Shortcuts on this page
r m x toggle line displays
j k next/prev highlighted chunk
0 (zero) top of page
1 (one) first highlighted chunk
1"""
2 pygments.lexers.lilypond
3 ~~~~~~~~~~~~~~~~~~~~~~~~
5 Lexer for LilyPond.
7 :copyright: Copyright 2006-2025 by the Pygments team, see AUTHORS.
8 :license: BSD, see LICENSE for details.
9"""
11import re
13from pygments.lexer import bygroups, default, inherit, words
14from pygments.lexers.lisp import SchemeLexer
15from pygments.lexers._lilypond_builtins import (
16 keywords, pitch_language_names, clefs, scales, repeat_types, units,
17 chord_modifiers, pitches, music_functions, dynamics, articulations,
18 music_commands, markup_commands, grobs, translators, contexts,
19 context_properties, grob_properties, scheme_functions, paper_variables,
20 header_variables
21)
22from pygments.token import Token
24__all__ = ["LilyPondLexer"]
26# In LilyPond, (unquoted) name tokens only contain letters, hyphens,
27# and underscores, where hyphens and underscores must not start or end
28# a name token.
29#
30# Note that many of the entities listed as LilyPond built-in keywords
31# (in file `_lilypond_builtins.py`) are only valid if surrounded by
32# double quotes, for example, 'hufnagel-fa1'. This means that
33# `NAME_END_RE` doesn't apply to such entities in valid LilyPond code.
34NAME_END_RE = r"(?=\d|[^\w\-]|[\-_][\W\d])"
36def builtin_words(names, backslash, suffix=NAME_END_RE):
37 prefix = r"[\-_^]?"
38 if backslash == "mandatory":
39 prefix += r"\\"
40 elif backslash == "optional":
41 prefix += r"\\?"
42 else:
43 assert backslash == "disallowed"
44 return words(names, prefix, suffix)
47class LilyPondLexer(SchemeLexer):
48 """
49 Lexer for input to LilyPond, a text-based music typesetter.
51 .. important::
53 This lexer is meant to be used in conjunction with the ``lilypond`` style.
54 """
55 name = 'LilyPond'
56 url = 'https://lilypond.org'
57 aliases = ['lilypond']
58 filenames = ['*.ly']
59 mimetypes = []
60 version_added = '2.11'
62 flags = re.DOTALL | re.MULTILINE
64 # Because parsing LilyPond input is very tricky (and in fact
65 # impossible without executing LilyPond when there is Scheme
66 # code in the file), this lexer does not try to recognize
67 # lexical modes. Instead, it catches the most frequent pieces
68 # of syntax, and, above all, knows about many kinds of builtins.
70 # In order to parse embedded Scheme, this lexer subclasses the SchemeLexer.
71 # It redefines the 'root' state entirely, and adds a rule for #{ #}
72 # to the 'value' state. The latter is used to parse a Scheme expression
73 # after #.
75 def get_tokens_unprocessed(self, text):
76 """Highlight Scheme variables as LilyPond builtins when applicable."""
77 for index, token, value in super().get_tokens_unprocessed(text):
78 if token is Token.Name.Function or token is Token.Name.Variable:
79 if value in scheme_functions:
80 token = Token.Name.Builtin.SchemeFunction
81 elif token is Token.Name.Builtin:
82 token = Token.Name.Builtin.SchemeBuiltin
83 yield index, token, value
85 tokens = {
86 "root": [
87 # Whitespace.
88 (r"\s+", Token.Text.Whitespace),
90 # Multi-line comments. These are non-nestable.
91 (r"%\{.*?%\}", Token.Comment.Multiline),
93 # Simple comments.
94 (r"%.*?$", Token.Comment.Single),
96 # End of embedded LilyPond in Scheme.
97 (r"#\}", Token.Punctuation, "#pop"),
99 # Embedded Scheme, starting with # ("delayed"),
100 # or $ (immediate). #@ and and $@ are the lesser known
101 # "list splicing operators".
102 (r"[#$]@?", Token.Punctuation, "value"),
104 # Any kind of punctuation:
105 # - sequential music: { },
106 # - parallel music: << >>,
107 # - voice separator: << \\ >>,
108 # - chord: < >,
109 # - bar check: |,
110 # - dot in nested properties: \revert NoteHead.color,
111 # - equals sign in assignments and lists for various commands:
112 # \override Stem.color = red,
113 # - comma as alternative syntax for lists: \time 3,3,2 4/4,
114 # - colon in tremolos: c:32,
115 # - double hyphen and underscore in lyrics: li -- ly -- pond __
116 # (which must be preceded by ASCII whitespace)
117 (r"""(?x)
118 \\\\
119 | (?<= \s ) (?: -- | __ )
120 | [{}<>=.,:|]
121 """, Token.Punctuation),
123 # Pitches, with optional octavation marks, octave check,
124 # and forced or cautionary accidental.
125 (words(pitches, suffix=r"=?[',]*!?\??" + NAME_END_RE), Token.Pitch),
127 # Strings, optionally with direction specifier.
128 (r'[\-_^]?"', Token.String, "string"),
130 # Numbers.
