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__
= "0.13.0"
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
):
383 self
, item
: _Item
, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
385 super().__init
__(text_loc
)
386 _ExprMixin
.__init
__(self
, expr_str
, expr
)
395 return "_Cond({}, {}, {}, {})".format(
397 repr(self
._expr
_str
),
399 repr(self
._text
_loc
),
403 # Macro definition item.
404 class _MacroDef(_Item
):
406 self
, name
: str, param_names
: List
[str], group
: _Group
, text_loc
: TextLocation
408 super().__init
__(text_loc
)
410 self
._param
_names
= param_names
420 def param_names(self
):
421 return self
._param
_names
429 return "_MacroDef({}, {}, {}, {})".format(
431 repr(self
._param
_names
),
433 repr(self
._text
_loc
),
437 # Macro expansion parameter.
438 class _MacroExpParam
:
439 def __init__(self
, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
):
440 self
._expr
_str
= expr_str
442 self
._text
_loc
= text_loc
447 return self
._expr
_str
454 # Source text location.
457 return self
._text
_loc
460 return "_MacroExpParam({}, {}, {})".format(
461 repr(self
._expr
_str
), repr(self
._expr
), repr(self
._text
_loc
)
465 # Macro expansion item.
466 class _MacroExp(_Item
, _RepableItem
):
470 params
: List
[_MacroExpParam
],
471 text_loc
: TextLocation
,
473 super().__init
__(text_loc
)
475 self
._params
= params
488 return "_MacroExp({}, {}, {})".format(
491 repr(self
._text
_loc
),
495 # A parsing error containing a message and a text location.
496 class ParseError(RuntimeError):
498 def _create(cls
, msg
: str, text_loc
: TextLocation
):
499 self
= cls
.__new
__(cls
)
500 self
._init
(msg
, text_loc
)
503 def __init__(self
, *args
, **kwargs
): # type: ignore
504 raise NotImplementedError
506 def _init(self
, msg
: str, text_loc
: TextLocation
):
507 super().__init
__(msg
)
508 self
._text
_loc
= text_loc
510 # Source text location.
513 return self
._text
_loc
516 # Raises a parsing error, forwarding the parameters to the constructor.
517 def _raise_error(msg
: str, text_loc
: TextLocation
) -> NoReturn
:
518 raise ParseError
._create
(msg
, text_loc
) # pyright: ignore[reportPrivateUsage]
521 # Variables dictionary type (for type hints).
522 VariablesT
= Dict
[str, Union
[int, float]]
525 # Labels dictionary type (for type hints).
526 LabelsT
= Dict
[str, int]
529 # Python name pattern.
530 _py_name_pat
= re
.compile(r
"[a-zA-Z_][a-zA-Z0-9_]*")
533 # Macro definition dictionary.
534 _MacroDefsT
= Dict
[str, _MacroDef
]
539 # The constructor accepts a Normand input. After building, use the `res`
540 # property to get the resulting main group.
542 # Builds a parser to parse the Normand input `normand`, parsing
544 def __init__(self
, normand
: str, variables
: VariablesT
, labels
: LabelsT
):
545 self
._normand
= normand
549 self
._label
_names
= set(labels
.keys())
550 self
._var
_names
= set(variables
.keys())
551 self
._macro
_defs
= {} # type: _MacroDefsT
554 # Result (main group).
561 def macro_defs(self
):
562 return self
._macro
_defs
564 # Current text location.
567 return TextLocation
._create
( # pyright: ignore[reportPrivateUsage]
568 self
._line
_no
, self
._col
_no
571 # Returns `True` if this parser is done parsing.
573 return self
._at
== len(self
._normand
)
575 # Returns `True` if this parser isn't done parsing.
576 def _isnt_done(self
):
577 return not self
._is
_done
()
579 # Raises a parse error, creating it using the message `msg` and the
580 # current text location.
581 def _raise_error(self
, msg
: str) -> NoReturn
:
582 _raise_error(msg
, self
._text
_loc
)
584 # Tries to make the pattern `pat` match the current substring,
585 # returning the match object and updating `self._at`,
586 # `self._line_no`, and `self._col_no` on success.
587 def _try_parse_pat(self
, pat
: Pattern
[str]):
588 m
= pat
.match(self
._normand
, self
._at
)
593 # Skip matched string
594 self
._at
+= len(m
.group(0))
597 self
._line
_no
+= m
.group(0).count("\n")
599 # Update column number
600 for i
in reversed(range(self
._at
)):
601 if self
._normand
[i
] == "\n" or i
== 0:
603 self
._col
_no
= self
._at
+ 1
605 self
._col
_no
= self
._at
- i
609 # Return match object
612 # Expects the pattern `pat` to match the current substring,
613 # returning the match object and updating `self._at`,
614 # `self._line_no`, and `self._col_no` on success, or raising a parse
615 # error with the message `error_msg` on error.
616 def _expect_pat(self
, pat
: Pattern
[str], error_msg
: str):
618 m
= self
._try
_parse
_pat
(pat
)
622 self
._raise
_error
(error_msg
)
624 # Return match object
627 # Pattern for _skip_ws_and_comments()
628 _ws_or_syms_or_comments_pat
= re
.compile(
629 r
"(?:[\s/\\?&:;.,[\]_=|-]|#[^#]*?(?:\n|#))*"
632 # Skips as many whitespaces, insignificant symbol characters, and
633 # comments as possible.
634 def _skip_ws_and_comments(self
):
635 self
._try
_parse
_pat
(self
._ws
_or
_syms
_or
_comments
_pat
)
637 # Pattern for _skip_ws()
638 _ws_pat
= re
.compile(r
"\s*")
640 # Skips as many whitespaces as possible.
642 self
._try
_parse
_pat
(self
._ws
_pat
)
644 # Pattern for _try_parse_hex_byte()
645 _nibble_pat
= re
.compile(r
"[A-Fa-f0-9]")
647 # Tries to parse a hexadecimal byte, returning a byte item on
649 def _try_parse_hex_byte(self
):
650 begin_text_loc
= self
._text
_loc
652 # Match initial nibble
653 m_high
= self
._try
_parse
_pat
(self
._nibble
_pat
)
659 # Expect another nibble
660 self
._skip
_ws
_and
_comments
()
661 m_low
= self
._expect
_pat
(
662 self
._nibble
_pat
, "Expecting another hexadecimal nibble"
666 return _Byte(int(m_high
.group(0) + m_low
.group(0), 16), begin_text_loc
)
668 # Patterns for _try_parse_bin_byte()
669 _bin_byte_bit_pat
= re
.compile(r
"[01]")
670 _bin_byte_prefix_pat
= re
.compile(r
"%")
672 # Tries to parse a binary byte, returning a byte item on success.
