Commit | Line | Data |
---|---|---|
dea4d055 MW |
1 | ;;; -*-lisp-*- |
2 | ;;; | |
3 | ;;; Implementation of lexical analysis protocol. | |
4 | ;;; | |
5 | ;;; (c) 2009 Straylight/Edgeware | |
6 | ;;; | |
7 | ||
8 | ;;;----- Licensing notice --------------------------------------------------- | |
9 | ;;; | |
10 | ;;; This file is part of the Sensble Object Design, an object system for C. | |
11 | ;;; | |
12 | ;;; SOD is free software; you can redistribute it and/or modify | |
13 | ;;; it under the terms of the GNU General Public License as published by | |
14 | ;;; the Free Software Foundation; either version 2 of the License, or | |
15 | ;;; (at your option) any later version. | |
16 | ;;; | |
17 | ;;; SOD is distributed in the hope that it will be useful, | |
18 | ;;; but WITHOUT ANY WARRANTY; without even the implied warranty of | |
19 | ;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
20 | ;;; GNU General Public License for more details. | |
21 | ;;; | |
22 | ;;; You should have received a copy of the GNU General Public License | |
23 | ;;; along with SOD; if not, write to the Free Software Foundation, | |
24 | ;;; Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. | |
25 | ||
26 | (cl:in-package #:sod) | |
27 | ||
28 | ;;;-------------------------------------------------------------------------- | |
239fa5bd | 29 | ;;; Class implementation. |
dea4d055 | 30 | |
239fa5bd MW |
31 | (defmethod shared-initialize :after |
32 | ((scanner sod-token-scanner) slot-names &key) | |
33 | (default-slot (scanner 'sod-parser::filename slot-names) | |
34 | (scanner-filename (token-scanner-char-scanner scanner)))) | |
dea4d055 | 35 | |
239fa5bd MW |
36 | (defmethod make-scanner-stream ((scanner sod-token-scanner)) |
37 | (make-scanner-stream (token-scanner-char-scanner scanner))) | |
dea4d055 MW |
38 | |
39 | ;;;-------------------------------------------------------------------------- | |
239fa5bd | 40 | ;;; Indicators and error messages. |
dea4d055 | 41 | |
239fa5bd MW |
42 | (defun show-char (stream char &optional colonp atsignp) |
43 | "Format CHAR to STREAM in a readable way. | |
dea4d055 | 44 | |
239fa5bd MW |
45 | Usable in `format''s ~/.../ command." |
46 | (declare (ignore colonp atsignp)) | |
47 | (cond ((null char) (write-string "<eof>" stream)) | |
48 | ((and (graphic-char-p char) (char/= char #\space)) | |
49 | (format stream "`~C'" char)) | |
50 | (t (format stream "<~(~:C~)>" char)))) | |
dea4d055 | 51 | |
048d0b2d MW |
52 | (defun skip-until (scanner token-types &key keep-end) |
53 | "This is the implementation of the `skip-until' parser." | |
54 | (do ((consumedp nil t)) | |
55 | ((member (token-type scanner) token-types) | |
56 | (unless keep-end (scanner-step scanner)) | |
57 | (values nil t (or keep-end consumedp))) | |
58 | (when (scanner-at-eof-p scanner) | |
59 | (return (values token-types nil consumedp))) | |
60 | (scanner-step scanner))) | |
61 | ||
62 | (defun parse-error-recover (scanner parser recover) | |
63 | "This is the implementation of the `error' parser." | |
64 | (multiple-value-bind (result win consumedp) (funcall parser) | |
65 | (cond ((or win (and (not consumedp) (scanner-at-eof-p scanner))) | |
66 | ;; If we succeeded then there's nothing for us to do here. On the | |
67 | ;; other hand, if we failed, didn't consume any tokens, and we're | |
68 | ;; at end-of-file, then there's not much hope of making onward | |
69 | ;; progress, so in this case we propagate the failure rather than | |
70 | ;; trying to recover. And we assume that the continuation will | |
71 | ;; somehow arrange to report the problem, and avoid inundating the | |
72 | ;; user with error reports. | |
73 | (values result win consumedp)) | |
74 | (t | |
75 | ;; Now we have to do some kind of sensible error recovery. The | |
76 | ;; important thing to do here is to make sure that we make some | |
