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__
= "0.15.0"
56 from typing
import Any
, Set
, Dict
, List
, Union
, Pattern
, Callable
, NoReturn
, Optional
59 # Text location (line and column numbers).
62 def _create(cls
, line_no
: int, col_no
: int):
63 self
= cls
.__new
__(cls
)
64 self
._init
(line_no
, col_no
)
67 def __init__(*args
, **kwargs
): # type: ignore
68 raise NotImplementedError
70 def _init(self
, line_no
: int, col_no
: int):
71 self
._line
_no
= line_no
85 return "TextLocation({}, {})".format(self
._line
_no
, self
._col
_no
)
90 def __init__(self
, text_loc
: TextLocation
):
91 self
._text
_loc
= text_loc
93 # Source text location.
100 class _ScalarItem(_Item
):
101 # Returns the size, in bytes, of this item.
104 def size(self
) -> int:
114 class _Byte(_ScalarItem
, _RepableItem
):
115 def __init__(self
, val
: int, text_loc
: TextLocation
):
116 super().__init
__(text_loc
)
129 return "_Byte({}, {})".format(hex(self
._val
), repr(self
._text
_loc
))
133 class _Str(_ScalarItem
, _RepableItem
):
134 def __init__(self
, data
: bytes
, text_loc
: TextLocation
):
135 super().__init
__(text_loc
)
145 return len(self
._data
)
148 return "_Str({}, {})".format(repr(self
._data
), repr(self
._text
_loc
))
153 class ByteOrder(enum
.Enum
):
161 # Byte order setting.
163 def __init__(self
, bo
: ByteOrder
, text_loc
: TextLocation
):
164 super().__init
__(text_loc
)
172 return "_SetBo({}, {})".format(repr(self
._bo
), repr(self
._text
_loc
))
177 def __init__(self
, name
: str, text_loc
: TextLocation
):
178 super().__init
__(text_loc
)
187 return "_Label({}, {})".format(repr(self
._name
), repr(self
._text
_loc
))
191 class _SetOffset(_Item
):
192 def __init__(self
, val
: int, text_loc
: TextLocation
):
193 super().__init
__(text_loc
)
196 # Offset value (bytes).
202 return "_SetOffset({}, {})".format(repr(self
._val
), repr(self
._text
_loc
))
206 class _AlignOffset(_Item
):
207 def __init__(self
, val
: int, pad_val
: int, text_loc
: TextLocation
):
208 super().__init
__(text_loc
)
210 self
._pad
_val
= pad_val
212 # Alignment value (bits).
217 # Padding byte value.
223 return "_AlignOffset({}, {}, {})".format(
224 repr(self
._val
), repr(self
._pad
_val
), repr(self
._text
_loc
)
228 # Mixin of containing an AST expression and its string.
230 def __init__(self
, expr_str
: str, expr
: ast
.Expression
):
231 self
._expr
_str
= expr_str
237 return self
._expr
_str
239 # Expression node to evaluate.
245 # Fill until some offset.
246 class _FillUntil(_Item
, _ExprMixin
):
248 self
, expr_str
: str, expr
: ast
.Expression
, pad_val
: int, text_loc
: TextLocation
250 super().__init
__(text_loc
)
251 _ExprMixin
.__init
__(self
, expr_str
, expr
)
252 self
._pad
_val
= pad_val
254 # Padding byte value.
260 return "_FillUntil({}, {}, {}, {})".format(
261 repr(self
._expr
_str
),
264 repr(self
._text
_loc
),
268 # Variable assignment.
269 class _VarAssign(_Item
, _ExprMixin
):
271 self
, name
: str, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
273 super().__init
__(text_loc
)
274 _ExprMixin
.__init
__(self
, expr_str
, expr
)
283 return "_VarAssign({}, {}, {}, {})".format(
285 repr(self
._expr
_str
),
287 repr(self
._text
_loc
),
291 # Fixed-length number, possibly needing more than one byte.
