earlgrey 0.4.1

A library for parsing context-free grammars using Earley algorithm
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288

#![deny(warnings)]

use super::ebnf_tokenizer::EbnfTokenizer;
use crate::earley::{
    EarleyForest, EarleyParser, Grammar, GrammarBuilder
};
use std::cell::RefCell;

macro_rules! debug {
    ($($args:tt)*) => (if cfg!(feature="debug") { eprintln!($($args)*); })
}

#[derive(Clone,Debug)]
enum G {
    VariantList(Vec<Vec<String>>),
    Variant(Vec<String>),
    Atom(String),
    Nop
}

// use to destructure G enum into a specific alternative
macro_rules! pull {
    ($p:path, $e:expr) => (match $e {
        $p(value) => value,
        n => panic!("Bad pull match={:?}", n)
    })
}

// https://en.wikipedia.org/wiki/Extended_Backus%E2%80%93Naur_form
fn ebnf_grammar() -> Grammar {
    GrammarBuilder::default()
      .terminal("<Id>", move |s|
                s.chars().enumerate().all(|(i, c)|
                    i == 0 && c.is_alphabetic() ||
                    i > 0 && (c.is_alphanumeric() || c == '_')))
      .terminal("<Chars>", move |s| s.chars().all(|c| !c.is_control()))
      .terminal("@<Tag>", move |s|
                s.chars().enumerate().all(|(i, c)|
                    i == 0 && c == '@' ||
                    i == 1 && c.is_alphabetic() ||
                    i > 1 && (c.is_alphanumeric() || c == '_')))
      .terminal(":=", |s| s == ":=")
      .terminal(";", |s| s == ";")
      .terminal("[", |s| s == "[")
      .terminal("]", |s| s == "]")
      .terminal("{", |s| s == "{")
      .terminal("}", |s| s == "}")
      .terminal("(", |s| s == "(")
      .terminal(")", |s| s == ")")
      .terminal("|", |s| s == "|")
      .terminal("'", |s| s == "'")
      .terminal("\"", |s| s == "\"")
      .nonterm("<RuleList>")
      .nonterm("<Rule>")
      .nonterm("<VariantList>")
      .nonterm("<Variant>")
      .nonterm("<Atom>")
      .rule("<RuleList>", &["<RuleList>", "<Rule>"])
      .rule("<RuleList>", &["<Rule>"])
      .rule("<Rule>", &["<Id>", ":=", "<VariantList>", ";"])
      .rule("<VariantList>", &["<VariantList>", "|", "<Variant>"])
      .rule("<VariantList>", &["<Variant>"])
      .rule("<Variant>", &["<Variant>", "<Atom>"])
      .rule("<Variant>", &["<Atom>"])
      .rule("<Atom>", &["<Id>"])
      .rule("<Atom>", &["'", "<Chars>", "'"])
      .rule("<Atom>", &["\"", "<Chars>", "\""])
      .rule("<Atom>", &["[", "<VariantList>", "]"])
      .rule("<Atom>", &["{", "<VariantList>", "}"])
      .rule("<Atom>", &["(", "<VariantList>", ")"])
      .rule("<Atom>", &["[", "<VariantList>", "]", "@<Tag>"])
      .rule("<Atom>", &["{", "<VariantList>", "}", "@<Tag>"])
      .rule("<Atom>", &["(", "<VariantList>", ")", "@<Tag>"])
      .into_grammar("<RuleList>")
      .expect("Bad EBNF Grammar")
}

fn ebnf_terminal_parser(
    user_grammar_builder: &RefCell<GrammarBuilder>
) -> impl Fn(&str, &str) -> G + '_ {
    move |symbol, token| {
        match symbol {
            "<Id>" => {
                debug!("Adding non-term {:?}", token);
                user_grammar_builder.borrow_mut().nonterm_try(token);
            },
            "@<Tag>" => {
                debug!("Adding non-term {:?}", token);
                user_grammar_builder.borrow_mut().nonterm_try(token);
            },
            "<Chars>" => {
                debug!("Adding terminal {:?}", token);
                let tok = token.to_string();
                user_grammar_builder.borrow_mut()
                    .terminal_try(token, move |s| s == tok);
            },
            _ => ()
        }
        G::Atom(token.to_string())
    }
}

