Commit | Line | Data |
---|---|---|
71aaa3f7 PP |
1 | # The MIT License (MIT) |
2 | # | |
3 | # Copyright (c) 2023 Philippe Proulx <eeppeliteloop@gmail.com> | |
4 | # | |
5 | # Permission is hereby granted, free of charge, to any person obtaining | |
6 | # a copy of this software and associated documentation files (the | |
7 | # "Software"), to deal in the Software without restriction, including | |
8 | # without limitation the rights to use, copy, modify, merge, publish, | |
9 | # distribute, sublicense, and/or sell copies of the Software, and to | |
10 | # permit persons to whom the Software is furnished to do so, subject to | |
11 | # the following conditions: | |
12 | # | |
13 | # The above copyright notice and this permission notice shall be | |
14 | # included in all copies or substantial portions of the Software. | |
15 | # | |
16 | # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, | |
17 | # EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF | |
18 | # MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. | |
19 | # IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY | |
20 | # CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, | |
21 | # TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE | |
22 | # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. | |
23 | ||
5a993698 PP |
24 | # This module is the portable Normand processor. It offers both the |
25 | # parse() function and the command-line tool (run the module itself) | |
26 | # without external dependencies except a `typing` module for Python 3.4. | |
27 | # | |
28 | # Feel free to copy this module file to your own project to use Normand. | |
29 | # | |
30 | # Upstream repository: <https://github.com/efficios/normand>. | |
31 | ||
71aaa3f7 | 32 | __author__ = "Philippe Proulx" |
676f6189 | 33 | __version__ = "0.7.0" |
71aaa3f7 PP |
34 | __all__ = [ |
35 | "ByteOrder", | |
36 | "parse", | |
37 | "ParseError", | |
38 | "ParseResult", | |
39 | "TextLoc", | |
1b8aa84a | 40 | "SymbolsT", |
71aaa3f7 PP |
41 | "__author__", |
42 | "__version__", | |
43 | ] | |
44 | ||
45 | import re | |
46 | import abc | |
47 | import ast | |
48 | import sys | |
49 | import enum | |
05f81895 | 50 | import math |
71aaa3f7 | 51 | import struct |
2adf4336 PP |
52 | from typing import ( |
53 | Any, | |
54 | Set, | |
55 | Dict, | |
56 | List, | |
2adf4336 PP |
57 | Union, |
58 | Pattern, | |
59 | Callable, | |
60 | NoReturn, | |
61 | Optional, | |
62 | ) | |
71aaa3f7 PP |
63 | |
64 | ||
65 | # Text location (line and column numbers). | |
66 | class TextLoc: | |
67 | @classmethod | |
68 | def _create(cls, line_no: int, col_no: int): | |
69 | self = cls.__new__(cls) | |
70 | self._init(line_no, col_no) | |
71 | return self | |
72 | ||
73 | def __init__(*args, **kwargs): # type: ignore | |
74 | raise NotImplementedError | |
75 | ||
76 | def _init(self, line_no: int, col_no: int): | |
77 | self._line_no = line_no | |
78 | self._col_no = col_no | |
79 | ||
80 | # Line number. | |
81 | @property | |
82 | def line_no(self): | |
83 | return self._line_no | |
84 | ||
85 | # Column number. | |
86 | @property | |
87 | def col_no(self): | |
88 | return self._col_no | |
89 | ||
2adf4336 PP |
90 | def __repr__(self): |
91 | return "TextLoc({}, {})".format(self._line_no, self._col_no) | |
92 | ||
71aaa3f7 PP |
93 | |
94 | # Any item. | |
95 | class _Item: | |
96 | def __init__(self, text_loc: TextLoc): | |
97 | self._text_loc = text_loc | |
98 | ||
99 | # Source text location. | |
100 | @property | |
101 | def text_loc(self): | |
102 | return self._text_loc | |
103 | ||
2adf4336 PP |
104 | |
105 | # Scalar item. | |
106 | class _ScalarItem(_Item): | |
71aaa3f7 PP |
107 | # Returns the size, in bytes, of this item. |
108 | @property | |
109 | @abc.abstractmethod | |
110 | def size(self) -> int: | |
111 | ... | |
112 | ||
113 | ||
114 | # A repeatable item. | |
2adf4336 | 115 | class _RepableItem: |
71aaa3f7 PP |
116 | pass |
117 | ||
118 | ||
119 | # Single byte. | |
2adf4336 | 120 | class _Byte(_ScalarItem, _RepableItem): |
71aaa3f7 PP |
121 | def __init__(self, val: int, text_loc: TextLoc): |
122 | super().__init__(text_loc) | |
123 | self._val = val | |
124 | ||
125 | # Byte value. | |
126 | @property | |
127 | def val(self): | |
128 | return self._val | |
129 | ||
130 | @property | |
131 | def size(self): | |
132 | return 1 | |
133 | ||
134 | def __repr__(self): | |
676f6189 | 135 | return "_Byte({}, {})".format(hex(self._val), repr(self._text_loc)) |
71aaa3f7 PP |
136 | |
137 | ||
138 | # String. | |
2adf4336 | 139 | class _Str(_ScalarItem, _RepableItem): |
71aaa3f7 PP |
140 | def __init__(self, data: bytes, text_loc: TextLoc): |
141 | super().__init__(text_loc) | |
142 | self._data = data | |
143 | ||
144 | # Encoded bytes. | |
145 | @property | |
146 | def data(self): | |
147 | return self._data | |
148 | ||
149 | @property | |
150 | def size(self): | |
151 | return len(self._data) | |
152 | ||
153 | def __repr__(self): | |
676f6189 | 154 | return "_Str({}, {})".format(repr(self._data), repr(self._text_loc)) |
71aaa3f7 PP |
155 | |
156 | ||
157 | # Byte order. | |
158 | @enum.unique | |
159 | class ByteOrder(enum.Enum): | |
160 | # Big endian. | |
161 | BE = "be" | |
162 | ||
163 | # Little endian. | |
164 | LE = "le" | |
165 | ||
166 | ||
2adf4336 PP |
167 | # Byte order setting. |
168 | class _SetBo(_Item): | |
0e8e3169 PP |
169 | def __init__(self, bo: ByteOrder, text_loc: TextLoc): |
170 | super().__init__(text_loc) | |
71aaa3f7 PP |
171 | self._bo = bo |
172 | ||
173 | @property | |
174 | def bo(self): | |
175 | return self._bo | |
176 | ||
2adf4336 | 177 | def __repr__(self): |
676f6189 | 178 | return "_SetBo({}, {})".format(repr(self._bo), repr(self._text_loc)) |
71aaa3f7 PP |
179 | |
180 | ||
181 | # Label. | |
182 | class _Label(_Item): | |
183 | def __init__(self, name: str, text_loc: TextLoc): | |
184 | super().__init__(text_loc) | |
185 | self._name = name | |
186 | ||
187 | # Label name. | |
188 | @property | |
189 | def name(self): | |
190 | return self._name | |
191 | ||
71aaa3f7 | 192 | def __repr__(self): |
676f6189 | 193 | return "_Label({}, {})".format(repr(self._name), repr(self._text_loc)) |
71aaa3f7 PP |
194 | |
195 | ||
2adf4336 PP |
196 | # Offset setting. |
197 | class _SetOffset(_Item): | |
71aaa3f7 PP |
198 | def __init__(self, val: int, text_loc: TextLoc): |
199 | super().__init__(text_loc) | |
200 | self._val = val | |
201 | ||
676f6189 | 202 | # Offset value (bytes). |
71aaa3f7 PP |
203 | @property |
204 | def val(self): | |
205 | return self._val | |
206 | ||
71aaa3f7 | 207 | def __repr__(self): |
676f6189 PP |
208 | return "_SetOffset({}, {})".format(repr(self._val), repr(self._text_loc)) |
209 | ||
210 | ||
211 | # Offset alignment. | |
212 | class _AlignOffset(_Item): | |
213 | def __init__(self, val: int, pad_val: int, text_loc: TextLoc): | |
214 | super().__init__(text_loc) | |
215 | self._val = val | |
216 | self._pad_val = pad_val | |
217 | ||
218 | # Alignment value (bits). | |
219 | @property | |
220 | def val(self): | |
221 | return self._val | |
222 | ||
223 | # Padding byte value. | |
224 | @property | |
225 | def pad_val(self): | |
226 | return self._pad_val | |
227 | ||
228 | def __repr__(self): | |
229 | return "_AlignOffset({}, {}, {})".format( | |
230 | repr(self._val), repr(self._pad_val), repr(self._text_loc) | |
231 | ) | |
71aaa3f7 PP |
232 | |
233 | ||
234 | # Mixin of containing an AST expression and its string. | |
235 | class _ExprMixin: | |
236 | def __init__(self, expr_str: str, expr: ast.Expression): | |
237 | self._expr_str = expr_str | |
238 | self._expr = expr | |
239 | ||
240 | # Expression string. | |
241 | @property | |
242 | def expr_str(self): | |
243 | return self._expr_str | |
244 | ||
245 | # Expression node to evaluate. | |
246 | @property | |
247 | def expr(self): | |
248 | return self._expr | |
249 | ||
250 | ||
2adf4336 PP |
251 | # Variable assignment. |
252 | class _VarAssign(_Item, _ExprMixin): | |
71aaa3f7 PP |
253 | def __init__( |
254 | self, name: str, expr_str: str, expr: ast.Expression, text_loc: TextLoc | |
255 | ): | |
256 | super().__init__(text_loc) | |
257 | _ExprMixin.__init__(self, expr_str, expr) | |
258 | self._name = name | |
259 | ||
260 | # Name. | |
261 | @property | |
262 | def name(self): | |
263 | return self._name | |
264 | ||
71aaa3f7 | 265 | def __repr__(self): |
2adf4336 | 266 | return "_VarAssign({}, {}, {}, {})".format( |
676f6189 PP |
267 | repr(self._name), |
268 | repr(self._expr_str), | |
269 | repr(self._expr), | |
270 | repr(self._text_loc), | |
71aaa3f7 PP |
271 | ) |
272 | ||
273 | ||
269f6eb3 PP |
274 | # Fixed-length number, possibly needing more than one byte. |
275 | class _FlNum(_ScalarItem, _RepableItem, _ExprMixin): | |
71aaa3f7 PP |
276 | def __init__( |
277 | self, expr_str: str, expr: ast.Expression, len: int, text_loc: TextLoc | |
278 | ): | |
279 | super().__init__(text_loc) | |
280 | _ExprMixin.__init__(self, expr_str, expr) | |
281 | self._len = len | |
282 | ||
283 | # Length (bits). | |
284 | @property | |
285 | def len(self): | |
286 | return self._len | |
287 | ||
288 | @property | |
289 | def size(self): | |
290 | return self._len // 8 | |
291 | ||
292 | def __repr__(self): | |
269f6eb3 | 293 | return "_FlNum({}, {}, {}, {})".format( |
676f6189 PP |
294 | repr(self._expr_str), |
295 | repr(self._expr), | |
296 | repr(self._len), | |
297 | repr(self._text_loc), | |
71aaa3f7 PP |
298 | ) |
299 | ||
300 | ||
05f81895 PP |
301 | # LEB128 integer. |
