Commit | Line | Data |
---|---|---|
a55c5245 | 1 | /* dquote.c |
04e98a4d | 2 | * |
a55c5245 JH |
3 | * This file contains functions that are related to |
4 | * parsing double-quotish expressions. | |
04e98a4d | 5 | * |
04e98a4d AD |
6 | */ |
7 | ||
a55c5245 JH |
8 | #include "EXTERN.h" |
9 | #define PERL_IN_DQUOTE_C | |
10 | #include "perl.h" | |
881ffab6 | 11 | |
68b355dd | 12 | /* XXX Add documentation after final interface and behavior is decided */ |
f7e03a10 | 13 | /* May want to show context for error, so would pass S_grok_bslash_c(pTHX_ const char* current, const char* start, const bool output_warning) |
68b355dd | 14 | U8 source = *current; |
68b355dd KW |
15 | */ |
16 | ||
a55c5245 JH |
17 | char |
18 | Perl_grok_bslash_c(pTHX_ const char source, const bool output_warning) | |
68b355dd KW |
19 | { |
20 | ||
21 | U8 result; | |
22 | ||
421e43ba | 23 | if (! isPRINT_A(source)) { |
7357bd17 KW |
24 | Perl_croak(aTHX_ "%s", |
25 | "Character following \"\\c\" must be printable ASCII"); | |
68b355dd | 26 | } |
421e43ba | 27 | else if (source == '{') { |
a27ed980 KW |
28 | const char control = toCTRL('{'); |
29 | if (isPRINT_A(control)) { | |
30 | /* diag_listed_as: Use "%s" instead of "%s" */ | |
31 | Perl_croak(aTHX_ "Use \"%c\" instead of \"\\c{\"", control); | |
32 | } | |
33 | else { | |
34 | Perl_croak(aTHX_ "Sequence \"\\c{\" invalid"); | |
35 | } | |
421e43ba | 36 | } |
68b355dd KW |
37 | |
38 | result = toCTRL(source); | |
5e784d58 | 39 | if (output_warning && isPRINT_A(result)) { |
4d8be631 KW |
40 | U8 clearer[3]; |
41 | U8 i = 0; | |
42 | if (! isWORDCHAR(result)) { | |
43 | clearer[i++] = '\\'; | |
44 | } | |
45 | clearer[i++] = result; | |
46 | clearer[i++] = '\0'; | |
68b355dd | 47 | |
4d8be631 KW |
48 | Perl_ck_warner(aTHX_ packWARN(WARN_SYNTAX), |
49 | "\"\\c%c\" is more clearly written simply as \"%s\"", | |
50 | source, | |
51 | clearer); | |
68b355dd KW |
52 | } |
53 | ||
54 | return result; | |
55 | } | |
56 | ||
a55c5245 JH |
57 | bool |
58 | Perl_grok_bslash_o(pTHX_ char **s, UV *uv, const char** error_msg, | |
80f4111b | 59 | const bool output_warning, const bool strict, |
17896a85 | 60 | const bool silence_non_portable, |
80f4111b | 61 | const bool UTF) |
db30362b KW |
62 | { |
63 | ||
64 | /* Documentation to be supplied when interface nailed down finally | |
65 | * This returns FALSE if there is an error which the caller need not recover | |
61caa4da DM |
66 | * from; otherwise TRUE. In either case the caller should look at *len [???]. |
67 | * It guarantees that the returned codepoint, *uv, when expressed as | |
7a4ca5b4 | 68 | * utf8 bytes, would fit within the skipped "\o{...}" bytes. |
db30362b | 69 | * On input: |
00ce5563 KW |
70 | * s is the address of a pointer to a NULL terminated string that begins |
71 | * with 'o', and the previous character was a backslash. At exit, *s | |
72 | * will be advanced to the byte just after those absorbed by this | |
73 | * function. Hence the caller can continue parsing from there. In | |
74 | * the case of an error, this routine has generally positioned *s to | |
75 | * point just to the right of the first bad spot, so that a message | |
76 | * that has a "<--" to mark the spot will be correctly positioned. | |
db30362b KW |
77 | * uv points to a UV that will hold the output value, valid only if the |
78 | * return from the function is TRUE | |
db30362b KW |
79 | * error_msg is a pointer that will be set to an internal buffer giving an |
80 | * error message upon failure (the return is FALSE). Untouched if | |
81 | * function succeeds | |
82 | * output_warning says whether to output any warning messages, or suppress | |
83 | * them | |
80f4111b KW |
84 | * strict is true if this should fail instead of warn if there are |
85 | * non-octal digits within the braces | |
17896a85 KW |
86 | * silence_non_portable is true if to suppress warnings about the code |
