yarp 0.12.0 → 0.13.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/CHANGELOG.md +29 -8
- data/CONTRIBUTING.md +2 -2
- data/Makefile +5 -5
- data/README.md +11 -12
- data/config.yml +6 -2
- data/docs/build_system.md +21 -21
- data/docs/building.md +4 -4
- data/docs/configuration.md +25 -21
- data/docs/design.md +2 -2
- data/docs/encoding.md +17 -17
- data/docs/fuzzing.md +4 -4
- data/docs/heredocs.md +3 -3
- data/docs/mapping.md +94 -94
- data/docs/ripper.md +4 -4
- data/docs/ruby_api.md +11 -11
- data/docs/serialization.md +17 -16
- data/docs/testing.md +6 -6
- data/ext/prism/api_node.c +4725 -0
- data/ext/{yarp → prism}/api_pack.c +82 -82
- data/ext/{yarp → prism}/extconf.rb +13 -13
- data/ext/{yarp → prism}/extension.c +175 -168
- data/ext/prism/extension.h +18 -0
- data/include/prism/ast.h +1932 -0
- data/include/prism/defines.h +45 -0
- data/include/prism/diagnostic.h +231 -0
- data/include/{yarp/enc/yp_encoding.h → prism/enc/pm_encoding.h} +40 -40
- data/include/prism/node.h +41 -0
- data/include/prism/pack.h +141 -0
- data/include/{yarp → prism}/parser.h +143 -142
- data/include/prism/regexp.h +19 -0
- data/include/prism/unescape.h +48 -0
- data/include/prism/util/pm_buffer.h +51 -0
- data/include/{yarp/util/yp_char.h → prism/util/pm_char.h} +20 -20
- data/include/{yarp/util/yp_constant_pool.h → prism/util/pm_constant_pool.h} +26 -22
- data/include/{yarp/util/yp_list.h → prism/util/pm_list.h} +21 -21
- data/include/prism/util/pm_memchr.h +14 -0
- data/include/{yarp/util/yp_newline_list.h → prism/util/pm_newline_list.h} +11 -11
- data/include/prism/util/pm_state_stack.h +24 -0
- data/include/{yarp/util/yp_string.h → prism/util/pm_string.h} +20 -20
- data/include/prism/util/pm_string_list.h +25 -0
- data/include/{yarp/util/yp_strpbrk.h → prism/util/pm_strpbrk.h} +7 -7
- data/include/prism/version.h +4 -0
- data/include/prism.h +82 -0
- data/lib/prism/compiler.rb +465 -0
- data/lib/prism/debug.rb +157 -0
- data/lib/{yarp/desugar_visitor.rb → prism/desugar_compiler.rb} +4 -2
- data/lib/prism/dispatcher.rb +2051 -0
- data/lib/prism/dsl.rb +750 -0
- data/lib/{yarp → prism}/ffi.rb +66 -67
- data/lib/{yarp → prism}/lex_compat.rb +40 -43
- data/lib/{yarp/mutation_visitor.rb → prism/mutation_compiler.rb} +3 -3
- data/lib/{yarp → prism}/node.rb +2012 -2593
- data/lib/prism/node_ext.rb +55 -0
- data/lib/prism/node_inspector.rb +68 -0
- data/lib/{yarp → prism}/pack.rb +1 -1
- data/lib/{yarp → prism}/parse_result/comments.rb +1 -1
- data/lib/{yarp → prism}/parse_result/newlines.rb +1 -1
- data/lib/prism/parse_result.rb +266 -0
- data/lib/{yarp → prism}/pattern.rb +14 -14
- data/lib/{yarp → prism}/ripper_compat.rb +5 -5
- data/lib/{yarp → prism}/serialize.rb +12 -7
- data/lib/prism/visitor.rb +470 -0
- data/lib/prism.rb +64 -0
- data/lib/yarp.rb +2 -614
- data/src/diagnostic.c +213 -208
- data/src/enc/pm_big5.c +52 -0
- data/src/enc/pm_euc_jp.c +58 -0
- data/src/enc/{yp_gbk.c → pm_gbk.c} +16 -16
- data/src/enc/pm_shift_jis.c +56 -0
- data/src/enc/{yp_tables.c → pm_tables.c} +69 -69
- data/src/enc/{yp_unicode.c → pm_unicode.c} +40 -40
- data/src/enc/pm_windows_31j.c +56 -0
- data/src/node.c +1293 -1233
- data/src/pack.c +247 -247
- data/src/prettyprint.c +1479 -1479
- data/src/{yarp.c → prism.c} +5205 -5083
- data/src/regexp.c +132 -132
- data/src/serialize.c +1121 -1121
- data/src/token_type.c +169 -167
- data/src/unescape.c +106 -87
- data/src/util/pm_buffer.c +103 -0
- data/src/util/{yp_char.c → pm_char.c} +72 -72
- data/src/util/{yp_constant_pool.c → pm_constant_pool.c} +85 -64
- data/src/util/{yp_list.c → pm_list.c} +10 -10
- data/src/util/{yp_memchr.c → pm_memchr.c} +6 -4
- data/src/util/{yp_newline_list.c → pm_newline_list.c} +21 -21
