Mercurial > repo

comparison perl-5.22.2/dquote_static.c @ 8045:a16537d2fe07

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
<xfix> tar xf perl-5.22.2.tar.gz # Ah, whatever, I\'m doing it anyway
author HackBot
date Sat, 14 May 2016 14:54:38 +0000
parents
children
comparison
equal deleted inserted replaced
8044:711c038a7dce 8045:a16537d2fe07
1 /* dquote_static.c
2 *
3 * This file contains static functions that are related to
4 * parsing double-quotish expressions, but are used in more than
5 * one file.
6 *
7 * It is currently #included by regcomp.c and toke.c.
8 */
9
10 #define PERL_IN_DQUOTE_STATIC_C
11 #include "embed.h"
12
13 /*
14 - regcurly - a little FSA that accepts {\d+,?\d*}
15 Pulled from regcomp.c.
16 */
17 PERL_STATIC_INLINE I32
18 S_regcurly(const char *s)
19 {
20 PERL_ARGS_ASSERT_REGCURLY;
21
22 if (*s++ != '{')
23 return FALSE;
24 if (!isDIGIT(*s))
25 return FALSE;
26 while (isDIGIT(*s))
27 s++;
28 if (*s == ',') {
29 s++;
30 while (isDIGIT(*s))
31 s++;
32 }
33
34 return *s == '}';
35 }
36
37 /* XXX Add documentation after final interface and behavior is decided */
38 /* May want to show context for error, so would pass Perl_bslash_c(pTHX_ const char* current, const char* start, const bool output_warning)
39 U8 source = *current;
40 */
41
42 STATIC char
43 S_grok_bslash_c(pTHX_ const char source, const bool output_warning)
44 {
45
46 U8 result;
47
48 if (! isPRINT_A(source)) {
49 Perl_croak(aTHX_ "%s",
50 "Character following \"\\c\" must be printable ASCII");
51 }
52 else if (source == '{') {
53 const char control = toCTRL('{');
54 if (isPRINT_A(control)) {
55 /* diag_listed_as: Use "%s" instead of "%s" */
56 Perl_croak(aTHX_ "Use \"%c\" instead of \"\\c{\"", control);
57 }
58 else {
59 Perl_croak(aTHX_ "Sequence \"\\c{\" invalid");
60 }
61 }
62
63 result = toCTRL(source);
64 if (output_warning && isPRINT_A(result)) {
65 U8 clearer[3];
66 U8 i = 0;
67 if (! isWORDCHAR(result)) {
68 clearer[i++] = '\\';
69 }
70 clearer[i++] = result;
71 clearer[i++] = '\0';
72
73 Perl_ck_warner(aTHX_ packWARN(WARN_SYNTAX),
74 "\"\\c%c\" is more clearly written simply as \"%s\"",
75 source,
76 clearer);
77 }
78
79 return result;
80 }
81
82 STATIC bool
83 S_grok_bslash_o(pTHX_ char **s, UV *uv, const char** error_msg,
84 const bool output_warning, const bool strict,
85 const bool silence_non_portable,
86 const bool UTF)
87 {
88
89 /* Documentation to be supplied when interface nailed down finally
90 * This returns FALSE if there is an error which the caller need not recover
91 * from; otherwise TRUE. In either case the caller should look at *len [???].
92 * It guarantees that the returned codepoint, *uv, when expressed as
93 * utf8 bytes, would fit within the skipped "\o{...}" bytes.
94 * On input:
95 * s is the address of a pointer to a NULL terminated string that begins
96 * with 'o', and the previous character was a backslash. At exit, *s
97 * will be advanced to the byte just after those absorbed by this
98 * function. Hence the caller can continue parsing from there. In
99 * the case of an error, this routine has generally positioned *s to
100 * point just to the right of the first bad spot, so that a message
101 * that has a "<--" to mark the spot will be correctly positioned.
102 * uv points to a UV that will hold the output value, valid only if the
103 * return from the function is TRUE
104 * error_msg is a pointer that will be set to an internal buffer giving an
105 * error message upon failure (the return is FALSE). Untouched if
106 * function succeeds
107 * output_warning says whether to output any warning messages, or suppress
108 * them
109 * strict is true if this should fail instead of warn if there are
110 * non-octal digits within the braces
111 * silence_non_portable is true if to suppress warnings about the code
112 * point returned being too large to fit on all platforms.