fn ebnf_rule_action<'a>(ev: &mut EarleyForest<'a, G>, gb: &'a RefCell<GrammarBuilder>) {
    ev.action("<Rule> -> <Id> := <VariantList> ;", move |mut n| {
        let id = pull!(G::Atom, n.remove(0));
        let body = pull!(G::VariantList, n.remove(1));
        let mut t_gb = gb.borrow_mut();
        for rule in body {
            debug!("Adding rule {:?} -> {:?}", id, rule);
            t_gb.rule_try(&id, &rule.iter().map(|s| s.as_str()).collect::<Vec<&str>>());
        }
        G::Nop
    });
}

fn ebnf_variantlist_action(ev: &mut EarleyForest<'_, G>) {
    ev.action("<VariantList> -> <VariantList> | <Variant>", |mut n| {
        let mut body = pull!(G::VariantList, n.remove(0));
        body.push(pull!(G::Variant, n.remove(1)));
        G::VariantList(body)
    });
    ev.action("<VariantList> -> <Variant>", |mut n| {
        let part = pull!(G::Variant, n.remove(0));
        G::VariantList(vec!(part))
    });
}

fn ebnf_variant_action(ev: &mut EarleyForest<'_, G>) {
    ev.action("<Variant> -> <Variant> <Atom>", |mut n| {
        let mut part = pull!(G::Variant, n.remove(0));
        part.push(pull!(G::Atom, n.remove(0)));
        G::Variant(part)
    });
    ev.action("<Variant> -> <Atom>", |mut n| {
        G::Variant(vec!(pull!(G::Atom, n.remove(0))))
    });
}

fn ebnf_grouping_action<'a>(ev: &mut EarleyForest<'a, G>, gb: &'a RefCell<GrammarBuilder>) {
    ev.action("<Atom> -> ( <VariantList> )", move |mut n| {
        let aux = gb.borrow().unique_symbol_name();
        debug!("Adding non-term {:?}", aux);
        let mut t_gb = gb.borrow_mut();
        t_gb.nonterm_try(&aux);
        let body = pull!(G::VariantList, n.remove(1));
        for rule in body {
            debug!("Adding rule {:?} -> {:?}", aux, rule);
            t_gb.rule_try(&aux, &rule.iter().map(|s| s.as_str()).collect::<Vec<&str>>());
        }
        G::Atom(aux)
    });
    ev.action("<Atom> -> ( <VariantList> ) @<Tag>", move |mut n| {
        let aux = pull!(G::Atom, n.remove(3));
        debug!("Adding non-term {:?}", aux);
        let mut t_gb = gb.borrow_mut();
        t_gb.nonterm_try(&aux);
        let body = pull!(G::VariantList, n.remove(1));
        for rule in body {
            debug!("Adding rule {:?} -> {:?}", aux, rule);
            t_gb.rule_try(&aux, &rule.iter().map(|s| s.as_str()).collect::<Vec<&str>>());
        }
        G::Atom(aux)
    });
}

fn ebnf_optional_action<'a>(ev: &mut EarleyForest<'a, G>, gb: &'a RefCell<GrammarBuilder>) {
    ev.action("<Atom> -> [ <VariantList> ]", move |mut n| {
        // <Atom> -> aux ; aux -> <e> | <VariantList> ;
        let aux = gb.borrow().unique_symbol_name();
        debug!("Adding non-term {:?}", aux);
        let mut t_gb = gb.borrow_mut();
        t_gb.nonterm_try(&aux);
        let body = pull!(G::VariantList, n.remove(1));
        for rule in body {
            debug!("Adding rule {:?} -> {:?}", aux, rule);
            t_gb.rule_try(&aux, &rule.iter().map(|s| s.as_str()).collect::<Vec<&str>>());
            debug!("Adding rule {:?} -> []", aux);
            t_gb.rule_try(&aux, &[]);
        }
        G::Atom(aux)
    });
    ev.action("<Atom> -> [ <VariantList> ] @<Tag>", move |mut n| {
        let aux = pull!(G::Atom, n.remove(3));
        debug!("Adding non-term {:?}", aux);
        let mut t_gb = gb.borrow_mut();
        t_gb.nonterm_try(&aux);
        let body = pull!(G::VariantList, n.remove(1));
        for rule in body {
            debug!("Adding rule {:?} -> {:?}", aux, rule);
            t_gb.rule_try(&aux, &rule.iter().map(|s| s.as_str()).collect::<Vec<&str>>());
            debug!("Adding rule {:?} -> []", aux);
            t_gb.rule_try(&aux, &[]);
        }
        G::Atom(aux)
    });
}