77 | ;; progress. If we consumed any tokens then we're fine, and we'll | |
78 | ;; just try the provided recovery strategy. Otherwise, if we're | |
79 | ;; not at EOF, then we can ensure progress by discarding the | |
80 | ;; current token. Finally, if we are at EOF then our best bet is | |
81 | ;; simply to propagate the current failure back to the caller, but | |
82 | ;; we handled that case above. | |
83 | (syntax-error scanner result :continuep t) | |
84 | (unless consumedp (scanner-step scanner)) | |
85 | (funcall recover))))) | |
86 | ||
239fa5bd MW |
87 | ;;;-------------------------------------------------------------------------- |
88 | ;;; Token scanning. | |
89 | ||
90 | (defmethod scanner-token ((scanner sod-token-scanner)) | |
91 | (with-slots (char-scanner line column) scanner | |
92 | (with-parser-context (character-scanner-context :scanner char-scanner) | |
93 | ||
94 | (flet ((scan-digits (&key (radix 10) (min 1) (init 0)) | |
95 | ;; Scan and return a sequence of digits. | |
96 | (parse (many (acc init (+ (* acc radix) it) :min min) | |
97 | (label (list :digit radix) | |
98 | (filter (lambda (ch) | |
99 | (digit-char-p ch radix)))))))) | |
100 | ||
101 | ;; Skip initial junk, and remember the place. | |
102 | (loop | |
103 | (setf (scanner-line scanner) (scanner-line char-scanner) | |
104 | (scanner-column scanner) (scanner-column char-scanner)) | |
105 | (cond-parse (:consumedp cp :expected exp) | |
106 | ((satisfies whitespace-char-p) (parse :whitespace)) | |
107 | ((scan-comment char-scanner)) | |
108 | (t (if cp (lexer-error char-scanner exp cp) (return))))) | |
109 | ||
110 | ;; Now parse something. | |
111 | (cond-parse (:consumedp cp :expected exp) | |
112 | ||
113 | ;; Alphanumerics mean we read an identifier. | |
114 | ((or #\_ (satisfies alpha-char-p)) | |
115 | (values :id (with-output-to-string (out) | |
116 | (write-char it out) | |
117 | (parse (many (nil nil (write-char it out)) | |
118 | (or #\_ (satisfies alphanumericp))))))) | |
119 | ||
120 | ;; Quotes introduce a literal. | |
121 | ((seq ((quote (or #\" #\')) | |
122 | (contents (many (out (make-string-output-stream) | |
123 | (progn (write-char it out) out) | |
124 | :final (get-output-stream-string out)) | |
125 | (or (and #\\ :any) (not quote)))) | |
126 | (nil (char quote))) | |
127 | (ecase quote | |
128 | (#\" contents) | |
129 | (#\' (case (length contents) | |
130 | (1 (char contents 0)) | |
131 | (0 (cerror* "Empty character literal") #\?) | |
132 | (t (cerror* "Too many characters in literal") | |
133 | (char contents 0)))))) | |
134 | (values (etypecase it | |
135 | (character :char) | |
136 | (string :string)) | |
137 | it)) | |
138 | ||
139 | ;; Zero introduces a chosen-radix integer. | |
140 | ((and #\0 | |
141 | (or (and (or #\b #\B) (scan-digits :radix 2)) | |
142 | (and (or #\o #\O) (scan-digits :radix 8)) | |
143 | (and (or #\x #\X) (scan-digits :radix 16)) | |
144 | (scan-digits :radix 8 :min 0))) | |
145 | (values :int it)) | |
146 | ||
147 | ;; Any other digit forces radix-10. | |
148 | ((seq ((d (filter digit-char-p)) | |
149 | (i (scan-digits :radix 10 :min 0 :init d))) | |
150 | i) | |
151 | (values :int it)) | |
152 | ||
153 | ;; Some special punctuation sequences are single tokens. | |
154 | ("..." (values :ellipsis nil)) | |
155 | ||
156 | ;; Any other character is punctuation. | |
157 | (:any (values it nil)) | |
158 | ||
159 | ;; End of file means precisely that. | |
160 | (:eof (values :eof nil)) | |
161 | ||
162 | ;; Report errors and try again. Because we must have consumed some | |
163 | ;; input in order to get here (we've matched both :any and :eof) we | |
164 | ;; must make progress on every call. | |
165 | (t | |
166 | (assert cp) | |
167 | (lexer-error char-scanner exp cp) | |
168 | (scanner-token scanner))))))) | |
dea4d055 MW |
169 | |
170 | ;;;----- That's all, folks -------------------------------------------------- |