292 class _FlNum(_ScalarItem
, _RepableItem
, _ExprMixin
):
294 self
, expr_str
: str, expr
: ast
.Expression
, len: int, text_loc
: TextLocation
296 super().__init
__(text_loc
)
297 _ExprMixin
.__init
__(self
, expr_str
, expr
)
307 return self
._len
// 8
310 return "_FlNum({}, {}, {}, {})".format(
311 repr(self
._expr
_str
),
314 repr(self
._text
_loc
),
319 class _Leb128Int(_Item
, _RepableItem
, _ExprMixin
):
320 def __init__(self
, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
):
321 super().__init
__(text_loc
)
322 _ExprMixin
.__init
__(self
, expr_str
, expr
)
325 return "{}({}, {}, {})".format(
326 self
.__class
__.__name
__,
327 repr(self
._expr
_str
),
329 repr(self
._text
_loc
),
333 # Unsigned LEB128 integer.
334 class _ULeb128Int(_Leb128Int
, _RepableItem
, _ExprMixin
):
338 # Signed LEB128 integer.
339 class _SLeb128Int(_Leb128Int
, _RepableItem
, _ExprMixin
):
344 class _Group(_Item
, _RepableItem
):
345 def __init__(self
, items
: List
[_Item
], text_loc
: TextLocation
):
346 super().__init
__(text_loc
)
355 return "_Group({}, {})".format(repr(self
._items
), repr(self
._text
_loc
))
359 class _Rep(_Item
, _ExprMixin
):
361 self
, item
: _Item
, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
363 super().__init
__(text_loc
)
364 _ExprMixin
.__init
__(self
, expr_str
, expr
)
373 return "_Rep({}, {}, {}, {})".format(
375 repr(self
._expr
_str
),
377 repr(self
._text
_loc
),
382 class _Cond(_Item
, _ExprMixin
):
388 expr
: ast
.Expression
,
389 text_loc
: TextLocation
,
391 super().__init
__(text_loc
)
392 _ExprMixin
.__init
__(self
, expr_str
, expr
)
393 self
._true
_item
= true_item
394 self
._false
_item
= false_item
396 # Item when condition is true.
399 return self
._true
_item
401 # Item when condition is false.
403 def false_item(self
):
404 return self
._false
_item
407 return "_Cond({}, {}, {}, {}, {})".format(
408 repr(self
._true
_item
),
409 repr(self
._false
_item
),
410 repr(self
._expr
_str
),
412 repr(self
._text
_loc
),
416 # Macro definition item.
417 class _MacroDef(_Item
):
419 self
, name
: str, param_names
: List
[str], group
: _Group
, text_loc
: TextLocation
421 super().__init
__(text_loc
)
423 self
._param
_names
= param_names
433 def param_names(self
):
434 return self
._param
_names
442 return "_MacroDef({}, {}, {}, {})".format(
444 repr(self
._param
_names
),
446 repr(self
._text
_loc
),
450 # Macro expansion parameter.
451 class _MacroExpParam
:
452 def __init__(self
, expr_str
: str, expr
: ast
.Expression
, text_loc
: TextLocation
):
453 self
._expr
_str
= expr_str
455 self
._text
_loc
= text_loc
460 return self
._expr
_str
467 # Source text location.
470 return self
._text
_loc
473 return "_MacroExpParam({}, {}, {})".format(
474 repr(self
._expr
_str
), repr(self
._expr
), repr(self
._text
_loc
)
478 # Macro expansion item.
479 class _MacroExp(_Item
, _RepableItem
):
483 params
: List
[_MacroExpParam
],
484 text_loc
: TextLocation
,
486 super().__init
__(text_loc
)
488 self
._params
= params
501 return "_MacroExp({}, {}, {})".format(
504 repr(self
._text
_loc
),
508 # A parsing error message: a string and a text location.
