lexington 0.3.0

A very simple library for lexing / parsing
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
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
use std::marker::PhantomData;
use crate::Token;

/// A _shift-reduce parser_ processes input one token at a time
/// (currently without backtracking).  Tokens are transformed into
/// terms using two rules: _shift_ and _reduce_.  These operate over a
/// stack of (partially complete) terms.
pub trait ShiftReduceRule : Sized {
    /// Identifies the kind of tokens being processed by this rule.
    type Kind;
    /// Identifies the type of terms being produced by this rule.
    type Term;

    /// A _reduction rule combines a sub-term into its enclosing term.
    /// For example, when parsing arithmetic expressions we might reduce
    /// `1 + 2*x` with `y` to form `1 + 2*x + y`.    
    fn reduce(&self, lhs: Self::Term, rhs: Self::Term) -> Result<Self::Term,()>;

    /// A _shift rule_ reads the next token and decides what to do.  For
    /// example, it can return a completed term or it can begin a new
    /// term.    
    fn shift(&mut self, token: Token<Self::Kind>, stack: &mut Vec<Self::Term>) -> Result<Option<Self::Term>,()>;

    /// Apply a given reduction as necessary when reducing terms.
    fn apply<F:Fn(Self::Term,Self::Term)->Result<Self::Term,()>>(self, rule: F) -> ShiftApplyRule<Self::Kind,F,Self> {
        ShiftApplyRule{rule, rest: self, dummy: PhantomData }
    }

    fn first(self, term: Self::Term) -> ShiftFirstRule<Self::Kind,Self::Term,Self> {
        ShiftFirstRule{ term: Some(term), rest: self, dummy: PhantomData }
    }
    
    /// Handle a terminating token kind using a given parser.  For
    /// example, a token representing a number can be handled directly
    /// using an appropriate parser (e.g. `usize::parse()`).
    fn terminate<F>(self,kind: Self::Kind, rule: F) -> ShiftTerminalRule<Self::Kind,F,Self>
    where F:Fn(Token<Self::Kind>)->Self::Term
    {
        ShiftTerminalRule{kind, rule, rest: self}         
    }

    /// Skip all occurrences of the given token.  For example, if our
    /// language supports `WhiteSpace` tokens then we might want to
    /// skip them and, likewise, for comments, etc.
    fn skip(self,kind: Self::Kind) -> ShiftSkipRule<Self::Kind,Self::Term,Self> {
        ShiftSkipRule{kind, dummy: PhantomData, rest: self}
    }

    /// Begin parsing a new compound term when the given token is
    /// encountered.  The new term is initialised with a given default
    /// value.  For example, when parsing an S-expression (i.e. lisp),
    /// then upon encountering a `(` we start a new empty list.    
    fn open(self,kind: Self::Kind, default: Self::Term) -> ShiftOpenRule<Self::Kind,Self::Term,Self> {
        ShiftOpenRule{kind, default, rest: self}
    }
    
    /// Finish parsing a compound a compound term when the given token
    /// is encountered.  When this happens, the current stack element
    /// is popped off the stack and either reduced into the element
    /// below, or returned.  For example, when parsing an S-expression
    /// (i.e. lisp), the token `)` signifies the end of a list.    
    fn close(self,kind: Self::Kind) -> ShiftCloseRule<Self::Kind,fn(Self::Term)->Self::Term,Self>
    {
        ShiftCloseRule{kind, rule: |t| t, rest: self}
    }

    /// Finish parsing a compound a compound term when the given token
    /// is encountered.  When this happens, the current stack element
    /// is popped off the stack and either reduced into the element
    /// below, or returned.  For example, when parsing an S-expression
    /// (i.e. lisp), the token `)` signifies the end of a list.    
    fn close_with<F>(self,kind: Self::Kind, rule: F) -> ShiftCloseRule<Self::Kind,F,Self>
    where F:Fn(Self::Term)->Self::Term
    {
        ShiftCloseRule{kind, rule, rest: self}
    }
    