302 | class _Leb128Int(_Item, _RepableItem, _ExprMixin): | |
303 | def __init__(self, expr_str: str, expr: ast.Expression, text_loc: TextLoc): | |
304 | super().__init__(text_loc) | |
305 | _ExprMixin.__init__(self, expr_str, expr) | |
306 | ||
307 | def __repr__(self): | |
308 | return "{}({}, {}, {})".format( | |
309 | self.__class__.__name__, | |
310 | repr(self._expr_str), | |
311 | repr(self._expr), | |
676f6189 | 312 | repr(self._text_loc), |
05f81895 PP |
313 | ) |
314 | ||
315 | ||
316 | # Unsigned LEB128 integer. | |
317 | class _ULeb128Int(_Leb128Int, _RepableItem, _ExprMixin): | |
318 | pass | |
319 | ||
320 | ||
321 | # Signed LEB128 integer. | |
322 | class _SLeb128Int(_Leb128Int, _RepableItem, _ExprMixin): | |
323 | pass | |
324 | ||
325 | ||
71aaa3f7 | 326 | # Group of items. |
2adf4336 | 327 | class _Group(_Item, _RepableItem): |
71aaa3f7 PP |
328 | def __init__(self, items: List[_Item], text_loc: TextLoc): |
329 | super().__init__(text_loc) | |
330 | self._items = items | |
71aaa3f7 PP |
331 | |
332 | # Contained items. | |
333 | @property | |
334 | def items(self): | |
335 | return self._items | |
336 | ||
71aaa3f7 | 337 | def __repr__(self): |
676f6189 | 338 | return "_Group({}, {})".format(repr(self._items), repr(self._text_loc)) |
71aaa3f7 PP |
339 | |
340 | ||
341 | # Repetition item. | |
2adf4336 PP |
342 | class _Rep(_Item, _ExprMixin): |
343 | def __init__( | |
344 | self, item: _Item, expr_str: str, expr: ast.Expression, text_loc: TextLoc | |
345 | ): | |
71aaa3f7 | 346 | super().__init__(text_loc) |
2adf4336 | 347 | _ExprMixin.__init__(self, expr_str, expr) |
71aaa3f7 | 348 | self._item = item |
71aaa3f7 PP |
349 | |
350 | # Item to repeat. | |
351 | @property | |
352 | def item(self): | |
353 | return self._item | |
354 | ||
71aaa3f7 | 355 | def __repr__(self): |
2adf4336 | 356 | return "_Rep({}, {}, {}, {})".format( |
676f6189 PP |
357 | repr(self._item), |
358 | repr(self._expr_str), | |
359 | repr(self._expr), | |
360 | repr(self._text_loc), | |
71aaa3f7 PP |
361 | ) |
362 | ||
363 | ||
2adf4336 | 364 | # Expression item type. |
269f6eb3 | 365 | _ExprItemT = Union[_FlNum, _Leb128Int, _VarAssign, _Rep] |
2adf4336 PP |
366 | |
367 | ||
71aaa3f7 PP |
368 | # A parsing error containing a message and a text location. |
369 | class ParseError(RuntimeError): | |
370 | @classmethod | |
371 | def _create(cls, msg: str, text_loc: TextLoc): | |
372 | self = cls.__new__(cls) | |
373 | self._init(msg, text_loc) | |
374 | return self | |
375 | ||
376 | def __init__(self, *args, **kwargs): # type: ignore | |
377 | raise NotImplementedError | |
378 | ||
379 | def _init(self, msg: str, text_loc: TextLoc): | |
380 | super().__init__(msg) | |
381 | self._text_loc = text_loc | |
382 | ||
383 | # Source text location. | |
384 | @property | |
385 | def text_loc(self): | |
386 | return self._text_loc | |
387 | ||
388 | ||
389 | # Raises a parsing error, forwarding the parameters to the constructor. | |
390 | def _raise_error(msg: str, text_loc: TextLoc) -> NoReturn: | |
391 | raise ParseError._create(msg, text_loc) # pyright: ignore[reportPrivateUsage] | |
392 | ||
393 | ||
2adf4336 | 394 | # Variable/label dictionary type. |
82425ef0 | 395 | SymbolsT = Dict[str, Union[int, float]] |
71aaa3f7 PP |
396 | |
397 | ||
398 | # Python name pattern. | |
399 | _py_name_pat = re.compile(r"[a-zA-Z_][a-zA-Z0-9_]*") | |
400 | ||
401 | ||
402 | # Normand parser. | |
403 | # | |
404 | # The constructor accepts a Normand input. After building, use the `res` | |
405 | # property to get the resulting main group. | |
406 | class _Parser: | |
407 | # Builds a parser to parse the Normand input `normand`, parsing | |
408 | # immediately. | |
1b8aa84a | 409 | def __init__(self, normand: str, variables: SymbolsT, labels: SymbolsT): |
71aaa3f7 PP |
410 | self._normand = normand |
411 | self._at = 0 | |
412 | self._line_no = 1 | |
413 | self._col_no = 1 | |
414 | self._label_names = set(labels.keys()) | |
415 | self._var_names = set(variables.keys()) | |
416 | self._parse() | |
417 | ||
418 | # Result (main group). | |
419 | @property | |
420 | def res(self): | |
421 | return self._res | |
422 | ||
423 | # Current text location. | |
424 | @property | |
425 | def _text_loc(self): | |
426 | return TextLoc._create( # pyright: ignore[reportPrivateUsage] | |
427 | self._line_no, self._col_no | |
428 | ) | |
429 | ||
430 | # Returns `True` if this parser is done parsing. | |
431 | def _is_done(self): | |
432 | return self._at == len(self._normand) | |
433 | ||
434 | # Returns `True` if this parser isn't done parsing. | |
435 | def _isnt_done(self): | |
436 | return not self._is_done() | |
437 | ||
438 | # Raises a parse error, creating it using the message `msg` and the | |
439 | # current text location. | |
440 | def _raise_error(self, msg: str) -> NoReturn: | |
441 | _raise_error(msg, self._text_loc) | |
442 | ||
443 | # Tries to make the pattern `pat` match the current substring, | |
444 | # returning the match object and updating `self._at`, | |
445 | # `self._line_no`, and `self._col_no` on success. | |
446 | def _try_parse_pat(self, pat: Pattern[str]): | |
447 | m = pat.match(self._normand, self._at) | |
448 | ||
449 | if m is None: | |
450 | return | |
451 | ||
452 | # Skip matched string | |
453 | self._at += len(m.group(0)) | |
454 | ||
455 | # Update line number | |
456 | self._line_no += m.group(0).count("\n") | |
457 | ||
458 | # Update column number | |
459 | for i in reversed(range(self._at)): | |
460 | if self._normand[i] == "\n" or i == 0: | |
461 | if i == 0: | |
462 | self._col_no = self._at + 1 | |
463 | else: | |
464 | self._col_no = self._at - i | |
465 | ||
466 | break | |
467 | ||
468 | # Return match object | |
469 | return m | |
470 | ||
471 | # Expects the pattern `pat` to match the current substring, | |
472 | # returning the match object and updating `self._at`, | |
473 | # `self._line_no`, and `self._col_no` on success, or raising a parse | |
474 | # error with the message `error_msg` on error. | |
475 | def _expect_pat(self, pat: Pattern[str], error_msg: str): | |
476 | # Match | |
477 | m = self._try_parse_pat(pat) | |
478 | ||
479 | if m is None: | |
480 | # No match: error | |
481 | self._raise_error(error_msg) | |
482 | ||
483 | # Return match object | |
484 | return m | |
485 | ||
486 | # Pattern for _skip_ws_and_comments() | |
487 | _ws_or_syms_or_comments_pat = re.compile( | |
676f6189 | 488 | r"(?:[\s!/\\?&:;.,+[\]_=|-]|#[^#]*?(?:\n|#))*" |
71aaa3f7 PP |
489 | ) |
490 | ||
491 | # Skips as many whitespaces, insignificant symbol characters, and | |
492 | # comments as possible. | |
493 | def _skip_ws_and_comments(self): | |
494 | self._try_parse_pat(self._ws_or_syms_or_comments_pat) | |
495 | ||
496 | # Pattern for _try_parse_hex_byte() | |
497 | _nibble_pat = re.compile(r"[A-Fa-f0-9]") | |
498 | ||
499 | # Tries to parse a hexadecimal byte, returning a byte item on | |
500 | # success. | |
501 | def _try_parse_hex_byte(self): | |
0e8e3169 PP |
502 | begin_text_loc = self._text_loc |
503 | ||
71aaa3f7 PP |
504 | # Match initial nibble |
505 | m_high = self._try_parse_pat(self._nibble_pat) | |
506 | ||
507 | if m_high is None: | |
508 | # No match | |
509 | return | |
510 | ||
511 | # Expect another nibble | |
512 | self._skip_ws_and_comments() | |
513 | m_low = self._expect_pat( | |
514 | self._nibble_pat, "Expecting another hexadecimal nibble" | |
515 | ) | |
516 | ||
517 | # Return item | |
0e8e3169 | 518 | return _Byte(int(m_high.group(0) + m_low.group(0), 16), begin_text_loc) |
71aaa3f7 PP |
519 | |
520 | # Patterns for _try_parse_bin_byte() | |
521 | _bin_byte_bit_pat = re.compile(r"[01]") | |
522 | _bin_byte_prefix_pat = re.compile(r"%") | |
523 | ||
524 | # Tries to parse a binary byte, returning a byte item on success. | |
525 | def _try_parse_bin_byte(self): | |
0e8e3169 PP |
526 | begin_text_loc = self._text_loc |
527 | ||
71aaa3f7 PP |
528 | # Match prefix |
529 | if self._try_parse_pat(self._bin_byte_prefix_pat) is None: | |
530 | # No match | |
531 | return | |
532 | ||
533 | # Expect eight bits | |
534 | bits = [] # type: List[str] | |
535 | ||
536 | for _ in range(8): | |
537 | self._skip_ws_and_comments() | |
538 | m = self._expect_pat(self._bin_byte_bit_pat, "Expecting a bit (`0` or `1`)") | |
539 | bits.append(m.group(0)) | |
540 | ||
541 | # Return item | |
0e8e3169 | 542 | return _Byte(int("".join(bits), 2), begin_text_loc) |
71aaa3f7 PP |
543 | |
544 | # Patterns for _try_parse_dec_byte() | |
545 | _dec_byte_prefix_pat = re.compile(r"\$\s*") | |
546 | _dec_byte_val_pat = re.compile(r"(?P<neg>-?)(?P<val>\d+)") | |
547 | ||
548 | # Tries to parse a decimal byte, returning a byte item on success. | |
549 | def _try_parse_dec_byte(self): | |
0e8e3169 PP |
550 | begin_text_loc = self._text_loc |
551 | ||
71aaa3f7 PP |
552 | # Match prefix |
553 | if self._try_parse_pat(self._dec_byte_prefix_pat) is None: | |
554 | # No match | |
555 | return | |
556 | ||
557 | # Expect the value | |
558 | m = self._expect_pat(self._dec_byte_val_pat, "Expecting a decimal constant") | |
559 | ||
560 | # Compute value | |
561 | val = int(m.group("val")) * (-1 if m.group("neg") == "-" else 1) | |
562 | ||
563 | # Validate | |
564 | if val < -128 or val > 255: | |
0e8e3169 | 565 | _raise_error("Invalid decimal byte value {}".format(val), begin_text_loc) |
71aaa3f7 PP |
566 | |
567 | # Two's complement | |
05f81895 | 568 | val %= 256 |
71aaa3f7 PP |
569 | |
570 | # Return item | |
0e8e3169 | 571 | return _Byte(val, begin_text_loc) |
71aaa3f7 PP |
