]> git.donarmstrong.com Git - lilypond.git/blobdiff - lily/lexer.ll
Merge branch 'master' into lilypond/translation
[lilypond.git] / lily / lexer.ll
index 47c39793364a086c000b15d057fe0a5c26dc47ba..5743f44964d9e7ba99a393903d53dcd5a0faaeeb 100644 (file)
@@ -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 <hanwen@xs4all.nl>
+                 Jan Nieuwenhuizen <janneke@gnu.org>
 
-  (c) 1996--2007 Han-Wen Nienhuys <hanwen@xs4all.nl>
-           Jan Nieuwenhuizen <janneke@gnu.org>
-*/
+  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 <http://www.gnu.org/licenses/>.
+*/
 
 /*
   backup rules
 using namespace std;
 
 #include "context-def.hh"
+#include "duration.hh"
 #include "identifier-smob.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 +88,10 @@ 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 ());
 
@@ -122,7 +135,6 @@ PUNCT               [?!:'`]
 ACCENT         \\[`'"^]
 NATIONAL       [\001-\006\021-\027\031\036]
 TEX            {AA}|-|{PUNCT}|{ACCENT}|{NATIONAL}
-WORD           {A}{AN}*
 DASHED_WORD            {A}({AN}|-)*
 DASHED_KEY_WORD                \\{DASHED_WORD}
 
@@ -135,7 +147,6 @@ 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 +163,7 @@ BOM_UTF8    \357\273\277
 
 
 <*>\r          {
-       // windows-suck-suck-suck
+       // swallow and ignore carriage returns
 }
 
 <extratoken>{ANY_CHAR} {
@@ -174,26 +185,25 @@ BOM_UTF8  \357\273\277
    /* Use the trailing context feature. Otherwise, the BOM will not be
       found if the file starts with an identifier definition. */
 <INITIAL,chords,lyrics,figures,notes>{BOM_UTF8}/.* {
-  if (this->lexloc->line_number () != 1 || this->lexloc->column_number () != 0)
+  if (this->lexloc_->line_number () != 1 || this->lexloc_->column_number () != 0)
     {
       LexerError (_ ("stray UTF-8 BOM encountered").c_str ());
       exit (1);
     }
-  if (be_verbose_global)
-     message (_ ("Skipping UTF-8 BOM"));
+  debug_output (_ ("Skipping UTF-8 BOM"));
 }
 
 <INITIAL,chords,figures,incl,lyrics,markup,notes>{
   "%{" {
        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}+     {
 
@@ -251,15 +261,15 @@ BOM_UTF8  \357\273\277
        this->here_input ().get_source_file ()->set_line (here_input ().start (), i);
 }
 
-<version>    {
+<version>{ANY_CHAR}    {
        LexerError (_ ("quoted string expected after \\version").c_str ());
        yy_pop_state ();
 }
-<sourcefilename>     {
+<sourcefilename>{ANY_CHAR}     {
        LexerError (_ ("quoted string expected after \\sourcefilename").c_str ());
        yy_pop_state ();
 }
-<sourcefileline>     {
+<sourcefileline>{ANY_CHAR}     {
        LexerError (_ ("integer expected after \\sourcefileline").c_str ());
        yy_pop_state ();
 }
@@ -272,12 +282,6 @@ BOM_UTF8   \357\273\277
        "%"+"}"         {
                yy_pop_state ();
        }
-       <<EOF>>         {
-               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
-       }
 }
 
 
@@ -301,7 +305,7 @@ BOM_UTF8    \357\273\277
        new_input (s, sources_);
        yy_pop_state ();
 }
-<incl>\\{BLACK}*{WHITE} { /* got the include identifier */
+<incl>\\{BLACK}*{WHITE}? { /* got the include identifier */
        string s = YYText () + 1;
        strip_trailing_white (s);
        if (s.length () && (s[s.length () - 1] == ';'))
@@ -320,7 +324,7 @@ BOM_UTF8    \357\273\277
            scm_display (sid, err);
          }
 }
-<incl>\"[^"]*   { // backup rule
+<incl,version,sourcefilename>\"[^"]*   { // backup rule
        error (_ ("end quote missing"));
        exit (1);
 }
@@ -356,6 +360,11 @@ BOM_UTF8   \357\273\277
                yylval.scm = unpack_identifier(sval);
                return identifier_type (yylval.scm);
        }
+
+       for (size_t i = 0; i < pending_string_includes_.size (); i++)
+               new_input ("<included string>", pending_string_includes_[i],
+                          parser_->sources_);
+       pending_string_includes_.clear ();
                