673 def _try_parse_bin_byte(self
):
674 begin_text_loc
= self
._text
_loc
677 if self
._try
_parse
_pat
(self
._bin
_byte
_prefix
_pat
) is None:
682 bits
= [] # type: List[str]
685 self
._skip
_ws
_and
_comments
()
686 m
= self
._expect
_pat
(self
._bin
_byte
_bit
_pat
, "Expecting a bit (`0` or `1`)")
687 bits
.append(m
.group(0))
690 return _Byte(int("".join(bits
), 2), begin_text_loc
)
692 # Patterns for _try_parse_dec_byte()
693 _dec_byte_prefix_pat
= re
.compile(r
"\$")
694 _dec_byte_val_pat
= re
.compile(r
"(?P<neg>-?)(?P<val>\d+)")
696 # Tries to parse a decimal byte, returning a byte item on success.
697 def _try_parse_dec_byte(self
):
698 begin_text_loc
= self
._text
_loc
701 if self
._try
_parse
_pat
(self
._dec
_byte
_prefix
_pat
) is None:
707 m
= self
._expect
_pat
(self
._dec
_byte
_val
_pat
, "Expecting a decimal constant")
710 val
= int(m
.group("val")) * (-1 if m
.group("neg") == "-" else 1)
713 if val
< -128 or val
> 255:
714 _raise_error("Invalid decimal byte value {}".format(val
), begin_text_loc
)
720 return _Byte(val
, begin_text_loc
)
722 # Tries to parse a byte, returning a byte item on success.
723 def _try_parse_byte(self
):
725 item
= self
._try
_parse
_hex
_byte
()
731 item
= self
._try
_parse
_bin
_byte
()
737 item
= self
._try
_parse
_dec
_byte
()
742 # Patterns for _try_parse_str()
743 _str_prefix_pat
= re
.compile(r
'(?:u(?P<len>16|32)(?P<bo>be|le))?\s*"')
744 _str_suffix_pat
= re
.compile(r
'"')
745 _str_str_pat
= re
.compile(r
'(?:(?:\\.)|[^"])*')
747 # Strings corresponding to escape sequence characters
748 _str_escape_seq_strs
= {
762 # Tries to parse a string, returning a string item on success.
763 def _try_parse_str(self
):
764 begin_text_loc
= self
._text
_loc
767 m
= self
._try
_parse
_pat
(self
._str
_prefix
_pat
)
776 if m
.group("len") is not None:
777 encoding
= "utf_{}_{}".format(m
.group("len"), m
.group("bo"))
780 m
= self
._expect
_pat
(self
._str
_str
_pat
, "Expecting a literal string")
782 # Expect end of string
783 self
._expect
_pat
(self
._str
_suffix
_pat
, 'Expecting `"` (end of literal string)')
785 # Replace escape sequences
788 for ec
in '0abefnrtv"\\':
789 val
= val
.replace(r
"\{}".format(ec
), self
._str
_escape
_seq
_strs
[ec
])
792 data
= val
.encode(encoding
)
795 return _Str(data
, begin_text_loc
)
797 # Common right parenthesis pattern
798 _right_paren_pat
= re
.compile(r
"\)")
800 # Patterns for _try_parse_group()
801 _group_prefix_pat
= re
.compile(r
"\(|!g(?:roup)?\b")
803 # Tries to parse a group, returning a group item on success.
804 def _try_parse_group(self
):
805 begin_text_loc
= self
._text
_loc
808 m_open
= self
._try
_parse
_pat
(self
._group
_prefix
_pat
)
815 items
= self
._parse
_items
()
817 # Expect end of group
818 self
._skip
_ws
_and
_comments
()
820 if m_open
.group(0) == "(":
821 pat
= self
._right
_paren
_pat
824 pat
= self
._block
_end
_pat
827 self
._expect
_pat
(pat
, "Expecting an item or `{}` (end of group)".format(exp
))
830 return _Group(items
, begin_text_loc
)
832 # Returns a stripped expression string and an AST expression node
833 # from the expression string `expr_str` at text location `text_loc`.
834 def _ast_expr_from_str(self
, expr_str
: str, text_loc
: TextLocation
):
835 # Create an expression node from the expression string
836 expr_str
= expr_str
.strip().replace("\n", " ")
839 expr
= ast
.parse(expr_str
, mode
="eval")
842 "Invalid expression `{}`: invalid syntax".format(expr_str
),
846 return expr_str
, expr
848 # Patterns for _try_parse_num_and_attr()
849 _val_expr_pat
= re
.compile(r
"([^}:]+):\s*")
850 _fl_num_len_attr_pat
= re
.compile(r
"8|16|24|32|40|48|56|64")
851 _leb128_int_attr_pat
= re
.compile(r
"(u|s)leb128")
853 # Tries to parse a value and attribute (fixed length in bits or
854 # `leb128`), returning a value item on success.
855 def _try_parse_num_and_attr(self
):
856 begin_text_loc
= self
._text
_loc
859 m_expr
= self
._try
_parse
_pat
(self
._val
_expr
_pat
)
865 # Create an expression node from the expression string
866 expr_str
, expr
= self
._ast
_expr
_from
_str
(m_expr
.group(1), begin_text_loc
)
869 m_attr
= self
._try
_parse
_pat
(self
._fl
_num
_len
_attr
_pat
)
873 m_attr
= self
._try
_parse
_pat
(self
._leb
128_int
_attr
_pat
)
876 # At this point it's invalid
878 "Expecting a length (multiple of eight bits), `uleb128`, or `sleb128`"
881 # Return LEB128 integer item
882 cls
= _ULeb128Int
if m_attr
.group(1) == "u" else _SLeb128Int
883 return cls(expr_str
, expr
, begin_text_loc
)
885 # Return fixed-length number item
889 int(m_attr
.group(0)),
893 # Patterns for _try_parse_var_assign()
894 _var_assign_name_equal_pat
= re
.compile(r
"({})\s*=".format(_py_name_pat
.pattern
))
895 _var_assign_expr_pat
= re
.compile(r
"[^}]+")
897 # Tries to parse a variable assignment, returning a variable
898 # assignment item on success.
899 def _try_parse_var_assign(self
):
900 begin_text_loc
= self
._text
_loc
903 m
= self
._try
_parse
_pat
(self
._var
_assign
_name
_equal
_pat
)
912 if name
== _icitte_name
:
914 "`{}` is a reserved variable name".format(_icitte_name
), begin_text_loc
917 if name
in self
._label
_names
:
918 _raise_error("Existing label named `{}`".format(name
), begin_text_loc
)
920 # Expect an expression
922 m
= self
._expect
_pat
(self
._var
_assign
_expr
_pat
, "Expecting an expression")
924 # Create an expression node from the expression string
925 expr_str
, expr
= self
._ast
_expr
_from
_str
(m
.group(0), begin_text_loc
)
927 # Add to known variable names
928 self
._var
_names
.add(name
)
938 # Pattern for _try_parse_set_bo()
939 _bo_pat
= re
.compile(r
"[bl]e")
941 # Tries to parse a byte order name, returning a byte order setting
943 def _try_parse_set_bo(self
):
944 begin_text_loc
= self
._text
_loc
947 m
= self
._try
_parse
_pat
(self
._bo
_pat
)
953 # Return corresponding item
954 if m
.group(0) == "be":
955 return _SetBo(ByteOrder
.BE
, begin_text_loc
)
957 assert m
.group(0) == "le"
958 return _SetBo(ByteOrder
.LE
, begin_text_loc
)
960 # Patterns for _try_parse_val_or_bo()
961 _val_var_assign_set_bo_prefix_pat
= re
.compile(r
"\{")
962 _val_var_assign_set_bo_suffix_pat
= re
.compile(r
"\}")
964 # Tries to parse a value, a variable assignment, or a byte order
965 # setting, returning an item on success.
