1 # The MIT License (MIT)
3 # Copyright (c) 2023 Philippe Proulx <eeppeliteloop@gmail.com>
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:
13 # The above copyright notice and this permission notice shall be
14 # included in all copies or substantial portions of the Software.
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.
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.
28 # Feel free to copy this module file to your own project to use Normand.
30 # Upstream repository: <https://github.com/efficios/normand>.
32 __author__
= "Philippe Proulx"
33 __version__
= "0.14.0"
55 from typing
import Any
, Set
, Dict
, List
, Union
, Pattern
, Callable
, NoReturn
, Optional
58 # Text location (line and column numbers).
61 def _create(cls
, line_no
: int, col_no
: int):
62 self
= cls
.__new
__(cls
)
63 self
._init
(line_no
, col_no
)
66 def __init__(*args
, **kwargs
): # type: ignore
67 raise NotImplementedError
69 def _init(self
, line_no
: int, col_no
: int):
70 self
._line
_no
= line_no
84 return "TextLocation({}, {})".format(self
._line
_no
, self
._col
_no
)
89 def __init__(self
, text_loc
: TextLocation
):
90 self
._text
_loc
= text_loc
92 # Source text location.
99 class _ScalarItem(_Item
):
100 # Returns the size, in bytes, of this item.
103 def size(self
) -> int:
113 class _Byte(_ScalarItem
, _RepableItem
):
114 def __init__(self
, val
: int, text_loc
: TextLocation
):
115 super().__init
__(text_loc
)
128 return "_Byte({}, {})".format(hex(self
._val
), repr(self
._text
_loc
))
132 class _Str(_ScalarItem
, _RepableItem
):
133 def __init__(self
, data
: bytes
, text_loc
: TextLocation
):
134 super().__init
__(text_loc
)
144 return len(self
._data
)
147 return "_Str({}, {})".format(repr(self
._data
), repr(self
._text
_loc
))
152 class ByteOrder(enum
.Enum
):
160 # Byte order setting.
162 def __init__(self
, bo
: ByteOrder
, text_loc
: TextLocation
):
163 super().__init
__(text_loc
)
171 return "_SetBo({}, {})".format(repr(self
._bo
), repr(self
._text
_loc
))
176 def __init__(self
, name
: str, text_loc
: TextLocation
):
177 super().__init
__(text_loc
)
186 return "_Label({}, {})".format(repr(self
._name
), repr(self
._text
_loc
))
190 class _SetOffset(_Item
):
191 def __init__(self
, val
: int, text_loc
: TextLocation
):
192 super().__init
__(text_loc
)
195 # Offset value (bytes).
201 return "_SetOffset({}, {})".format(repr(self
._val
), repr(self
._text
_loc
))
205 class _AlignOffset(_Item
):
206 def __init__(self
, val
: int, pad_val
: int, text_loc
: TextLocation
):
207 super().__init
__(text_loc
)
209 self
._pad
_val
= pad_val
211 # Alignment value (bits).
216 # Padding byte value.
222 return "_AlignOffset({}, {}, {})".format(
223 repr(self
._val
), repr(self
._pad
_val
), repr(self
._text
_loc
)
227 # Mixin of containing an AST expression and its string.
229 def __init__(self
, expr_str
: str, expr
: ast
.Expression
):
230 self
._expr
_str
= expr_str
236 return self
._expr
_str
238 # Expression node to evaluate.
244 # Fill until some offset.
245 class _FillUntil(_Item
, _ExprMixin
):
247 self
, expr_str
: str, expr
: ast
.Expression
, pad_val
: int, text_loc
: TextLocation
249 super().__init
__(text_loc
)
250 _ExprMixin
.__init
__(self
, expr_str
, expr
)
251 self
._pad
_val
= pad_val
253 # Padding byte value.
259 return "_FillUntil({}, {}, {}, {})".format(
260 repr(self
._expr
_str
),
263 repr(self
._text
_loc
),
267 # Variable assignment.
268 class _VarAssign(_Item
, _ExprMixin
):
270 self
, name
: str, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
272 super().__init
__(text_loc
)
273 _ExprMixin
.__init
__(self
, expr_str
, expr
)
282 return "_VarAssign({}, {}, {}, {})".format(
284 repr(self
._expr
_str
),
286 repr(self
._text
_loc
),
290 # Fixed-length number, possibly needing more than one byte.
291 class _FlNum(_ScalarItem
, _RepableItem
, _ExprMixin
):
293 self
, expr_str
: str, expr
: ast
.Expression
, len: int, text_loc
: TextLocation
295 super().__init
__(text_loc
)
296 _ExprMixin
.__init
__(self
, expr_str
, expr
)
306 return self
._len
// 8
309 return "_FlNum({}, {}, {}, {})".format(
310 repr(self
._expr
_str
),
313 repr(self
._text
_loc
),
318 class _Leb128Int(_Item
, _RepableItem
, _ExprMixin
):
319 def __init__(self
, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
):
320 super().__init
__(text_loc
)
321 _ExprMixin
.__init
__(self
, expr_str
, expr
)
324 return "{}({}, {}, {})".format(
325 self
.__class
__.__name
__,
326 repr(self
._expr
_str
),
328 repr(self
._text
_loc
),
332 # Unsigned LEB128 integer.
333 class _ULeb128Int(_Leb128Int
, _RepableItem
, _ExprMixin
):
337 # Signed LEB128 integer.
338 class _SLeb128Int(_Leb128Int
, _RepableItem
, _ExprMixin
):
343 class _Group(_Item
, _RepableItem
):
344 def __init__(self
, items
: List
[_Item
], text_loc
: TextLocation
):
345 super().__init
__(text_loc
)
354 return "_Group({}, {})".format(repr(self
._items
), repr(self
._text
_loc
))
358 class _Rep(_Item
, _ExprMixin
):
360 self
, item
: _Item
, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
362 super().__init
__(text_loc
)
363 _ExprMixin
.__init
__(self
, expr_str
, expr
)
372 return "_Rep({}, {}, {}, {})".format(
374 repr(self
._expr
_str
),
376 repr(self
._text
_loc
),
381 class _Cond(_Item
, _ExprMixin
):
387 expr
: ast
.Expression
,
388 text_loc
: TextLocation
,
390 super().__init
__(text_loc
)
391 _ExprMixin
.__init
__(self
, expr_str
, expr
)
392 self
._true
_item
= true_item
393 self
._false
_item
= false_item
395 # Item when condition is true.
398 return self
._true
_item
400 # Item when condition is false.
402 def false_item(self
):
403 return self
._false
_item
406 return "_Cond({}, {}, {}, {}, {})".format(
407 repr(self
._true
_item
),
408 repr(self
._false
_item
),
409 repr(self
._expr
_str
),
411 repr(self
._text
_loc
),
415 # Macro definition item.
416 class _MacroDef(_Item
):
418 self
, name
: str, param_names
: List
[str], group
: _Group
, text_loc
: TextLocation
420 super().__init
__(text_loc
)
422 self
._param
_names
= param_names
432 def param_names(self
):
433 return self
._param
_names
441 return "_MacroDef({}, {}, {}, {})".format(
443 repr(self
._param
_names
),
445 repr(self
._text
_loc
),
449 # Macro expansion parameter.
450 class _MacroExpParam
:
451 def __init__(self
, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
):
452 self
._expr
_str
= expr_str
454 self
._text
_loc
= text_loc
459 return self
._expr
_str
466 # Source text location.
469 return self
._text
_loc
472 return "_MacroExpParam({}, {}, {})".format(
473 repr(self
._expr
_str
), repr(self
._expr
), repr(self
._text
_loc
)
477 # Macro expansion item.
478 class _MacroExp(_Item
, _RepableItem
):
482 params
: List
[_MacroExpParam
],
483 text_loc
: TextLocation
,
485 super().__init
__(text_loc
)
487 self
._params
= params
500 return "_MacroExp({}, {}, {})".format(
503 repr(self
._text
_loc
),
507 # A parsing error containing a message and a text location.