- data/src/util/{yp_state_stack.c → pm_state_stack.c} +4 -4
- data/src/util/{yp_string.c → pm_string.c} +38 -38
- data/src/util/pm_string_list.c +29 -0
- data/src/util/{yp_strncasecmp.c → pm_strncasecmp.c} +1 -1
- data/src/util/{yp_strpbrk.c → pm_strpbrk.c} +8 -8
- data/yarp.gemspec +68 -59
- metadata +70 -61
- data/ext/yarp/api_node.c +0 -4728
- data/ext/yarp/extension.h +0 -18
- data/include/yarp/ast.h +0 -1929
- data/include/yarp/defines.h +0 -45
- data/include/yarp/diagnostic.h +0 -226
- data/include/yarp/node.h +0 -42
- data/include/yarp/pack.h +0 -141
- data/include/yarp/regexp.h +0 -19
- data/include/yarp/unescape.h +0 -44
- data/include/yarp/util/yp_buffer.h +0 -51
- data/include/yarp/util/yp_memchr.h +0 -14
- data/include/yarp/util/yp_state_stack.h +0 -24
- data/include/yarp/util/yp_string_list.h +0 -25
- data/include/yarp/version.h +0 -4
- data/include/yarp.h +0 -82
- data/src/enc/yp_big5.c +0 -52
- data/src/enc/yp_euc_jp.c +0 -58
- data/src/enc/yp_shift_jis.c +0 -56
- data/src/enc/yp_windows_31j.c +0 -56
- data/src/util/yp_buffer.c +0 -101
- data/src/util/yp_string_list.c +0 -29
data/src/unescape.c
CHANGED
@@ -1,13 +1,13 @@
|
|
1
|
-
#include "
|
1
|
+
#include "prism.h"
|
2
2
|
|
3
3
|
/******************************************************************************/
|
4
4
|
/* Character checks */
|
5
5
|
/******************************************************************************/
|
6
6
|
|
7
7
|
static inline bool
|
8
|
-
|
8
|
+
pm_char_is_hexadecimal_digits(const uint8_t *string, size_t length) {
|
9
9
|
for (size_t index = 0; index < length; index++) {
|
10
|
-
if (!
|
10
|
+
if (!pm_char_is_hexadecimal_digit(string[index])) {
|
11
11
|
return false;
|
12
12
|
}
|
13
13
|
}
|
@@ -18,7 +18,7 @@ yp_char_is_hexadecimal_digits(const uint8_t *string, size_t length) {
|
|
18
18
|
// expensive to go through the indirection of the function pointer. Instead we
|
19
19
|
// provide a fast path that will check if we can just return 1.
|
20
20
|
static inline size_t
|
21
|
-
|
21
|
+
pm_char_width(pm_parser_t *parser, const uint8_t *start, const uint8_t *end) {
|
22
22
|
if (parser->encoding_changed || (*start >= 0x80)) {
|
23
23
|
return parser->encoding.char_width(start, end - start);
|
24
24
|
} else {
|
@@ -71,11 +71,11 @@ char_is_ascii_printable(const uint8_t b) {
|
|
71
71
|
static inline size_t
|
72
72
|
unescape_octal(const uint8_t *backslash, uint8_t *value, const uint8_t *end) {
|
73
73
|
*value = (uint8_t) (backslash[1] - '0');
|
74
|
-
if (backslash + 2 >= end || !
|
74
|
+
if (backslash + 2 >= end || !pm_char_is_octal_digit(backslash[2])) {
|
75
75
|
return 2;
|
76
76
|
}
|
77
77
|
*value = (uint8_t) ((*value << 3) | (backslash[2] - '0'));
|
78
|
-
if (backslash + 3 >= end || !
|
78
|
+
if (backslash + 3 >= end || !pm_char_is_octal_digit(backslash[3])) {
|
79
79
|
return 3;
|
80
80
|
}
|
81
81
|
*value = (uint8_t) ((*value << 3) | (backslash[3] - '0'));
|
@@ -91,14 +91,14 @@ unescape_hexadecimal_digit(const uint8_t value) {
|
|
91
91
|
// Scan the 1-2 digits of hexadecimal into the value. Returns the number of
|
92
92
|
// digits scanned.
|
93
93
|
static inline size_t
|
94
|
-
unescape_hexadecimal(const uint8_t *backslash, uint8_t *value, const uint8_t *end,
|
94
|
+
unescape_hexadecimal(const uint8_t *backslash, uint8_t *value, const uint8_t *end, pm_list_t *error_list) {
|
95
95
|
*value = 0;
|
96
|
-
if (backslash + 2 >= end || !
|
97
|
-
if (error_list)
|
96
|
+
if (backslash + 2 >= end || !pm_char_is_hexadecimal_digit(backslash[2])) {
|
97
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 2, PM_ERR_ESCAPE_INVALID_HEXADECIMAL);
|
98
98
|
return 2;
|
99
99
|
}
|
100
100
|
*value = unescape_hexadecimal_digit(backslash[2]);
|
101
|
-
if (backslash + 3 >= end || !