966 def _try_parse_val_or_var_assign_or_set_bo(self
):
968 if self
._try
_parse
_pat
(self
._val
_var
_assign
_set
_bo
_prefix
_pat
) is None:
974 # Variable assignment item?
975 item
= self
._try
_parse
_var
_assign
()
979 item
= self
._try
_parse
_num
_and
_attr
()
982 # Byte order setting item?
983 item
= self
._try
_parse
_set
_bo
()
986 # At this point it's invalid
988 "Expecting a fixed-length number, a variable assignment, or a byte order setting"
993 self
._expect
_pat
(self
._val
_var
_assign
_set
_bo
_suffix
_pat
, "Expecting `}`")
996 # Returns a normalized version (so as to be parseable by int()) of
997 # the constant integer string `s`, possibly negative, dealing with
1000 def _norm_const_int(s
: str):
1004 if s
.startswith("-"):
1009 if pos
.startswith("0" + r
):
1025 for suf
in asm_suf_base
:
1027 s
= "{}0{}{}".format(neg
, asm_suf_base
[suf
], pos
.rstrip(suf
))
1031 # Common constant integer patterns
1032 _pos_const_int_pat
= re
.compile(
1033 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+"
1035 _const_int_pat
= re
.compile(r
"(?P<neg>-)?(?:{})".format(_pos_const_int_pat
.pattern
))
1037 # Tries to parse an offset setting value (after the initial `<`),
1038 # returning an offset item on success.
1039 def _try_parse_set_offset_val(self
):
1040 begin_text_loc
= self
._text
_loc
1043 m
= self
._try
_parse
_pat
(self
._pos
_const
_int
_pat
)
1050 return _SetOffset(int(self
._norm
_const
_int
(m
.group(0)), 0), begin_text_loc
)
1052 # Tries to parse a label name (after the initial `<`), returning a
1053 # label item on success.
1054 def _try_parse_label_name(self
):
1055 begin_text_loc
= self
._text
_loc
1058 m
= self
._try
_parse
_pat
(_py_name_pat
)
1067 if name
== _icitte_name
:
1069 "`{}` is a reserved label name".format(_icitte_name
), begin_text_loc
1072 if name
in self
._label
_names
:
1073 _raise_error("Duplicate label name `{}`".format(name
), begin_text_loc
)
1075 if name
in self
._var
_names
:
1076 _raise_error("Existing variable named `{}`".format(name
), begin_text_loc
)
1078 # Add to known label names
1079 self
._label
_names
.add(name
)
1082 return _Label(name
, begin_text_loc
)
1084 # Patterns for _try_parse_label_or_set_offset()
1085 _label_set_offset_prefix_pat
= re
.compile(r
"<")
1086 _label_set_offset_suffix_pat
= re
.compile(r
">")
1088 # Tries to parse a label or an offset setting, returning an item on
1090 def _try_parse_label_or_set_offset(self
):
1092 if self
._try
_parse
_pat
(self
._label
_set
_offset
_prefix
_pat
) is None:
1096 # Offset setting item?
1098 item
= self
._try
_parse
_set
_offset
_val
()
1102 item
= self
._try
_parse
_label
_name
()
1105 # At this point it's invalid
1106 self
._raise
_error
("Expecting a label name or an offset setting value")
1110 self
._expect
_pat
(self
._label
_set
_offset
_suffix
_pat
, "Expecting `>`")
1113 # Pattern for _parse_pad_val()
1114 _pad_val_prefix_pat
= re
.compile(r
"~")
1116 # Tries to parse a padding value, returning the padding value, or 0
1118 def _parse_pad_val(self
):
1123 if self
._try
_parse
_pat
(self
._pad
_val
_prefix
_pat
) is not None:
1125 pad_val_text_loc
= self
._text
_loc
1126 m
= self
._expect
_pat
(
1127 self
._pos
_const
_int
_pat
,
1128 "Expecting a positive constant integer (byte value)",
1132 pad_val
= int(self
._norm
_const
_int
(m
.group(0)), 0)
1136 "Invalid padding byte value {}".format(pad_val
),
1142 # Patterns for _try_parse_align_offset()
1143 _align_offset_prefix_pat
= re
.compile(r
"@")
1144 _align_offset_val_pat
= re
.compile(r
"\d+")
1146 # Tries to parse an offset alignment, returning an offset alignment
1148 def _try_parse_align_offset(self
):
1149 begin_text_loc
= self
._text
_loc
1152 if self
._try
_parse
_pat
(self
._align
_offset
_prefix
_pat
) is None:
1156 # Expect an alignment
1158 align_text_loc
= self
._text
_loc
1159 m
= self
._expect
_pat
(
1160 self
._align
_offset
_val
_pat
,
1161 "Expecting an alignment (positive multiple of eight bits)",
1164 # Validate alignment
1165 val
= int(m
.group(0))
1167 if val
<= 0 or (val
% 8) != 0:
1169 "Invalid alignment value {} (not a positive multiple of eight)".format(
1176 pad_val
= self
._parse
_pad
_val
()
1179 return _AlignOffset(val
, pad_val
, begin_text_loc
)
1181 # Patterns for _try_parse_fill_until()
1182 _fill_until_prefix_pat
= re
.compile(r
"\+")
1183 _fill_until_pad_val_prefix_pat
= re
.compile(r
"~")
1185 # Tries to parse a filling, returning a filling item on success.
1186 def _try_parse_fill_until(self
):
1187 begin_text_loc
= self
._text
_loc
1190 if self
._try
_parse
_pat
(self
._fill
_until
_prefix
_pat
) is None:
1196 expr_str
, expr
= self
._expect
_const
_int
_name
_expr
(True)
1199 pad_val
= self
._parse
_pad
_val
()
1202 return _FillUntil(expr_str
, expr
, pad_val
, begin_text_loc
)
1204 # Patterns for _expect_rep_mul_expr()
1205 _inner_expr_prefix_pat
= re
.compile(r
"\{")
1206 _inner_expr_pat
= re
.compile(r
"[^}]+")
1207 _inner_expr_suffix_pat
= re
.compile(r
"\}")
1209 # Parses a constant integer if `accept_const_int` is `True`
1210 # (possibly negative if `allow_neg` is `True`), a name, or an
1211 # expression within `{` and `}`.