508 class ParseError(RuntimeError):
510 def _create(cls
, msg
: str, text_loc
: TextLocation
):
511 self
= cls
.__new
__(cls
)
512 self
._init
(msg
, text_loc
)
515 def __init__(self
, *args
, **kwargs
): # type: ignore
516 raise NotImplementedError
518 def _init(self
, msg
: str, text_loc
: TextLocation
):
519 super().__init
__(msg
)
520 self
._text
_loc
= text_loc
522 # Source text location.
525 return self
._text
_loc
528 # Raises a parsing error, forwarding the parameters to the constructor.
529 def _raise_error(msg
: str, text_loc
: TextLocation
) -> NoReturn
:
530 raise ParseError
._create
(msg
, text_loc
) # pyright: ignore[reportPrivateUsage]
533 # Variables dictionary type (for type hints).
534 VariablesT
= Dict
[str, Union
[int, float]]
537 # Labels dictionary type (for type hints).
538 LabelsT
= Dict
[str, int]
541 # Python name pattern.
542 _py_name_pat
= re
.compile(r
"[a-zA-Z_][a-zA-Z0-9_]*")
545 # Macro definition dictionary.
546 _MacroDefsT
= Dict
[str, _MacroDef
]
551 # The constructor accepts a Normand input. After building, use the `res`
552 # property to get the resulting main group.
554 # Builds a parser to parse the Normand input `normand`, parsing
556 def __init__(self
, normand
: str, variables
: VariablesT
, labels
: LabelsT
):
557 self
._normand
= normand
561 self
._label
_names
= set(labels
.keys())
562 self
._var
_names
= set(variables
.keys())
563 self
._macro
_defs
= {} # type: _MacroDefsT
566 # Result (main group).
573 def macro_defs(self
):
574 return self
._macro
_defs
576 # Current text location.
579 return TextLocation
._create
( # pyright: ignore[reportPrivateUsage]
580 self
._line
_no
, self
._col
_no
583 # Returns `True` if this parser is done parsing.
585 return self
._at
== len(self
._normand
)
587 # Returns `True` if this parser isn't done parsing.
588 def _isnt_done(self
):
589 return not self
._is
_done
()
591 # Raises a parse error, creating it using the message `msg` and the
592 # current text location.
593 def _raise_error(self
, msg
: str) -> NoReturn
:
594 _raise_error(msg
, self
._text
_loc
)
596 # Tries to make the pattern `pat` match the current substring,
597 # returning the match object and updating `self._at`,
598 # `self._line_no`, and `self._col_no` on success.
599 def _try_parse_pat(self
, pat
: Pattern
[str]):
600 m
= pat
.match(self
._normand
, self
._at
)
605 # Skip matched string
606 self
._at
+= len(m
.group(0))
609 self
._line
_no
+= m
.group(0).count("\n")
611 # Update column number
612 for i
in reversed(range(self
._at
)):
613 if self
._normand
[i
] == "\n" or i
== 0:
615 self
._col
_no
= self
._at
+ 1
617 self
._col
_no
= self
._at
- i
621 # Return match object
624 # Expects the pattern `pat` to match the current substring,
625 # returning the match object and updating `self._at`,
626 # `self._line_no`, and `self._col_no` on success, or raising a parse
627 # error with the message `error_msg` on error.
628 def _expect_pat(self
, pat
: Pattern
[str], error_msg
: str):
630 m
= self
._try
_parse
_pat
(pat
)
634 self
._raise
_error
(error_msg
)
636 # Return match object
639 # Pattern for _skip_ws_and_comments()
640 _ws_or_syms_or_comments_pat
= re
.compile(
641 r
"(?:[\s/\\?&:;.,[\]_=|-]|#[^#]*?(?:\n|#))*"
644 # Skips as many whitespaces, insignificant symbol characters, and
645 # comments as possible.
646 def _skip_ws_and_comments(self
):
647 self
._try
_parse
_pat
(self
._ws
_or
_syms
_or
_comments
_pat
)
649 # Pattern for _skip_ws()
650 _ws_pat
= re
.compile(r
"\s*")
652 # Skips as many whitespaces as possible.
654 self
._try
_parse
_pat
(self
._ws
_pat
)
656 # Pattern for _try_parse_hex_byte()
657 _nibble_pat
= re
.compile(r
"[A-Fa-f0-9]")
659 # Tries to parse a hexadecimal byte, returning a byte item on
661 def _try_parse_hex_byte(self
):
662 begin_text_loc
= self
._text
_loc
664 # Match initial nibble
665 m_high
= self
._try
_parse
_pat
(self
._nibble
_pat
)
671 # Expect another nibble
672 self
._skip
_ws
_and
_comments
()
673 m_low
= self
._expect
_pat
(
674 self
._nibble
_pat
, "Expecting another hexadecimal nibble"
678 return _Byte(int(m_high
.group(0) + m_low
.group(0), 16), begin_text_loc
)
680 # Patterns for _try_parse_bin_byte()
681 _bin_byte_bit_pat
= re
.compile(r
"[01]")
682 _bin_byte_prefix_pat
= re
.compile(r
"%")
684 # Tries to parse a binary byte, returning a byte item on success.
685 def _try_parse_bin_byte(self
):
686 begin_text_loc
= self
._text
_loc
689 if self
._try
_parse
_pat
(self
._bin
_byte
_prefix
_pat
) is None:
694 bits
= [] # type: List[str]
697 self
._skip
_ws
_and
_comments
()
698 m
= self
._expect
_pat
(self
._bin
_byte
_bit
_pat
, "Expecting a bit (`0` or `1`)")
699 bits
.append(m
.group(0))
702 return _Byte(int("".join(bits
), 2), begin_text_loc
)
704 # Patterns for _try_parse_dec_byte()
705 _dec_byte_prefix_pat
= re
.compile(r
"\$")
706 _dec_byte_val_pat
= re
.compile(r
"(?P<neg>-?)(?P<val>\d+)")
708 # Tries to parse a decimal byte, returning a byte item on success.
709 def _try_parse_dec_byte(self
):
710 begin_text_loc
= self
._text
_loc
713 if self
._try
_parse
_pat
(self
._dec
_byte
_prefix
_pat
) is None:
719 m
= self
._expect
_pat
(self
._dec
_byte
_val
_pat
, "Expecting a decimal constant")
722 val
= int(m
.group("val")) * (-1 if m
.group("neg") == "-" else 1)
725 if val
< -128 or val
> 255:
726 _raise_error("Invalid decimal byte value {}".format(val
), begin_text_loc
)
732 return _Byte(val
, begin_text_loc
)
734 # Tries to parse a byte, returning a byte item on success.
735 def _try_parse_byte(self
):
737 item
= self
._try
_parse
_hex
_byte
()
743 item
= self
._try
_parse
_bin
_byte
()
749 item
= self
._try
_parse
_dec
_byte
()
754 # Patterns for _try_parse_str()
755 _str_prefix_pat
= re
.compile(r
'(?:u(?P<len>16|32)(?P<bo>be|le))?\s*"')
756 _str_suffix_pat
= re
.compile(r
'"')
757 _str_str_pat
= re
.compile(r
'(?:(?:\\.)|[^"])*')
759 # Strings corresponding to escape sequence characters
760 _str_escape_seq_strs
= {
774 # Tries to parse a string, returning a string item on success.