572 | |
573 | # Tries to parse a byte, returning a byte item on success. | |
574 | def _try_parse_byte(self): | |
575 | # Hexadecimal | |
576 | item = self._try_parse_hex_byte() | |
577 | ||
578 | if item is not None: | |
579 | return item | |
580 | ||
581 | # Binary | |
582 | item = self._try_parse_bin_byte() | |
583 | ||
584 | if item is not None: | |
585 | return item | |
586 | ||
587 | # Decimal | |
588 | item = self._try_parse_dec_byte() | |
589 | ||
590 | if item is not None: | |
591 | return item | |
592 | ||
593 | # Patterns for _try_parse_str() | |
594 | _str_prefix_pat = re.compile(r'(?:u(?P<len>16|32)(?P<bo>be|le))?\s*"') | |
595 | _str_suffix_pat = re.compile(r'"') | |
596 | _str_str_pat = re.compile(r'(?:(?:\\.)|[^"])*') | |
597 | ||
598 | # Strings corresponding to escape sequence characters | |
599 | _str_escape_seq_strs = { | |
600 | "0": "\0", | |
601 | "a": "\a", | |
602 | "b": "\b", | |
603 | "e": "\x1b", | |
604 | "f": "\f", | |
605 | "n": "\n", | |
606 | "r": "\r", | |
607 | "t": "\t", | |
608 | "v": "\v", | |
609 | "\\": "\\", | |
610 | '"': '"', | |
611 | } | |
612 | ||
613 | # Tries to parse a string, returning a string item on success. | |
614 | def _try_parse_str(self): | |
0e8e3169 PP |
615 | begin_text_loc = self._text_loc |
616 | ||
71aaa3f7 PP |
617 | # Match prefix |
618 | m = self._try_parse_pat(self._str_prefix_pat) | |
619 | ||
620 | if m is None: | |
621 | # No match | |
622 | return | |
623 | ||
624 | # Get encoding | |
625 | encoding = "utf8" | |
626 | ||
627 | if m.group("len") is not None: | |
628 | encoding = "utf_{}_{}".format(m.group("len"), m.group("bo")) | |
629 | ||
630 | # Actual string | |
631 | m = self._expect_pat(self._str_str_pat, "Expecting a literal string") | |
632 | ||
633 | # Expect end of string | |
634 | self._expect_pat(self._str_suffix_pat, 'Expecting `"` (end of literal string)') | |
635 | ||
636 | # Replace escape sequences | |
637 | val = m.group(0) | |
638 | ||
639 | for ec in '0abefnrtv"\\': | |
640 | val = val.replace(r"\{}".format(ec), self._str_escape_seq_strs[ec]) | |
641 | ||
642 | # Encode | |
643 | data = val.encode(encoding) | |
644 | ||
645 | # Return item | |
0e8e3169 | 646 | return _Str(data, begin_text_loc) |
71aaa3f7 PP |
647 | |
648 | # Patterns for _try_parse_group() | |
649 | _group_prefix_pat = re.compile(r"\(") | |
650 | _group_suffix_pat = re.compile(r"\)") | |
651 | ||
652 | # Tries to parse a group, returning a group item on success. | |
653 | def _try_parse_group(self): | |
0e8e3169 PP |
654 | begin_text_loc = self._text_loc |
655 | ||
71aaa3f7 PP |
656 | # Match prefix |
657 | if self._try_parse_pat(self._group_prefix_pat) is None: | |
658 | # No match | |
659 | return | |
660 | ||
661 | # Parse items | |
662 | items = self._parse_items() | |
663 | ||
664 | # Expect end of group | |
665 | self._skip_ws_and_comments() | |
666 | self._expect_pat( | |
667 | self._group_suffix_pat, "Expecting an item or `)` (end of group)" | |
668 | ) | |
669 | ||
670 | # Return item | |
0e8e3169 | 671 | return _Group(items, begin_text_loc) |
71aaa3f7 PP |
672 | |
673 | # Returns a stripped expression string and an AST expression node | |
674 | # from the expression string `expr_str` at text location `text_loc`. | |
675 | def _ast_expr_from_str(self, expr_str: str, text_loc: TextLoc): | |
676 | # Create an expression node from the expression string | |
677 | expr_str = expr_str.strip().replace("\n", " ") | |
678 | ||
679 | try: | |
680 | expr = ast.parse(expr_str, mode="eval") | |
681 | except SyntaxError: | |
682 | _raise_error( | |
683 | "Invalid expression `{}`: invalid syntax".format(expr_str), | |
684 | text_loc, | |
685 | ) | |
686 | ||
687 | return expr_str, expr | |
688 | ||
269f6eb3 | 689 | # Patterns for _try_parse_num_and_attr() |
05f81895 | 690 | _val_expr_pat = re.compile(r"([^}:]+):\s*") |
269f6eb3 | 691 | _fl_num_len_attr_pat = re.compile(r"8|16|24|32|40|48|56|64") |
05f81895 | 692 | _leb128_int_attr_pat = re.compile(r"(u|s)leb128") |
71aaa3f7 | 693 | |
05f81895 PP |
694 | # Tries to parse a value and attribute (fixed length in bits or |
695 | # `leb128`), returning a value item on success. | |
269f6eb3 | 696 | def _try_parse_num_and_attr(self): |
71aaa3f7 PP |
697 | begin_text_loc = self._text_loc |
698 | ||
699 | # Match | |
700 | m_expr = self._try_parse_pat(self._val_expr_pat) | |
701 | ||
702 | if m_expr is None: | |
703 | # No match | |
704 | return | |
705 | ||
71aaa3f7 PP |
706 | # Create an expression node from the expression string |
707 | expr_str, expr = self._ast_expr_from_str(m_expr.group(1), begin_text_loc) | |
708 | ||
05f81895 | 709 | # Length? |
269f6eb3 | 710 | m_attr = self._try_parse_pat(self._fl_num_len_attr_pat) |
05f81895 PP |
711 | |
712 | if m_attr is None: | |
713 | # LEB128? | |
714 | m_attr = self._try_parse_pat(self._leb128_int_attr_pat) | |
715 | ||
716 | if m_attr is None: | |
717 | # At this point it's invalid | |
718 | self._raise_error( | |
719 | "Expecting a length (multiple of eight bits), `uleb128`, or `sleb128`" | |
720 | ) | |
721 | ||
722 | # Return LEB128 integer item | |
723 | cls = _ULeb128Int if m_attr.group(1) == "u" else _SLeb128Int | |
724 | return cls(expr_str, expr, begin_text_loc) | |
725 | else: | |
269f6eb3 PP |
726 | # Return fixed-length number item |
727 | return _FlNum( | |
05f81895 PP |
728 | expr_str, |
729 | expr, | |
730 | int(m_attr.group(0)), | |
731 | begin_text_loc, | |
732 | ) | |
71aaa3f7 | 733 | |
269f6eb3 | 734 | # Patterns for _try_parse_num_and_attr() |
2adf4336 | 735 | _var_assign_pat = re.compile( |
71aaa3f7 PP |
736 | r"(?P<name>{})\s*=\s*(?P<expr>[^}}]+)".format(_py_name_pat.pattern) |
737 | ) | |
738 | ||
2adf4336 PP |
739 | # Tries to parse a variable assignment, returning a variable |
740 | # assignment item on success. | |
741 | def _try_parse_var_assign(self): | |
71aaa3f7 PP |
742 | begin_text_loc = self._text_loc |
743 | ||
744 | # Match | |
2adf4336 | 745 | m = self._try_parse_pat(self._var_assign_pat) |
71aaa3f7 PP |
746 | |
747 | if m is None: | |
748 | # No match | |
749 | return | |
750 | ||
751 | # Validate name | |
752 | name = m.group("name") | |
753 | ||
754 | if name == _icitte_name: | |
0e8e3169 PP |
755 | _raise_error( |
756 | "`{}` is a reserved variable name".format(_icitte_name), begin_text_loc | |
757 | ) | |
71aaa3f7 PP |
758 | |
759 | if name in self._label_names: | |
0e8e3169 | 760 | _raise_error("Existing label named `{}`".format(name), begin_text_loc) |
71aaa3f7 PP |
761 | |
762 | # Add to known variable names | |
763 | self._var_names.add(name) | |
764 | ||
765 | # Create an expression node from the expression string | |
766 | expr_str, expr = self._ast_expr_from_str(m.group("expr"), begin_text_loc) | |
767 | ||
768 | # Return item | |
2adf4336 | 769 | return _VarAssign( |
71aaa3f7 PP |
770 | name, |
771 | expr_str, | |
772 | expr, | |
0e8e3169 | 773 | begin_text_loc, |
71aaa3f7 PP |
774 | ) |
775 | ||
2adf4336 | 776 | # Pattern for _try_parse_set_bo() |
71aaa3f7 PP |
777 | _bo_pat = re.compile(r"[bl]e") |
778 | ||
2adf4336 PP |
779 | # Tries to parse a byte order name, returning a byte order setting |
780 | # item on success. | |
781 | def _try_parse_set_bo(self): | |
0e8e3169 PP |
782 | begin_text_loc = self._text_loc |
783 | ||
71aaa3f7 PP |
784 | # Match |
785 | m = self._try_parse_pat(self._bo_pat) | |
786 | ||
787 | if m is None: | |
788 | # No match | |
789 | return | |
790 | ||
791 | # Return corresponding item | |
792 | if m.group(0) == "be": | |
2adf4336 | 793 | return _SetBo(ByteOrder.BE, begin_text_loc) |
71aaa3f7 PP |
794 | else: |
795 | assert m.group(0) == "le" | |
2adf4336 | 796 | return _SetBo(ByteOrder.LE, begin_text_loc) |
71aaa3f7 PP |
797 | |
798 | # Patterns for _try_parse_val_or_bo() | |
2adf4336 PP |
799 | _val_var_assign_set_bo_prefix_pat = re.compile(r"\{\s*") |
800 | _val_var_assign_set_bo_suffix_pat = re.compile(r"\s*}") | |
71aaa3f7 | 801 | |
2adf4336 PP |
802 | # Tries to parse a value, a variable assignment, or a byte order |
803 | # setting, returning an item on success. | |
804 | def _try_parse_val_or_var_assign_or_set_bo(self): | |
71aaa3f7 | 805 | # Match prefix |
2adf4336 | 806 | if self._try_parse_pat(self._val_var_assign_set_bo_prefix_pat) is None: |
71aaa3f7 PP |
807 | # No match |
808 | return | |
809 | ||
2adf4336 PP |
810 | # Variable assignment item? |
811 | item = self._try_parse_var_assign() | |
71aaa3f7 PP |
812 | |
813 | if item is None: | |
269f6eb3 PP |
814 | # Number item? |
815 | item = self._try_parse_num_and_attr() | |
71aaa3f7 PP |
816 | |
817 | if item is None: | |
2adf4336 PP |
818 | # Byte order setting item? |
819 | item = self._try_parse_set_bo() | |
71aaa3f7 PP |
820 | |
821 | if item is None: | |
822 | # At this point it's invalid | |
2adf4336 | 823 | self._raise_error( |
269f6eb3 | 824 | "Expecting a fixed-length number, a variable assignment, or a byte order setting" |
2adf4336 | 825 | ) |
71aaa3f7 PP |
826 | |
827 | # Expect suffix | |
2adf4336 | 828 | self._expect_pat(self._val_var_assign_set_bo_suffix_pat, "Expecting `}`") |
71aaa3f7 PP |
829 | return item |
830 | ||
2adf4336 | 831 | # Pattern for _try_parse_set_offset_val() and _try_parse_rep() |
71aaa3f7 PP |
832 | _pos_const_int_pat = re.compile(r"0[Xx][A-Fa-f0-9]+|\d+") |
833 | ||
2adf4336 PP |
834 | # Tries to parse an offset setting value (after the initial `<`), |
835 | # returning an offset item on success. | |
836 | def _try_parse_set_offset_val(self): | |
0e8e3169 PP |
837 | begin_text_loc = self._text_loc |
838 | ||
71aaa3f7 PP |