1212 def _expect_const_int_name_expr(
1213 self
, accept_const_int
: bool, allow_neg
: bool = False
1215 expr_text_loc
= self
._text
_loc
1220 if accept_const_int
:
1221 m
= self
._try
_parse
_pat
(self
._const
_int
_pat
)
1225 m
= self
._try
_parse
_pat
(_py_name_pat
)
1229 if self
._try
_parse
_pat
(self
._inner
_expr
_prefix
_pat
) is None:
1230 pos_msg
= "" if allow_neg
else "positive "
1232 if accept_const_int
:
1233 mid_msg
= "a {}constant integer, a name, or `{{`".format(
1237 mid_msg
= "a name or `{`"
1239 # At this point it's invalid
1240 self
._raise
_error
("Expecting {}".format(mid_msg
))
1242 # Expect an expression
1244 expr_text_loc
= self
._text
_loc
1245 m
= self
._expect
_pat
(self
._inner
_expr
_pat
, "Expecting an expression")
1246 expr_str
= m
.group(0)
1250 self
._expect
_pat
(self
._inner
_expr
_suffix
_pat
, "Expecting `}`")
1252 expr_str
= m
.group(0)
1254 if m
.group("neg") == "-" and not allow_neg
:
1255 _raise_error("Expecting a positive constant integer", expr_text_loc
)
1257 expr_str
= self
._norm
_const
_int
(m
.group(0))
1259 return self
._ast
_expr
_from
_str
(expr_str
, expr_text_loc
)
1261 # Parses the multiplier expression of a repetition (block or
1262 # post-item) and returns the expression string and AST node.
1263 def _expect_rep_mul_expr(self
):
1264 return self
._expect
_const
_int
_name
_expr
(True)
1266 # Common block end pattern
1267 _block_end_pat
= re
.compile(r
"!end\b")
1269 # Pattern for _try_parse_rep_block()
1270 _rep_block_prefix_pat
= re
.compile(r
"!r(?:epeat)?\b")
1272 # Tries to parse a repetition block, returning a repetition item on
1274 def _try_parse_rep_block(self
):
1275 begin_text_loc
= self
._text
_loc
1278 if self
._try
_parse
_pat
(self
._rep
_block
_prefix
_pat
) is None:
1283 self
._skip
_ws
_and
_comments
()
1284 expr_str
, expr
= self
._expect
_rep
_mul
_expr
()
1287 self
._skip
_ws
_and
_comments
()
1288 items_text_loc
= self
._text
_loc
1289 items
= self
._parse
_items
()
1291 # Expect end of block
1292 self
._skip
_ws
_and
_comments
()
1294 self
._block
_end
_pat
, "Expecting an item or `!end` (end of repetition block)"
1298 return _Rep(_Group(items
, items_text_loc
), expr_str
, expr
, begin_text_loc
)
1300 # Pattern for _try_parse_cond_block()
1301 _cond_block_prefix_pat
= re
.compile(r
"!if\b")
1303 # Tries to parse a conditional block, returning a conditional item
1305 def _try_parse_cond_block(self
):
1306 begin_text_loc
= self
._text
_loc
1309 if self
._try
_parse
_pat
(self
._cond
_block
_prefix
_pat
) is None:
1314 self
._skip
_ws
_and
_comments
()
1315 expr_str
, expr
= self
._expect
_const
_int
_name
_expr
(False)
1318 self
._skip
_ws
_and
_comments
()
1319 items_text_loc
= self
._text
_loc
1320 items
= self
._parse
_items
()
1322 # Expect end of block
1323 self
._skip
_ws
_and
_comments
()
1325 self
._block
_end
_pat
,
1326 "Expecting an item or `!end` (end of conditional block)",
1330 return _Cond(_Group(items
, items_text_loc
), expr_str
, expr
, begin_text_loc
)
1332 # Common left parenthesis pattern
1333 _left_paren_pat
= re
.compile(r
"\(")
1335 # Patterns for _try_parse_macro_def() and _try_parse_macro_exp()
1336 _macro_params_comma_pat
= re
.compile(",")
1338 # Patterns for _try_parse_macro_def()
1339 _macro_def_prefix_pat
= re
.compile(r
"!m(?:acro)?\b")
1341 # Tries to parse a macro definition, adding it to `self._macro_defs`
1342 # and returning `True` on success.
1343 def _try_parse_macro_def(self
):
1344 begin_text_loc
= self
._text
_loc
1347 if self
._try
_parse
_pat
(self
._macro
_def
_prefix
_pat
) is None:
1353 name_text_loc
= self
._text
_loc
1354 m
= self
._expect
_pat
(_py_name_pat
, "Expecting a valid macro name")
1359 if name
in self
._macro
_defs
:
1360 _raise_error("Duplicate macro named `{}`".format(name
), name_text_loc
)
1364 self
._expect
_pat
(self
._left
_paren
_pat
, "Expecting `(`")
1366 # Try to parse comma-separated parameter names
1367 param_names
= [] # type: List[str]
1368 expect_comma
= False
1374 if self
._try
_parse
_pat
(self
._right
_paren
_pat
) is not None:
1380 self
._expect
_pat
(self
._macro
_params
_comma
_pat
, "Expecting `,`")
1382 # Expect parameter name
1384 param_text_loc
= self
._text
_loc
1385 m
= self
._expect
_pat
(_py_name_pat
, "Expecting valid parameter name")
1387 if m
.group(0) in param_names
:
1389 "Duplicate macro parameter named `{}`".format(m
.group(0)),
1393 param_names
.append(m
.group(0))
1397 self
._skip
_ws
_and
_comments
()
1398 items_text_loc
= self
._text
_loc
1399 old_var_names
= self
._var
_names
.copy()
1400 old_label_names
= self
._label
_names
.copy()
1401 self
._var
_names
= set() # type: Set[str]
1402 self
._label
_names
= set() # type: Set[str]
1403 items
= self
._parse
_items
()
1404 self
._var
_names
= old_var_names
1405 self
._label
_names
= old_label_names
1409 self
._block
_end
_pat
, "Expecting an item or `!end` (end of macro block)"
1413 self
._macro
_defs
[name
] = _MacroDef(
1414 name
, param_names
, _Group(items
, items_text_loc
), begin_text_loc
1419 # Patterns for _try_parse_macro_exp()
1420 _macro_exp_prefix_pat
= re
.compile(r
"m\b")
1421 _macro_exp_colon_pat
= re
.compile(r
":")
1423 # Tries to parse a macro expansion, returning a macro expansion item
1425 def _try_parse_macro_exp(self
):
1426 begin_text_loc
= self
._text
_loc
1429 if self
._try
_parse
_pat
(self
._macro
_exp
_prefix
_pat
) is None:
1435 self
._expect
_pat
(self
._macro
_exp
_colon
_pat
, "Expecting `:`")
1437 # Expect a macro name
1439 name_text_loc
= self
._text
_loc
1440 m
= self
._expect
_pat
(_py_name_pat
, "Expecting a valid macro name")
1444 macro_def
= self
._macro
_defs
.get(name
)
1446 if macro_def
is None:
1447 _raise_error("Unknown macro name `{}`".format(name
), name_text_loc
)
1451 self
._expect
_pat
(self
._left
_paren
_pat
, "Expecting `(`")
1453 # Try to parse comma-separated parameter values
1454 params_text_loc
= self
._text
_loc
1455 params
= [] # type: List[_MacroExpParam]
1456 expect_comma
= False
1462 if self
._try
_parse
_pat
(self
._right
_paren
_pat
) is not None:
1468 self
._expect
_pat
(self
._macro
_params
_comma
_pat
, "Expecting `,`")
1471 param_text_loc
= self
._text
_loc
1474 *self
._expect
_const
_int
_name
_expr
(True, True), param_text_loc
1479 # Validate parameter values
1480 if len(params
) != len(macro_def
.param_names
):
1481 sing_plur
= "" if len(params
) == 1 else "s"
1483 "Macro expansion passes {} parameter{} while the definition expects {}".format(
1484 len(params
), sing_plur
, len(macro_def
.param_names
)
1490 return _MacroExp(name
, params
, begin_text_loc
)
1492 # Tries to parse a base item (anything except a repetition),
1493 # returning it on success.