    /// Update the current state (i.e. topmost stack term) based on a
    /// given token.
    fn update<F:Fn(Self::Term)->Self::Term>(self, kind: Self::Kind, rule: F) -> ShiftUpdateRule<Self::Kind,F,Self> {
        ShiftUpdateRule{kind, rule, rest: self}
    }

    /// Update the current state (i.e. topmost stack item) based on a
    /// given token.  If the given rule returns `None`, then no action
    /// is taken.  Otherwise, the topmost stack item is replaced with
    /// that returned.
    fn update_with<F:Fn(Self::Term)->Result<(bool,Self::Term),()>>(self, kind: Self::Kind, rule: F) -> ShiftUpdateWithRule<Self::Kind,F,Self> {
        ShiftUpdateWithRule{kind, rule, rest: self}
    }    

    /// Update the current state (i.e. topmost stack item) based on a
    /// given token.  If the given rule returns `false`, then the rule
    /// did not apply and other rules should be considered.
    fn update_as<F:Fn(&mut Self::Term)->Result<bool,()>>(self, kind: Self::Kind, rule: F) -> ShiftUpdateAsRule<Self::Kind,F,Self> {
        ShiftUpdateAsRule{kind, rule, rest: self}
    }    
    
    /// Parse a given token stream and construct a corresponding term
    /// (or report an error).
    fn parse<I>(&mut self, input: I) -> Result<Self::Term,()>
    where I: IntoIterator<Item=Token<Self::Kind>>
    {
        let mut input = input.into_iter();
        // Construct initially empty stack
        let mut stack = Vec::new();
        // Continue reading tokens until nothing is left.
        while let Some(token) = input.next() {
            // Perform shift
            if let Some(t2) = self.shift(token, &mut stack)? {
                match stack.pop() {
                    None => {
                        // Parsing is complete!
                        return Ok(t2);
                    }
                    Some(t1) => {
                        // Perform reduction
                        let reduction = self.reduce(t1,t2)?;
                        // Process reduction                        
                        stack.push(reduction);
                    }
                }
            }
        }
        // If we get here then we reached the end-of-file before
        // parsing was complete.
        Err(())
    }    
}

// ===================================================================
// Empty Shift
// ===================================================================

/// The empty shift rule simply fails immediately with an error.  The
/// intention is that, starting from this rule, we can build up a
/// complete shift rule.
pub struct ShiftReduceParser<K,T>{
    dummy: PhantomData<(K,T)>,
}

impl<K,T> ShiftReduceParser<K,T> {
    pub fn new() -> Self {
        ShiftReduceParser{ dummy: PhantomData}
    }
}

impl<K:PartialEq,T> ShiftReduceRule for ShiftReduceParser<K,T>
{
    type Kind = K;
    type Term = T;
    
    fn shift(&mut self, _token: Token<K>, _stack: &mut Vec<T>) -> Result<Option<T>,()> {        
        Err(())
    }

    fn reduce(&self, _parent: T, _child: T) -> Result<T,()> {
        Err(())
    }    
}

// ===================================================================
// Shift Apply
// ===================================================================

pub struct ShiftApplyRule<K,F,SR> { rule: F, rest: SR, dummy: PhantomData<K>  }

impl<K:PartialEq,T,F:Fn(T,T)->Result<T,()>,SR> ShiftReduceRule for ShiftApplyRule<K,F,SR>
where SR: ShiftReduceRule<Kind=K,Term=T>
{
    type Kind = K;
    type Term = T;
        
    fn shift(&mut self, token: Token<K>, stack: &mut Vec<T>) -> Result<Option<T>,()> {
        self.rest.shift(token,stack)
    }

    fn reduce(&self, parent: T, child: T) -> Result<T,()> {
        (self.rule)(parent,child)
    }
}