87 | * point returned being too large to fit on all platforms. | |
80f4111b | 88 | * UTF is true iff the string *s is encoded in UTF-8. |
db30362b | 89 | */ |
00ce5563 | 90 | char* e; |
db30362b KW |
91 | STRLEN numbers_len; |
92 | I32 flags = PERL_SCAN_ALLOW_UNDERSCORES | |
93 | | PERL_SCAN_DISALLOW_PREFIX | |
94 | /* XXX Until the message is improved in grok_oct, handle errors | |
95 | * ourselves */ | |
96 | | PERL_SCAN_SILENT_ILLDIGIT; | |
97 | ||
61caa4da DM |
98 | #ifdef DEBUGGING |
99 | char *start = *s - 1; | |
100 | assert(*start == '\\'); | |
101 | #endif | |
102 | ||
db30362b KW |
103 | PERL_ARGS_ASSERT_GROK_BSLASH_O; |
104 | ||
105 | ||
00ce5563 KW |
106 | assert(**s == 'o'); |
107 | (*s)++; | |
db30362b | 108 | |
00ce5563 | 109 | if (**s != '{') { |
db30362b KW |
110 | *error_msg = "Missing braces on \\o{}"; |
111 | return FALSE; | |
112 | } | |
113 | ||
00ce5563 | 114 | e = strchr(*s, '}'); |
db30362b | 115 | if (!e) { |
00ce5563 | 116 | (*s)++; /* Move past the '{' */ |
b8de99ca KW |
117 | while (isOCTAL(**s)) { /* Position beyond the legal digits */ |
118 | (*s)++; | |
119 | } | |
00ce5563 | 120 | *error_msg = "Missing right brace on \\o{"; |
db30362b KW |
121 | return FALSE; |
122 | } | |
123 | ||
00ce5563 KW |
124 | (*s)++; /* Point to expected first digit (could be first byte of utf8 |
125 | sequence if not a digit) */ | |
126 | numbers_len = e - *s; | |
db30362b | 127 | if (numbers_len == 0) { |
00ce5563 | 128 | (*s)++; /* Move past the } */ |
db30362b KW |
129 | *error_msg = "Number with no digits"; |
130 | return FALSE; | |
131 | } | |
132 | ||
17896a85 KW |
133 | if (silence_non_portable) { |
134 | flags |= PERL_SCAN_SILENT_NON_PORTABLE; | |
135 | } | |
136 | ||
00ce5563 | 137 | *uv = grok_oct(*s, &numbers_len, &flags, NULL); |
db30362b KW |
138 | /* Note that if has non-octal, will ignore everything starting with that up |
139 | * to the '}' */ | |
140 | ||
80f4111b KW |
141 | if (numbers_len != (STRLEN) (e - *s)) { |
142 | if (strict) { | |
143 | *s += numbers_len; | |
144 | *s += (UTF) ? UTF8SKIP(*s) : (STRLEN) 1; | |
145 | *error_msg = "Non-octal character"; | |
146 | return FALSE; | |
147 | } | |
148 | else if (output_warning) { | |
b67d718a KW |
149 | Perl_ck_warner(aTHX_ packWARN(WARN_DIGIT), |
150 | /* diag_listed_as: Non-octal character '%c'. Resolved as "%s" */ | |
151 | "Non-octal character '%c'. Resolved as \"\\o{%.*s}\"", | |
152 | *(*s + numbers_len), | |
153 | (int) numbers_len, | |
154 | *s); | |
80f4111b | 155 | } |
db30362b KW |
156 | } |
157 | ||
00ce5563 KW |
158 | /* Return past the '}' */ |
159 | *s = e + 1; | |
160 | ||
db30362b KW |
161 | return TRUE; |
162 | } | |
163 | ||
a55c5245 JH |
164 | char* |
165 | Perl_form_short_octal_warning(pTHX_ | |
5e0a247b KW |
166 | const char * const s, /* Points to first non-octal */ |
167 | const STRLEN len /* Length of octals string, so | |
168 | (s-len) points to first | |
169 | octal */ | |
170 | ) { | |
171 | /* Return a character string consisting of a warning message for when a | |
172 | * string constant in octal is weird, like "\078". */ | |
173 | ||
174 | const char * sans_leading_zeros = s - len; | |
175 | ||
176 | PERL_ARGS_ASSERT_FORM_SHORT_OCTAL_WARNING; | |
177 | ||
178 | assert(*s == '8' || *s == '9'); | |
179 | ||
180 | /* Remove the leading zeros, retaining one zero so won't be zero length */ | |
181 | while (*sans_leading_zeros == '0') sans_leading_zeros++; | |
182 | if (sans_leading_zeros == s) { | |
183 | sans_leading_zeros--; | |
184 | } | |
185 | ||
186 | return Perl_form(aTHX_ | |
187 | "'%.*s' resolved to '\\o{%.*s}%c'", | |
188 | (int) (len + 2), s - len - 1, | |
189 | (int) (s - sans_leading_zeros), sans_leading_zeros, | |
190 | *s); | |
191 | } | |
192 | ||
04e98a4d | 193 | /* |
14d04a33 | 194 | * ex: set ts=8 sts=4 sw=4 et: |
04e98a4d | 195 | */ |