1494 def _try_parse_base_item(self
):
1496 item
= self
._try
_parse
_byte
()
1498 if item
is not None:
1502 item
= self
._try
_parse
_str
()
1504 if item
is not None:
1507 # Value, variable assignment, or byte order setting item?
1508 item
= self
._try
_parse
_val
_or
_var
_assign
_or
_set
_bo
()
1510 if item
is not None:
1513 # Label or offset setting item?
1514 item
= self
._try
_parse
_label
_or
_set
_offset
()
1516 if item
is not None:
1519 # Offset alignment item?
1520 item
= self
._try
_parse
_align
_offset
()
1522 if item
is not None:
1526 item
= self
._try
_parse
_fill
_until
()
1528 if item
is not None:
1532 item
= self
._try
_parse
_group
()
1534 if item
is not None:
1537 # Repetition block item?
1538 item
= self
._try
_parse
_rep
_block
()
1540 if item
is not None:
1543 # Conditional block item?
1544 item
= self
._try
_parse
_cond
_block
()
1546 if item
is not None:
1550 item
= self
._try
_parse
_macro
_exp
()
1552 if item
is not None:
1555 # Pattern for _try_parse_rep_post()
1556 _rep_post_prefix_pat
= re
.compile(r
"\*")
1558 # Tries to parse a post-item repetition, returning the expression
1559 # string and AST expression node on success.
1560 def _try_parse_rep_post(self
):
1562 if self
._try
_parse
_pat
(self
._rep
_post
_prefix
_pat
) is None:
1566 # Return expression string and AST expression
1567 self
._skip
_ws
_and
_comments
()
1568 return self
._expect
_rep
_mul
_expr
()
1570 # Tries to parse an item, possibly followed by a repetition,
1571 # returning `True` on success.
1573 # Appends any parsed item to `items`.
1574 def _try_append_item(self
, items
: List
[_Item
]):
1575 self
._skip
_ws
_and
_comments
()
1578 item
= self
._try
_parse
_base
_item
()
1583 # Parse repetition if the base item is repeatable
1584 if isinstance(item
, _RepableItem
):
1585 self
._skip
_ws
_and
_comments
()
1586 rep_text_loc
= self
._text
_loc
1587 rep_ret
= self
._try
_parse
_rep
_post
()
1589 if rep_ret
is not None:
1590 item
= _Rep(item
, *rep_ret
, rep_text_loc
)
1595 # Parses and returns items, skipping whitespaces, insignificant
1596 # symbols, and comments when allowed, and stopping at the first
1597 # unknown character.
1599 # Accepts and registers macro definitions if `accept_macro_defs`
1601 def _parse_items(self
, accept_macro_defs
: bool = False) -> List
[_Item
]:
1602 items
= [] # type: List[_Item]
1604 while self
._isnt
_done
():
1605 # Try to append item
1606 if not self
._try
_append
_item
(items
):
1607 if accept_macro_defs
and self
._try
_parse
_macro
_def
():
1610 # Unknown at this point
1615 # Parses the whole Normand input, setting `self._res` to the main
1616 # group item on success.
1618 if len(self
._normand
.strip()) == 0:
1619 # Special case to make sure there's something to consume
1620 self
._res
= _Group([], self
._text
_loc
)
1623 # Parse first level items
1624 items
= self
._parse
_items
(True)
1626 # Make sure there's nothing left
1627 self
._skip
_ws
_and
_comments
()
1629 if self
._isnt
_done
():
1631 "Unexpected character `{}`".format(self
._normand
[self
._at
])
1634 # Set main group item
1635 self
._res
= _Group(items
, self
._text
_loc
)
1638 # The return type of parse().
1644 variables
: VariablesT
,
1647 bo
: Optional
[ByteOrder
],
1649 self
= cls
.__new
__(cls
)
1650 self
._init
(data
, variables
, labels
, offset
, bo
)
1653 def __init__(self
, *args
, **kwargs
): # type: ignore
1654 raise NotImplementedError
1659 variables
: VariablesT
,
1662 bo
: Optional
[ByteOrder
],
1665 self
._vars
= variables
1666 self
._labels
= labels
1667 self
._offset
= offset
1675 # Dictionary of updated variable names to their last computed value.
1677 def variables(self
):
1680 # Dictionary of updated main group label names to their computed
1691 # Updated byte order.
1693 def byte_order(self
):
1697 # Raises a parse error for the item `item`, creating it using the
1699 def _raise_error_for_item(msg
: str, item
: _Item
) -> NoReturn
:
1700 _raise_error(msg
, item
.text_loc
)
1703 # The `ICITTE` reserved name.
1704 _icitte_name
= "ICITTE"
1707 # Base node visitor.
1709 # Calls the _visit_name() method for each name node which isn't the name
1711 class _NodeVisitor(ast
.NodeVisitor
):
1713 self
._parent
_is
_call
= False
1715 def generic_visit(self
, node
: ast
.AST
):
1716 if type(node
) is ast
.Call
:
1717 self
._parent
_is
_call
= True
1718 elif type(node
) is ast
.Name
and not self
._parent
_is
_call
:
1719 self
._visit
_name
(node
.id)
1721 super().generic_visit(node
)
1722 self
._parent
_is
_call
= False
1725 def _visit_name(self
, name
: str):
1729 # Expression validator: validates that all the names within the
1730 # expression are allowed.
1731 class _ExprValidator(_NodeVisitor
):
1732 def __init__(self
, expr_str
: str, text_loc
: TextLocation
, allowed_names
: Set
[str]):
1734 self
._expr
_str
= expr_str
1735 self
._text
_loc
= text_loc
1736 self
._allowed
_names
= allowed_names
1738 def _visit_name(self
, name
: str):
1739 # Make sure the name refers to a known and reachable
1740 # variable/label name.