// ===================================================================
// Shift First
// ===================================================================

pub struct ShiftFirstRule<K,T,SR> { term: Option<T>, dummy: PhantomData<K>, rest: SR }

impl<K:PartialEq,T,SR> ShiftReduceRule for ShiftFirstRule<K,T,SR>
    where SR: ShiftReduceRule<Kind=K,Term=T> {
    type Kind = K;
    type Term = T;
    
    fn shift(&mut self, token: Token<K>, stack: &mut Vec<T>) -> Result<Option<T>,()> {
        // Push term if first time
        if self.term.is_some() { stack.push(self.term.take().unwrap()); }
        // Continue with whatever comes next
        self.rest.shift(token,stack)            
    }

    fn reduce(&self, parent: T, child: T) -> Result<T,()> {
        self.rest.reduce(parent,child)
    }    
}

// ===================================================================
// Shift Terminate
// ===================================================================

/// Handle tokens directly using a given parser.  For example, a
/// number `123` can be handled directly using `parse()`, etc.
pub struct ShiftTerminalRule<K,F,SR> { kind: K, rule: F, rest: SR  }

impl<K:PartialEq,T,F:Fn(Token<K>)->T,SR> ShiftReduceRule for ShiftTerminalRule<K,F,SR>
where SR: ShiftReduceRule<Kind=K,Term=T>
{
    type Kind = K;
    type Term = T;
        
    fn shift(&mut self, token: Token<K>, stack: &mut Vec<T>) -> Result<Option<T>,()> {
        if token.kind == self.kind {
            Ok(Some((self.rule)(token)))
        } else {
            self.rest.shift(token,stack)
        }
    }

    fn reduce(&self, parent: T, child: T) -> Result<T,()> {
        self.rest.reduce(parent,child)
    }
}

// ===================================================================
// Shift Up
// ===================================================================

/// The given token indicates the start of a new compound element.  As
/// such, a new default element is pushed onto the stack.  For
/// example, when parsing an S-expression (i.e. lisp), then upon
/// encountering a `(` we start a new empty list.
pub struct ShiftOpenRule<K,T,SR> { kind: K, default: T, rest: SR }

impl<K:PartialEq,T:Clone,SR> ShiftReduceRule for ShiftOpenRule<K,T,SR>
where SR: ShiftReduceRule<Kind=K,Term=T>
{
    type Kind = K;
    type Term = T;
    
    fn shift(&mut self, token: Token<K>, stack: &mut Vec<T>) -> Result<Option<T>,()> {
        if token.kind == self.kind {
            // Push up the default value
            stack.push(self.default.clone());
            // Done
            Ok(None)
        } else {
            self.rest.shift(token,stack)
        }
    }

    fn reduce(&self, parent: T, child: T) -> Result<T,()> {
        self.rest.reduce(parent,child)
    }    
}

// ===================================================================
// Shift Close
// ===================================================================

/// The given token indicates the end of a compound element.  As such,
/// current stack element is popped off the stack and either reduced
/// into the element below, or returned.  For example, when parsing an
/// S-expression (i.e. lisp), the token `)` signifies the end of a
/// list.
pub struct ShiftCloseRule<K,F,SR> { kind: K, rule: F, rest: SR }

impl<K:PartialEq,T,F,SR> ShiftReduceRule for ShiftCloseRule<K,F,SR>
where F:Fn(T)->T,
      SR: ShiftReduceRule<Kind=K,Term=T> {
    
    type Kind =K;
    type Term = T;
    
    fn shift(&mut self, token: Token<K>, stack: &mut Vec<T>) -> Result<Option<T>,()> {
        if token.kind == self.kind && !stack.is_empty() {
            // Extract term to close.
            let term = stack.pop().unwrap();
            // Close term.
            let closed = (self.rule)(term);
            // Done
            Ok(Some(closed))
        } else {
            self.rest.shift(token,stack)
        }
    }

    fn reduce(&self, parent: T, child: T) -> Result<T,()> {
        self.rest.reduce(parent,child)
    }    
}

// ===================================================================
// Shift Skip
// ===================================================================