113 * UTF is true iff the string *s is encoded in UTF-8.
114 */
115 char* e;
116 STRLEN numbers_len;
117 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
118 | PERL_SCAN_DISALLOW_PREFIX
119 /* XXX Until the message is improved in grok_oct, handle errors
120 * ourselves */
121 | PERL_SCAN_SILENT_ILLDIGIT;
122
123 #ifdef DEBUGGING
124 char *start = *s - 1;
125 assert(*start == '\\');
126 #endif
127
128 PERL_ARGS_ASSERT_GROK_BSLASH_O;
129
130
131 assert(**s == 'o');
132 (*s)++;
133
134 if (**s != '{') {
135 *error_msg = "Missing braces on \\o{}";
136 return FALSE;
137 }
138
139 e = strchr(*s, '}');
140 if (!e) {
141 (*s)++; /* Move past the '{' */
142 while (isOCTAL(**s)) { /* Position beyond the legal digits */
143 (*s)++;
144 }
145 *error_msg = "Missing right brace on \\o{";
146 return FALSE;
147 }
148
149 (*s)++; /* Point to expected first digit (could be first byte of utf8
150 sequence if not a digit) */
151 numbers_len = e - *s;
152 if (numbers_len == 0) {
153 (*s)++; /* Move past the } */
154 *error_msg = "Number with no digits";
155 return FALSE;
156 }
157
158 if (silence_non_portable) {
159 flags |= PERL_SCAN_SILENT_NON_PORTABLE;
160 }
161
162 *uv = grok_oct(*s, &numbers_len, &flags, NULL);
163 /* Note that if has non-octal, will ignore everything starting with that up
164 * to the '}' */
165
166 if (numbers_len != (STRLEN) (e - *s)) {
167 if (strict) {
168 *s += numbers_len;
169 *s += (UTF) ? UTF8SKIP(*s) : (STRLEN) 1;
170 *error_msg = "Non-octal character";
171 return FALSE;
172 }
173 else if (output_warning) {
174 Perl_ck_warner(aTHX_ packWARN(WARN_DIGIT),
175 /* diag_listed_as: Non-octal character '%c'. Resolved as "%s" */
176 "Non-octal character '%c'. Resolved as \"\\o{%.*s}\"",
177 *(*s + numbers_len),
178 (int) numbers_len,
179 *s);
180 }
181 }
182
183 /* Return past the '}' */
184 *s = e + 1;
185
186 /* guarantee replacing "\o{...}" with utf8 bytes fits within
187 * existing space */
188 assert(OFFUNISKIP(*uv) < *s - start);
189
190 return TRUE;
191 }
192
193 PERL_STATIC_INLINE bool
194 S_grok_bslash_x(pTHX_ char **s, UV *uv, const char** error_msg,
195 const bool output_warning, const bool strict,
196 const bool silence_non_portable,
197 const bool UTF)
198 {
199
200 /* Documentation to be supplied when interface nailed down finally
201 * This returns FALSE if there is an error which the caller need not recover
202 * from; otherwise TRUE.
203 * It guarantees that the returned codepoint, *uv, when expressed as
204 * utf8 bytes, would fit within the skipped "\x{...}" bytes.
205 *
206 * On input:
207 * s is the address of a pointer to a NULL terminated string that begins
208 * with 'x', and the previous character was a backslash. At exit, *s
209 * will be advanced to the byte just after those absorbed by this
210 * function. Hence the caller can continue parsing from there. In
211 * the case of an error, this routine has generally positioned *s to
212 * point just to the right of the first bad spot, so that a message
213 * that has a "<--" to mark the spot will be correctly positioned.
214 * uv points to a UV that will hold the output value, valid only if the
215 * return from the function is TRUE
216 * error_msg is a pointer that will be set to an internal buffer giving an
217 * error message upon failure (the return is FALSE). Untouched if
218 * function succeeds
219 * output_warning says whether to output any warning messages, or suppress
220 * them
221 * strict is true if anything out of the ordinary should cause this to
222 * fail instead of warn or be silent. For example, it requires
223 * exactly 2 digits following the \x (when there are no braces).