|
101
|
+
if (backslash + 3 >= end || !pm_char_is_hexadecimal_digit(backslash[3])) {
|
102
102
|
return 3;
|
103
103
|
}
|
104
104
|
*value = (uint8_t) ((*value << 4) | unescape_hexadecimal_digit(backslash[3]));
|
@@ -121,7 +121,7 @@ unescape_unicode(const uint8_t *string, size_t length, uint32_t *value) {
|
|
121
121
|
// 32-bit value to write. Writes the UTF-8 representation of the value to the
|
122
122
|
// string and returns the number of bytes written.
|
123
123
|
static inline size_t
|
124
|
-
unescape_unicode_write(uint8_t *dest, uint32_t value, const uint8_t *start, const uint8_t *end,
|
124
|
+
unescape_unicode_write(uint8_t *dest, uint32_t value, const uint8_t *start, const uint8_t *end, pm_list_t *error_list) {
|
125
125
|
if (value <= 0x7F) {
|
126
126
|
// 0xxxxxxx
|
127
127
|
dest[0] = (uint8_t) value;
|
@@ -157,7 +157,7 @@ unescape_unicode_write(uint8_t *dest, uint32_t value, const uint8_t *start, cons
|
|
157
157
|
// If we get here, then the value is too big. This is an error, but we don't
|
158
158
|
// want to just crash, so instead we'll add an error to the error list and put
|
159
159
|
// in a replacement character instead.
|
160
|
-
if (error_list)
|
160
|
+
if (error_list) pm_diagnostic_list_append(error_list, start, end, PM_ERR_ESCAPE_INVALID_UNICODE);
|
161
161
|
dest[0] = 0xEF;
|
162
162
|
dest[1] = 0xBF;
|
163
163
|
dest[2] = 0xBD;
|
@@ -165,20 +165,20 @@ unescape_unicode_write(uint8_t *dest, uint32_t value, const uint8_t *start, cons
|
|
165
165
|
}
|
166
166
|
|
167
167
|
typedef enum {
|
168
|
-
|
169
|
-
|
170
|
-
|
171
|
-
|
172
|
-
}
|
168
|
+
PM_UNESCAPE_FLAG_NONE = 0,
|
169
|
+
PM_UNESCAPE_FLAG_CONTROL = 1,
|
170
|
+
PM_UNESCAPE_FLAG_META = 2,
|
171
|
+
PM_UNESCAPE_FLAG_EXPECT_SINGLE = 4
|
172
|
+
} pm_unescape_flag_t;
|
173
173
|
|
174
174
|
// Unescape a single character value based on the given flags.
|
175
175
|
static inline uint8_t
|
176
176
|
unescape_char(uint8_t value, const uint8_t flags) {
|
177
|
-
if (flags &
|
177
|
+
if (flags & PM_UNESCAPE_FLAG_CONTROL) {
|
178
178
|
value &= 0x1f;
|
179
179
|
}
|
180
180
|
|
181
|
-
if (flags &
|
181
|
+
if (flags & PM_UNESCAPE_FLAG_META) {
|
182
182
|
value |= 0x80;
|
183
183
|
}
|
184
184
|
|
@@ -188,13 +188,13 @@ unescape_char(uint8_t value, const uint8_t flags) {
|
|
188
188
|
// Read a specific escape sequence into the given destination.
|
189
189
|
static const uint8_t *
|
190
190
|
unescape(
|
191
|
-
|
191
|
+
pm_parser_t *parser,
|
192
192
|
uint8_t *dest,
|
193
193
|
size_t *dest_length,
|
194
194
|
const uint8_t *backslash,
|
195
195
|
const uint8_t *end,
|
196
196
|
const uint8_t flags,
|
197
|
-
|
197
|
+
pm_list_t *error_list
|
198
198
|
) {
|
199
199
|
switch (backslash[1]) {
|
200
200
|
case 'a':
|
@@ -234,8 +234,8 @@ unescape(
|
|
234
234
|
// \u{nnnn ...} Unicode character(s), where each nnnn is 1-6 hexadecimal digits ([0-9a-fA-F])
|
235
235
|
// \unnnn Unicode character, where nnnn is exactly 4 hexadecimal digits ([0-9a-fA-F])
|
236
236
|
case 'u': {
|
237
|
-
if ((flags &
|
238
|
-
if (error_list)
|
237
|
+
if ((flags & PM_UNESCAPE_FLAG_CONTROL) | (flags & PM_UNESCAPE_FLAG_META)) {
|
238
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 2, PM_ERR_ESCAPE_INVALID_UNICODE_CM_FLAGS);
|
239
239
|
return backslash + 2;
|
240
240
|
}
|
241
241
|
|
@@ -244,26 +244,26 @@ unescape(
|
|
244
244
|
const uint8_t *extra_codepoints_start = NULL;
|
245
245
|
int codepoints_count = 0;
|
246
246
|
|
247
|
-
unicode_cursor +=
|
247
|
+
unicode_cursor += pm_strspn_whitespace(unicode_cursor, end - unicode_cursor);
|
248
248
|
|
249
249
|
while ((unicode_cursor < end) && (*unicode_cursor != '}')) {
|
250
250
|
const uint8_t *unicode_start = unicode_cursor;
|
251
|
-
size_t hexadecimal_length =
|
251
|
+
size_t hexadecimal_length = pm_strspn_hexadecimal_digit(unicode_cursor, end - unicode_cursor);