775 def _try_parse_str(self
):
776 begin_text_loc
= self
._text
_loc
779 m
= self
._try
_parse
_pat
(self
._str
_prefix
_pat
)
788 if m
.group("len") is not None:
789 encoding
= "utf_{}_{}".format(m
.group("len"), m
.group("bo"))
792 m
= self
._expect
_pat
(self
._str
_str
_pat
, "Expecting a literal string")
794 # Expect end of string
795 self
._expect
_pat
(self
._str
_suffix
_pat
, 'Expecting `"` (end of literal string)')
797 # Replace escape sequences
800 for ec
in '0abefnrtv"\\':
801 val
= val
.replace(r
"\{}".format(ec
), self
._str
_escape
_seq
_strs
[ec
])
804 data
= val
.encode(encoding
)
807 return _Str(data
, begin_text_loc
)
809 # Common right parenthesis pattern
810 _right_paren_pat
= re
.compile(r
"\)")
812 # Patterns for _try_parse_group()
813 _group_prefix_pat
= re
.compile(r
"\(|!g(?:roup)?\b")
815 # Tries to parse a group, returning a group item on success.
816 def _try_parse_group(self
):
817 begin_text_loc
= self
._text
_loc
820 m_open
= self
._try
_parse
_pat
(self
._group
_prefix
_pat
)
827 items
= self
._parse
_items
()
829 # Expect end of group
830 self
._skip
_ws
_and
_comments
()
832 if m_open
.group(0) == "(":
833 pat
= self
._right
_paren
_pat
836 pat
= self
._block
_end
_pat
839 self
._expect
_pat
(pat
, "Expecting an item or `{}` (end of group)".format(exp
))
842 return _Group(items
, begin_text_loc
)
844 # Returns a stripped expression string and an AST expression node
845 # from the expression string `expr_str` at text location `text_loc`.
846 def _ast_expr_from_str(self
, expr_str
: str, text_loc
: TextLocation
):
847 # Create an expression node from the expression string
848 expr_str
= expr_str
.strip().replace("\n", " ")
851 expr
= ast
.parse(expr_str
, mode
="eval")
854 "Invalid expression `{}`: invalid syntax".format(expr_str
),
858 return expr_str
, expr
860 # Patterns for _try_parse_num_and_attr()
861 _val_expr_pat
= re
.compile(r
"([^}:]+):\s*")
862 _fl_num_len_attr_pat
= re
.compile(r
"8|16|24|32|40|48|56|64")
863 _leb128_int_attr_pat
= re
.compile(r
"(u|s)leb128")
865 # Tries to parse a value and attribute (fixed length in bits or
866 # `leb128`), returning a value item on success.
867 def _try_parse_num_and_attr(self
):
868 begin_text_loc
= self
._text
_loc
871 m_expr
= self
._try
_parse
_pat
(self
._val
_expr
_pat
)
877 # Create an expression node from the expression string
878 expr_str
, expr
= self
._ast
_expr
_from
_str
(m_expr
.group(1), begin_text_loc
)
881 m_attr
= self
._try
_parse
_pat
(self
._fl
_num
_len
_attr
_pat
)
885 m_attr
= self
._try
_parse
_pat
(self
._leb
128_int
_attr
_pat
)
888 # At this point it's invalid
890 "Expecting a length (multiple of eight bits), `uleb128`, or `sleb128`"
893 # Return LEB128 integer item
894 cls
= _ULeb128Int
if m_attr
.group(1) == "u" else _SLeb128Int
895 return cls(expr_str
, expr
, begin_text_loc
)
897 # Return fixed-length number item
901 int(m_attr
.group(0)),
905 # Patterns for _try_parse_var_assign()
906 _var_assign_name_equal_pat
= re
.compile(r
"({})\s*=".format(_py_name_pat
.pattern
))
907 _var_assign_expr_pat
= re
.compile(r
"[^}]+")
909 # Tries to parse a variable assignment, returning a variable
910 # assignment item on success.
911 def _try_parse_var_assign(self
):
912 begin_text_loc
= self
._text
_loc
915 m
= self
._try
_parse
_pat
(self
._var
_assign
_name
_equal
_pat
)
924 if name
== _icitte_name
:
926 "`{}` is a reserved variable name".format(_icitte_name
), begin_text_loc
929 if name
in self
._label
_names
:
930 _raise_error("Existing label named `{}`".format(name
), begin_text_loc
)
932 # Expect an expression
934 m
= self
._expect
_pat
(self
._var
_assign
_expr
_pat
, "Expecting an expression")
936 # Create an expression node from the expression string
937 expr_str
, expr
= self
._ast
_expr
_from
_str
(m
.group(0), begin_text_loc
)
939 # Add to known variable names
940 self
._var
_names
.add(name
)
950 # Pattern for _try_parse_set_bo()
951 _bo_pat
= re
.compile(r
"[bl]e")
953 # Tries to parse a byte order name, returning a byte order setting
955 def _try_parse_set_bo(self
):
956 begin_text_loc
= self
._text
_loc
959 m
= self
._try
_parse
_pat
(self
._bo
_pat
)
965 # Return corresponding item
966 if m
.group(0) == "be":
967 return _SetBo(ByteOrder
.BE
, begin_text_loc
)
969 assert m
.group(0) == "le"
970 return _SetBo(ByteOrder
.LE
, begin_text_loc
)
972 # Patterns for _try_parse_val_or_bo()
973 _val_var_assign_set_bo_prefix_pat
= re
.compile(r
"\{")
974 _val_var_assign_set_bo_suffix_pat
= re
.compile(r
"\}")
976 # Tries to parse a value, a variable assignment, or a byte order
977 # setting, returning an item on success.
978 def _try_parse_val_or_var_assign_or_set_bo(self
):
980 if self
._try
_parse
_pat
(self
._val
_var
_assign
_set
_bo
_prefix
_pat
) is None:
986 # Variable assignment item?
987 item
= self
._try
_parse
_var
_assign
()
991 item
= self
._try
_parse
_num
_and
_attr
()
994 # Byte order setting item?
995 item
= self
._try
_parse
_set
_bo
()
998 # At this point it's invalid
1000 "Expecting a fixed-length number, a variable assignment, or a byte order setting"
1005 self
._expect
_pat
(self
._val
_var
_assign
_set
_bo
_suffix
_pat
, "Expecting `}`")
1008 # Returns a normalized version (so as to be parseable by int()) of
1009 # the constant integer string `s`, possibly negative, dealing with
1012 def _norm_const_int(s
: str):
1016 if s
.startswith("-"):
1021 if pos
.startswith("0" + r
):
1037 for suf
in asm_suf_base
:
1039 s
= "{}0{}{}".format(neg
, asm_suf_base
[suf
], pos
.rstrip(suf
))
1043 # Common constant integer patterns
1044 _pos_const_int_pat
= re
.compile(
1045 r
"0[Xx][A-Fa-f0-9]+|0[Oo][0-7]+|0[Bb][01]+|[A-Fa-f0-9]+[hH]|[0-7]+[qQoO]|[01]+[bB]|\d+"
1047 _const_int_pat
= re
.compile(r
"(?P<neg>-)?(?:{})".format(_pos_const_int_pat
.pattern
))
1049 # Tries to parse an offset setting value (after the initial `<`),
1050 # returning an offset item on success.
1051 def _try_parse_set_offset_val(self
):
1052 begin_text_loc
= self
._text
_loc
1055 m
= self
._try
_parse
_pat
(self
._pos
_const
_int
_pat
)
1062 return _SetOffset(int(self
._norm
_const
_int
(m
.group(0)), 0), begin_text_loc
)
1064 # Tries to parse a label name (after the initial `<`), returning a
1065 # label item on success.