131 (r"-?\d+\.\d+", Token.Number.Float), # 5. and .5 are not allowed
132 (r"-?\d+/\d+", Token.Number.Fraction),
133 # Integers, or durations with optional augmentation dots.
134 # We have no way to distinguish these, so we highlight
135 # them all as numbers.
136 #
137 # Normally, there is a space before the integer (being an
138 # argument to a music function), which we check here. The
139 # case without a space is handled below (as a fingering
140 # number).
141 (r"""(?x)
142 (?<= \s ) -\d+
143 | (?: (?: \d+ | \\breve | \\longa | \\maxima )
144 \.* )
145 """, Token.Number),
146 # Separates duration and duration multiplier highlighted as fraction.
147 (r"\*", Token.Number),
149 # Ties, slurs, manual beams.
150 (r"[~()[\]]", Token.Name.Builtin.Articulation),
152 # Predefined articulation shortcuts. A direction specifier is
153 # required here.
154 (r"[\-_^][>^_!.\-+]", Token.Name.Builtin.Articulation),
156 # Fingering numbers, string numbers.
157 (r"[\-_^]?\\?\d+", Token.Name.Builtin.Articulation),
159 # Builtins.
160 (builtin_words(keywords, "mandatory"), Token.Keyword),
161 (builtin_words(pitch_language_names, "disallowed"), Token.Name.PitchLanguage),
162 (builtin_words(clefs, "disallowed"), Token.Name.Builtin.Clef),
163 (builtin_words(scales, "mandatory"), Token.Name.Builtin.Scale),
164 (builtin_words(repeat_types, "disallowed"), Token.Name.Builtin.RepeatType),
165 (builtin_words(units, "mandatory"), Token.Number),
166 (builtin_words(chord_modifiers, "disallowed"), Token.ChordModifier),
167 (builtin_words(music_functions, "mandatory"), Token.Name.Builtin.MusicFunction),
168 (builtin_words(dynamics, "mandatory"), Token.Name.Builtin.Dynamic),
169 # Those like slurs that don't take a backslash are covered above.
170 (builtin_words(articulations, "mandatory"), Token.Name.Builtin.Articulation),
171 (builtin_words(music_commands, "mandatory"), Token.Name.Builtin.MusicCommand),
172 (builtin_words(markup_commands, "mandatory"), Token.Name.Builtin.MarkupCommand),
173 (builtin_words(grobs, "disallowed"), Token.Name.Builtin.Grob),
174 (builtin_words(translators, "disallowed"), Token.Name.Builtin.Translator),
175 # Optional backslash because of \layout { \context { \Score ... } }.
176 (builtin_words(contexts, "optional"), Token.Name.Builtin.Context),
177 (builtin_words(context_properties, "disallowed"), Token.Name.Builtin.ContextProperty),
178 (builtin_words(grob_properties, "disallowed"),
179 Token.Name.Builtin.GrobProperty,
180 "maybe-subproperties"),
181 # Optional backslashes here because output definitions are wrappers
182 # around modules. Concretely, you can do, e.g.,
183 # \paper { oddHeaderMarkup = \evenHeaderMarkup }
184 (builtin_words(paper_variables, "optional"), Token.Name.Builtin.PaperVariable),
185 (builtin_words(header_variables, "optional"), Token.Name.Builtin.HeaderVariable),
187 # Other backslashed-escaped names (like dereferencing a
188 # music variable), possibly with a direction specifier.
189 (r"[\-_^]?\\.+?" + NAME_END_RE, Token.Name.BackslashReference),
191 # Definition of a variable. Support assignments to alist keys
192 # (myAlist.my-key.my-nested-key = \markup \spam \eggs).
193 (r"""(?x)
194 (?: [^\W\d] | - )+
195 (?= (?: [^\W\d] | [\-.] )* \s* = )
196 """, Token.Name.Lvalue),
198 # Virtually everything can appear in markup mode, so we highlight
199 # as text. Try to get a complete word, or we might wrongly lex
200 # a suffix that happens to be a builtin as a builtin (e.g., "myStaff").
201 (r"([^\W\d]|-)+?" + NAME_END_RE, Token.Text),
202 (r".", Token.Text),
203 ],
204 "string": [
205 (r'"', Token.String, "#pop"),
206 (r'\\.', Token.String.Escape),
207 (r'[^\\"]+', Token.String),
208 ],
209 "value": [
210 # Scan a LilyPond value, then pop back since we had a
211 # complete expression.
212 (r"#\{", Token.Punctuation, ("#pop", "root")),
213 inherit,
214 ],
215 # Grob subproperties are undeclared and it would be tedious
216 # to maintain them by hand. Instead, this state allows recognizing
217 # everything that looks like a-known-property.foo.bar-baz as
218 # one single property name.
219 "maybe-subproperties": [
220 (r"\s+", Token.Text.Whitespace),
221 (r"(\.)((?:[^\W\d]|-)+?)" + NAME_END_RE,
222 bygroups(Token.Punctuation, Token.Name.Builtin.GrobProperty)),
223 default("#pop"),
224 ]
225 }