839 | # Match |
840 | m = self._try_parse_pat(self._pos_const_int_pat) | |
841 | ||
842 | if m is None: | |
843 | # No match | |
844 | return | |
845 | ||
846 | # Return item | |
2adf4336 | 847 | return _SetOffset(int(m.group(0), 0), begin_text_loc) |
71aaa3f7 PP |
848 | |
849 | # Tries to parse a label name (after the initial `<`), returning a | |
850 | # label item on success. | |
851 | def _try_parse_label_name(self): | |
0e8e3169 PP |
852 | begin_text_loc = self._text_loc |
853 | ||
71aaa3f7 PP |
854 | # Match |
855 | m = self._try_parse_pat(_py_name_pat) | |
856 | ||
857 | if m is None: | |
858 | # No match | |
859 | return | |
860 | ||
861 | # Validate | |
862 | name = m.group(0) | |
863 | ||
864 | if name == _icitte_name: | |
0e8e3169 PP |
865 | _raise_error( |
866 | "`{}` is a reserved label name".format(_icitte_name), begin_text_loc | |
867 | ) | |
71aaa3f7 PP |
868 | |
869 | if name in self._label_names: | |
0e8e3169 | 870 | _raise_error("Duplicate label name `{}`".format(name), begin_text_loc) |
71aaa3f7 PP |
871 | |
872 | if name in self._var_names: | |
0e8e3169 | 873 | _raise_error("Existing variable named `{}`".format(name), begin_text_loc) |
71aaa3f7 PP |
874 | |
875 | # Add to known label names | |
876 | self._label_names.add(name) | |
877 | ||
878 | # Return item | |
0e8e3169 | 879 | return _Label(name, begin_text_loc) |
71aaa3f7 | 880 | |
2adf4336 PP |
881 | # Patterns for _try_parse_label_or_set_offset() |
882 | _label_set_offset_prefix_pat = re.compile(r"<\s*") | |
883 | _label_set_offset_suffix_pat = re.compile(r"\s*>") | |
71aaa3f7 | 884 | |
2adf4336 PP |
885 | # Tries to parse a label or an offset setting, returning an item on |
886 | # success. | |
887 | def _try_parse_label_or_set_offset(self): | |
71aaa3f7 | 888 | # Match prefix |
2adf4336 | 889 | if self._try_parse_pat(self._label_set_offset_prefix_pat) is None: |
71aaa3f7 PP |
890 | # No match |
891 | return | |
892 | ||
2adf4336 PP |
893 | # Offset setting item? |
894 | item = self._try_parse_set_offset_val() | |
71aaa3f7 PP |
895 | |
896 | if item is None: | |
897 | # Label item? | |
898 | item = self._try_parse_label_name() | |
899 | ||
900 | if item is None: | |
901 | # At this point it's invalid | |
2adf4336 | 902 | self._raise_error("Expecting a label name or an offset setting value") |
71aaa3f7 PP |
903 | |
904 | # Expect suffix | |
2adf4336 | 905 | self._expect_pat(self._label_set_offset_suffix_pat, "Expecting `>`") |
71aaa3f7 PP |
906 | return item |
907 | ||
676f6189 PP |
908 | # Patterns for _try_parse_align_offset() |
909 | _align_offset_prefix_pat = re.compile(r"@\s*") | |
910 | _align_offset_val_pat = re.compile(r"(\d+)\s*") | |
911 | _align_offset_pad_val_prefix_pat = re.compile(r"~\s*") | |
912 | ||
913 | # Tries to parse an offset alignment, returning an offset alignment | |
914 | # item on success. | |
915 | def _try_parse_align_offset(self): | |
916 | begin_text_loc = self._text_loc | |
917 | ||
918 | # Match prefix | |
919 | if self._try_parse_pat(self._align_offset_prefix_pat) is None: | |
920 | # No match | |
921 | return | |
922 | ||
923 | align_text_loc = self._text_loc | |
924 | m = self._expect_pat( | |
925 | self._align_offset_val_pat, | |
926 | "Expecting an alignment (positive multiple of eight bits)", | |
927 | ) | |
928 | ||
929 | # Validate alignment | |
930 | val = int(m.group(1)) | |
931 | ||
932 | if val <= 0 or (val % 8) != 0: | |
933 | _raise_error( | |
934 | "Invalid alignment value {} (not a positive multiple of eight)".format( | |
935 | val | |
936 | ), | |
937 | align_text_loc, | |
938 | ) | |
939 | ||
940 | # Padding value? | |
941 | pad_val = 0 | |
942 | ||
943 | if self._try_parse_pat(self._align_offset_pad_val_prefix_pat) is not None: | |
944 | pad_val_text_loc = self._text_loc | |
945 | m = self._expect_pat(self._pos_const_int_pat, "Expecting a byte value") | |
946 | ||
947 | # Validate | |
948 | pad_val = int(m.group(0), 0) | |
949 | ||
950 | if pad_val > 255: | |
951 | _raise_error( | |
952 | "Invalid padding byte value {}".format(pad_val), | |
953 | pad_val_text_loc, | |
954 | ) | |
955 | ||
956 | # Return item | |
957 | return _AlignOffset(val, pad_val, begin_text_loc) | |
958 | ||
71aaa3f7 PP |
959 | # Tries to parse a base item (anything except a repetition), |
960 | # returning it on success. | |
961 | def _try_parse_base_item(self): | |
962 | # Byte item? | |
963 | item = self._try_parse_byte() | |
964 | ||
965 | if item is not None: | |
966 | return item | |
967 | ||
968 | # String item? | |
969 | item = self._try_parse_str() | |
970 | ||
971 | if item is not None: | |
972 | return item | |
973 | ||
2adf4336 PP |
974 | # Value, variable assignment, or byte order setting item? |
975 | item = self._try_parse_val_or_var_assign_or_set_bo() | |
71aaa3f7 PP |
976 | |
977 | if item is not None: | |
978 | return item | |
979 | ||
2adf4336 PP |
980 | # Label or offset setting item? |
981 | item = self._try_parse_label_or_set_offset() | |
71aaa3f7 | 982 | |
676f6189 PP |
983 | if item is not None: |
984 | return item | |
985 | ||
986 | # Offset alignment item? | |
987 | item = self._try_parse_align_offset() | |
988 | ||
71aaa3f7 PP |
989 | if item is not None: |
990 | return item | |
991 | ||
992 | # Group item? | |
993 | item = self._try_parse_group() | |
994 | ||
995 | if item is not None: | |
996 | return item | |
997 | ||
998 | # Pattern for _try_parse_rep() | |
999 | _rep_prefix_pat = re.compile(r"\*\s*") | |
2adf4336 PP |
1000 | _rep_expr_prefix_pat = re.compile(r"\{") |
1001 | _rep_expr_pat = re.compile(r"[^}p]+") | |
1002 | _rep_expr_suffix_pat = re.compile(r"\}") | |
71aaa3f7 | 1003 | |
2adf4336 PP |
1004 | # Tries to parse a repetition, returning the expression string and |
1005 | # AST expression node on success. | |
71aaa3f7 | 1006 | def _try_parse_rep(self): |
71aaa3f7 PP |
1007 | # Match prefix |
1008 | if self._try_parse_pat(self._rep_prefix_pat) is None: | |
1009 | # No match | |
2adf4336 | 1010 | return |
71aaa3f7 PP |
1011 | |
1012 | # Expect and return a decimal multiplier | |
1013 | self._skip_ws_and_comments() | |
2adf4336 PP |
1014 | |
1015 | # Integer? | |
1016 | m = self._try_parse_pat(self._pos_const_int_pat) | |
1017 | ||
1018 | if m is None: | |
1019 | # Expression? | |
1020 | if self._try_parse_pat(self._rep_expr_prefix_pat) is None: | |
1021 | # At this point it's invalid | |
1022 | self._raise_error("Expecting a positive integral multiplier or `{`") | |
1023 | ||
1024 | # Expect an expression | |
1025 | expr_str_loc = self._text_loc | |
1026 | m = self._expect_pat(self._rep_expr_pat, "Expecting an expression") | |
1027 | expr_str = self._ast_expr_from_str(m.group(0), expr_str_loc) | |
1028 | ||
1029 | # Expect `}` | |
1030 | self._expect_pat(self._rep_expr_suffix_pat, "Expecting `}`") | |
1031 | expr_str = m.group(0) | |
1032 | else: | |
1033 | expr_str_loc = self._text_loc | |
1034 | expr_str = m.group(0) | |
1035 | ||
1036 | return self._ast_expr_from_str(expr_str, expr_str_loc) | |
71aaa3f7 | 1037 | |
1ca7b5e1 PP |
1038 | # Tries to parse an item, possibly followed by a repetition, |
1039 | # returning `True` on success. | |
1040 | # | |
1041 | # Appends any parsed item to `items`. | |
1042 | def _try_append_item(self, items: List[_Item]): | |
71aaa3f7 PP |
1043 | self._skip_ws_and_comments() |
1044 | ||
1045 | # Parse a base item | |
1046 | item = self._try_parse_base_item() | |
1047 | ||
1048 | if item is None: | |
1049 | # No item | |
1ca7b5e1 | 1050 | return False |
71aaa3f7 PP |
1051 | |
1052 | # Parse repetition if the base item is repeatable | |
1053 | if isinstance(item, _RepableItem): | |
0e8e3169 PP |
1054 | self._skip_ws_and_comments() |
1055 | rep_text_loc = self._text_loc | |
2adf4336 | 1056 | rep_ret = self._try_parse_rep() |
71aaa3f7 | 1057 | |
2adf4336 PP |
1058 | if rep_ret is not None: |
1059 | item = _Rep(item, rep_ret[0], rep_ret[1], rep_text_loc) | |
71aaa3f7 | 1060 | |
1ca7b5e1 PP |
1061 | items.append(item) |
1062 | return True | |
71aaa3f7 PP |
1063 | |
1064 | # Parses and returns items, skipping whitespaces, insignificant | |
1065 | # symbols, and comments when allowed, and stopping at the first | |
1066 | # unknown character. | |
1067 | def _parse_items(self) -> List[_Item]: | |
1068 | items = [] # type: List[_Item] | |
1069 | ||
1070 | while self._isnt_done(): | |
1ca7b5e1 PP |
1071 | # Try to append item |
1072 | if not self._try_append_item(items): | |
1073 | # Unknown at this point | |
1074 | break | |
71aaa3f7 PP |
1075 | |
1076 | return items | |
1077 | ||
1078 | # Parses the whole Normand input, setting `self._res` to the main | |
1079 | # group item on success. | |
1080 | def _parse(self): | |
1081 | if len(self._normand.strip()) == 0: | |
1082 | # Special case to make sure there's something to consume | |
1083 | self._res = _Group([], self._text_loc) | |
1084 | return | |
1085 | ||
1086 | # Parse first level items | |
1087 | items = self._parse_items() | |
1088 | ||
1089 | # Make sure there's nothing left | |
1090 | self._skip_ws_and_comments() | |
1091 | ||
1092 | if self._isnt_done(): | |
1093 | self._raise_error( | |
1094 | "Unexpected character `{}`".format(self._normand[self._at]) | |
1095 | ) | |
1096 | ||
1097 | # Set main group item | |
1098 | self._res = _Group(items, self._text_loc) | |
1099 | ||
1100 | ||
1101 | # The return type of parse(). | |
1102 | class ParseResult: | |
1103 | @classmethod | |
1104 | def _create( | |
1105 | cls, | |
1106 | data: bytearray, | |
1b8aa84a PP |
1107 | variables: SymbolsT, |
1108 | labels: SymbolsT, | |