|
252
252
|
|
253
253
|
// \u{nnnn} character literal allows only 1-6 hexadecimal digits
|
254
254
|
if (hexadecimal_length > 6) {
|
255
|
-
if (error_list)
|
255
|
+
if (error_list) pm_diagnostic_list_append(error_list, unicode_cursor, unicode_cursor + hexadecimal_length, PM_ERR_ESCAPE_INVALID_UNICODE_LONG);
|
256
256
|
}
|
257
257
|
// there are not hexadecimal characters
|
258
258
|
else if (hexadecimal_length == 0) {
|
259
|
-
if (error_list)
|
259
|
+
if (error_list) pm_diagnostic_list_append(error_list, unicode_cursor, unicode_cursor + hexadecimal_length, PM_ERR_ESCAPE_INVALID_UNICODE);
|
260
260
|
return unicode_cursor;
|
261
261
|
}
|
262
262
|
|
263
263
|
unicode_cursor += hexadecimal_length;
|
264
264
|
|
265
265
|
codepoints_count++;
|
266
|
-
if (flags &
|
266
|
+
if (flags & PM_UNESCAPE_FLAG_EXPECT_SINGLE && codepoints_count == 2)
|
267
267
|
extra_codepoints_start = unicode_start;
|
268
268
|
|
269
269
|
uint32_t value;
|
@@ -272,23 +272,23 @@ unescape(
|
|
272
272
|
*dest_length += unescape_unicode_write(dest + *dest_length, value, unicode_start, unicode_cursor, error_list);
|
273
273
|
}
|
274
274
|
|
275
|
-
unicode_cursor +=
|
275
|
+
unicode_cursor += pm_strspn_whitespace(unicode_cursor, end - unicode_cursor);
|
276
276
|
}
|
277
277
|
|
278
278
|
// ?\u{nnnn} character literal should contain only one codepoint and cannot be like ?\u{nnnn mmmm}
|
279
|
-
if (flags &
|
280
|
-
if (error_list)
|
279
|
+
if (flags & PM_UNESCAPE_FLAG_EXPECT_SINGLE && codepoints_count > 1) {
|
280
|
+
if (error_list) pm_diagnostic_list_append(error_list, extra_codepoints_start, unicode_cursor - 1, PM_ERR_ESCAPE_INVALID_UNICODE_LITERAL);
|
281
281
|
}
|
282
282
|
|
283
283
|
if (unicode_cursor < end && *unicode_cursor == '}') {
|
284
284
|
unicode_cursor++;
|
285
285
|
} else {
|
286
|
-
if (error_list)
|
286
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, unicode_cursor, PM_ERR_ESCAPE_INVALID_UNICODE_TERM);
|
287
287
|
}
|
288
288
|
|
289
289
|
return unicode_cursor;
|
290
290
|
}
|
291
|
-
else if ((backslash + 5) < end &&
|
291
|
+
else if ((backslash + 5) < end && pm_char_is_hexadecimal_digits(backslash + 2, 4)) {
|
292
292
|
uint32_t value;
|
293
293
|
unescape_unicode(backslash + 2, 4, &value);
|
294
294
|
|
@@ -298,7 +298,7 @@ unescape(
|
|
298
298
|
return backslash + 6;
|
299
299
|
}
|
300
300
|
|
301
|
-
if (error_list)
|
301
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 2, PM_ERR_ESCAPE_INVALID_UNICODE);
|
302
302
|
return backslash + 2;
|
303
303
|
}
|
304
304
|
// \c\M-x meta control character, where x is an ASCII printable character
|
@@ -306,18 +306,18 @@ unescape(
|
|
306
306
|
// \cx control character, where x is an ASCII printable character
|
307
307
|
case 'c':
|
308
308
|
if (backslash + 2 >= end) {
|
309
|
-
if (error_list)
|
309
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 1, PM_ERR_ESCAPE_INVALID_CONTROL);
|
310
310
|
return end;
|
311
311
|
}
|
312
312
|
|
313
|
-
if (flags &
|
314
|
-
if (error_list)
|
313
|
+
if (flags & PM_UNESCAPE_FLAG_CONTROL) {
|
314
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 1, PM_ERR_ESCAPE_INVALID_CONTROL_REPEAT);
|
315
315
|
return backslash + 2;
|
316
316
|
}
|
317
317
|
|
318
318
|
switch (backslash[2]) {
|
319
319
|
case '\\':
|
320
|
-
return unescape(parser, dest, dest_length, backslash + 2, end, flags |
|
320
|
+
return unescape(parser, dest, dest_length, backslash + 2, end, flags | PM_UNESCAPE_FLAG_CONTROL, error_list);
|
321
321
|
case '?':
|
322
322
|
if (dest) {
|
323
323
|
dest[(*dest_length)++] = unescape_char(0x7f, flags);
|
@@ -325,12 +325,12 @@ unescape(
|
|
325
325
|
return backslash + 3;
|
326
326
|
default: {
|
327
327
|
if (!char_is_ascii_printable(backslash[2])) {
|
328
|
-
if (error_list)
|
328
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 1, PM_ERR_ESCAPE_INVALID_CONTROL);