1066 def _try_parse_label_name(self
):
1067 begin_text_loc
= self
._text
_loc
1070 m
= self
._try
_parse
_pat
(_py_name_pat
)
1079 if name
== _icitte_name
:
1081 "`{}` is a reserved label name".format(_icitte_name
), begin_text_loc
1084 if name
in self
._label
_names
:
1085 _raise_error("Duplicate label name `{}`".format(name
), begin_text_loc
)
1087 if name
in self
._var
_names
:
1088 _raise_error("Existing variable named `{}`".format(name
), begin_text_loc
)
1090 # Add to known label names
1091 self
._label
_names
.add(name
)
1094 return _Label(name
, begin_text_loc
)
1096 # Patterns for _try_parse_label_or_set_offset()
1097 _label_set_offset_prefix_pat
= re
.compile(r
"<")
1098 _label_set_offset_suffix_pat
= re
.compile(r
">")
1100 # Tries to parse a label or an offset setting, returning an item on
1102 def _try_parse_label_or_set_offset(self
):
1104 if self
._try
_parse
_pat
(self
._label
_set
_offset
_prefix
_pat
) is None:
1108 # Offset setting item?
1110 item
= self
._try
_parse
_set
_offset
_val
()
1114 item
= self
._try
_parse
_label
_name
()
1117 # At this point it's invalid
1118 self
._raise
_error
("Expecting a label name or an offset setting value")
1122 self
._expect
_pat
(self
._label
_set
_offset
_suffix
_pat
, "Expecting `>`")
1125 # Pattern for _parse_pad_val()
1126 _pad_val_prefix_pat
= re
.compile(r
"~")
1128 # Tries to parse a padding value, returning the padding value, or 0
1130 def _parse_pad_val(self
):
1135 if self
._try
_parse
_pat
(self
._pad
_val
_prefix
_pat
) is not None:
1137 pad_val_text_loc
= self
._text
_loc
1138 m
= self
._expect
_pat
(
1139 self
._pos
_const
_int
_pat
,
1140 "Expecting a positive constant integer (byte value)",
1144 pad_val
= int(self
._norm
_const
_int
(m
.group(0)), 0)
1148 "Invalid padding byte value {}".format(pad_val
),
1154 # Patterns for _try_parse_align_offset()
1155 _align_offset_prefix_pat
= re
.compile(r
"@")
1156 _align_offset_val_pat
= re
.compile(r
"\d+")
1158 # Tries to parse an offset alignment, returning an offset alignment
1160 def _try_parse_align_offset(self
):
1161 begin_text_loc
= self
._text
_loc
1164 if self
._try
_parse
_pat
(self
._align
_offset
_prefix
_pat
) is None:
1168 # Expect an alignment
1170 align_text_loc
= self
._text
_loc
1171 m
= self
._expect
_pat
(
1172 self
._align
_offset
_val
_pat
,
1173 "Expecting an alignment (positive multiple of eight bits)",
1176 # Validate alignment
1177 val
= int(m
.group(0))
1179 if val
<= 0 or (val
% 8) != 0:
1181 "Invalid alignment value {} (not a positive multiple of eight)".format(
1188 pad_val
= self
._parse
_pad
_val
()
1191 return _AlignOffset(val
, pad_val
, begin_text_loc
)
1193 # Patterns for _try_parse_fill_until()
1194 _fill_until_prefix_pat
= re
.compile(r
"\+")
1195 _fill_until_pad_val_prefix_pat
= re
.compile(r
"~")
1197 # Tries to parse a filling, returning a filling item on success.
1198 def _try_parse_fill_until(self
):
1199 begin_text_loc
= self
._text
_loc
1202 if self
._try
_parse
_pat
(self
._fill
_until
_prefix
_pat
) is None:
1208 expr_str
, expr
= self
._expect
_const
_int
_name
_expr
(True)
1211 pad_val
= self
._parse
_pad
_val
()
1214 return _FillUntil(expr_str
, expr
, pad_val
, begin_text_loc
)
1216 # Patterns for _expect_rep_mul_expr()
1217 _inner_expr_prefix_pat
= re
.compile(r
"\{")
1218 _inner_expr_pat
= re
.compile(r
"[^}]+")
1219 _inner_expr_suffix_pat
= re
.compile(r
"\}")
1221 # Parses a constant integer if `accept_const_int` is `True`
1222 # (possibly negative if `allow_neg` is `True`), a name, or an
1223 # expression within `{` and `}`.
1224 def _expect_const_int_name_expr(
1225 self
, accept_const_int
: bool, allow_neg
: bool = False
1227 expr_text_loc
= self
._text
_loc
1232 if accept_const_int
:
1233 m
= self
._try
_parse
_pat
(self
._const
_int
_pat
)
1237 m
= self
._try
_parse
_pat
(_py_name_pat
)
1241 if self
._try
_parse
_pat
(self
._inner
_expr
_prefix
_pat
) is None:
1242 pos_msg
= "" if allow_neg
else "positive "
1244 if accept_const_int
:
1245 mid_msg
= "a {}constant integer, a name, or `{{`".format(
1249 mid_msg
= "a name or `{`"
1251 # At this point it's invalid
1252 self
._raise
_error
("Expecting {}".format(mid_msg
))
1254 # Expect an expression
1256 expr_text_loc
= self
._text
_loc
1257 m
= self
._expect
_pat
(self
._inner
_expr
_pat
, "Expecting an expression")
1258 expr_str
= m
.group(0)
1262 self
._expect
_pat
(self
._inner
_expr
_suffix
_pat
, "Expecting `}`")
1264 expr_str
= m
.group(0)
1266 if m
.group("neg") == "-" and not allow_neg
:
1267 _raise_error("Expecting a positive constant integer", expr_text_loc
)
1269 expr_str
= self
._norm
_const
_int
(m
.group(0))
1271 return self
._ast
_expr
_from
_str
(expr_str
, expr_text_loc
)
1273 # Parses the multiplier expression of a repetition (block or
1274 # post-item) and returns the expression string and AST node.
1275 def _expect_rep_mul_expr(self
):
1276 return self
._expect
_const
_int
_name
_expr
(True)
1278 # Common block end pattern
1279 _block_end_pat
= re
.compile(r
"!end\b")
1281 # Pattern for _try_parse_rep_block()
1282 _rep_block_prefix_pat
= re
.compile(r
"!r(?:epeat)?\b")
1284 # Tries to parse a repetition block, returning a repetition item on
1286 def _try_parse_rep_block(self
):
1287 begin_text_loc
= self
._text
_loc
1290 if self
._try
_parse
_pat
(self
._rep
_block
_prefix
_pat
) is None:
1295 self
._skip
_ws
_and
_comments
()
1296 expr_str
, expr
= self
._expect
_rep
_mul
_expr
()
1299 self
._skip
_ws
_and
_comments
()
1300 items_text_loc
= self
._text
_loc
1301 items
= self
._parse
_items
()
1303 # Expect end of block
1304 self
._skip
_ws
_and
_comments
()
1306 self
._block
_end
_pat
, "Expecting an item or `!end` (end of repetition block)"
1310 return _Rep(_Group(items
, items_text_loc
), expr_str
, expr
, begin_text_loc
)
1312 # Pattern for _try_parse_cond_block()
1313 _cond_block_prefix_pat
= re
.compile(r
"!if\b")
1314 _cond_block_else_pat
= re
.compile(r
"!else\b")
1316 # Tries to parse a conditional block, returning a conditional item
1318 def _try_parse_cond_block(self
):
1319 begin_text_loc
= self
._text
_loc
1322 if self
._try
_parse
_pat
(self
._cond
_block
_prefix
_pat
) is None:
1327 self
._skip
_ws
_and
_comments
()
1328 expr_str
, expr
= self
._expect
_const
_int
_name
_expr
(False)
1330 # Parse "true" items
1331 self
._skip
_ws
_and
_comments
()
1332 true_items_text_loc
= self
._text
_loc
1333 true_items
= self
._parse
_items
()
1334 false_items
= [] # type: List[_Item]
1335 false_items_text_loc
= begin_text_loc
1338 self
._skip
_ws
_and
_comments
()
1340 if self
._try
_parse
_pat
(self
._cond
_block
_else
_pat
) is not None:
1341 # Parse "false" items
1342 self
._skip
_ws
_and
_comments
()
1343 false_items_text_loc
= self
._text
_loc
1344 false_items
= self
._parse
_items
()
1346 # Expect end of block
1348 self
._block
_end
_pat
,
1349 "Expecting an item, `!else`, or `!end` (end of conditional block)",
1354 _Group(true_items
, true_items_text_loc
),
1355 _Group(false_items
, false_items_text_loc
),
1361 # Common left parenthesis pattern
1362 _left_paren_pat
= re
.compile(r
"\(")
1364 # Patterns for _try_parse_macro_def() and _try_parse_macro_exp()
1365 _macro_params_comma_pat
= re
.compile(",")
1367 # Patterns for _try_parse_macro_def()
1368 _macro_def_prefix_pat
= re
.compile(r
"!m(?:acro)?\b")
1370 # Tries to parse a macro definition, adding it to `self._macro_defs`
1371 # and returning `True` on success.