71aaa3f7 PP |
1109 | offset: int, |
1110 | bo: Optional[ByteOrder], | |
1111 | ): | |
1112 | self = cls.__new__(cls) | |
1113 | self._init(data, variables, labels, offset, bo) | |
1114 | return self | |
1115 | ||
1116 | def __init__(self, *args, **kwargs): # type: ignore | |
1117 | raise NotImplementedError | |
1118 | ||
1119 | def _init( | |
1120 | self, | |
1121 | data: bytearray, | |
1b8aa84a PP |
1122 | variables: SymbolsT, |
1123 | labels: SymbolsT, | |
71aaa3f7 PP |
1124 | offset: int, |
1125 | bo: Optional[ByteOrder], | |
1126 | ): | |
1127 | self._data = data | |
1128 | self._vars = variables | |
1129 | self._labels = labels | |
1130 | self._offset = offset | |
1131 | self._bo = bo | |
1132 | ||
1133 | # Generated data. | |
1134 | @property | |
1135 | def data(self): | |
1136 | return self._data | |
1137 | ||
1138 | # Dictionary of updated variable names to their last computed value. | |
1139 | @property | |
1140 | def variables(self): | |
1141 | return self._vars | |
1142 | ||
1143 | # Dictionary of updated main group label names to their computed | |
1144 | # value. | |
1145 | @property | |
1146 | def labels(self): | |
1147 | return self._labels | |
1148 | ||
1149 | # Updated offset. | |
1150 | @property | |
1151 | def offset(self): | |
1152 | return self._offset | |
1153 | ||
1154 | # Updated byte order. | |
1155 | @property | |
1156 | def byte_order(self): | |
1157 | return self._bo | |
1158 | ||
1159 | ||
1160 | # Raises a parse error for the item `item`, creating it using the | |
1161 | # message `msg`. | |
1162 | def _raise_error_for_item(msg: str, item: _Item) -> NoReturn: | |
1163 | _raise_error(msg, item.text_loc) | |
1164 | ||
1165 | ||
1166 | # The `ICITTE` reserved name. | |
1167 | _icitte_name = "ICITTE" | |
1168 | ||
1169 | ||
2adf4336 PP |
1170 | # Base node visitor. |
1171 | # | |
1172 | # Calls the _visit_name() method for each name node which isn't the name | |
1173 | # of a call. | |
1174 | class _NodeVisitor(ast.NodeVisitor): | |
1175 | def __init__(self): | |
71aaa3f7 PP |
1176 | self._parent_is_call = False |
1177 | ||
1178 | def generic_visit(self, node: ast.AST): | |
1179 | if type(node) is ast.Call: | |
1180 | self._parent_is_call = True | |
1181 | elif type(node) is ast.Name and not self._parent_is_call: | |
2adf4336 | 1182 | self._visit_name(node.id) |
71aaa3f7 PP |
1183 | |
1184 | super().generic_visit(node) | |
1185 | self._parent_is_call = False | |
1186 | ||
2adf4336 PP |
1187 | @abc.abstractmethod |
1188 | def _visit_name(self, name: str): | |
1189 | ... | |
1190 | ||
71aaa3f7 | 1191 | |
2adf4336 PP |
1192 | # Expression validator: validates that all the names within the |
1193 | # expression are allowed. | |
1194 | class _ExprValidator(_NodeVisitor): | |
1195 | def __init__(self, item: _ExprItemT, allowed_names: Set[str], icitte_allowed: bool): | |
1196 | super().__init__() | |
1197 | self._item = item | |
1198 | self._allowed_names = allowed_names | |
1199 | self._icitte_allowed = icitte_allowed | |
1200 | ||
1201 | def _visit_name(self, name: str): | |
1202 | # Make sure the name refers to a known and reachable | |
1203 | # variable/label name. | |
1204 | if name == _icitte_name and not self._icitte_allowed: | |
1205 | _raise_error( | |
1206 | "Illegal reserved name `{}` in expression `{}`".format( | |
1207 | _icitte_name, self._item.expr_str | |
1208 | ), | |
1209 | self._item.text_loc, | |
1210 | ) | |
1211 | elif name != _icitte_name and name not in self._allowed_names: | |
1212 | msg = "Illegal (unknown or unreachable) variable/label name `{}` in expression `{}`".format( | |
1213 | name, self._item.expr_str | |
1214 | ) | |
1215 | ||
05f81895 | 1216 | allowed_names = self._allowed_names.copy() |
2adf4336 | 1217 | |
05f81895 PP |
1218 | if self._icitte_allowed: |
1219 | allowed_names.add(_icitte_name) | |
2adf4336 | 1220 | |
05f81895 | 1221 | if len(allowed_names) > 0: |
2adf4336 PP |
1222 | allowed_names_str = ", ".join( |
1223 | sorted(["`{}`".format(name) for name in allowed_names]) | |
1224 | ) | |
1225 | msg += "; the legal names are {{{}}}".format(allowed_names_str) | |
1226 | ||
1227 | _raise_error( | |
1228 | msg, | |
1229 | self._item.text_loc, | |
1230 | ) | |
1231 | ||
1232 | ||
1233 | # Expression visitor getting all the contained names. | |
1234 | class _ExprNamesVisitor(_NodeVisitor): | |
71aaa3f7 | 1235 | def __init__(self): |
2adf4336 PP |
1236 | self._parent_is_call = False |
1237 | self._names = set() # type: Set[str] | |
1238 | ||
1239 | @property | |
1240 | def names(self): | |
1241 | return self._names | |
71aaa3f7 | 1242 | |
2adf4336 PP |
1243 | def _visit_name(self, name: str): |
1244 | self._names.add(name) | |
71aaa3f7 | 1245 | |
71aaa3f7 | 1246 | |
2adf4336 PP |
1247 | # Generator state. |
1248 | class _GenState: | |
1249 | def __init__( | |
1b8aa84a PP |
1250 | self, |
1251 | variables: SymbolsT, | |
1252 | labels: SymbolsT, | |
1253 | offset: int, | |
1254 | bo: Optional[ByteOrder], | |
2adf4336 PP |
1255 | ): |
1256 | self.variables = variables.copy() | |
1257 | self.labels = labels.copy() | |
1258 | self.offset = offset | |
1259 | self.bo = bo | |
71aaa3f7 PP |
1260 | |
1261 | ||
2adf4336 | 1262 | # Generator of data and final state from a group item. |
71aaa3f7 PP |
1263 | # |
1264 | # Generation happens in memory at construction time. After building, use | |
1265 | # the `data`, `variables`, `labels`, `offset`, and `bo` properties to | |
1266 | # get the resulting context. | |
2adf4336 PP |
1267 | # |
1268 | # The steps of generation are: | |
1269 | # | |
05f81895 PP |
1270 | # 1. Validate that each repetition and LEB128 integer expression uses |
1271 | # only reachable names and not `ICITTE`. | |
2adf4336 | 1272 | # |
05f81895 PP |
1273 | # 2. Compute and keep the effective repetition count and LEB128 integer |
1274 | # value for each repetition and LEB128 integer instance. | |
2adf4336 PP |
1275 | # |
1276 | # 3. Generate bytes, updating the initial state as it goes which becomes | |
1277 | # the final state after the operation. | |
1278 | # | |
05f81895 PP |
1279 | # During the generation, when handling a `_Rep` or `_Leb128Int` item, |
1280 | # we already have the effective repetition count or value of the | |
1281 | # instance. | |
2adf4336 PP |
1282 | # |
1283 | # When handling a `_Group` item, first update the current labels with | |
1284 | # all the immediate (not nested) labels, and then handle each | |
1285 | # contained item. This gives contained item access to "future" outer | |
1286 | # labels. Then remove the immediate labels from the state so that | |
1287 | # outer items don't have access to inner labels. | |
71aaa3f7 PP |
1288 | class _Gen: |
1289 | def __init__( | |
1290 | self, | |
1291 | group: _Group, | |
1b8aa84a PP |
1292 | variables: SymbolsT, |
1293 | labels: SymbolsT, | |
71aaa3f7 PP |
1294 | offset: int, |
1295 | bo: Optional[ByteOrder], | |
1296 | ): | |
05f81895 PP |
1297 | self._validate_vl_exprs(group, set(variables.keys()), set(labels.keys())) |
1298 | self._vl_instance_vals = self._compute_vl_instance_vals( | |
2adf4336 PP |
1299 | group, _GenState(variables, labels, offset, bo) |
1300 | ) | |
1301 | self._gen(group, _GenState(variables, labels, offset, bo)) | |
71aaa3f7 PP |
1302 | |
1303 | # Generated bytes. | |
1304 | @property | |
1305 | def data(self): | |
1306 | return self._data | |
1307 | ||
1308 | # Updated variables. | |
1309 | @property | |
1310 | def variables(self): | |
2adf4336 | 1311 | return self._final_state.variables |
71aaa3f7 PP |
1312 | |
1313 | # Updated main group labels. | |
1314 | @property | |
1315 | def labels(self): | |
2adf4336 | 1316 | return self._final_state.labels |
71aaa3f7 PP |
1317 | |
1318 | # Updated offset. | |
1319 | @property | |
1320 | def offset(self): | |
2adf4336 | 1321 | return self._final_state.offset |
71aaa3f7 PP |
1322 | |
1323 | # Updated byte order. | |
1324 | @property | |
1325 | def bo(self): | |
2adf4336 PP |
1326 | return self._final_state.bo |
1327 | ||
1328 | # Returns the set of used, non-called names within the AST | |
1329 | # expression `expr`. | |
1330 | @staticmethod | |
1331 | def _names_of_expr(expr: ast.Expression): | |
1332 | visitor = _ExprNamesVisitor() | |
1333 | visitor.visit(expr) | |
1334 | return visitor.names | |
1335 | ||
05f81895 PP |
1336 | # Validates that all the repetition and LEB128 integer expressions |
1337 | # within `group` don't refer, directly or indirectly, to subsequent | |
1338 | # labels. | |
71aaa3f7 | 1339 | # |
2adf4336 PP |
1340 | # The strategy here is to keep a set of allowed label names, per |
1341 | # group, initialized to `allowed_label_names`, and a set of allowed | |
1342 | # variable names initialized to `allowed_variable_names`. | |
1343 | # | |
1344 | # Then, depending on the type of `item`: | |
1345 | # | |
1346 | # `_Label`: | |
1347 | # Add its name to the local allowed label names: a label | |
1348 | # occurring before a repetition, and not within a nested group, | |
1349 | # is always reachable. | |
1350 | # | |
1351 | # `_VarAssign`: | |
1352 | # If all the names within its expression are allowed, then add | |
1353 | # its name to the allowed variable names. | |
1354 | # | |
1355 | # Otherwise, remove its name from the allowed variable names (if | |
1356 | # it's in there): a variable which refers to an unreachable name | |
1357 | # is unreachable itself. | |
1358 | # | |
05f81895 | 1359 | # `_Rep` and `_Leb128`: |
2adf4336 PP |
1360 | # Make sure all the names within its expression are allowed. |
1361 | # | |