|
329
329
|
return backslash + 2;
|
330
330
|
}
|
331
331
|
|
332
332
|
if (dest) {
|
333
|
-
dest[(*dest_length)++] = unescape_char(backslash[2], flags |
|
333
|
+
dest[(*dest_length)++] = unescape_char(backslash[2], flags | PM_UNESCAPE_FLAG_CONTROL);
|
334
334
|
}
|
335
335
|
return backslash + 3;
|
336
336
|
}
|
@@ -339,23 +339,23 @@ unescape(
|
|
339
339
|
// \C-? delete, ASCII 7Fh (DEL)
|
340
340
|
case 'C':
|
341
341
|
if (backslash + 3 >= end) {
|
342
|
-
if (error_list)
|
342
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 1, PM_ERR_ESCAPE_INVALID_CONTROL);
|
343
343
|
return end;
|
344
344
|
}
|
345
345
|
|
346
|
-
if (flags &
|
347
|
-
if (error_list)
|
346
|
+
if (flags & PM_UNESCAPE_FLAG_CONTROL) {
|
347
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 1, PM_ERR_ESCAPE_INVALID_CONTROL_REPEAT);
|
348
348
|
return backslash + 2;
|
349
349
|
}
|
350
350
|
|
351
351
|
if (backslash[2] != '-') {
|
352
|
-
if (error_list)
|
352
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 1, PM_ERR_ESCAPE_INVALID_CONTROL);
|
353
353
|
return backslash + 2;
|
354
354
|
}
|
355
355
|
|
356
356
|
switch (backslash[3]) {
|
357
357
|
case '\\':
|
358
|
-
return unescape(parser, dest, dest_length, backslash + 3, end, flags |
|
358
|
+
return unescape(parser, dest, dest_length, backslash + 3, end, flags | PM_UNESCAPE_FLAG_CONTROL, error_list);
|
359
359
|
case '?':
|
360
360
|
if (dest) {
|
361
361
|
dest[(*dest_length)++] = unescape_char(0x7f, flags);
|
@@ -363,12 +363,12 @@ unescape(
|
|
363
363
|
return backslash + 4;
|
364
364
|
default:
|
365
365
|
if (!char_is_ascii_printable(backslash[3])) {
|
366
|
-
if (error_list)
|
366
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 2, PM_ERR_ESCAPE_INVALID_CONTROL);
|
367
367
|
return backslash + 2;
|
368
368
|
}
|
369
369
|
|
370
370
|
if (dest) {
|
371
|
-
dest[(*dest_length)++] = unescape_char(backslash[3], flags |
|
371
|
+
dest[(*dest_length)++] = unescape_char(backslash[3], flags | PM_UNESCAPE_FLAG_CONTROL);
|
372
372
|
}
|
373
373
|
return backslash + 4;
|
374
374
|
}
|
@@ -377,32 +377,32 @@ unescape(
|
|
377
377
|
// \M-x meta character, where x is an ASCII printable character
|
378
378
|
case 'M': {
|
379
379
|
if (backslash + 3 >= end) {
|
380
|
-
if (error_list)
|
380
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 1, PM_ERR_ESCAPE_INVALID_META);
|
381
381
|
return end;
|
382
382
|
}
|
383
383
|
|
384
|
-
if (flags &
|
385
|
-
if (error_list)
|
384
|
+
if (flags & PM_UNESCAPE_FLAG_META) {
|
385
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 2, PM_ERR_ESCAPE_INVALID_META_REPEAT);
|
386
386
|
return backslash + 2;
|
387
387
|
}
|
388
388
|
|
389
389
|
if (backslash[2] != '-') {
|
390
|
-
if (error_list)
|
390
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 2, PM_ERR_ESCAPE_INVALID_META);
|
391
391
|
return backslash + 2;
|
392
392
|
}
|
393
393
|
|
394
394
|
if (backslash[3] == '\\') {
|
395
|
-
return unescape(parser, dest, dest_length, backslash + 3, end, flags |
|
395
|
+
return unescape(parser, dest, dest_length, backslash + 3, end, flags | PM_UNESCAPE_FLAG_META, error_list);
|
396
396
|
}
|
397
397
|
|
398
398
|
if (char_is_ascii_printable(backslash[3])) {
|
399
399
|
if (dest) {
|
400
|
-
dest[(*dest_length)++] = unescape_char(backslash[3], flags |
|
400
|
+
dest[(*dest_length)++] = unescape_char(backslash[3], flags | PM_UNESCAPE_FLAG_META);
|
401
401
|
}
|
402
402
|
return backslash + 4;
|
403
403
|
}
|
404
404
|
|
405
|
-
if (error_list)
|
405
|
+
if (error_list) pm_diagnostic_list_append(error_list, backslash, backslash + 2, PM_ERR_ESCAPE_INVALID_META);
|
406
406
|
return backslash + 3;
|
407
407
|
}
|
408
408
|
// \n
|
@@ -416,7 +416,7 @@ unescape(
|
|
416
416
|
/* fallthrough */
|
417
417
|
// In this case we're escaping something that doesn't need escaping.