fn ebnf_repeat_action<'a>(ev: &mut EarleyForest<'a, G>, gb: &'a RefCell<GrammarBuilder>) {
    ev.action("<Atom> -> { <VariantList> }", move |mut n| {
        // <Atom> -> aux ; aux -> <e> | <VariantList> aux ;
        let aux = gb.borrow().unique_symbol_name();
        debug!("Adding non-term {:?}", aux);
        let mut t_gb = gb.borrow_mut();
        t_gb.nonterm_try(&aux);
        let body = pull!(G::VariantList, n.remove(1));
        for mut rule in body {
            rule.push(aux.clone());
            debug!("Adding rule {:?} -> {:?}", aux, rule);
            t_gb.rule_try(&aux, &rule.iter().map(|s| s.as_str()).collect::<Vec<&str>>());
            debug!("Adding rule {:?} -> []", aux);
            t_gb.rule_try(&aux, &[]);
        }
        G::Atom(aux)
    });
    ev.action("<Atom> -> { <VariantList> } @<Tag>", move |mut n| {
        // <Atom> -> aux ; aux -> <e> | <VariantList> aux ;
        let aux = pull!(G::Atom, n.remove(3));
        debug!("Adding non-term {:?}", aux);
        let mut t_gb = gb.borrow_mut();
        t_gb.nonterm_try(&aux);
        let body = pull!(G::VariantList, n.remove(1));
        for mut rule in body {
            rule.push(aux.clone());
            debug!("Adding rule {:?} -> {:?}", aux, rule);
            t_gb.rule_try(&aux, &rule.iter().map(|s| s.as_str()).collect::<Vec<&str>>());
            debug!("Adding rule {:?} -> []", aux);
            t_gb.rule_try(&aux, &[]);
        }
        G::Atom(aux)
    });
}

pub struct EbnfGrammarParser {
    start: String,
    grammar: String,
    grammar_builder: GrammarBuilder,
}

impl EbnfGrammarParser {
    // Parse a user grammar into a builder where we can plug terminal matchers
    pub fn new(grammar: &str, start: &str) -> Self {
        Self{
            start: start.to_string(),
            grammar: grammar.to_string(),
            grammar_builder: GrammarBuilder::default()
        }
    }

    // Plug-in functions that parse Terminals before we build the grammar
    pub fn plug_terminal(mut self, name: &str, pred: impl Fn(&str) -> bool + 'static) -> Self {
        debug!("Adding terminal {:?}", name);
        self.grammar_builder.terminal_try(&name, pred);
        self
    }

    pub fn into_grammar(self) -> Result<Grammar, String> {
        // Need to move grammar_builder into a refcell because ebnf
        // semantic actions need mutable access to add rules and symbols.
        // These grammar-builder changes are executed while the ebnf-parser
        // is evaluating semantic actions, ie: at `eval_all` line.
        let grammar_builder = RefCell::new(self.grammar_builder);
        {
            let mut user_semanter = EarleyForest::new(
                ebnf_terminal_parser(&grammar_builder)
            );
            user_semanter.action("<RuleList> -> <RuleList> <Rule>", |_| G::Nop);
            user_semanter.action("<RuleList> -> <Rule>", |_| G::Nop);
            ebnf_rule_action(&mut user_semanter, &grammar_builder);
            ebnf_variantlist_action(&mut user_semanter);
            ebnf_variant_action(&mut user_semanter);
            ebnf_grouping_action(&mut user_semanter, &grammar_builder);
            ebnf_optional_action(&mut user_semanter, &grammar_builder);
            ebnf_repeat_action(&mut user_semanter, &grammar_builder);
            user_semanter.action("<Atom> -> <Id>", |mut n| n.remove(0));
            user_semanter.action("<Atom> -> ' <Chars> '", |mut n| n.remove(1));
            user_semanter.action("<Atom> -> \" <Chars> \"", |mut n| n.remove(1));

            // Create a parser for EBNF which we'll use to parse input grammar
            let parsed_user_grammar = EarleyParser::new(ebnf_grammar())
                .parse(EbnfTokenizer::new(self.grammar.chars()))?;
            // 
            if user_semanter.eval_all(&parsed_user_grammar)?.len() != 1 {
                panic!("BUG: EBNF grammar shouldn't be ambiguous!");
            }
        }
        grammar_builder.into_inner().into_grammar(&self.start)
    }
}