1362 | # `_Group`: | |
1363 | # Call this function for each contained item with a _copy_ of | |
1364 | # the current allowed label names and the same current allowed | |
1365 | # variable names. | |
1366 | @staticmethod | |
05f81895 | 1367 | def _validate_vl_exprs( |
2adf4336 PP |
1368 | item: _Item, allowed_variable_names: Set[str], allowed_label_names: Set[str] |
1369 | ): | |
1370 | if type(item) is _Label: | |
1371 | allowed_label_names.add(item.name) | |
1372 | elif type(item) is _VarAssign: | |
1373 | # Check if this variable name is allowed | |
1374 | allowed = True | |
1375 | ||
1376 | for name in _Gen._names_of_expr(item.expr): | |
1377 | if name not in ( | |
1378 | allowed_label_names | allowed_variable_names | {_icitte_name} | |
1379 | ): | |
1380 | # Not allowed | |
1381 | allowed = False | |
1382 | break | |
1383 | ||
1384 | if allowed: | |
1385 | allowed_variable_names.add(item.name) | |
1386 | elif item.name in allowed_variable_names: | |
1387 | allowed_variable_names.remove(item.name) | |
05f81895 PP |
1388 | elif isinstance(item, _Leb128Int): |
1389 | # Validate the expression (`ICITTE` allowed) | |
1390 | _ExprValidator( | |
1391 | item, allowed_label_names | allowed_variable_names, True | |
1392 | ).visit(item.expr) | |
71aaa3f7 | 1393 | elif type(item) is _Rep: |
05f81895 | 1394 | # Validate the expression first (`ICITTE` not allowed) |
2adf4336 PP |
1395 | _ExprValidator( |
1396 | item, allowed_label_names | allowed_variable_names, False | |
1397 | ).visit(item.expr) | |
1398 | ||
1399 | # Validate inner item | |
05f81895 | 1400 | _Gen._validate_vl_exprs( |
2adf4336 PP |
1401 | item.item, allowed_variable_names, allowed_label_names |
1402 | ) | |
1403 | elif type(item) is _Group: | |
1404 | # Copy `allowed_label_names` so that this frame cannot | |
1405 | # access the nested label names. | |
1406 | group_allowed_label_names = allowed_label_names.copy() | |
71aaa3f7 | 1407 | |
2adf4336 | 1408 | for subitem in item.items: |
05f81895 | 1409 | _Gen._validate_vl_exprs( |
2adf4336 PP |
1410 | subitem, allowed_variable_names, group_allowed_label_names |
1411 | ) | |
71aaa3f7 | 1412 | |
2adf4336 PP |
1413 | # Evaluates the expression of `item` considering the current |
1414 | # generation state `state`. | |
1415 | # | |
1416 | # If `allow_icitte` is `True`, then the `ICITTE` name is available | |
1417 | # for the expression to evaluate. | |
269f6eb3 PP |
1418 | # |
1419 | # If `allow_float` is `True`, then the type of the result may be | |
1420 | # `float` too. | |
2adf4336 | 1421 | @staticmethod |
269f6eb3 PP |
1422 | def _eval_item_expr( |
1423 | item: _ExprItemT, | |
1424 | state: _GenState, | |
1425 | allow_icitte: bool, | |
1426 | allow_float: bool = False, | |
1427 | ): | |
2adf4336 | 1428 | syms = state.labels.copy() |
71aaa3f7 | 1429 | |
2adf4336 PP |
1430 | # Set the `ICITTE` name to the current offset, if any |
1431 | if allow_icitte: | |
1432 | syms[_icitte_name] = state.offset | |
71aaa3f7 PP |
1433 | |
1434 | # Add the current variables | |
2adf4336 | 1435 | syms.update(state.variables) |
71aaa3f7 PP |
1436 | |
1437 | # Validate the node and its children | |
2adf4336 | 1438 | _ExprValidator(item, set(syms.keys()), True).visit(item.expr) |
71aaa3f7 PP |
1439 | |
1440 | # Compile and evaluate expression node | |
1441 | try: | |
1442 | val = eval(compile(item.expr, "", "eval"), None, syms) | |
1443 | except Exception as exc: | |
1444 | _raise_error_for_item( | |
1445 | "Failed to evaluate expression `{}`: {}".format(item.expr_str, exc), | |
1446 | item, | |
1447 | ) | |
1448 | ||
269f6eb3 PP |
1449 | # Validate result type |
1450 | expected_types = {int} # type: Set[type] | |
1451 | type_msg = "`int`" | |
1452 | ||
1453 | if allow_float: | |
1454 | expected_types.add(float) | |
1455 | type_msg += " or `float`" | |
1456 | ||
1457 | if type(val) not in expected_types: | |
71aaa3f7 | 1458 | _raise_error_for_item( |
269f6eb3 PP |
1459 | "Invalid expression `{}`: expecting result type {}, not `{}`".format( |
1460 | item.expr_str, type_msg, type(val).__name__ | |
71aaa3f7 PP |
1461 | ), |
1462 | item, | |
1463 | ) | |
1464 | ||
1465 | return val | |
1466 | ||
05f81895 PP |
1467 | # Returns the size, in bytes, required to encode the value `val` |
1468 | # with LEB128 (signed version if `is_signed` is `True`). | |
1469 | @staticmethod | |
1470 | def _leb128_size_for_val(val: int, is_signed: bool): | |
1471 | if val < 0: | |
1472 | # Equivalent upper bound. | |
1473 | # | |
1474 | # For example, if `val` is -128, then the full integer for | |
1475 | # this number of bits would be [-128, 127]. | |
1476 | val = -val - 1 | |
1477 | ||
1478 | # Number of bits (add one for the sign if needed) | |
1479 | bits = val.bit_length() + int(is_signed) | |
1480 | ||
1481 | if bits == 0: | |
1482 | bits = 1 | |
1483 | ||
1484 | # Seven bits per byte | |
1485 | return math.ceil(bits / 7) | |
1486 | ||
676f6189 PP |
1487 | # Returns the offset `offset` aligned according to `item`. |
1488 | @staticmethod | |
1489 | def _align_offset(offset: int, item: _AlignOffset): | |
1490 | align_bytes = item.val // 8 | |
1491 | return (offset + align_bytes - 1) // align_bytes * align_bytes | |
1492 | ||
05f81895 PP |
1493 | # Computes the effective value for each repetition and LEB128 |
1494 | # integer instance, filling `instance_vals` (if not `None`) and | |
1495 | # returning `instance_vals`. | |
2adf4336 | 1496 | # |
05f81895 PP |
1497 | # At this point it must be known that, for a given variable-length |
1498 | # item, its expression only contains reachable names. | |
2adf4336 PP |
1499 | # |
1500 | # When handling a `_Rep` item, this function appends its effective | |
1501 | # multiplier to `instance_vals` _before_ handling its repeated item. | |
1502 | # | |
05f81895 PP |
1503 | # When handling a `_VarAssign` item, this function only evaluates it |
1504 | # if all its names are reachable. | |
2adf4336 | 1505 | @staticmethod |
05f81895 | 1506 | def _compute_vl_instance_vals( |
2adf4336 PP |
1507 | item: _Item, state: _GenState, instance_vals: Optional[List[int]] = None |
1508 | ): | |
1509 | if instance_vals is None: | |
1510 | instance_vals = [] | |
1511 | ||
1512 | if isinstance(item, _ScalarItem): | |
1513 | state.offset += item.size | |
1514 | elif type(item) is _Label: | |
1515 | state.labels[item.name] = state.offset | |
1516 | elif type(item) is _VarAssign: | |
1517 | # Check if all the names are reachable | |
1518 | do_eval = True | |
1519 | ||
1520 | for name in _Gen._names_of_expr(item.expr): | |
1521 | if ( | |
1522 | name != _icitte_name | |
1523 | and name not in state.variables | |
1524 | and name not in state.labels | |
1525 | ): | |
1526 | # A name is unknown: cannot evaluate | |
1527 | do_eval = False | |
1528 | break | |
1529 | ||
1530 | if do_eval: | |
1531 | # Evaluate the expression and keep the result | |
269f6eb3 PP |
1532 | state.variables[item.name] = _Gen._eval_item_expr( |
1533 | item, state, True, True | |
1534 | ) | |
2adf4336 PP |
1535 | elif type(item) is _SetOffset: |
1536 | state.offset = item.val | |
676f6189 PP |
1537 | elif type(item) is _AlignOffset: |
1538 | state.offset = _Gen._align_offset(state.offset, item) | |
05f81895 PP |
1539 | elif isinstance(item, _Leb128Int): |
1540 | # Evaluate the expression | |
1541 | val = _Gen._eval_item_expr(item, state, True) | |
1542 | ||
1543 | # Validate result | |
1544 | if type(item) is _ULeb128Int and val < 0: | |
1545 | _raise_error_for_item( | |
1546 | "Invalid expression `{}`: unexpected negative result {:,} for a ULEB128 encoding".format( | |
1547 | item.expr_str, val | |
1548 | ), | |
1549 | item, | |
1550 | ) | |
1551 | ||
1552 | # Add the evaluation result to the to variable-length item | |
1553 | # instance values. | |
1554 | instance_vals.append(val) | |
1555 | ||
1556 | # Update offset | |
1557 | state.offset += _Gen._leb128_size_for_val(val, type(item) is _SLeb128Int) | |
2adf4336 PP |
1558 | elif type(item) is _Rep: |
1559 | # Evaluate the expression and keep the result | |
1560 | val = _Gen._eval_item_expr(item, state, False) | |
1561 | ||
1562 | # Validate result | |
1563 | if val < 0: | |
1564 | _raise_error_for_item( | |
1565 | "Invalid expression `{}`: unexpected negative result {:,}".format( | |
1566 | item.expr_str, val | |
1567 | ), | |
1568 | item, | |
1569 | ) | |
1570 | ||
1571 | # Add to repetition instance values | |
1572 | instance_vals.append(val) | |
1573 | ||
1574 | # Process the repeated item `val` times | |
1575 | for _ in range(val): | |
05f81895 | 1576 | _Gen._compute_vl_instance_vals(item.item, state, instance_vals) |
2adf4336 PP |
1577 | elif type(item) is _Group: |
1578 | prev_labels = state.labels.copy() | |
1579 | ||
1580 | # Process each item | |
1581 | for subitem in item.items: | |
05f81895 | 1582 | _Gen._compute_vl_instance_vals(subitem, state, instance_vals) |
2adf4336 PP |
1583 | |
1584 | state.labels = prev_labels | |
1585 | ||
1586 | return instance_vals | |
1587 | ||
676f6189 PP |
1588 | def _update_offset_noop(self, item: _Item, state: _GenState, next_vl_instance: int): |
1589 | return next_vl_instance | |
05f81895 | 1590 | |
676f6189 PP |
1591 | def _dry_handle_scalar_item( |
1592 | self, item: _ScalarItem, state: _GenState, next_vl_instance: int | |
1593 | ): | |
1594 | state.offset += item.size | |
1595 | return next_vl_instance | |
2adf4336 | 1596 | |
676f6189 PP |
1597 | def _dry_handle_leb128_int_item( |
1598 | self, item: _Leb128Int, state: _GenState, next_vl_instance: int | |
1599 | ): | |
05f81895 PP |