|
418
418
|
default: {
|
419
|
-
size_t width =
|
419
|
+
size_t width = pm_char_width(parser, backslash + 1, end);
|
420
420
|
|
421
421
|
if (dest) {
|
422
422
|
memcpy(dest + *dest_length, backslash + 1, width);
|
@@ -457,13 +457,13 @@ unescape(
|
|
457
457
|
// \c? or \C-? delete, ASCII 7Fh (DEL)
|
458
458
|
//
|
459
459
|
static void
|
460
|
-
|
461
|
-
if (unescape_type ==
|
460
|
+
pm_unescape_manipulate_string_or_char_literal(pm_parser_t *parser, pm_string_t *string, pm_unescape_type_t unescape_type, bool expect_single_codepoint) {
|
461
|
+
if (unescape_type == PM_UNESCAPE_NONE) {
|
462
462
|
// If we're not unescaping then we can reference the source directly.
|
463
463
|
return;
|
464
464
|
}
|
465
465
|
|
466
|
-
const uint8_t *backslash =
|
466
|
+
const uint8_t *backslash = pm_memchr(string->source, '\\', string->length, parser->encoding_changed, &parser->encoding);
|
467
467
|
|
468
468
|
if (backslash == NULL) {
|
469
469
|
// Here there are no escapes, so we can reference the source directly.
|
@@ -474,7 +474,7 @@ yp_unescape_manipulate_string_or_char_literal(yp_parser_t *parser, yp_string_t *
|
|
474
474
|
// within the string.
|
475
475
|
uint8_t *allocated = malloc(string->length);
|
476
476
|
if (allocated == NULL) {
|
477
|
-
|
477
|
+
pm_diagnostic_list_append(&parser->error_list, string->source, string->source + string->length, PM_ERR_MALLOC_FAILED);
|
478
478
|
return;
|
479
479
|
}
|
480
480
|
|
@@ -509,7 +509,17 @@ yp_unescape_manipulate_string_or_char_literal(yp_parser_t *parser, yp_string_t *
|
|
509
509
|
cursor = backslash + 2;
|
510
510
|
break;
|
511
511
|
default:
|
512
|
-
if (unescape_type ==
|
512
|
+
if (unescape_type == PM_UNESCAPE_WHITESPACE) {
|
513
|
+
if (backslash[1] == '\r' && backslash[2] == '\n') {
|
514
|
+
cursor = backslash + 2;
|
515
|
+
break;
|
516
|
+
}
|
517
|
+
if (pm_strspn_whitespace(backslash + 1, 1)) {
|
518
|
+
cursor = backslash + 1;
|
519
|
+
break;
|
520
|
+
}
|
521
|
+
}
|
522
|
+
if (unescape_type == PM_UNESCAPE_WHITESPACE || unescape_type == PM_UNESCAPE_MINIMAL) {
|
513
523
|
// In this case we're escaping something that doesn't need escaping.
|
514
524
|
dest[dest_length++] = '\\';
|
515
525
|
cursor = backslash + 1;
|
@@ -518,11 +528,11 @@ yp_unescape_manipulate_string_or_char_literal(yp_parser_t *parser, yp_string_t *
|
|
518
528
|
|
519
529
|
// This is the only type of unescaping left. In this case we need to
|
520
530
|
// handle all of the different unescapes.
|
521
|
-
assert(unescape_type ==
|
531
|
+
assert(unescape_type == PM_UNESCAPE_ALL);
|
522
532
|
|
523
|
-
uint8_t flags =
|
533
|
+
uint8_t flags = PM_UNESCAPE_FLAG_NONE;
|
524
534
|
if (expect_single_codepoint) {
|
525
|
-
flags |=
|
535
|
+
flags |= PM_UNESCAPE_FLAG_EXPECT_SINGLE;
|
526
536
|
}
|
527
537
|
|
528
538
|
cursor = unescape(parser, dest, &dest_length, backslash, end, flags, &parser->error_list);
|
@@ -530,7 +540,7 @@ yp_unescape_manipulate_string_or_char_literal(yp_parser_t *parser, yp_string_t *
|
|
530
540
|
}
|
531
541
|
|
532
542
|
if (end > cursor) {
|
533
|
-
backslash =
|
543
|
+
backslash = pm_memchr(cursor, '\\', (size_t) (end - cursor), parser->encoding_changed, &parser->encoding);
|
534
544
|
} else {
|
535
545
|
backslash = NULL;
|
536
546
|
}
|
@@ -545,30 +555,30 @@ yp_unescape_manipulate_string_or_char_literal(yp_parser_t *parser, yp_string_t *
|
|
545
555
|
|
546
556
|
// If the string was already allocated, then we need to free that memory
|
547
557
|
// here. That's because we're about to override it with the escaped string.