1741 if name
!= _icitte_name
and name
not in self
._allowed
_names
:
1742 msg
= "Illegal (unknown or unreachable) variable/label name `{}` in expression `{}`".format(
1743 name
, self
._expr
_str
1746 allowed_names
= self
._allowed
_names
.copy()
1747 allowed_names
.add(_icitte_name
)
1749 if len(allowed_names
) > 0:
1750 allowed_names_str
= ", ".join(
1751 sorted(["`{}`".format(name
) for name
in allowed_names
])
1753 msg
+= "; the legal names are {{{}}}".format(allowed_names_str
)
1765 variables
: VariablesT
,
1768 bo
: Optional
[ByteOrder
],
1770 self
.variables
= variables
.copy()
1771 self
.labels
= labels
.copy()
1772 self
.offset
= offset
1776 return "_GenState({}, {}, {}, {})".format(
1777 repr(self
.variables
), repr(self
.labels
), repr(self
.offset
), repr(self
.bo
)
1781 # Fixed-length number item instance.
1782 class _FlNumItemInst
:
1783 def __init__(self
, item
: _FlNum
, offset_in_data
: int, state
: _GenState
):
1785 self
._offset
_in
_data
= offset_in_data
1793 def offset_in_data(self
):
1794 return self
._offset
_in
_data
1801 # Generator of data and final state from a group item.
1803 # Generation happens in memory at construction time. After building, use
1804 # the `data`, `variables`, `labels`, `offset`, and `bo` properties to
1805 # get the resulting context.
1807 # The steps of generation are:
1809 # 1. Handle each item in prefix order.
1811 # The handlers append bytes to `self._data` and update some current
1812 # state object (`_GenState` instance).
1814 # When handling a fixed-length number item, try to evaluate its
1815 # expression using the current state. If this fails, then it might be
1816 # because the expression refers to a "future" label: save the current
1817 # offset in `self._data` (generated data) and a snapshot of the
1818 # current state within `self._fl_num_item_insts` (`_FlNumItemInst`
1819 # object). _gen_fl_num_item_insts() will deal with this later.
1821 # When handling the items of a group, keep a map of immediate label
1822 # names to their offset. Then, after having processed all the items,
1823 # update the relevant saved state snapshots in
1824 # `self._fl_num_item_insts` with those immediate label values.
1825 # _gen_fl_num_item_insts() will deal with this later.
1827 # 2. Handle all the fixed-length number item instances of which the
1828 # expression evaluation failed before.
1830 # At this point, `self._fl_num_item_insts` contains everything that's
1831 # needed to evaluate the expressions, including the values of
1832 # "future" labels from the point of view of some fixed-length number
1835 # If an evaluation fails at this point, then it's a user error.
1840 macro_defs
: _MacroDefsT
,
1841 variables
: VariablesT
,
1844 bo
: Optional
[ByteOrder
],
1846 self
._macro
_defs
= macro_defs
1847 self
._fl
_num
_item
_insts
= [] # type: List[_FlNumItemInst]
1848 self
._gen
(group
, _GenState(variables
, labels
, offset
, bo
))
1855 # Updated variables.
1857 def variables(self
):
1858 return self
._final
_state
.variables
1860 # Updated main group labels.
1863 return self
._final
_state
.labels
1868 return self
._final
_state
.offset
1870 # Updated byte order.
1873 return self
._final
_state
.bo
1875 # Evaluates the expression `expr` of which the original string is
1876 # `expr_str` at the location `text_loc` considering the current
1877 # generation state `state`.
1879 # If `allow_float` is `True`, then the type of the result may be
1884 expr
: ast
.Expression
,
1885 text_loc
: TextLocation
,
1887 allow_float
: bool = False,
1889 syms
= {} # type: VariablesT
1890 syms
.update(state
.labels
)
1892 # Set the `ICITTE` name to the current offset
1893 syms
[_icitte_name
] = state
.offset
1895 # Add the current variables
1896 syms
.update(state
.variables
)
1898 # Validate the node and its children
1899 _ExprValidator(expr_str
, text_loc
, set(syms
.keys())).visit(expr
)
1901 # Compile and evaluate expression node
1903 val
= eval(compile(expr
, "", "eval"), None, syms
)
1904 except Exception as exc
:
1906 "Failed to evaluate expression `{}`: {}".format(expr_str
, exc
),
1910 # Convert `bool` result type to `int` to normalize
1911 if type(val
) is bool:
1914 # Validate result type
1915 expected_types
= {int}
# type: Set[type]
1919 expected_types
.add(float)
1920 type_msg
+= " or `float`"
1922 if type(val
) not in expected_types
:
1924 "Invalid expression `{}`: expecting result type {}, not `{}`".format(
1925 expr_str
, type_msg
, type(val
).__name
__
1932 # Evaluates the expression of `item` considering the current
1933 # generation state `state`.
1935 # If `allow_float` is `True`, then the type of the result may be
1938 def _eval_item_expr(
1939 item
: Union
[_FlNum
, _Leb128Int
, _FillUntil
, _VarAssign
, _Rep
, _Cond
],
1941 allow_float
: bool = False,
1943 return _Gen
._eval
_expr
(
1944 item
.expr_str
, item
.expr
, item
.text_loc
, state
, allow_float
1947 # Handles the byte item `item`.
1948 def _handle_byte_item(self
, item
: _Byte
, state
: _GenState
):
1949 self
._data
.append(item
.val
)
1950 state
.offset
+= item
.size
1952 # Handles the string item `item`.
1953 def _handle_str_item(self
, item
: _Str
, state
: _GenState
):
1954 self
._data
+= item
.data
1955 state
.offset
+= item
.size
1957 # Handles the byte order setting item `item`.
1958 def _handle_set_bo_item(self
, item
: _SetBo
, state
: _GenState
):
1959 # Update current byte order
1962 # Handles the variable assignment item `item`.
1963 def _handle_var_assign_item(self
, item
: _VarAssign
, state
: _GenState
):
1965 state
.variables
[item
.name
] = self
._eval
_item
_expr
(item
, state
, True)
1967 # Handles the fixed-length number item `item`.
1968 def _handle_fl_num_item(self
, item
: _FlNum
, state
: _GenState
):
1969 # Validate current byte order
1970 if state
.bo
is None and item
.len > 8:
1971 _raise_error_for_item(
1972 "Current byte order isn't defined at first fixed-length number (`{}`) to encode on more than 8 bits".format(
1978 # Try an immediate evaluation. If it fails, then keep everything
1979 # needed to (try to) generate the bytes of this item later.
1981 data
= self
._gen
_fl
_num
_item
_inst
_data
(item
, state
)
1983 self
._fl
_num
_item
_insts
.append(
1984 _FlNumItemInst(item
, len(self
._data
), copy
.deepcopy(state
))
1987 # Reserve space in `self._data` for this instance
1988 data
= bytes([0] * (item
.len // 8))
1994 state
.offset
+= len(data
)
1996 # Returns the size, in bytes, required to encode the value `val`
1997 # with LEB128 (signed version if `is_signed` is `True`).