1372 def _try_parse_macro_def(self
):
1373 begin_text_loc
= self
._text
_loc
1376 if self
._try
_parse
_pat
(self
._macro
_def
_prefix
_pat
) is None:
1382 name_text_loc
= self
._text
_loc
1383 m
= self
._expect
_pat
(_py_name_pat
, "Expecting a valid macro name")
1388 if name
in self
._macro
_defs
:
1389 _raise_error("Duplicate macro named `{}`".format(name
), name_text_loc
)
1393 self
._expect
_pat
(self
._left
_paren
_pat
, "Expecting `(`")
1395 # Try to parse comma-separated parameter names
1396 param_names
= [] # type: List[str]
1397 expect_comma
= False
1403 if self
._try
_parse
_pat
(self
._right
_paren
_pat
) is not None:
1409 self
._expect
_pat
(self
._macro
_params
_comma
_pat
, "Expecting `,`")
1411 # Expect parameter name
1413 param_text_loc
= self
._text
_loc
1414 m
= self
._expect
_pat
(_py_name_pat
, "Expecting valid parameter name")
1416 if m
.group(0) in param_names
:
1418 "Duplicate macro parameter named `{}`".format(m
.group(0)),
1422 param_names
.append(m
.group(0))
1426 self
._skip
_ws
_and
_comments
()
1427 items_text_loc
= self
._text
_loc
1428 old_var_names
= self
._var
_names
.copy()
1429 old_label_names
= self
._label
_names
.copy()
1430 self
._var
_names
= set() # type: Set[str]
1431 self
._label
_names
= set() # type: Set[str]
1432 items
= self
._parse
_items
()
1433 self
._var
_names
= old_var_names
1434 self
._label
_names
= old_label_names
1438 self
._block
_end
_pat
, "Expecting an item or `!end` (end of macro block)"
1442 self
._macro
_defs
[name
] = _MacroDef(
1443 name
, param_names
, _Group(items
, items_text_loc
), begin_text_loc
1448 # Patterns for _try_parse_macro_exp()
1449 _macro_exp_prefix_pat
= re
.compile(r
"m\b")
1450 _macro_exp_colon_pat
= re
.compile(r
":")
1452 # Tries to parse a macro expansion, returning a macro expansion item
1454 def _try_parse_macro_exp(self
):
1455 begin_text_loc
= self
._text
_loc
1458 if self
._try
_parse
_pat
(self
._macro
_exp
_prefix
_pat
) is None:
1464 self
._expect
_pat
(self
._macro
_exp
_colon
_pat
, "Expecting `:`")
1466 # Expect a macro name
1468 name_text_loc
= self
._text
_loc
1469 m
= self
._expect
_pat
(_py_name_pat
, "Expecting a valid macro name")
1473 macro_def
= self
._macro
_defs
.get(name
)
1475 if macro_def
is None:
1476 _raise_error("Unknown macro name `{}`".format(name
), name_text_loc
)
1480 self
._expect
_pat
(self
._left
_paren
_pat
, "Expecting `(`")
1482 # Try to parse comma-separated parameter values
1483 params_text_loc
= self
._text
_loc
1484 params
= [] # type: List[_MacroExpParam]
1485 expect_comma
= False
1491 if self
._try
_parse
_pat
(self
._right
_paren
_pat
) is not None:
1497 self
._expect
_pat
(self
._macro
_params
_comma
_pat
, "Expecting `,`")
1500 param_text_loc
= self
._text
_loc
1503 *self
._expect
_const
_int
_name
_expr
(True, True), param_text_loc
1508 # Validate parameter values
1509 if len(params
) != len(macro_def
.param_names
):
1510 sing_plur
= "" if len(params
) == 1 else "s"
1512 "Macro expansion passes {} parameter{} while the definition expects {}".format(
1513 len(params
), sing_plur
, len(macro_def
.param_names
)
1519 return _MacroExp(name
, params
, begin_text_loc
)
1521 # Tries to parse a base item (anything except a repetition),
1522 # returning it on success.
1523 def _try_parse_base_item(self
):
1525 item
= self
._try
_parse
_byte
()
1527 if item
is not None:
1531 item
= self
._try
_parse
_str
()
1533 if item
is not None:
1536 # Value, variable assignment, or byte order setting item?
1537 item
= self
._try
_parse
_val
_or
_var
_assign
_or
_set
_bo
()
1539 if item
is not None:
1542 # Label or offset setting item?
1543 item
= self
._try
_parse
_label
_or
_set
_offset
()
1545 if item
is not None:
1548 # Offset alignment item?
1549 item
= self
._try
_parse
_align
_offset
()
1551 if item
is not None:
1555 item
= self
._try
_parse
_fill
_until
()
1557 if item
is not None:
1561 item
= self
._try
_parse
_group
()
1563 if item
is not None:
1566 # Repetition block item?
1567 item
= self
._try
_parse
_rep
_block
()
1569 if item
is not None:
1572 # Conditional block item?
1573 item
= self
._try
_parse
_cond
_block
()
1575 if item
is not None:
1579 item
= self
._try
_parse
_macro
_exp
()
1581 if item
is not None:
1584 # Pattern for _try_parse_rep_post()
1585 _rep_post_prefix_pat
= re
.compile(r
"\*")
1587 # Tries to parse a post-item repetition, returning the expression
1588 # string and AST expression node on success.
1589 def _try_parse_rep_post(self
):
1591 if self
._try
_parse
_pat
(self
._rep
_post
_prefix
_pat
) is None:
1595 # Return expression string and AST expression
1596 self
._skip
_ws
_and
_comments
()
1597 return self
._expect
_rep
_mul
_expr
()
1599 # Tries to parse an item, possibly followed by a repetition,
1600 # returning `True` on success.
1602 # Appends any parsed item to `items`.
1603 def _try_append_item(self
, items
: List
[_Item
]):
1604 self
._skip
_ws
_and
_comments
()
1607 item
= self
._try
_parse
_base
_item
()
1612 # Parse repetition if the base item is repeatable
1613 if isinstance(item
, _RepableItem
):
1614 self
._skip
_ws
_and
_comments
()
1615 rep_text_loc
= self
._text
_loc
1616 rep_ret
= self
._try
_parse
_rep
_post
()
1618 if rep_ret
is not None:
1619 item
= _Rep(item
, *rep_ret
, rep_text_loc
)
1624 # Parses and returns items, skipping whitespaces, insignificant
1625 # symbols, and comments when allowed, and stopping at the first
1626 # unknown character.
1628 # Accepts and registers macro definitions if `accept_macro_defs`
1630 def _parse_items(self
, accept_macro_defs
: bool = False) -> List
[_Item
]:
1631 items
= [] # type: List[_Item]
1633 while self
._isnt
_done
():
1634 # Try to append item
1635 if not self
._try
_append
_item
(items
):
1636 if accept_macro_defs
and self
._try
_parse
_macro
_def
():
1639 # Unknown at this point
1644 # Parses the whole Normand input, setting `self._res` to the main
1645 # group item on success.