509 class ParseErrorMessage
:
511 def _create(cls
, text
: str, text_loc
: TextLocation
):
512 self
= cls
.__new
__(cls
)
513 self
._init
(text
, text_loc
)
516 def __init__(self
, *args
, **kwargs
): # type: ignore
517 raise NotImplementedError
519 def _init(self
, text
: str, text_loc
: TextLocation
):
521 self
._text
_loc
= text_loc
528 # Source text location.
530 def text_location(self
):
531 return self
._text
_loc
534 # A parsing error containing one or more messages (`ParseErrorMessage`).
535 class ParseError(RuntimeError):
537 def _create(cls
, msg
: str, text_loc
: TextLocation
):
538 self
= cls
.__new
__(cls
)
539 self
._init
(msg
, text_loc
)
542 def __init__(self
, *args
, **kwargs
): # type: ignore
543 raise NotImplementedError
545 def _init(self
, msg
: str, text_loc
: TextLocation
):
546 super().__init
__(msg
)
547 self
._msgs
= [] # type: List[ParseErrorMessage]
548 self
._add
_msg
(msg
, text_loc
)
550 def _add_msg(self
, msg
: str, text_loc
: TextLocation
):
552 ParseErrorMessage
._create
( # pyright: ignore[reportPrivateUsage]
557 # Parsing error messages.
559 # The first message is the most specific one.
565 # Raises a parsing error, forwarding the parameters to the constructor.
566 def _raise_error(msg
: str, text_loc
: TextLocation
) -> NoReturn
:
567 raise ParseError
._create
(msg
, text_loc
) # pyright: ignore[reportPrivateUsage]
570 # Adds a message to the parsing error `exc`.
571 def _add_error_msg(exc
: ParseError
, msg
: str, text_loc
: TextLocation
):
572 exc
._add
_msg
(msg
, text_loc
) # pyright: ignore[reportPrivateUsage]
575 # Appends a message to the parsing error `exc` and reraises it.
576 def _augment_error(exc
: ParseError
, msg
: str, text_loc
: TextLocation
) -> NoReturn
:
577 _add_error_msg(exc
, msg
, text_loc
)
581 # Variables dictionary type (for type hints).
582 VariablesT
= Dict
[str, Union
[int, float]]
585 # Labels dictionary type (for type hints).
586 LabelsT
= Dict
[str, int]
589 # Python name pattern.
590 _py_name_pat
= re
.compile(r
"[a-zA-Z_][a-zA-Z0-9_]*")
593 # Macro definition dictionary.
594 _MacroDefsT
= Dict
[str, _MacroDef
]
599 # The constructor accepts a Normand input. After building, use the `res`
600 # property to get the resulting main group.
602 # Builds a parser to parse the Normand input `normand`, parsing
604 def __init__(self
, normand
: str, variables
: VariablesT
, labels
: LabelsT
):
605 self
._normand
= normand
609 self
._label
_names
= set(labels
.keys())
610 self
._var
_names
= set(variables
.keys())
611 self
._macro
_defs
= {} # type: _MacroDefsT
614 # Result (main group).
621 def macro_defs(self
):
622 return self
._macro
_defs
624 # Current text location.
627 return TextLocation
._create
( # pyright: ignore[reportPrivateUsage]
628 self
._line
_no
, self
._col
_no
631 # Returns `True` if this parser is done parsing.
633 return self
._at
== len(self
._normand
)
635 # Returns `True` if this parser isn't done parsing.
636 def _isnt_done(self
):
637 return not self
._is
_done
()
639 # Raises a parse error, creating it using the message `msg` and the
640 # current text location.
641 def _raise_error(self
, msg
: str) -> NoReturn
:
642 _raise_error(msg
, self
._text
_loc
)
644 # Tries to make the pattern `pat` match the current substring,
645 # returning the match object and updating `self._at`,
646 # `self._line_no`, and `self._col_no` on success.