|
548
|
-
|
558
|
+
pm_string_free(string);
|
549
559
|
|
550
560
|
// We also need to update the length at the end. This is because every escape
|
551
561
|
// reduces the length of the final string, and we don't want garbage at the
|
552
562
|
// end.
|
553
|
-
|
563
|
+
pm_string_owned_init(string, allocated, dest_length + ((size_t) (end - cursor)));
|
554
564
|
}
|
555
565
|
|
556
|
-
|
557
|
-
|
558
|
-
|
566
|
+
PRISM_EXPORTED_FUNCTION void
|
567
|
+
pm_unescape_manipulate_string(pm_parser_t *parser, pm_string_t *string, pm_unescape_type_t unescape_type) {
|
568
|
+
pm_unescape_manipulate_string_or_char_literal(parser, string, unescape_type, false);
|
559
569
|
}
|
560
570
|
|
561
571
|
void
|
562
|
-
|
563
|
-
|
572
|
+
pm_unescape_manipulate_char_literal(pm_parser_t *parser, pm_string_t *string, pm_unescape_type_t unescape_type) {
|
573
|
+
pm_unescape_manipulate_string_or_char_literal(parser, string, unescape_type, true);
|
564
574
|
}
|
565
575
|
|
566
|
-
// This function is similar to
|
576
|
+
// This function is similar to pm_unescape_manipulate_string, except it doesn't
|
567
577
|
// actually perform any string manipulations. Instead, it calculates how long
|
568
578
|
// the unescaped character is, and returns that value
|
569
579
|
size_t
|
570
|
-
|
571
|
-
assert(unescape_type !=
|
580
|
+
pm_unescape_calculate_difference(pm_parser_t *parser, const uint8_t *backslash, pm_unescape_type_t unescape_type, bool expect_single_codepoint) {
|
581
|
+
assert(unescape_type != PM_UNESCAPE_NONE);
|
572
582
|
|
573
583
|
if (backslash + 1 >= parser->end) {
|
574
584
|
return 0;
|
@@ -579,17 +589,26 @@ yp_unescape_calculate_difference(yp_parser_t *parser, const uint8_t *backslash,
|
|
579
589
|
case '\'':
|
580
590
|
return 2;
|
581
591
|
default: {
|
582
|
-
if (unescape_type ==
|
583
|
-
|
592
|
+
if (unescape_type == PM_UNESCAPE_WHITESPACE) {
|
593
|
+
if (backslash[1] == '\r' && backslash[2] == '\n') {
|
594
|
+
return 2;
|
595
|
+
}
|
596
|
+
size_t whitespace = pm_strspn_whitespace(backslash + 1, 1);
|
597
|
+
if (whitespace > 0) {
|
598
|
+
return whitespace;
|
599
|
+
}
|
600
|
+
}
|
601
|
+
if (unescape_type == PM_UNESCAPE_WHITESPACE || unescape_type == PM_UNESCAPE_MINIMAL) {
|
602
|
+
return 1 + pm_char_width(parser, backslash + 1, parser->end);
|
584
603
|
}
|
585
604
|
|
586
605
|
// This is the only type of unescaping left. In this case we need to
|
587
606
|
// handle all of the different unescapes.
|
588
|
-
assert(unescape_type ==
|
607
|
+
assert(unescape_type == PM_UNESCAPE_ALL);
|
589
608
|
|
590
|
-
uint8_t flags =
|
609
|
+
uint8_t flags = PM_UNESCAPE_FLAG_NONE;
|
591
610
|
if (expect_single_codepoint) {
|
592
|
-
flags |=
|
611
|
+
flags |= PM_UNESCAPE_FLAG_EXPECT_SINGLE;
|
593
612
|
}
|
594
613
|
|
595
614
|
const uint8_t *cursor = unescape(parser, NULL, 0, backslash, parser->end, flags, NULL);
|
@@ -603,16 +622,16 @@ yp_unescape_calculate_difference(yp_parser_t *parser, const uint8_t *backslash,
|
|
603
622
|
// This is one of the main entry points into the extension. It accepts a source
|
604
623
|
// string, a type of unescaping, and a pointer to a result string. It returns a
|
605
624
|
// boolean indicating whether or not the unescaping was successful.
|
606
|
-
|
607
|
-
|
608
|
-
|
609
|
-
|
625
|
+
PRISM_EXPORTED_FUNCTION bool
|
626
|
+
pm_unescape_string(const uint8_t *start, size_t length, pm_unescape_type_t unescape_type, pm_string_t *result) {
|
627
|
+
pm_parser_t parser;
|
628
|
+
pm_parser_init(&parser, start, length, NULL);
|
610
629
|
|
611
|
-
|
612
|
-
|
630
|
+
pm_string_shared_init(result, start, start + length);
|
631
|
+
pm_unescape_manipulate_string(&parser, result, unescape_type);
|
613
632
|
|
614
|
-
bool success =
|
615
|
-
|
633
|
+
bool success = pm_list_empty_p(&parser.error_list);
|
634
|
+
pm_parser_free(&parser);
|
616
635
|
|
617
636
|
return success;
|
618
637
|
}
|
@@ -0,0 +1,103 @@
|
|
1
|
+
#include "prism/util/pm_buffer.h"
|
2
|
+
|
3
|
+
#define PRISM_BUFFER_INITIAL_SIZE 1024
|
4
|
+
|
5
|
+
// Return the size of the pm_buffer_t struct.