/// The given token should be ignored.  For example, if it indicates
/// whitespace or comments, etc.
pub struct ShiftSkipRule<K,T,SR> { kind: K, dummy: PhantomData<T>, rest: SR }

impl<K:PartialEq,T,SR> ShiftReduceRule for ShiftSkipRule<K,T,SR>
    where SR: ShiftReduceRule<Kind=K,Term=T> {
    type Kind = K;
    type Term = T;
    
    fn shift(&mut self, token: Token<K>, stack: &mut Vec<T>) -> Result<Option<T>,()> {
        if token.kind == self.kind {            
            Ok(None)
        } else {
            self.rest.shift(token,stack)
        }
    }

    fn reduce(&self, parent: T, child: T) -> Result<T,()> {
        self.rest.reduce(parent,child)
    }    
}

// ===================================================================
// Shift Update
// ===================================================================

pub struct ShiftUpdateRule<K,F,SR> { kind: K, rule: F, rest: SR  }

impl<K:PartialEq,T,F:Fn(T)->T,SR> ShiftReduceRule for ShiftUpdateRule<K,F,SR>
where SR: ShiftReduceRule<Kind=K,Term=T>
{
    type Kind = K;
    type Term = T;
        
    fn shift(&mut self, token: Token<K>, stack: &mut Vec<T>) -> Result<Option<T>,()> {
        if token.kind == self.kind && !stack.is_empty(){
            // Extract topmost item
            let top = stack.pop().unwrap();
            // Apply the update rule
            stack.push((self.rule)(top));
            // Done
            Ok(None)
        } else {
            self.rest.shift(token,stack)
        }
    }

    fn reduce(&self, parent: T, child: T) -> Result<T,()> {
        self.rest.reduce(parent,child)
    }
}

// ===================================================================
// Shift Update With
// ===================================================================

pub struct ShiftUpdateWithRule<K,F,SR> { kind: K, rule: F, rest: SR  }

impl<K:PartialEq,T,F:Fn(T)->Result<(bool,T),()>,SR> ShiftReduceRule for ShiftUpdateWithRule<K,F,SR>
where SR: ShiftReduceRule<Kind=K,Term=T>
{
    type Kind = K;
    type Term = T;
        
    fn shift(&mut self, token: Token<K>, stack: &mut Vec<T>) -> Result<Option<T>,()> {
        if token.kind == self.kind && !stack.is_empty(){
            // Extract topmost item
            let top = stack.pop().unwrap();
            // Apply the update rule
            let (ok,item) = (self.rule)(top)?;
	    // Put what we have back
	    stack.push(item);
            // Apply the update rule
	    if ok { return Ok(None); }
        } 
        self.rest.shift(token,stack)
    }

    fn reduce(&self, parent: T, child: T) -> Result<T,()> {
        self.rest.reduce(parent,child)
    }
}


// ===================================================================
// Shift Update As
// ===================================================================

pub struct ShiftUpdateAsRule<K,F,SR> { kind: K, rule: F, rest: SR  }

impl<K:PartialEq,T,F:Fn(&mut T)->Result<bool,()>,SR> ShiftReduceRule for ShiftUpdateAsRule<K,F,SR>
where SR: ShiftReduceRule<Kind=K,Term=T>
{
    type Kind = K;
    type Term = T;
        
    fn shift(&mut self, token: Token<K>, stack: &mut Vec<T>) -> Result<Option<T>,()> {
        if token.kind == self.kind {
	    // Extract topmost item
	    match stack.last_mut() {
		Some(item) => {
		    // Apply the update rule		    
		    if (self.rule)(item)? {
			// Successfully applied
			return Ok(None)
		    }
		}
		_ => {}
	    }
        }
	// Rule didn't apply, so continue trying to shift.
        self.rest.shift(token,stack)
    }

    fn reduce(&self, parent: T, child: T) -> Result<T,()> {
        self.rest.reduce(parent,child)
    }
}