647 def _try_parse_pat(self
, pat
: Pattern
[str]):
648 m
= pat
.match(self
._normand
, self
._at
)
653 # Skip matched string
654 self
._at
+= len(m
.group(0))
657 self
._line
_no
+= m
.group(0).count("\n")
659 # Update column number
660 for i
in reversed(range(self
._at
)):
661 if self
._normand
[i
] == "\n" or i
== 0:
663 self
._col
_no
= self
._at
+ 1
665 self
._col
_no
= self
._at
- i
669 # Return match object
672 # Expects the pattern `pat` to match the current substring,
673 # returning the match object and updating `self._at`,
674 # `self._line_no`, and `self._col_no` on success, or raising a parse
675 # error with the message `error_msg` on error.
676 def _expect_pat(self
, pat
: Pattern
[str], error_msg
: str):
678 m
= self
._try
_parse
_pat
(pat
)
682 self
._raise
_error
(error_msg
)
684 # Return match object
687 # Pattern for _skip_ws_and_comments()
688 _ws_or_syms_or_comments_pat
= re
.compile(
689 r
"(?:[\s/\\?&:;.,[\]_=|-]|#[^#]*?(?:\n|#))*"
692 # Skips as many whitespaces, insignificant symbol characters, and
693 # comments as possible.
694 def _skip_ws_and_comments(self
):
695 self
._try
_parse
_pat
(self
._ws
_or
_syms
_or
_comments
_pat
)
697 # Pattern for _skip_ws()
698 _ws_pat
= re
.compile(r
"\s*")
700 # Skips as many whitespaces as possible.
702 self
._try
_parse
_pat
(self
._ws
_pat
)
704 # Pattern for _try_parse_hex_byte()
705 _nibble_pat
= re
.compile(r
"[A-Fa-f0-9]")
707 # Tries to parse a hexadecimal byte, returning a byte item on
709 def _try_parse_hex_byte(self
):
710 begin_text_loc
= self
._text
_loc
712 # Match initial nibble
713 m_high
= self
._try
_parse
_pat
(self
._nibble
_pat
)
719 # Expect another nibble
720 self
._skip
_ws
_and
_comments
()
721 m_low
= self
._expect
_pat
(
722 self
._nibble
_pat
, "Expecting another hexadecimal nibble"
726 return _Byte(int(m_high
.group(0) + m_low
.group(0), 16), begin_text_loc
)
728 # Patterns for _try_parse_bin_byte()
729 _bin_byte_bit_pat
= re
.compile(r
"[01]")
730 _bin_byte_prefix_pat
= re
.compile(r
"%")
732 # Tries to parse a binary byte, returning a byte item on success.
733 def _try_parse_bin_byte(self
):
734 begin_text_loc
= self
._text
_loc
737 if self
._try
_parse
_pat
(self
._bin
_byte
_prefix
_pat
) is None:
742 bits
= [] # type: List[str]
745 self
._skip
_ws
_and
_comments
()
746 m
= self
._expect
_pat
(self
._bin
_byte
_bit
_pat
, "Expecting a bit (`0` or `1`)")
747 bits
.append(m
.group(0))
750 return _Byte(int("".join(bits
), 2), begin_text_loc
)
752 # Patterns for _try_parse_dec_byte()
753 _dec_byte_prefix_pat
= re
.compile(r
"\$")
754 _dec_byte_val_pat
= re
.compile(r
"(?P<neg>-?)(?P<val>\d+)")
756 # Tries to parse a decimal byte, returning a byte item on success.
757 def _try_parse_dec_byte(self
):
758 begin_text_loc
= self
._text
_loc
761 if self
._try
_parse
_pat
(self
._dec
_byte
_prefix
_pat
) is None:
767 m
= self
._expect
_pat
(self
._dec
_byte
_val
_pat
, "Expecting a decimal constant")
770 val
= int(m
.group("val")) * (-1 if m
.group("neg") == "-" else 1)
773 if val
< -128 or val
> 255:
774 _raise_error("Invalid decimal byte value {}".format(val
), begin_text_loc
)
780 return _Byte(val
, begin_text_loc
)
782 # Tries to parse a byte, returning a byte item on success.
