_WORD \\{DASHED_WORD} ALPHAWORD {A}+ -DIGIT {N} UNSIGNED {N}+ E_UNSIGNED \\{N}+ FRACTION {N}+\/{N}+ INT -?{UNSIGNED} REAL ({INT}\.{N}*)|(-?\.{N}+) -KEYWORD \\{WORD} WHITE [ \n\t\f\r] HORIZONTALWHITE [ \t] BLACK [^ \n\t\f\r] @@ -152,7 +162,7 @@ BOM_UTF8 \357\273\277 <*>\r { - // windows-suck-suck-suck + // swallow and ignore carriage returns } {ANY_CHAR} { @@ -163,35 +173,50 @@ BOM_UTF8 \357\273\277 yyless (0); /* produce requested token */ - int type = extra_token_types_.back (); - extra_token_types_.pop_back (); - if (extra_token_types_.empty ()) + 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; } -{BOM_UTF8} { - if (this->lexloc->line_number () != 1 || this->lexloc->column_number () != 0) +<> { + /* 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}/.* { + if (this->lexloc_->line_number () != 1 || this->lexloc_->column_number () != 0) { - LexerError (_ ("stray UTF-8 BOM encountered").c_str ()); - exit (1); + LexerWarning (_ ("stray UTF-8 BOM encountered").c_str ()); + // exit (1); } - if (be_verbose_global) - message (_ ("Skipping UTF-8 BOM")); + debug_output (_ ("Skipping UTF-8 BOM")); } { "%{" { yy_push_state (longcomment); } - %[^{\n\r].*[\n\r] { + %[^{\n\r][^\n\r]*[\n\r] { } %[^{\n\r] { // backup rule } %[\n\r] { } - %[^{\n\r].* { + %[^{\n\r][^\n\r]* { } {WHITE}+ { @@ -249,15 +274,15 @@ BOM_UTF8 \357\273\277 this->here_input ().get_source_file ()->set_line (here_input ().start (), i); } -. { +{ANY_CHAR} { LexerError (_ ("quoted string expected after \\version").c_str ()); yy_pop_state (); } -. { +{ANY_CHAR} { LexerError (_ ("quoted string expected after \\sourcefilename").c_str ()); yy_pop_state (); } -. { +{ANY_CHAR} { LexerError (_ ("integer expected after \\sourcefileline").c_str ()); yy_pop_state (); } @@ -270,12 +295,6 @@ BOM_UTF8 \357\273\277 "%"+"}" { yy_pop_state (); } - <> { - 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 - } } @@ -292,14 +311,14 @@ BOM_UTF8 \357\273\277 \\include { yy_push_state (incl); } -\"[^"]*\" { /* got the include file name */ +\"[^""]*\" { /* got the include file name */ string s (YYText ()+1); s = s.substr (0, s.rfind ('"')); new_input (s, sources_); yy_pop_state (); } -\\{BLACK}*{WHITE} { /* got the include identifier */ +\\{BLACK}*{WHITE}? { /* got the include identifier */ string s = YYText () + 1; strip_trailing_white (s); if (s.length () && (s[s.length () - 1] == ';')) @@ -309,7 +328,7 @@ BOM_UTF8 \357\273\277 if (scm_is_string (sid)) { new_input (ly_scm2string (sid), sources_); yy_pop_state (); - } else { + } else { string msg (_f ("wrong or undefined identifier: `%s'", s )); LexerError (msg.c_str ()); @@ -318,7 +337,34 @@ BOM_UTF8 \357\273\277 scm_display (sid, err); } } -\"[^"]* { // backup rule +(\$|#) { // 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); + + for (int i = 0; i < n; i++) + { + yyinput (); + } + char_count_stack_.back () += n; + + if (scm_is_string (sval)) { + new_input (ly_scm2string (sval), sources_); + yy_pop_state (); + } else { + LexerError (_ ("string expected after \\include").c_str ()); + if (sval != SCM_UNDEFINED) { + SCM err = scm_current_error_port (); + scm_puts ("This value was found instead: ", err); + scm_display (sval, err); + } + } +} + +\"[^""]* { // backup rule error (_ ("end quote missing")); exit (1); } @@ -335,13 +381,10 @@ BOM_UTF8 \357\273\277 Input hi = here_input(); hi.step_forward (); SCM sval = ly_parse_scm (hi.start (), &n, hi, - be_safe_global && is_main_input_); + be_safe_global && is_main_input_, parser_); if (sval == SCM_UNDEFINED) - { - sval = SCM_UNSPECIFIED; error_level_ = 1; - } for (int i = 0; i < n; i++) { @@ -349,15 +392,30 @@ BOM_UTF8 \357\273\277 } char_count_stack_.back () += n; - if (unpack_identifier (sval) != SCM_UNDEFINED) + yylval.scm = 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_); + + for (int i = 0; i < n; i++) { - yylval.scm = unpack_identifier(sval); - return identifier_type (yylval.scm); + yyinput (); } + char_count_stack_.back () += n; + + sval = eval_scm (sval); - yylval.scm = sval; - return SCM_TOKEN; + int token = scan_scm_id (sval); + if (!scm_is_eq (yylval.scm, SCM_UNSPECIFIED)) + return token; } + { \<\< { return DOUBLE_ANGLE_OPEN; @@ -400,13 +458,9 @@ BOM_UTF8 \357\273\277 yylval.scm = scan_fraction (YYText ()); return FRACTION; } - - {DIGIT} { - yylval.i = String_convert::dec2int (string (YYText ())); - return DIGIT; - } + {UNSIGNED}/\/ | // backup rule {UNSIGNED} { - yylval.i = String_convert::dec2int (string (YYText ())); + yylval.scm = scm_c_read_string (YYText ()); return UNSIGNED; } {E_UNSIGNED} { @@ -419,7 +473,7 @@ BOM_UTF8 \357\273\277 \\{ESCAPED} { *yylval.string += to_string (escaped_char (YYText ()[1])); } - [^\\"]+ { + [^\\""]+ { *yylval.string += YYText (); } \" { @@ -445,8 +499,13 @@ BOM_UTF8 \357\273\277 yylval.scm = scan_fraction (YYText ()); return FRACTION; } + {UNSIGNED}/\/[^0-9] { // backup rule + yylval.scm = scm_c_read_string (YYText ()); + return UNSIGNED; + } + {UNSIGNED}/\/ | // backup rule {UNSIGNED} { - yylval.i = String_convert::dec2int (string (YYText ())); + yylval.scm = scm_c_read_string (YYText ()); return UNSIGNED; } {NOTECOMMAND} { @@ -485,8 +544,13 @@ BOM_UTF8 \357\273\277 yylval.scm = scan_fraction (YYText ()); return FRACTION; } + {UNSIGNED}/\/[^0-9] { // backup rule + yylval.scm = scm_c_read_string (YYText ()); + return UNSIGNED; + } + {UNSIGNED}/\/ | // backup rule {UNSIGNED} { - yylval.i = String_convert::dec2int (string (YYText ())); + yylval.scm = scm_c_read_string (YYText ()); return UNSIGNED; } - { @@ -516,61 +580,58 @@ BOM_UTF8 \357\273\277 } {MARKUPCOMMAND} { string str (YYText () + 1); + + int token_type = MARKUP_FUNCTION; SCM s = lookup_markup_command (str); - SCM s2 = lookup_markup_list_command (str); - if (scm_is_pair (s) && scm_is_symbol (scm_cdr (s)) ) { - yylval.scm = scm_car(s); - SCM tag = scm_cdr(s); - if (tag == ly_symbol2scm("markup0")) - return MARKUP_HEAD_MARKUP0; - if (tag == ly_symbol2scm("empty")) - return MARKUP_HEAD_EMPTY; - else if (tag == ly_symbol2scm ("markup0-markup1")) - return MARKUP_HEAD_MARKUP0_MARKUP1; - else if (tag == ly_symbol2scm ("markup-list0")) - return MARKUP_HEAD_LIST0; - else if (tag == ly_symbol2scm ("scheme0")) - return MARKUP_HEAD_SCM0; - else if (tag == ly_symbol2scm ("scheme0-scheme1")) - return MARKUP_HEAD_SCM0_SCM1; - else if (tag == ly_symbol2scm ("scheme0-markup1")) - return MARKUP_HEAD_SCM0_MARKUP1; - else if (tag == ly_symbol2scm ("scheme0-scheme1-markup2")) - return MARKUP_HEAD_SCM0_SCM1_MARKUP2; - else if (tag == ly_symbol2scm ("scheme0-markup1-markup2")) - return MARKUP_HEAD_SCM0_MARKUP1_MARKUP2; - else if (tag == ly_symbol2scm ("scheme0-scheme1-scheme2")) - return MARKUP_HEAD_SCM0_SCM1_SCM2; - else { - programming_error ("no parser tag defined for this markup signature"); - ly_display_scm (s); - assert(false); - } - } else if (scm_is_pair (s2) && scm_is_symbol (scm_cdr (s2))) { - yylval.scm = scm_car(s2); - SCM tag = scm_cdr(s2); - if (tag == ly_symbol2scm("empty")) - return MARKUP_LIST_HEAD_EMPTY; - else if (tag == ly_symbol2scm ("scheme0")) - return MARKUP_LIST_HEAD_SCM0; - else if (tag == ly_symbol2scm ("markup-list0")) - return MARKUP_LIST_HEAD_LIST0; - else if (tag == ly_symbol2scm ("scheme0-markup-list1")) - return MARKUP_LIST_HEAD_SCM0_LIST1; - else if (tag == ly_symbol2scm ("scheme0-scheme1-markup-list2")) - return MARKUP_LIST_HEAD_SCM0_SCM1_LIST2; - else { - programming_error ("no parser tag defined for this markup list signature"); - ly_display_scm (s); - assert(false); - } - } else - return scan_escaped_word (str); + + // lookup-markup-command returns a pair with the car + // being the function to call, and the cdr being the + // call signature specified to define-markup-command, + // a list of predicates. + + if (!scm_is_pair (s)) { + // If lookup-markup-command was not successful, we + // try lookup-markup-list-command instead. + // If this fails as well, we just scan and return + // the escaped word. + s = lookup_markup_list_command (str); + if (scm_is_pair (s)) + token_type = MARKUP_LIST_FUNCTION; + else + return scan_escaped_word (str); + } + + // If the list of predicates is, say, + // (number? number? markup?), then tokens + // EXPECT_MARKUP EXPECT_SCM EXPECT_SCM EXPECT_NO_MORE_ARGS + // will be generated. Note that we have to push them + // in reverse order, so the first token pushed in the + // loop will be EXPECT_NO_MORE_ARGS. + + yylval.scm = 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); + 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); + else if (predicate == ly_lily_module_constant ("markup?")) + push_extra_token(EXPECT_MARKUP); + else + push_extra_token(EXPECT_SCM, predicate); + } + return token_type; } [{}] { return YYText ()[0]; } - [^#{}"\\ \t\n\r\f]+ { + [^$#{}\"\\ \t\n\r\f]+ { string s (YYText ()); char c = s[s.length () - 1]; @@ -588,8 +649,14 @@ BOM_UTF8 \357\273\277 } } -<*><> { - if (is_main_input_) +<> { + 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 + } + +<> { if (is_main_input_) { /* 2 = init.ly + current file. > because we're before closing, but is_main_input_ should @@ -614,24 +681,18 @@ BOM_UTF8 \357\273\277 } } -{WORD} { - return scan_bare_word (YYText ()); -} -{KEYWORD} { - return scan_escaped_word (YYText () + 1); -} +-{UNSIGNED} | // backup rule {REAL} { - Real r; - int cnv = sscanf (YYText (), "%lf", &r); - assert (cnv == 1); - (void) cnv; - - yylval.scm = scm_from_double (r); + yylval.scm = scm_c_read_string (YYText ()); + return REAL; +} +-\. { // backup rule + yylval.scm = scm_from_double (0.0); return REAL; } {UNSIGNED} { - yylval.i = String_convert::dec2int (string (YYText ())); + yylval.scm = scm_c_read_string (YYText ()); return UNSIGNED; } @@ -691,15 +752,15 @@ 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) +Lily_lexer::push_extra_token (int token_type, SCM scm) { - if (extra_token_types_.empty ()) + if (scm_is_null (extra_tokens_)) { if (YY_START != extratoken) hidden_state_ = YY_START; yy_push_state (extratoken); } - extra_token_types_.push_back (token_type); + extra_tokens_ = scm_acons (scm_from_int (token_type), scm, extra_tokens_); } void @@ -771,37 +832,72 @@ Lily_lexer::scan_escaped_word (string str) return i; SCM sid = lookup_identifier (str); + 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); + + return STRING; +} + +int +Lily_lexer::scan_scm_id (SCM sid) +{ if (is_music_function (sid)) { + int funtype = SCM_FUNCTION; + yylval.scm = get_music_function_transform (sid); SCM s = scm_object_property (yylval.scm, ly_symbol2scm ("music-function-signature")); - push_extra_token (EXPECT_NO_MORE_ARGS); - for (; scm_is_pair (s); s = scm_cdr (s)) + SCM cs = scm_car (s); + + if (scm_is_pair (cs)) { - if (scm_car (s) == ly_music_p_proc) - push_extra_token (EXPECT_MUSIC); - else if (scm_car (s) == ly_lily_module_constant ("markup?")) - push_extra_token (EXPECT_MARKUP); - else if (ly_is_procedure (scm_car (s))) - push_extra_token (EXPECT_SCM); - else programming_error ("Function parameter without type-checking predicate"); + cs = SCM_CAR (cs); } - return MUSIC_FUNCTION; - } - if (sid != SCM_UNDEFINED) - { - yylval.scm = sid; - return identifier_type (sid); - } - - string msg (_f ("unknown escaped string: `\\%s'", str)); - LexerError (msg.c_str ()); + if (scm_is_eq (cs, ly_lily_module_constant ("ly:music?"))) + funtype = MUSIC_FUNCTION; + else if (scm_is_eq (cs, ly_lily_module_constant ("ly:event?"))) + funtype = EVENT_FUNCTION; + else if (ly_is_procedure (cs)) + funtype = SCM_FUNCTION; + else programming_error ("Bad syntax function predicate"); - yylval.scm = ly_string2scm (str); + push_extra_token (EXPECT_NO_MORE_ARGS); + for (s = scm_cdr (s); scm_is_pair (s); s = scm_cdr (s)) + { + SCM optional = SCM_UNDEFINED; + cs = scm_car (s); - return STRING; + if (scm_is_pair (cs)) + { + optional = SCM_CDR (cs); + 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); + else + { + programming_error ("Function parameter without type-checking predicate"); + continue; + } + if (!scm_is_eq (optional, SCM_UNDEFINED)) + push_extra_token (EXPECT_OPTIONAL, optional); + } + return funtype; + } + yylval.scm = sid; + return identifier_type (sid); } int @@ -820,11 +916,15 @@ Lily_lexer::scan_bare_word (string str) else if (scm_is_symbol (yylval.scm)) return DRUM_PITCH; } - else if ((handle = scm_hashq_get_handle (chordmodifier_tab_, sym))!= SCM_BOOL_F) + else if ((YYSTATE == chords) + && (handle = scm_hashq_get_handle (chordmodifier_tab_, sym))!= SCM_BOOL_F) { yylval.scm = scm_cdr (handle); return CHORD_MODIFIER; } + if ((chord_repetition_.repetition_symbol_ != SCM_EOL) + && to_boolean (scm_equal_p (chord_repetition_.repetition_symbol_, sym))) + return CHORD_REPETITION; } yylval.scm = ly_string2scm (str); @@ -864,6 +964,29 @@ Lily_lexer::is_figure_state () const return get_state () == figures; } +SCM +Lily_lexer::eval_scm (SCM readerdata) +{ + SCM sval = SCM_UNDEFINED; + + if (!SCM_UNBNDP (readerdata)) + { + sval = ly_eval_scm (scm_car (readerdata), + *unsmob_input (scm_cdr (readerdata)), + be_safe_global && is_main_input_, + parser_); + } + + if (SCM_UNBNDP (sval)) + { + error_level_ = 1; + return SCM_UNSPECIFIED; + } + return sval; +} + + + /* urg, belong to string (_convert) and should be generalised