1999 def _leb128_size_for_val(val
: int, is_signed
: bool):
2001 # Equivalent upper bound.
2003 # For example, if `val` is -128, then the full integer for
2004 # this number of bits would be [-128, 127].
2007 # Number of bits (add one for the sign if needed)
2008 bits
= val
.bit_length() + int(is_signed
)
2013 # Seven bits per byte
2014 return math
.ceil(bits
/ 7)
2016 # Handles the LEB128 integer item `item`.
2017 def _handle_leb128_int_item(self
, item
: _Leb128Int
, state
: _GenState
):
2019 val
= self
._eval
_item
_expr
(item
, state
, False)
2022 size
= self
._leb
128_size
_for
_val
(val
, type(item
) is _SLeb128Int
)
2025 for _
in range(size
):
2026 # Seven LSBs, MSB of the byte set (continue)
2027 self
._data
.append((val
& 0x7F) |
0x80)
2030 # Clear MSB of last byte (stop)
2031 self
._data
[-1] &= ~
0x80
2034 state
.offset
+= size
2036 # Handles the group item `item`, removing the immediate labels from
2037 # `state` at the end if `remove_immediate_labels` is `True`.
2038 def _handle_group_item(
2039 self
, item
: _Group
, state
: _GenState
, remove_immediate_labels
: bool = True
2041 first_fl_num_item_inst_index
= len(self
._fl
_num
_item
_insts
)
2042 immediate_labels
= {} # type: LabelsT
2045 for subitem
in item
.items
:
2046 if type(subitem
) is _Label
:
2047 # Add to local immediate labels
2048 immediate_labels
[subitem
.name
] = state
.offset
2050 self
._handle
_item
(subitem
, state
)
2052 # Remove immediate labels from current state if needed
2053 if remove_immediate_labels
:
2054 for name
in immediate_labels
:
2055 del state
.labels
[name
]
2057 # Add all immediate labels to all state snapshots since
2058 # `first_fl_num_item_inst_index`.
2059 for inst
in self
._fl
_num
_item
_insts
[first_fl_num_item_inst_index
:]:
2060 inst
.state
.labels
.update(immediate_labels
)
2062 # Handles the repetition item `item`.
2063 def _handle_rep_item(self
, item
: _Rep
, state
: _GenState
):
2064 # Compute the repetition count
2065 mul
= _Gen
._eval
_item
_expr
(item
, state
)
2069 _raise_error_for_item(
2070 "Invalid expression `{}`: unexpected negative result {:,}".format(
2076 # Generate item data `mul` times
2077 for _
in range(mul
):
2078 self
._handle
_item
(item
.item
, state
)
2080 # Handles the conditional item `item`.
2081 def _handle_cond_item(self
, item
: _Rep
, state
: _GenState
):
2082 # Compute the conditional value
2083 val
= _Gen
._eval
_item
_expr
(item
, state
)
2085 # Generate item data if needed
2087 self
._handle
_item
(item
.item
, state
)
2089 # Evaluates the parameters of the macro expansion item `item`
2090 # considering the initial state `init_state` and returns a new state
2091 # to handle the items of the macro.
2092 def _eval_macro_exp_params(self
, item
: _MacroExp
, init_state
: _GenState
):
2094 exp_state
= _GenState({}, {}, init_state
.offset
, init_state
.bo
)
2096 # Evaluate the parameter expressions
2097 macro_def
= self
._macro
_defs
[item
.name
]
2099 for param_name
, param
in zip(macro_def
.param_names
, item
.params
):
2100 exp_state
.variables
[param_name
] = _Gen
._eval
_expr
(
2101 param
.expr_str
, param
.expr
, param
.text_loc
, init_state
, True
2106 # Handles the macro expansion item `item`.
2107 def _handle_macro_exp_item(self
, item
: _MacroExp
, state
: _GenState
):
2109 exp_state
= self
._eval
_macro
_exp
_params
(item
, state
)
2111 # Process the contained group
2112 init_data_size
= len(self
._data
)
2113 self
._handle
_item
(self
._macro
_defs
[item
.name
].group
, exp_state
)
2115 # Update state offset and return
2116 state
.offset
+= len(self
._data
) - init_data_size
2118 # Handles the offset setting item `item`.
2119 def _handle_set_offset_item(self
, item
: _SetOffset
, state
: _GenState
):
2120 state
.offset
= item
.val
2122 # Handles the offset alignment item `item` (adds padding).
2123 def _handle_align_offset_item(self
, item
: _AlignOffset
, state
: _GenState
):
2124 init_offset
= state
.offset
2125 align_bytes
= item
.val
// 8
2126 state
.offset
= (state
.offset
+ align_bytes
- 1) // align_bytes
* align_bytes
2127 self
._data
+= bytes([item
.pad_val
] * (state
.offset
- init_offset
))
2129 # Handles the filling item `item` (adds padding).
2130 def _handle_fill_until_item(self
, item
: _FillUntil
, state
: _GenState
):
2131 # Compute the new offset
2132 new_offset
= _Gen
._eval
_item
_expr
(item
, state
)
2134 # Validate the new offset
2135 if new_offset
< state
.offset
:
2136 _raise_error_for_item(
2137 "Invalid expression `{}`: new offset {:,} is less than current offset {:,}".format(
2138 item
.expr_str
, new_offset
, state
.offset
2144 self
._data
+= bytes([item
.pad_val
] * (new_offset
- state
.offset
))
2147 state
.offset
= new_offset
2149 # Handles the label item `item`.
2150 def _handle_label_item(self
, item
: _Label
, state
: _GenState
):
2151 state
.labels
[item
.name
] = state
.offset
2153 # Handles the item `item`, returning the updated next repetition
2155 def _handle_item(self
, item
: _Item
, state
: _GenState
):
2156 return self
._item
_handlers
[type(item
)](item
, state
)
2158 # Generates the data for a fixed-length integer item instance having
2159 # the value `val` and returns it.
2160 def _gen_fl_int_item_inst_data(self
, val
: int, item
: _FlNum
, state
: _GenState
):
2162 if val
< -(2 ** (item
.len - 1)) or val
> 2**item
.len - 1:
2163 _raise_error_for_item(
2164 "Value {:,} is outside the {}-bit range when evaluating expression `{}`".format(
2165 val
, item
.len, item
.expr_str
2170 # Encode result on 64 bits (to extend the sign bit whatever the
2171 # value of `item.len`).
2174 ">" if state
.bo
in (None, ByteOrder
.BE
) else "<",
2175 "Q" if val
>= 0 else "q",
2180 # Keep only the requested length
2181 len_bytes
= item
.len // 8
2183 if state
.bo
in (None, ByteOrder
.BE
):
2184 # Big endian: keep last bytes
2185 data
= data
[-len_bytes
:]
2187 # Little endian: keep first bytes
2188 assert state
.bo
== ByteOrder
.LE
2189 data
= data
[:len_bytes
]
2194 # Generates the data for a fixed-length floating point number item
2195 # instance having the value `val` and returns it.