1647 if len(self
._normand
.strip()) == 0:
1648 # Special case to make sure there's something to consume
1649 self
._res
= _Group([], self
._text
_loc
)
1652 # Parse first level items
1653 items
= self
._parse
_items
(True)
1655 # Make sure there's nothing left
1656 self
._skip
_ws
_and
_comments
()
1658 if self
._isnt
_done
():
1660 "Unexpected character `{}`".format(self
._normand
[self
._at
])
1663 # Set main group item
1664 self
._res
= _Group(items
, self
._text
_loc
)
1667 # The return type of parse().
1673 variables
: VariablesT
,
1676 bo
: Optional
[ByteOrder
],
1678 self
= cls
.__new
__(cls
)
1679 self
._init
(data
, variables
, labels
, offset
, bo
)
1682 def __init__(self
, *args
, **kwargs
): # type: ignore
1683 raise NotImplementedError
1688 variables
: VariablesT
,
1691 bo
: Optional
[ByteOrder
],
1694 self
._vars
= variables
1695 self
._labels
= labels
1696 self
._offset
= offset
1704 # Dictionary of updated variable names to their last computed value.
1706 def variables(self
):
1709 # Dictionary of updated main group label names to their computed
1720 # Updated byte order.
1722 def byte_order(self
):
1726 # Raises a parse error for the item `item`, creating it using the
1728 def _raise_error_for_item(msg
: str, item
: _Item
) -> NoReturn
:
1729 _raise_error(msg
, item
.text_loc
)
1732 # The `ICITTE` reserved name.
1733 _icitte_name
= "ICITTE"
1736 # Base node visitor.
1738 # Calls the _visit_name() method for each name node which isn't the name
1740 class _NodeVisitor(ast
.NodeVisitor
):
1742 self
._parent
_is
_call
= False
1744 def generic_visit(self
, node
: ast
.AST
):
1745 if type(node
) is ast
.Call
:
1746 self
._parent
_is
_call
= True
1747 elif type(node
) is ast
.Name
and not self
._parent
_is
_call
:
1748 self
._visit
_name
(node
.id)
1750 super().generic_visit(node
)
1751 self
._parent
_is
_call
= False
1754 def _visit_name(self
, name
: str):
1758 # Expression validator: validates that all the names within the
1759 # expression are allowed.
1760 class _ExprValidator(_NodeVisitor
):
1761 def __init__(self
, expr_str
: str, text_loc
: TextLocation
, allowed_names
: Set
[str]):
1763 self
._expr
_str
= expr_str
1764 self
._text
_loc
= text_loc
1765 self
._allowed
_names
= allowed_names
1767 def _visit_name(self
, name
: str):
1768 # Make sure the name refers to a known and reachable
1769 # variable/label name.
1770 if name
!= _icitte_name
and name
not in self
._allowed
_names
:
1771 msg
= "Illegal (unknown or unreachable) variable/label name `{}` in expression `{}`".format(
1772 name
, self
._expr
_str
1775 allowed_names
= self
._allowed
_names
.copy()
1776 allowed_names
.add(_icitte_name
)
1778 if len(allowed_names
) > 0:
1779 allowed_names_str
= ", ".join(
1780 sorted(["`{}`".format(name
) for name
in allowed_names
])
1782 msg
+= "; the legal names are {{{}}}".format(allowed_names_str
)
1794 variables
: VariablesT
,
1797 bo
: Optional
[ByteOrder
],
1799 self
.variables
= variables
.copy()
1800 self
.labels
= labels
.copy()
1801 self
.offset
= offset
1805 return "_GenState({}, {}, {}, {})".format(
1806 repr(self
.variables
), repr(self
.labels
), repr(self
.offset
), repr(self
.bo
)
1810 # Fixed-length number item instance.
1811 class _FlNumItemInst
:
1812 def __init__(self
, item
: _FlNum
, offset_in_data
: int, state
: _GenState
):
1814 self
._offset
_in
_data
= offset_in_data
1822 def offset_in_data(self
):
1823 return self
._offset
_in
_data
1830 # Generator of data and final state from a group item.
1832 # Generation happens in memory at construction time. After building, use
1833 # the `data`, `variables`, `labels`, `offset`, and `bo` properties to
1834 # get the resulting context.
1836 # The steps of generation are:
1838 # 1. Handle each item in prefix order.
1840 # The handlers append bytes to `self._data` and update some current
1841 # state object (`_GenState` instance).
1843 # When handling a fixed-length number item, try to evaluate its
1844 # expression using the current state. If this fails, then it might be
1845 # because the expression refers to a "future" label: save the current
1846 # offset in `self._data` (generated data) and a snapshot of the
1847 # current state within `self._fl_num_item_insts` (`_FlNumItemInst`
1848 # object). _gen_fl_num_item_insts() will deal with this later.
1850 # When handling the items of a group, keep a map of immediate label
1851 # names to their offset. Then, after having processed all the items,
1852 # update the relevant saved state snapshots in
1853 # `self._fl_num_item_insts` with those immediate label values.
1854 # _gen_fl_num_item_insts() will deal with this later.
1856 # 2. Handle all the fixed-length number item instances of which the
1857 # expression evaluation failed before.
1859 # At this point, `self._fl_num_item_insts` contains everything that's
1860 # needed to evaluate the expressions, including the values of
1861 # "future" labels from the point of view of some fixed-length number
1864 # If an evaluation fails at this point, then it's a user error.
1869 macro_defs
: _MacroDefsT
,
1870 variables
: VariablesT
,
1873 bo
: Optional
[ByteOrder
],
1875 self
._macro
_defs
= macro_defs
1876 self
._fl
_num
_item
_insts
= [] # type: List[_FlNumItemInst]
1877 self
._gen
(group
, _GenState(variables
, labels
, offset
, bo
))
1884 # Updated variables.
1886 def variables(self
):
1887 return self
._final
_state
.variables
1889 # Updated main group labels.
1892 return self
._final
_state
.labels
1897 return self
._final
_state
.offset
1899 # Updated byte order.
1902 return self
._final
_state
.bo
1904 # Evaluates the expression `expr` of which the original string is
1905 # `expr_str` at the location `text_loc` considering the current
1906 # generation state `state`.
1908 # If `allow_float` is `True`, then the type of the result may be
1913 expr
: ast
.Expression
,
1914 text_loc
: TextLocation
,
1916 allow_float
: bool = False,
1918 syms
= {} # type: VariablesT
1919 syms
.update(state
.labels
)
1921 # Set the `ICITTE` name to the current offset
1922 syms
[_icitte_name
] = state
.offset
1924 # Add the current variables
1925 syms
.update(state
.variables
)
1927 # Validate the node and its children
1928 _ExprValidator(expr_str
, text_loc
, set(syms
.keys())).visit(expr
)
1930 # Compile and evaluate expression node
1932 val
= eval(compile(expr
, "", "eval"), None, syms
)
1933 except Exception as exc
:
1935 "Failed to evaluate expression `{}`: {}".format(expr_str
, exc
),
1939 # Convert `bool` result type to `int` to normalize
1940 if type(val
) is bool:
1943 # Validate result type
1944 expected_types
= {int}
# type: Set[type]
1948 expected_types
.add(float)
1949 type_msg
+= " or `float`"
1951 if type(val
) not in expected_types
:
1953 "Invalid expression `{}`: expecting result type {}, not `{}`".format(
1954 expr_str
, type_msg
, type(val
).__name
__
1961 # Evaluates the expression of `item` considering the current
1962 # generation state `state`.
1964 # If `allow_float` is `True`, then the type of the result may be
1967 def _eval_item_expr(
1968 item
: Union
[_FlNum
, _Leb128Int
, _FillUntil
, _VarAssign
, _Rep
, _Cond
],
1970 allow_float
: bool = False,
1972 return _Gen
._eval
_expr
(
1973 item
.expr_str
, item
.expr
, item
.text_loc
, state
, allow_float
1976 # Handles the byte item `item`.