1600 | # Get the value from `self._vl_instance_vals` _before_ |
1601 | # incrementing `next_vl_instance` to honor the order of | |
1602 | # _compute_vl_instance_vals(). | |
676f6189 PP |
1603 | state.offset += self._leb128_size_for_val( |
1604 | self._vl_instance_vals[next_vl_instance], type(item) is _SLeb128Int | |
05f81895 | 1605 | ) |
2adf4336 | 1606 | |
676f6189 | 1607 | return next_vl_instance + 1 |
2adf4336 | 1608 | |
676f6189 PP |
1609 | def _dry_handle_group_item( |
1610 | self, item: _Group, state: _GenState, next_vl_instance: int | |
1611 | ): | |
2adf4336 | 1612 | for subitem in item.items: |
676f6189 | 1613 | next_vl_instance = self._dry_handle_item(subitem, state, next_vl_instance) |
2adf4336 | 1614 | |
676f6189 | 1615 | return next_vl_instance |
2adf4336 | 1616 | |
676f6189 | 1617 | def _dry_handle_rep_item(self, item: _Rep, state: _GenState, next_vl_instance: int): |
05f81895 PP |
1618 | # Get the value from `self._vl_instance_vals` _before_ |
1619 | # incrementing `next_vl_instance` to honor the order of | |
1620 | # _compute_vl_instance_vals(). | |
1621 | mul = self._vl_instance_vals[next_vl_instance] | |
1622 | next_vl_instance += 1 | |
2adf4336 PP |
1623 | |
1624 | for _ in range(mul): | |
676f6189 | 1625 | next_vl_instance = self._dry_handle_item(item.item, state, next_vl_instance) |
2adf4336 | 1626 | |
676f6189 | 1627 | return next_vl_instance |
2adf4336 | 1628 | |
676f6189 PP |
1629 | def _dry_handle_align_offset_item( |
1630 | self, item: _AlignOffset, state: _GenState, next_vl_instance: int | |
1631 | ): | |
1632 | state.offset = self._align_offset(state.offset, item) | |
1633 | return next_vl_instance | |
1634 | ||
1635 | def _dry_handle_set_offset_item( | |
1636 | self, item: _SetOffset, state: _GenState, next_vl_instance: int | |
1637 | ): | |
1638 | state.offset = item.val | |
1639 | return next_vl_instance | |
1640 | ||
1641 | # Updates `state.offset` considering the generated data of `item`, | |
1642 | # without generating any, and returns the updated next | |
1643 | # variable-length item instance. | |
1644 | def _dry_handle_item(self, item: _Item, state: _GenState, next_vl_instance: int): | |
1645 | return self._dry_handle_item_funcs[type(item)](item, state, next_vl_instance) | |
2adf4336 PP |
1646 | |
1647 | # Handles the byte item `item`. | |
05f81895 | 1648 | def _handle_byte_item(self, item: _Byte, state: _GenState, next_vl_instance: int): |
2adf4336 PP |
1649 | self._data.append(item.val) |
1650 | state.offset += item.size | |
05f81895 | 1651 | return next_vl_instance |
2adf4336 PP |
1652 | |
1653 | # Handles the string item `item`. | |
05f81895 | 1654 | def _handle_str_item(self, item: _Str, state: _GenState, next_vl_instance: int): |
2adf4336 PP |
1655 | self._data += item.data |
1656 | state.offset += item.size | |
05f81895 | 1657 | return next_vl_instance |
2adf4336 PP |
1658 | |
1659 | # Handles the byte order setting item `item`. | |
1660 | def _handle_set_bo_item( | |
05f81895 | 1661 | self, item: _SetBo, state: _GenState, next_vl_instance: int |
2adf4336 PP |
1662 | ): |
1663 | # Update current byte order | |
1664 | state.bo = item.bo | |
05f81895 | 1665 | return next_vl_instance |
2adf4336 PP |
1666 | |
1667 | # Handles the variable assignment item `item`. | |
1668 | def _handle_var_assign_item( | |
05f81895 | 1669 | self, item: _VarAssign, state: _GenState, next_vl_instance: int |
2adf4336 | 1670 | ): |
71aaa3f7 | 1671 | # Update variable |
f04a3048 | 1672 | state.variables[item.name] = self._eval_item_expr(item, state, True, True) |
05f81895 | 1673 | return next_vl_instance |
71aaa3f7 | 1674 | |
05f81895 | 1675 | # Handles the fixed-length integer item `item`. |
269f6eb3 | 1676 | def _handle_fl_int_item(self, val: int, item: _FlNum, state: _GenState): |
71aaa3f7 PP |
1677 | # Validate range |
1678 | if val < -(2 ** (item.len - 1)) or val > 2**item.len - 1: | |
1679 | _raise_error_for_item( | |
1680 | "Value {:,} is outside the {}-bit range when evaluating expression `{}` at byte offset {:,}".format( | |
2adf4336 | 1681 | val, item.len, item.expr_str, state.offset |
71aaa3f7 PP |
1682 | ), |
1683 | item, | |
1684 | ) | |
1685 | ||
1686 | # Encode result on 64 bits (to extend the sign bit whatever the | |
1687 | # value of `item.len`). | |
71aaa3f7 PP |
1688 | data = struct.pack( |
1689 | "{}{}".format( | |
2adf4336 | 1690 | ">" if state.bo in (None, ByteOrder.BE) else "<", |
71aaa3f7 PP |
1691 | "Q" if val >= 0 else "q", |
1692 | ), | |
1693 | val, | |
1694 | ) | |
1695 | ||
1696 | # Keep only the requested length | |
1697 | len_bytes = item.len // 8 | |
1698 | ||
2adf4336 | 1699 | if state.bo in (None, ByteOrder.BE): |
71aaa3f7 PP |
1700 | # Big endian: keep last bytes |
1701 | data = data[-len_bytes:] | |
1702 | else: | |
1703 | # Little endian: keep first bytes | |
2adf4336 | 1704 | assert state.bo == ByteOrder.LE |
71aaa3f7 PP |
1705 | data = data[:len_bytes] |
1706 | ||
1707 | # Append to current bytes and update offset | |
1708 | self._data += data | |
269f6eb3 PP |
1709 | |
1710 | # Handles the fixed-length integer item `item`. | |
1711 | def _handle_fl_float_item(self, val: float, item: _FlNum, state: _GenState): | |
1712 | # Validate length | |
1713 | if item.len not in (32, 64): | |
1714 | _raise_error_for_item( | |
1715 | "Invalid {}-bit length for a fixed-length floating point number (value {:,})".format( | |
1716 | item.len, val | |
1717 | ), | |
1718 | item, | |
1719 | ) | |
1720 | ||
1721 | # Encode result | |
1722 | self._data += struct.pack( | |
1723 | "{}{}".format( | |
1724 | ">" if state.bo in (None, ByteOrder.BE) else "<", | |
1725 | "f" if item.len == 32 else "d", | |
1726 | ), | |
1727 | val, | |
1728 | ) | |
1729 | ||
1730 | # Handles the fixed-length number item `item`. | |
1731 | def _handle_fl_num_item( | |
1732 | self, item: _FlNum, state: _GenState, next_vl_instance: int | |
1733 | ): | |
1734 | # Compute value | |
1735 | val = self._eval_item_expr(item, state, True, True) | |
1736 | ||
1737 | # Validate current byte order | |
1738 | if state.bo is None and item.len > 8: | |
1739 | _raise_error_for_item( | |
1740 | "Current byte order isn't defined at first fixed-length number (`{}`) to encode on more than 8 bits".format( | |
1741 | item.expr_str | |
1742 | ), | |
1743 | item, | |
1744 | ) | |
1745 | ||
1746 | # Handle depending on type | |
1747 | if type(val) is int: | |
1748 | self._handle_fl_int_item(val, item, state) | |
1749 | else: | |
1750 | assert type(val) is float | |
1751 | self._handle_fl_float_item(val, item, state) | |
1752 | ||
1753 | # Update offset | |
1754 | state.offset += item.size | |
1755 | ||
05f81895 PP |
1756 | return next_vl_instance |
1757 | ||
1758 | # Handles the LEB128 integer item `item`. | |
1759 | def _handle_leb128_int_item( | |
1760 | self, item: _Leb128Int, state: _GenState, next_vl_instance: int | |
1761 | ): | |
1762 | # Get the precomputed value | |
1763 | val = self._vl_instance_vals[next_vl_instance] | |
1764 | ||
1765 | # Size in bytes | |
1766 | size = self._leb128_size_for_val(val, type(item) is _SLeb128Int) | |
1767 | ||
1768 | # For each byte | |
1769 | for _ in range(size): | |
1770 | # Seven LSBs, MSB of the byte set (continue) | |
1771 | self._data.append((val & 0x7F) | 0x80) | |
1772 | val >>= 7 | |
1773 | ||
1774 | # Clear MSB of last byte (stop) | |
1775 | self._data[-1] &= ~0x80 | |
1776 | ||
1777 | # Consumed this instance | |
1778 | return next_vl_instance + 1 | |
71aaa3f7 | 1779 | |
2adf4336 PP |
1780 | # Handles the group item `item`, only removing the immediate labels |
1781 | # from `state.labels` if `remove_immediate_labels` is `True`. | |
1782 | def _handle_group_item( | |
1783 | self, | |
1784 | item: _Group, | |
1785 | state: _GenState, | |
05f81895 | 1786 | next_vl_instance: int, |
2adf4336 PP |
1787 | remove_immediate_labels: bool = True, |
1788 | ): | |
1789 | # Compute the values of the immediate (not nested) labels. Those | |
1790 | # labels are reachable by any expression within the group. | |
676f6189 | 1791 | tmp_state = _GenState({}, {}, state.offset, None) |
2adf4336 | 1792 | immediate_label_names = set() # type: Set[str] |
05f81895 | 1793 | tmp_next_vl_instance = next_vl_instance |
71aaa3f7 | 1794 | |
2adf4336 | 1795 | for subitem in item.items: |
676f6189 | 1796 | if type(subitem) is _Label: |
2adf4336 | 1797 | # New immediate label |
676f6189 | 1798 | state.labels[subitem.name] = tmp_state.offset |
2adf4336 PP |
1799 | immediate_label_names.add(subitem.name) |
1800 | ||
676f6189 PP |
1801 | tmp_next_vl_instance = self._dry_handle_item( |
1802 | subitem, tmp_state, tmp_next_vl_instance | |
2adf4336 | 1803 | ) |
71aaa3f7 | 1804 | |
2adf4336 | 1805 | # Handle each item now with the actual state |
71aaa3f7 | 1806 | for subitem in item.items: |
05f81895 | 1807 | next_vl_instance = self._handle_item(subitem, state, next_vl_instance) |
2adf4336 PP |
1808 | |
1809 | # Remove immediate labels if required so that outer items won't | |
1810 | # reach inner labels. | |
1811 | if remove_immediate_labels: | |
1812 | for name in immediate_label_names: | |
1813 | del state.labels[name] | |
71aaa3f7 | 1814 | |
05f81895 | 1815 | return next_vl_instance |
71aaa3f7 | 1816 | |
2adf4336 | 1817 | # Handles the repetition item `item`. |
05f81895 PP |
1818 | def _handle_rep_item(self, item: _Rep, state: _GenState, next_vl_instance: int): |
1819 | # Get the precomputed repetition count | |
1820 | mul = self._vl_instance_vals[next_vl_instance] | |
1821 | ||
1822 | # Consumed this instance | |
1823 | next_vl_instance += 1 | |