|
6
|
+
size_t
|
7
|
+
pm_buffer_sizeof(void) {
|
8
|
+
return sizeof(pm_buffer_t);
|
9
|
+
}
|
10
|
+
|
11
|
+
// Initialize a pm_buffer_t with its default values.
|
12
|
+
bool
|
13
|
+
pm_buffer_init(pm_buffer_t *buffer) {
|
14
|
+
buffer->length = 0;
|
15
|
+
buffer->capacity = PRISM_BUFFER_INITIAL_SIZE;
|
16
|
+
|
17
|
+
buffer->value = (char *) malloc(PRISM_BUFFER_INITIAL_SIZE);
|
18
|
+
return buffer->value != NULL;
|
19
|
+
}
|
20
|
+
|
21
|
+
#undef PRISM_BUFFER_INITIAL_SIZE
|
22
|
+
|
23
|
+
// Return the value of the buffer.
|
24
|
+
char *
|
25
|
+
pm_buffer_value(pm_buffer_t *buffer) {
|
26
|
+
return buffer->value;
|
27
|
+
}
|
28
|
+
|
29
|
+
// Return the length of the buffer.
|
30
|
+
size_t
|
31
|
+
pm_buffer_length(pm_buffer_t *buffer) {
|
32
|
+
return buffer->length;
|
33
|
+
}
|
34
|
+
|
35
|
+
// Append the given amount of space to the buffer.
|
36
|
+
static inline void
|
37
|
+
pm_buffer_append_length(pm_buffer_t *buffer, size_t length) {
|
38
|
+
size_t next_length = buffer->length + length;
|
39
|
+
|
40
|
+
if (next_length > buffer->capacity) {
|
41
|
+
do {
|
42
|
+
buffer->capacity *= 2;
|
43
|
+
} while (next_length > buffer->capacity);
|
44
|
+
|
45
|
+
buffer->value = realloc(buffer->value, buffer->capacity);
|
46
|
+
}
|
47
|
+
|
48
|
+
buffer->length = next_length;
|
49
|
+
}
|
50
|
+
|
51
|
+
// Append a generic pointer to memory to the buffer.
|
52
|
+
static inline void
|
53
|
+
pm_buffer_append(pm_buffer_t *buffer, const void *source, size_t length) {
|
54
|
+
pm_buffer_append_length(buffer, length);
|
55
|
+
memcpy(buffer->value + (buffer->length - length), source, length);
|
56
|
+
}
|
57
|
+
|
58
|
+
// Append the given amount of space as zeroes to the buffer.
|
59
|
+
void
|
60
|
+
pm_buffer_append_zeroes(pm_buffer_t *buffer, size_t length) {
|
61
|
+
pm_buffer_append_length(buffer, length);
|
62
|
+
memset(buffer->value + (buffer->length - length), 0, length);
|
63
|
+
}
|
64
|
+
|
65
|
+
// Append a string to the buffer.
|
66
|
+
void
|
67
|
+
pm_buffer_append_str(pm_buffer_t *buffer, const char *value, size_t length) {
|
68
|
+
pm_buffer_append(buffer, value, length);
|
69
|
+
}
|
70
|
+
|
71
|
+
// Append a list of bytes to the buffer.
|
72
|
+
void
|
73
|
+
pm_buffer_append_bytes(pm_buffer_t *buffer, const uint8_t *value, size_t length) {
|
74
|
+
pm_buffer_append(buffer, (const char *) value, length);
|
75
|
+
}
|
76
|
+
|
77
|
+
// Append a single byte to the buffer.
|
78
|
+
void
|
79
|
+
pm_buffer_append_u8(pm_buffer_t *buffer, uint8_t value) {
|
80
|
+
const void *source = &value;
|
81
|
+
pm_buffer_append(buffer, source, sizeof(uint8_t));
|
82
|
+
}
|
83
|
+
|
84
|
+
// Append a 32-bit unsigned integer to the buffer.
|
85
|
+
void
|
86
|
+
pm_buffer_append_u32(pm_buffer_t *buffer, uint32_t value) {
|
87
|
+
if (value < 128) {
|
88
|
+
pm_buffer_append_u8(buffer, (uint8_t) value);
|
89
|
+
} else {
|
90
|
+
uint32_t n = value;
|
91
|
+
while (n >= 128) {
|
92
|
+
pm_buffer_append_u8(buffer, (uint8_t) (n | 128));
|
93
|
+
n >>= 7;
|
94
|
+
}
|
95
|
+
pm_buffer_append_u8(buffer, (uint8_t) n);
|
96
|
+
}
|
97
|
+
}
|
98
|
+
|
99
|
+
// Free the memory associated with the buffer.
|
100
|
+
void
|
101
|
+
pm_buffer_free(pm_buffer_t *buffer) {
|
102
|
+
free(buffer->value);
|
103
|
+
}
|