1977 def _handle_byte_item(self
, item
: _Byte
, state
: _GenState
):
1978 self
._data
.append(item
.val
)
1979 state
.offset
+= item
.size
1981 # Handles the string item `item`.
1982 def _handle_str_item(self
, item
: _Str
, state
: _GenState
):
1983 self
._data
+= item
.data
1984 state
.offset
+= item
.size
1986 # Handles the byte order setting item `item`.
1987 def _handle_set_bo_item(self
, item
: _SetBo
, state
: _GenState
):
1988 # Update current byte order
1991 # Handles the variable assignment item `item`.
1992 def _handle_var_assign_item(self
, item
: _VarAssign
, state
: _GenState
):
1994 state
.variables
[item
.name
] = self
._eval
_item
_expr
(item
, state
, True)
1996 # Handles the fixed-length number item `item`.
1997 def _handle_fl_num_item(self
, item
: _FlNum
, state
: _GenState
):
1998 # Validate current byte order
1999 if state
.bo
is None and item
.len > 8:
2000 _raise_error_for_item(
2001 "Current byte order isn't defined at first fixed-length number (`{}`) to encode on more than 8 bits".format(
2007 # Try an immediate evaluation. If it fails, then keep everything
2008 # needed to (try to) generate the bytes of this item later.
2010 data
= self
._gen
_fl
_num
_item
_inst
_data
(item
, state
)
2012 self
._fl
_num
_item
_insts
.append(
2013 _FlNumItemInst(item
, len(self
._data
), copy
.deepcopy(state
))
2016 # Reserve space in `self._data` for this instance
2017 data
= bytes([0] * (item
.len // 8))
2023 state
.offset
+= len(data
)
2025 # Returns the size, in bytes, required to encode the value `val`
2026 # with LEB128 (signed version if `is_signed` is `True`).
2028 def _leb128_size_for_val(val
: int, is_signed
: bool):
2030 # Equivalent upper bound.
2032 # For example, if `val` is -128, then the full integer for
2033 # this number of bits would be [-128, 127].
2036 # Number of bits (add one for the sign if needed)
2037 bits
= val
.bit_length() + int(is_signed
)
2042 # Seven bits per byte
2043 return math
.ceil(bits
/ 7)
2045 # Handles the LEB128 integer item `item`.
2046 def _handle_leb128_int_item(self
, item
: _Leb128Int
, state
: _GenState
):
2048 val
= self
._eval
_item
_expr
(item
, state
, False)
2051 size
= self
._leb
128_size
_for
_val
(val
, type(item
) is _SLeb128Int
)
2054 for _
in range(size
):
2055 # Seven LSBs, MSB of the byte set (continue)
2056 self
._data
.append((val
& 0x7F) |
0x80)
2059 # Clear MSB of last byte (stop)
2060 self
._data
[-1] &= ~
0x80
2063 state
.offset
+= size
2065 # Handles the group item `item`, removing the immediate labels from
2066 # `state` at the end if `remove_immediate_labels` is `True`.
2067 def _handle_group_item(
2068 self
, item
: _Group
, state
: _GenState
, remove_immediate_labels
: bool = True
2070 first_fl_num_item_inst_index
= len(self
._fl
_num
_item
_insts
)
2071 immediate_labels
= {} # type: LabelsT
2074 for subitem
in item
.items
:
2075 if type(subitem
) is _Label
:
2076 # Add to local immediate labels
2077 immediate_labels
[subitem
.name
] = state
.offset
2079 self
._handle
_item
(subitem
, state
)
2081 # Remove immediate labels from current state if needed
2082 if remove_immediate_labels
:
2083 for name
in immediate_labels
:
2084 del state
.labels
[name
]
2086 # Add all immediate labels to all state snapshots since
2087 # `first_fl_num_item_inst_index`.
2088 for inst
in self
._fl
_num
_item
_insts
[first_fl_num_item_inst_index
:]:
2089 inst
.state
.labels
.update(immediate_labels
)
2091 # Handles the repetition item `item`.
2092 def _handle_rep_item(self
, item
: _Rep
, state
: _GenState
):
2093 # Compute the repetition count
2094 mul
= _Gen
._eval
_item
_expr
(item
, state
)
2098 _raise_error_for_item(
2099 "Invalid expression `{}`: unexpected negative result {:,}".format(
2105 # Generate item data `mul` times
2106 for _
in range(mul
):
2107 self
._handle
_item
(item
.item
, state
)
2109 # Handles the conditional item `item`.
2110 def _handle_cond_item(self
, item
: _Cond
, state
: _GenState
):
2111 # Compute the conditional value
2112 val
= _Gen
._eval
_item
_expr
(item
, state
)
2114 # Generate item data if needed
2116 self
._handle
_item
(item
.true_item
, state
)
2118 self
._handle
_item
(item
.false_item
, state
)
2120 # Evaluates the parameters of the macro expansion item `item`
2121 # considering the initial state `init_state` and returns a new state
2122 # to handle the items of the macro.
2123 def _eval_macro_exp_params(self
, item
: _MacroExp
, init_state
: _GenState
):
2125 exp_state
= _GenState({}, {}, init_state
.offset
, init_state
.bo
)
2127 # Evaluate the parameter expressions
2128 macro_def
= self
._macro
_defs
[item
.name
]
2130 for param_name
, param
in zip(macro_def
.param_names
, item
.params
):
2131 exp_state
.variables
[param_name
] = _Gen
._eval
_expr
(
2132 param
.expr_str
, param
.expr
, param
.text_loc
, init_state
, True
2137 # Handles the macro expansion item `item`.
2138 def _handle_macro_exp_item(self
, item
: _MacroExp
, state
: _GenState
):
2140 exp_state
= self
._eval
_macro
_exp
_params
(item
, state
)
2142 # Process the contained group
2143 init_data_size
= len(self
._data
)
2144 self
._handle
_item
(self
._macro
_defs
[item
.name
].group
, exp_state
)
2146 # Update state offset and return
2147 state
.offset
+= len(self
._data
) - init_data_size
2149 # Handles the offset setting item `item`.
2150 def _handle_set_offset_item(self
, item
: _SetOffset
, state
: _GenState
):
2151 state
.offset
= item
.val
2153 # Handles the offset alignment item `item` (adds padding).
2154 def _handle_align_offset_item(self
, item
: _AlignOffset
, state
: _GenState
):
2155 init_offset
= state
.offset
2156 align_bytes
= item
.val
// 8
2157 state
.offset
= (state
.offset
+ align_bytes
- 1) // align_bytes
* align_bytes
2158 self
._data
+= bytes([item
.pad_val
] * (state
.offset
- init_offset
))
2160 # Handles the filling item `item` (adds padding).
2161 def _handle_fill_until_item(self
, item
: _FillUntil
, state
: _GenState
):
2162 # Compute the new offset
2163 new_offset
= _Gen
._eval
_item
_expr
(item
, state
)
2165 # Validate the new offset
2166 if new_offset
< state
.offset
:
2167 _raise_error_for_item(
2168 "Invalid expression `{}`: new offset {:,} is less than current offset {:,}".format(
2169 item
.expr_str
, new_offset
, state
.offset
2175 self
._data
+= bytes([item
.pad_val
] * (new_offset
- state
.offset
))
2178 state
.offset
= new_offset
2180 # Handles the label item `item`.
2181 def _handle_label_item(self
, item
: _Label
, state
: _GenState
):
2182 state
.labels
[item
.name
] = state
.offset
2184 # Handles the item `item`, returning the updated next repetition
2186 def _handle_item(self
, item
: _Item
, state
: _GenState
):
2187 return self
._item
_handlers
[type(item
)](item
, state
)
2189 # Generates the data for a fixed-length integer item instance having
2190 # the value `val` and returns it.