71aaa3f7 | 1824 | |
2adf4336 | 1825 | for _ in range(mul): |
05f81895 | 1826 | next_vl_instance = self._handle_item(item.item, state, next_vl_instance) |
71aaa3f7 | 1827 | |
05f81895 | 1828 | return next_vl_instance |
71aaa3f7 | 1829 | |
2adf4336 PP |
1830 | # Handles the offset setting item `item`. |
1831 | def _handle_set_offset_item( | |
05f81895 | 1832 | self, item: _SetOffset, state: _GenState, next_vl_instance: int |
2adf4336 PP |
1833 | ): |
1834 | state.offset = item.val | |
05f81895 | 1835 | return next_vl_instance |
2adf4336 | 1836 | |
676f6189 PP |
1837 | # Handles offset alignment item `item` (adds padding). |
1838 | def _handle_align_offset_item( | |
1839 | self, item: _AlignOffset, state: _GenState, next_vl_instance: int | |
1840 | ): | |
1841 | init_offset = state.offset | |
1842 | state.offset = self._align_offset(state.offset, item) | |
1843 | self._data += bytes([item.pad_val] * (state.offset - init_offset)) | |
1844 | return next_vl_instance | |
1845 | ||
2adf4336 | 1846 | # Handles the label item `item`. |
05f81895 PP |
1847 | def _handle_label_item(self, item: _Label, state: _GenState, next_vl_instance: int): |
1848 | return next_vl_instance | |
2adf4336 PP |
1849 | |
1850 | # Handles the item `item`, returning the updated next repetition | |
1851 | # instance. | |
05f81895 PP |
1852 | def _handle_item(self, item: _Item, state: _GenState, next_vl_instance: int): |
1853 | return self._item_handlers[type(item)](item, state, next_vl_instance) | |
2adf4336 PP |
1854 | |
1855 | # Generates the data (`self._data`) and final state | |
1856 | # (`self._final_state`) from `group` and the initial state `state`. | |
1857 | def _gen(self, group: _Group, state: _GenState): | |
71aaa3f7 PP |
1858 | # Initial state |
1859 | self._data = bytearray() | |
71aaa3f7 PP |
1860 | |
1861 | # Item handlers | |
1862 | self._item_handlers = { | |
676f6189 | 1863 | _AlignOffset: self._handle_align_offset_item, |
71aaa3f7 | 1864 | _Byte: self._handle_byte_item, |
269f6eb3 | 1865 | _FlNum: self._handle_fl_num_item, |
71aaa3f7 | 1866 | _Group: self._handle_group_item, |
2adf4336 | 1867 | _Label: self._handle_label_item, |
71aaa3f7 | 1868 | _Rep: self._handle_rep_item, |
2adf4336 PP |
1869 | _SetBo: self._handle_set_bo_item, |
1870 | _SetOffset: self._handle_set_offset_item, | |
05f81895 | 1871 | _SLeb128Int: self._handle_leb128_int_item, |
2adf4336 | 1872 | _Str: self._handle_str_item, |
05f81895 | 1873 | _ULeb128Int: self._handle_leb128_int_item, |
2adf4336 PP |
1874 | _VarAssign: self._handle_var_assign_item, |
1875 | } # type: Dict[type, Callable[[Any, _GenState, int], int]] | |
1876 | ||
676f6189 PP |
1877 | # Dry item handlers (only updates the state offset) |
1878 | self._dry_handle_item_funcs = { | |
1879 | _AlignOffset: self._dry_handle_align_offset_item, | |
1880 | _Byte: self._dry_handle_scalar_item, | |
1881 | _FlNum: self._dry_handle_scalar_item, | |
1882 | _Group: self._dry_handle_group_item, | |
1883 | _Label: self._update_offset_noop, | |
1884 | _Rep: self._dry_handle_rep_item, | |
1885 | _SetBo: self._update_offset_noop, | |
1886 | _SetOffset: self._dry_handle_set_offset_item, | |
1887 | _SLeb128Int: self._dry_handle_leb128_int_item, | |
1888 | _Str: self._dry_handle_scalar_item, | |
1889 | _ULeb128Int: self._dry_handle_leb128_int_item, | |
1890 | _VarAssign: self._update_offset_noop, | |
1891 | } # type: Dict[type, Callable[[Any, _GenState, int], int]] | |
2adf4336 PP |
1892 | |
1893 | # Handle the group item, _not_ removing the immediate labels | |
1894 | # because the `labels` property offers them. | |
1895 | self._handle_group_item(group, state, 0, False) | |
1896 | ||
1897 | # This is actually the final state | |
1898 | self._final_state = state | |
71aaa3f7 PP |
1899 | |
1900 | ||
1901 | # Returns a `ParseResult` instance containing the bytes encoded by the | |
1902 | # input string `normand`. | |
1903 | # | |
1904 | # `init_variables` is a dictionary of initial variable names (valid | |
1905 | # Python names) to integral values. A variable name must not be the | |
1906 | # reserved name `ICITTE`. | |
1907 | # | |
1908 | # `init_labels` is a dictionary of initial label names (valid Python | |
1909 | # names) to integral values. A label name must not be the reserved name | |
1910 | # `ICITTE`. | |
1911 | # | |
1912 | # `init_offset` is the initial offset. | |
1913 | # | |
1914 | # `init_byte_order` is the initial byte order. | |
1915 | # | |
1916 | # Raises `ParseError` on any parsing error. | |
1917 | def parse( | |
1918 | normand: str, | |
1b8aa84a PP |
1919 | init_variables: Optional[SymbolsT] = None, |
1920 | init_labels: Optional[SymbolsT] = None, | |
71aaa3f7 PP |
1921 | init_offset: int = 0, |
1922 | init_byte_order: Optional[ByteOrder] = None, | |
1923 | ): | |
1924 | if init_variables is None: | |
1925 | init_variables = {} | |
1926 | ||
1927 | if init_labels is None: | |
1928 | init_labels = {} | |
1929 | ||
1930 | gen = _Gen( | |
1931 | _Parser(normand, init_variables, init_labels).res, | |
1932 | init_variables, | |
1933 | init_labels, | |
1934 | init_offset, | |
1935 | init_byte_order, | |
1936 | ) | |
1937 | return ParseResult._create( # pyright: ignore[reportPrivateUsage] | |
1938 | gen.data, gen.variables, gen.labels, gen.offset, gen.bo | |
1939 | ) | |
1940 | ||
1941 | ||
1942 | # Parses the command-line arguments. | |
1943 | def _parse_cli_args(): | |
1944 | import argparse | |
1945 | ||
1946 | # Build parser | |
1947 | ap = argparse.ArgumentParser() | |
1948 | ap.add_argument( | |
1949 | "--offset", | |
1950 | metavar="OFFSET", | |
1951 | action="store", | |
1952 | type=int, | |
1953 | default=0, | |
1954 | help="initial offset (positive)", | |
1955 | ) | |
1956 | ap.add_argument( | |
1957 | "-b", | |
1958 | "--byte-order", | |
1959 | metavar="BO", | |
1960 | choices=["be", "le"], | |
1961 | type=str, | |
1962 | help="initial byte order (`be` or `le`)", | |
1963 | ) | |
1964 | ap.add_argument( | |
1965 | "--var", | |
1966 | metavar="NAME=VAL", | |
1967 | action="append", | |
1968 | help="add an initial variable (may be repeated)", | |
1969 | ) | |
1970 | ap.add_argument( | |
1971 | "-l", | |
1972 | "--label", | |
1973 | metavar="NAME=VAL", | |
1974 | action="append", | |
1975 | help="add an initial label (may be repeated)", | |
1976 | ) | |
1977 | ap.add_argument( | |
1978 | "--version", action="version", version="Normand {}".format(__version__) | |
1979 | ) | |
1980 | ap.add_argument( | |
1981 | "path", | |
1982 | metavar="PATH", | |
1983 | action="store", | |
1984 | nargs="?", | |
1985 | help="input path (none means standard input)", | |
1986 | ) | |
1987 | ||
1988 | # Parse | |
1989 | return ap.parse_args() | |
1990 | ||
1991 | ||
1992 | # Raises a command-line error with the message `msg`. | |
1993 | def _raise_cli_error(msg: str) -> NoReturn: | |
1994 | raise RuntimeError("Command-line error: {}".format(msg)) | |
1995 | ||
1996 | ||
1997 | # Returns a dictionary of string to integers from the list of strings | |
1998 | # `args` containing `NAME=VAL` entries. | |
1999 | def _dict_from_arg(args: Optional[List[str]]): | |
82425ef0 | 2000 | d = {} # type: SymbolsT |
71aaa3f7 PP |
2001 | |
2002 | if args is None: | |
2003 | return d | |
2004 | ||
2005 | for arg in args: | |
2006 | m = re.match(r"({})=(\d+)$".format(_py_name_pat.pattern), arg) | |
2007 | ||
2008 | if m is None: | |
2009 | _raise_cli_error("Invalid assignment {}".format(arg)) | |
2010 | ||
2e1c1acd PP |
2011 | d[m.group(1)] = int(m.group(2)) |
2012 | ||
71aaa3f7 PP |
2013 | return d |
2014 | ||
2015 | ||
2016 | # CLI entry point without exception handling. | |
2017 | def _try_run_cli(): | |
2018 | import os.path | |
2019 | ||
2020 | # Parse arguments | |
2021 | args = _parse_cli_args() | |
2022 | ||
2023 | # Read input | |
2024 | if args.path is None: | |
2025 | normand = sys.stdin.read() | |
2026 | else: | |
2027 | with open(args.path) as f: | |
2028 | normand = f.read() | |
2029 | ||
2030 | # Variables and labels | |
2031 | variables = _dict_from_arg(args.var) | |
2032 | labels = _dict_from_arg(args.label) | |
2033 | ||
2034 | # Validate offset | |
2035 | if args.offset < 0: | |
2036 | _raise_cli_error("Invalid negative offset {}") | |
2037 | ||
2038 | # Validate and set byte order | |
2039 | bo = None # type: Optional[ByteOrder] | |
2040 | ||
2041 | if args.byte_order is not None: | |
2042 | if args.byte_order == "be": | |
2043 | bo = ByteOrder.BE | |
2044 | else: | |
2045 | assert args.byte_order == "le" | |
2046 | bo = ByteOrder.LE | |
2047 | ||
2048 | # Parse | |
2049 | try: | |
2050 | res = parse(normand, variables, labels, args.offset, bo) | |
2051 | except ParseError as exc: | |
2052 | prefix = "" | |
2053 | ||
2054 | if args.path is not None: | |
2055 | prefix = "{}:".format(os.path.abspath(args.path)) | |
2056 | ||
2057 | _fail( | |
2058 | "{}{}:{} - {}".format( | |
2059 | prefix, exc.text_loc.line_no, exc.text_loc.col_no, str(exc) | |
2060 | ) | |
2061 | ) | |
2062 | ||
2063 | ||
2064 | sys.stdout.buffer.write(res.data) | |
2065 | ||
2066 | ||
2067 | # Prints the exception message `msg` and exits with status 1. | |
2068 | def _fail(msg: str) -> NoReturn: | |
2069 | if not msg.endswith("."): | |
2070 | msg += "." | |
2071 | ||
2072 | print(msg, file=sys.stderr) | |
2073 | sys.exit(1) | |
2074 | ||
2075 | ||
2076 | # CLI entry point. | |
2077 | def _run_cli(): | |
2078 | try: | |
2079 | _try_run_cli() | |
2080 | except Exception as exc: | |
2081 | _fail(str(exc)) | |
2082 | ||
2083 | ||
2084 | if __name__ == "__main__": | |
2085 | _run_cli() |