224 * 3 digits could be a mistake, so is forbidden in this mode.
225 * silence_non_portable is true if to suppress warnings about the code
226 * point returned being too large to fit on all platforms.
227 * UTF is true iff the string *s is encoded in UTF-8.
228 */
229 char* e;
230 STRLEN numbers_len;
231 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
232 #ifdef DEBUGGING
233 char *start = *s - 1;
234 assert(*start == '\\');
235 #endif
236
237 PERL_ARGS_ASSERT_GROK_BSLASH_X;
238
239 assert(**s == 'x');
240 (*s)++;
241
242 if (strict || ! output_warning) {
243 flags |= PERL_SCAN_SILENT_ILLDIGIT;
244 }
245
246 if (**s != '{') {
247 STRLEN len = (strict) ? 3 : 2;
248
249 *uv = grok_hex(*s, &len, &flags, NULL);
250 *s += len;
251 if (strict && len != 2) {
252 if (len < 2) {
253 *s += (UTF) ? UTF8SKIP(*s) : 1;
254 *error_msg = "Non-hex character";
255 }
256 else {
257 *error_msg = "Use \\x{...} for more than two hex characters";
258 }
259 return FALSE;
260 }
261 goto ok;
262 }
263
264 e = strchr(*s, '}');
265 if (!e) {
266 (*s)++; /* Move past the '{' */
267 while (isXDIGIT(**s)) { /* Position beyond the legal digits */
268 (*s)++;
269 }
270 /* XXX The corresponding message above for \o is just '\\o{'; other
271 * messages for other constructs include the '}', so are inconsistent.
272 */
273 *error_msg = "Missing right brace on \\x{}";
274 return FALSE;
275 }
276
277 (*s)++; /* Point to expected first digit (could be first byte of utf8
278 sequence if not a digit) */
279 numbers_len = e - *s;
280 if (numbers_len == 0) {
281 if (strict) {
282 (*s)++; /* Move past the } */
283 *error_msg = "Number with no digits";
284 return FALSE;
285 }
286 *s = e + 1;
287 *uv = 0;
288 goto ok;
289 }
290
291 flags |= PERL_SCAN_ALLOW_UNDERSCORES;
292 if (silence_non_portable) {
293 flags |= PERL_SCAN_SILENT_NON_PORTABLE;
294 }
295
296 *uv = grok_hex(*s, &numbers_len, &flags, NULL);
297 /* Note that if has non-hex, will ignore everything starting with that up
298 * to the '}' */
299
300 if (strict && numbers_len != (STRLEN) (e - *s)) {
301 *s += numbers_len;
302 *s += (UTF) ? UTF8SKIP(*s) : 1;
303 *error_msg = "Non-hex character";
304 return FALSE;
305 }
306
307 /* Return past the '}' */
308 *s = e + 1;
309
310 ok:
311 /* guarantee replacing "\x{...}" with utf8 bytes fits within
312 * existing space */
313 assert(OFFUNISKIP(*uv) < *s - start);
314 return TRUE;
315 }
316
317 STATIC char*
318 S_form_short_octal_warning(pTHX_
319 const char * const s, /* Points to first non-octal */
320 const STRLEN len /* Length of octals string, so
321 (s-len) points to first
322 octal */
323 ) {
324 /* Return a character string consisting of a warning message for when a
325 * string constant in octal is weird, like "\078". */
326
327 const char * sans_leading_zeros = s - len;
328
329 PERL_ARGS_ASSERT_FORM_SHORT_OCTAL_WARNING;
330
331 assert(*s == '8' || *s == '9');
332
333 /* Remove the leading zeros, retaining one zero so won't be zero length */
334 while (*sans_leading_zeros == '0') sans_leading_zeros++;
335 if (sans_leading_zeros == s) {
336 sans_leading_zeros--;
337 }
338
339 return Perl_form(aTHX_
340 "'%.*s' resolved to '\\o{%.*s}%c'",
341 (int) (len + 2), s - len - 1,
342 (int) (s - sans_leading_zeros), sans_leading_zeros,
343 *s);
344 }
345
346 /*
347 * ex: set ts=8 sts=4 sw=4 et:
348 */