2191 def _gen_fl_int_item_inst_data(self
, val
: int, item
: _FlNum
, state
: _GenState
):
2193 if val
< -(2 ** (item
.len - 1)) or val
> 2**item
.len - 1:
2194 _raise_error_for_item(
2195 "Value {:,} is outside the {}-bit range when evaluating expression `{}`".format(
2196 val
, item
.len, item
.expr_str
2201 # Encode result on 64 bits (to extend the sign bit whatever the
2202 # value of `item.len`).
2205 ">" if state
.bo
in (None, ByteOrder
.BE
) else "<",
2206 "Q" if val
>= 0 else "q",
2211 # Keep only the requested length
2212 len_bytes
= item
.len // 8
2214 if state
.bo
in (None, ByteOrder
.BE
):
2215 # Big endian: keep last bytes
2216 data
= data
[-len_bytes
:]
2218 # Little endian: keep first bytes
2219 assert state
.bo
== ByteOrder
.LE
2220 data
= data
[:len_bytes
]
2225 # Generates the data for a fixed-length floating point number item
2226 # instance having the value `val` and returns it.
2227 def _gen_fl_float_item_inst_data(self
, val
: float, item
: _FlNum
, state
: _GenState
):
2229 if item
.len not in (32, 64):
2230 _raise_error_for_item(
2231 "Invalid {}-bit length for a fixed-length floating point number (value {:,})".format(
2237 # Encode and return result
2240 ">" if state
.bo
in (None, ByteOrder
.BE
) else "<",
2241 "f" if item
.len == 32 else "d",
2246 # Generates the data for a fixed-length number item instance and
2248 def _gen_fl_num_item_inst_data(self
, item
: _FlNum
, state
: _GenState
):
2250 val
= self
._eval
_item
_expr
(item
, state
, True)
2252 # Handle depending on type
2253 if type(val
) is int:
2254 return self
._gen
_fl
_int
_item
_inst
_data
(val
, item
, state
)
2256 assert type(val
) is float
2257 return self
._gen
_fl
_float
_item
_inst
_data
(val
, item
, state
)
2259 # Generates the data for all the fixed-length number item instances
2260 # and writes it at the correct offset within `self._data`.
2261 def _gen_fl_num_item_insts(self
):
2262 for inst
in self
._fl
_num
_item
_insts
:
2264 data
= self
._gen
_fl
_num
_item
_inst
_data
(inst
.item
, inst
.state
)
2266 # Insert bytes into `self._data`
2267 self
._data
[inst
.offset_in_data
: inst
.offset_in_data
+ len(data
)] = data
2269 # Generates the data (`self._data`) and final state
2270 # (`self._final_state`) from `group` and the initial state `state`.
2271 def _gen(self
, group
: _Group
, state
: _GenState
):
2273 self
._data
= bytearray()
2276 self
._item
_handlers
= {
2277 _AlignOffset
: self
._handle
_align
_offset
_item
,
2278 _Byte
: self
._handle
_byte
_item
,
2279 _Cond
: self
._handle
_cond
_item
,
2280 _FillUntil
: self
._handle
_fill
_until
_item
,
2281 _FlNum
: self
._handle
_fl
_num
_item
,
2282 _Group
: self
._handle
_group
_item
,
2283 _Label
: self
._handle
_label
_item
,
2284 _MacroExp
: self
._handle
_macro
_exp
_item
,
2285 _Rep
: self
._handle
_rep
_item
,
2286 _SetBo
: self
._handle
_set
_bo
_item
,
2287 _SetOffset
: self
._handle
_set
_offset
_item
,
2288 _SLeb128Int
: self
._handle
_leb
128_int
_item
,
2289 _Str
: self
._handle
_str
_item
,
2290 _ULeb128Int
: self
._handle
_leb
128_int
_item
,
2291 _VarAssign
: self
._handle
_var
_assign
_item
,
2292 } # type: Dict[type, Callable[[Any, _GenState], None]]
2294 # Handle the group item, _not_ removing the immediate labels
2295 # because the `labels` property offers them.
2296 self
._handle
_group
_item
(group
, state
, False)
2298 # This is actually the final state
2299 self
._final
_state
= state
2301 # Generate all the fixed-length number bytes now that we know
2303 self
._gen
_fl
_num
_item
_insts
()
2306 # Returns a `ParseResult` instance containing the bytes encoded by the
2307 # input string `normand`.
2309 # `init_variables` is a dictionary of initial variable names (valid
2310 # Python names) to integral values. A variable name must not be the
2311 # reserved name `ICITTE`.
2313 # `init_labels` is a dictionary of initial label names (valid Python
2314 # names) to integral values. A label name must not be the reserved name
2317 # `init_offset` is the initial offset.
2319 # `init_byte_order` is the initial byte order.
2321 # Raises `ParseError` on any parsing error.
2324 init_variables
: Optional
[VariablesT
] = None,
2325 init_labels
: Optional
[LabelsT
] = None,
2326 init_offset
: int = 0,
2327 init_byte_order
: Optional
[ByteOrder
] = None,
2329 if init_variables
is None:
2332 if init_labels
is None:
2335 parser
= _Parser(normand
, init_variables
, init_labels
)
2344 return ParseResult
._create
( # pyright: ignore[reportPrivateUsage]
2345 gen
.data
, gen
.variables
, gen
.labels
, gen
.offset
, gen
.bo
2349 # Parses the command-line arguments.
2350 def _parse_cli_args():
2354 ap
= argparse
.ArgumentParser()
2361 help="initial offset (positive)",
2367 choices
=["be", "le"],
2369 help="initial byte order (`be` or `le`)",
2375 help="add an initial variable (may be repeated)",
2382 help="add an initial label (may be repeated)",
2385 "--version", action
="version", version
="Normand {}".format(__version__
)
2392 help="input path (none means standard input)",
2396 return ap
.parse_args()
2399 # Raises a command-line error with the message `msg`.
2400 def _raise_cli_error(msg
: str) -> NoReturn
:
2401 raise RuntimeError("Command-line error: {}".format(msg
))
2404 # Returns a dictionary of string to integers from the list of strings
2405 # `args` containing `NAME=VAL` entries.
2406 def _dict_from_arg(args
: Optional
[List
[str]]):
2407 d
= {} # type: LabelsT
2413 m
= re
.match(r
"({})=(\d+)$".format(_py_name_pat
.pattern
), arg
)
2416 _raise_cli_error("Invalid assignment {}".format(arg
))
2418 d
[m
.group(1)] = int(m
.group(2))
2423 # CLI entry point without exception handling.
2428 args
= _parse_cli_args()
2431 if args
.path
is None:
2432 normand
= sys
.stdin
.read()
2434 with
open(args
.path
) as f
:
2437 # Variables and labels
2438 variables
= typing
.cast(VariablesT
, _dict_from_arg(args
.var
))
2439 labels
= _dict_from_arg(args
.label
)
2443 _raise_cli_error("Invalid negative offset {}")
2445 # Validate and set byte order
2446 bo
= None # type: Optional[ByteOrder]
2448 if args
.byte_order
is not None:
2449 if args
.byte_order
== "be":
2452 assert args
.byte_order
== "le"
2457 res
= parse(normand
, variables
, labels
, args
.offset
, bo
)
2458 except ParseError
as exc
:
2461 if args
.path
is not None:
2462 prefix
= "{}:".format(os
.path
.abspath(args
.path
))
2465 "{}{}:{} - {}".format(
2466 prefix
, exc
.text_loc
.line_no
, exc
.text_loc
.col_no
, str(exc
)
2471 sys
.stdout
.buffer.write(res
.data
)
2474 # Prints the exception message `msg` and exits with status 1.
2475 def _fail(msg
: str) -> NoReturn
:
2476 if not msg
.endswith("."):
2479 print(msg
, file=sys
.stderr
)
2487 except Exception as exc
:
2491 if __name__
== "__main__":