2196 def _gen_fl_float_item_inst_data(self
, val
: float, item
: _FlNum
, state
: _GenState
):
2198 if item
.len not in (32, 64):
2199 _raise_error_for_item(
2200 "Invalid {}-bit length for a fixed-length floating point number (value {:,})".format(
2206 # Encode and return result
2209 ">" if state
.bo
in (None, ByteOrder
.BE
) else "<",
2210 "f" if item
.len == 32 else "d",
2215 # Generates the data for a fixed-length number item instance and
2217 def _gen_fl_num_item_inst_data(self
, item
: _FlNum
, state
: _GenState
):
2219 val
= self
._eval
_item
_expr
(item
, state
, True)
2221 # Handle depending on type
2222 if type(val
) is int:
2223 return self
._gen
_fl
_int
_item
_inst
_data
(val
, item
, state
)
2225 assert type(val
) is float
2226 return self
._gen
_fl
_float
_item
_inst
_data
(val
, item
, state
)
2228 # Generates the data for all the fixed-length number item instances
2229 # and writes it at the correct offset within `self._data`.
2230 def _gen_fl_num_item_insts(self
):
2231 for inst
in self
._fl
_num
_item
_insts
:
2233 data
= self
._gen
_fl
_num
_item
_inst
_data
(inst
.item
, inst
.state
)
2235 # Insert bytes into `self._data`
2236 self
._data
[inst
.offset_in_data
: inst
.offset_in_data
+ len(data
)] = data
2238 # Generates the data (`self._data`) and final state
2239 # (`self._final_state`) from `group` and the initial state `state`.
2240 def _gen(self
, group
: _Group
, state
: _GenState
):
2242 self
._data
= bytearray()
2245 self
._item
_handlers
= {
2246 _AlignOffset
: self
._handle
_align
_offset
_item
,
2247 _Byte
: self
._handle
_byte
_item
,
2248 _Cond
: self
._handle
_cond
_item
,
2249 _FillUntil
: self
._handle
_fill
_until
_item
,
2250 _FlNum
: self
._handle
_fl
_num
_item
,
2251 _Group
: self
._handle
_group
_item
,
2252 _Label
: self
._handle
_label
_item
,
2253 _MacroExp
: self
._handle
_macro
_exp
_item
,
2254 _Rep
: self
._handle
_rep
_item
,
2255 _SetBo
: self
._handle
_set
_bo
_item
,
2256 _SetOffset
: self
._handle
_set
_offset
_item
,
2257 _SLeb128Int
: self
._handle
_leb
128_int
_item
,
2258 _Str
: self
._handle
_str
_item
,
2259 _ULeb128Int
: self
._handle
_leb
128_int
_item
,
2260 _VarAssign
: self
._handle
_var
_assign
_item
,
2261 } # type: Dict[type, Callable[[Any, _GenState], None]]
2263 # Handle the group item, _not_ removing the immediate labels
2264 # because the `labels` property offers them.
2265 self
._handle
_group
_item
(group
, state
, False)
2267 # This is actually the final state
2268 self
._final
_state
= state
2270 # Generate all the fixed-length number bytes now that we know
2272 self
._gen
_fl
_num
_item
_insts
()
2275 # Returns a `ParseResult` instance containing the bytes encoded by the
2276 # input string `normand`.
2278 # `init_variables` is a dictionary of initial variable names (valid
2279 # Python names) to integral values. A variable name must not be the
2280 # reserved name `ICITTE`.
2282 # `init_labels` is a dictionary of initial label names (valid Python
2283 # names) to integral values. A label name must not be the reserved name
2286 # `init_offset` is the initial offset.
2288 # `init_byte_order` is the initial byte order.
2290 # Raises `ParseError` on any parsing error.
2293 init_variables
: Optional
[VariablesT
] = None,
2294 init_labels
: Optional
[LabelsT
] = None,
2295 init_offset
: int = 0,
2296 init_byte_order
: Optional
[ByteOrder
] = None,
2298 if init_variables
is None:
2301 if init_labels
is None:
2304 parser
= _Parser(normand
, init_variables
, init_labels
)
2313 return ParseResult
._create
( # pyright: ignore[reportPrivateUsage]
2314 gen
.data
, gen
.variables
, gen
.labels
, gen
.offset
, gen
.bo
2318 # Parses the command-line arguments.
2319 def _parse_cli_args():
2323 ap
= argparse
.ArgumentParser()
2330 help="initial offset (positive)",
2336 choices
=["be", "le"],
2338 help="initial byte order (`be` or `le`)",
2344 help="add an initial variable (may be repeated)",
2351 help="add an initial label (may be repeated)",
2354 "--version", action
="version", version
="Normand {}".format(__version__
)
2361 help="input path (none means standard input)",
2365 return ap
.parse_args()
2368 # Raises a command-line error with the message `msg`.
2369 def _raise_cli_error(msg
: str) -> NoReturn
:
2370 raise RuntimeError("Command-line error: {}".format(msg
))
2373 # Returns a dictionary of string to integers from the list of strings
2374 # `args` containing `NAME=VAL` entries.
2375 def _dict_from_arg(args
: Optional
[List
[str]]):
2376 d
= {} # type: LabelsT
2382 m
= re
.match(r
"({})=(\d+)$".format(_py_name_pat
.pattern
), arg
)
2385 _raise_cli_error("Invalid assignment {}".format(arg
))
2387 d
[m
.group(1)] = int(m
.group(2))
2392 # CLI entry point without exception handling.
2397 args
= _parse_cli_args()
2400 if args
.path
is None:
2401 normand
= sys
.stdin
.read()
2403 with
open(args
.path
) as f
:
2406 # Variables and labels
2407 variables
= typing
.cast(VariablesT
, _dict_from_arg(args
.var
))
2408 labels
= _dict_from_arg(args
.label
)
2412 _raise_cli_error("Invalid negative offset {}")
2414 # Validate and set byte order
2415 bo
= None # type: Optional[ByteOrder]
2417 if args
.byte_order
is not None:
2418 if args
.byte_order
== "be":
2421 assert args
.byte_order
== "le"
2426 res
= parse(normand
, variables
, labels
, args
.offset
, bo
)
2427 except ParseError
as exc
:
2430 if args
.path
is not None:
2431 prefix
= "{}:".format(os
.path
.abspath(args
.path
))
2434 "{}{}:{} - {}".format(
2435 prefix
, exc
.text_loc
.line_no
, exc
.text_loc
.col_no
, str(exc
)
2440 sys
.stdout
.buffer.write(res
.data
)
2443 # Prints the exception message `msg` and exits with status 1.
2444 def _fail(msg
: str) -> NoReturn
:
2445 if not msg
.endswith("."):
2448 print(msg
, file=sys
.stderr
)
2456 except Exception as exc
:
2460 if __name__
== "__main__":