1 | This package implements a PEG (parsing-expression grammar) |
---|
2 | parser-generator. PEGs are similar on the surface to CFGs, but avoid |
---|
3 | problems with ambiguity, &c. The syntax used is very similar to |
---|
4 | normal PEG syntax in a prefix style. |
---|
5 | |
---|
6 | |
---|
7 | Bases |
---|
8 | |
---|
9 | Characters and strings form the most basic base of any PE |
---|
10 | (parsing-expression). They match themselves at the beginning of the |
---|
11 | input: |
---|
12 | |
---|
13 | (if-matches "foobar" "foo" (next match)) ==> T |
---|
14 | (if-matches "foobar" "bar" (next match)) ==> NIL |
---|
15 | (if-matches "foobar" #\f (next match)) ==> T |
---|
16 | |
---|
17 | Regular expressions may also be used for matching. The CL-PPCRE |
---|
18 | library is used in this case. The regex is always anchored to the |
---|
19 | start of input. |
---|
20 | |
---|
21 | (if-matches "foobar" (^ "fo+") (next match)) ==> T |
---|
22 | (if-matches "foobar" (^ "bz?") (next match)) ==> NIL |
---|
23 | |
---|
24 | When these bases are used, input is expected to be a string. One |
---|
25 | other base may be used, and allows complete flexibility in both |
---|
26 | processing and input format. |
---|
27 | |
---|
28 | (defun test (input) |
---|
29 | (when (eq (first input) :a) (values t (rest input) :a))) |
---|
30 | |
---|
31 | (if-matches '(:a :b) test (next match)) ==> T |
---|
32 | |
---|
33 | The supplied function must return three values on success: T, the |
---|
34 | unconsumed input, and the tree that resulted from matching (this may |
---|
35 | be NIL). |
---|
36 | |
---|
37 | |
---|
38 | |
---|
39 | IF-MATCHES format |
---|
40 | |
---|
41 | When a match succeeds, the unconsumed input and the generated tree are |
---|
42 | bound to the variables specifies, and the THEN form is executed in |
---|
43 | this binding (the default is to evaluate to T). On failure, the ELSE |
---|
44 | form is executed (the default is NIL). |
---|
45 | |
---|
46 | (if-matches "foobar" "foo" (next match) (values next match) :fail) |
---|
47 | ==> "bar", "foo" |
---|
48 | |
---|
49 | (if-matches "foobar" "bar" (next match) (values next match) :fail) |
---|
50 | ==> :FAIL |
---|
51 | |
---|
52 | |
---|
53 | |
---|
54 | Combinations |
---|
55 | |
---|
56 | The bases may be combined in several ways. The first is in sequence, |
---|
57 | requiring each subsequent rule to match against what the previous left |
---|
58 | unconsumed. |
---|
59 | |
---|
60 | (if-matches "foobar" ("foo" "ba") (next match) (values next match)) |
---|
61 | ==> "r", ("foo" "ba") |
---|
62 | |
---|
63 | Next is the ordered choice. The first rule that matches determines |
---|
64 | the variable bindings, and no subsequent rule is attempted. |
---|
65 | |
---|
66 | (if-matches "foobar" (/ "foo" "f") (next match) (values next match)) |
---|
67 | ==> "bar", "foo" |
---|
68 | |
---|
69 | Next come the closure and semi-closures, * (0 or more), + (1 or more), |
---|
70 | and ? (0 or 1). |
---|
71 | |
---|
72 | (if-matches "foobar" (+ #\f) (next match) (values next match)) |
---|
73 | ==> "oobar", (#\f) |
---|
74 | |
---|
75 | A generalized count is also provided. |
---|
76 | |
---|
77 | (if-matches "oobar" ({} 1 3 #\o) (next match) (values next match)) |
---|
78 | ==> "bar", (#\o #\o) |
---|
79 | |
---|
80 | The closures may be defined in terms of the generalized count: |
---|
81 | |
---|
82 | (* rule) == ({} nil nil rule) |
---|
83 | (+ rule) == ({} 1 nil rule) |
---|
84 | (? rule) == ({} 0 1 rule) |
---|
85 | |
---|
86 | |
---|
87 | |
---|
88 | Nesting |
---|
89 | |
---|
90 | Rules may be nested to an arbitrary depth. |
---|
91 | |
---|
92 | (if-matches "foobar" (#\f ({} 1 3 #\o) "ba" (? #\x)) (next match) |
---|
93 | (values next match)) |
---|
94 | ==> "r", (#\f (#\o #\o) "ba") |
---|
95 | |
---|
96 | |
---|
97 | |
---|
98 | Greediness |
---|
99 | |
---|
100 | Matching is done in a greedy fashion. If a rule matches the input, |
---|
101 | the image of its match is consumed, even if this causes some |
---|
102 | containing or subsequent rule to fail later on. |
---|
103 | |
---|
104 | (if-matches "foobar" (#\f (+ #\o) "ob") (next match)) ==> NIL |
---|
105 | |
---|
106 | |
---|
107 | |
---|
108 | Predicates |
---|
109 | |
---|
110 | To help deal with greediness, two predicates are provided. The |
---|
111 | and-predicate requires its rule to match, and the not-predicate |
---|
112 | requires its rule to not match. Both predicates consume no input and |
---|
113 | generate no tree. The previous problem can be dealt with this way: |
---|
114 | |
---|
115 | (if-matches "foobar" (#\f (+ (#\o (! #\b))) "ob" (next match))) |
---|
116 | ==> T |
---|
117 | |
---|
118 | |
---|
119 | |
---|
120 | Efficiency |
---|
121 | |
---|
122 | This package does *not* generate packrat parsers (no memoization is |
---|
123 | done). Packratting allows parsing to execute in linear time, even |
---|
124 | when backtracking is necessary. How much a benefit packratting yields |
---|
125 | varies based both the language and the input, but for most computer |
---|
126 | languages backtracking is quite minimal and so packratting isn't |
---|
127 | needed. |
---|
128 | |
---|
129 | The generated code nonetheless should be quite efficient, is it |
---|
130 | expands into mostly IF-forms, and so the compiler is free to do many |
---|
131 | optimizations. In cases with static input (such as the examples), you |
---|
132 | may even see your compiler optimize away all calculations! |
---|
133 | |
---|
134 | If extra speed is needed, you may consider lexing your input |
---|
135 | beforehand. This won't necessarily eliminate backtracking, but will |
---|
136 | allow matching and backtracking to proceed in larger steps. (An |
---|
137 | example of this is forthcoming.) |
---|