        yylval.scm = sval;
        return SCM_TOKEN;
@@ -402,11 +411,15 @@ BOM_UTF8  \357\273\277
                yylval.scm =  scan_fraction (YYText ());
                return FRACTION;
        }
-
        {DIGIT}         {
                yylval.i = String_convert::dec2int (string (YYText ()));
                return DIGIT;
        }
+       {UNSIGNED}/\/[^0-9] { // backup rule
+               yylval.i = String_convert::dec2int (string (YYText ()));
+               return UNSIGNED;
+       }
+       {UNSIGNED}/\/   | // backup rule
        {UNSIGNED}              {
                yylval.i = String_convert::dec2int (string (YYText ()));
                return UNSIGNED;
@@ -421,7 +434,7 @@ BOM_UTF8    \357\273\277
        \\{ESCAPED}     {
                *yylval.string += to_string (escaped_char (YYText ()[1]));
        }
-       [^\\"]+ {
+       [^\\""]+        {
                *yylval.string += YYText ();
        }
        \"      {
@@ -447,6 +460,11 @@ BOM_UTF8   \357\273\277
                yylval.scm =  scan_fraction (YYText ());
                return FRACTION;
        }
+       {UNSIGNED}/\/[^0-9] { // backup rule
+               yylval.i = String_convert::dec2int (string (YYText ()));
+               return UNSIGNED;
+       }
+       {UNSIGNED}/\/   | // backup rule
        {UNSIGNED}              {
                yylval.i = String_convert::dec2int (string (YYText ()));
                return UNSIGNED;
@@ -487,6 +505,11 @@ BOM_UTF8   \357\273\277
                yylval.scm =  scan_fraction (YYText ());
                return FRACTION;
        }
+       {UNSIGNED}/\/[^0-9] { // backup rule
+               yylval.i = String_convert::dec2int (string (YYText ()));
+               return UNSIGNED;
+       }
+       {UNSIGNED}/\/   | // backup rule
        {UNSIGNED}              {
                yylval.i = String_convert::dec2int (string (YYText ()));
                return UNSIGNED;
@@ -518,58 +541,53 @@ 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-scheme1-markup2-markup3"))
-                               return MARKUP_HEAD_SCM0_SCM1_MARKUP2_MARKUP3;
-                       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);
+               }
+               return token_type;
        }
        [{}]    {
                return YYText ()[0];
@@ -593,7 +611,14 @@ BOM_UTF8   \357\273\277
 }
 
 <*><<EOF>> {
-       if (is_main_input_)
+       if (YY_START == longcomment)
+       {
+               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
+       }
+       else if (is_main_input_)
        {
                /* 2 = init.ly + current file.
                   > because we're before closing, but is_main_input_ should
@@ -618,12 +643,7 @@ 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);
@@ -633,6 +653,10 @@ BOM_UTF8   \357\273\277
        yylval.scm = scm_from_double (r);
        return REAL;
 }
+-\.    { // backup rule
+       yylval.scm = scm_from_double (0.0);
+       return REAL;
+}
 
 {UNSIGNED}     {
        yylval.i = String_convert::dec2int (string (YYText ()));
@@ -783,11 +807,17 @@ Lily_lexer::scan_escaped_word (string str)
                push_extra_token (EXPECT_NO_MORE_ARGS);
                for (; scm_is_pair (s); s = scm_cdr (s))
                {
-                       if (scm_car (s) == ly_music_p_proc)
+                       SCM cs = scm_car (s);
+                       
+                       if (cs == ly_music_p_proc)
                                push_extra_token (EXPECT_MUSIC);
-                       else if (scm_car (s) == ly_lily_module_constant ("markup?"))
+                       else 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 (cs == ly_lily_module_constant ("markup?"))
                                push_extra_token (EXPECT_MARKUP);
-                       else if (ly_is_procedure (scm_car (s)))
+                       else if (ly_is_procedure (cs))
                                push_extra_token (EXPECT_SCM);
                        else programming_error ("Function parameter without type-checking predicate");
                }
@@ -830,6 +860,9 @@ Lily_lexer::scan_bare_word (string str)
                    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);