783 def _try_parse_byte(self
):
785 item
= self
._try
_parse
_hex
_byte
()
791 item
= self
._try
_parse
_bin
_byte
()
797 item
= self
._try
_parse
_dec
_byte
()
802 # Patterns for _try_parse_str()
803 _str_prefix_pat
= re
.compile(r
'(?:u(?P<len>16|32)(?P<bo>be|le))?\s*"')
804 _str_suffix_pat
= re
.compile(r
'"')
805 _str_str_pat
= re
.compile(r
'(?:(?:\\.)|[^"])*')
807 # Strings corresponding to escape sequence characters
808 _str_escape_seq_strs
= {
822 # Tries to parse a string, returning a string item on success.
823 def _try_parse_str(self
):
824 begin_text_loc
= self
._text
_loc
827 m
= self
._try
_parse
_pat
(self
._str
_prefix
_pat
)
836 if m
.group("len") is not None:
837 encoding
= "utf_{}_{}".format(m
.group("len"), m
.group("bo"))
840 m
= self
._expect
_pat
(self
._str
_str
_pat
, "Expecting a literal string")
842 # Expect end of string
843 self
._expect
_pat
(self
._str
_suffix
_pat
, 'Expecting `"` (end of literal string)')
845 # Replace escape sequences
848 for ec
in '0abefnrtv"\\':
849 val
= val
.replace(r
"\{}".format(ec
), self
._str
_escape
_seq
_strs
[ec
])
852 data
= val
.encode(encoding
)
855 return _Str(data
, begin_text_loc
)
857 # Common right parenthesis pattern
858 _right_paren_pat
= re
.compile(r
"\)")
860 # Patterns for _try_parse_group()
861 _group_prefix_pat
= re
.compile(r
"\(|!g(?:roup)?\b")
863 # Tries to parse a group, returning a group item on success.
864 def _try_parse_group(self
):
865 begin_text_loc
= self
._text
_loc
868 m_open
= self
._try
_parse
_pat
(self
._group
_prefix
_pat
)
875 items
= self
._parse
_items
()
877 # Expect end of group
878 self
._skip
_ws
_and
_comments
()
880 if m_open
.group(0) == "(":
881 pat
= self
._right
_paren
_pat
884 pat
= self
._block
_end
_pat
887 self
._expect
_pat
(pat
, "Expecting an item or `{}` (end of group)".format(exp
))
890 return _Group(items
, begin_text_loc
)
892 # Returns a stripped expression string and an AST expression node
893 # from the expression string `expr_str` at text location `text_loc`.
894 def _ast_expr_from_str(self
, expr_str
: str, text_loc
: TextLocation
):
895 # Create an expression node from the expression string
896 expr_str
= expr_str
.strip().replace("\n", " ")
899 expr
= ast
.parse(expr_str
, mode
="eval")
902 "Invalid expression `{}`: invalid syntax".format(expr_str
),
906 return expr_str
, expr
908 # Patterns for _try_parse_num_and_attr()
909 _val_expr_pat
= re
.compile(r
"([^}:]+):\s*")
910 _fl_num_len_attr_pat
= re
.compile(r
"8|16|24|32|40|48|56|64")
911 _leb128_int_attr_pat
= re
.compile(r
"(u|s)leb128")
913 # Tries to parse a value and attribute (fixed length in bits or
914 # `leb128`), returning a value item on success.
915 def _try_parse_num_and_attr(self
):
916 begin_text_loc
= self
._text
_loc
919 m_expr
= self
._try
_parse
_pat
(self
._val
_expr
_pat
)
925 # Create an expression node from the expression string
926 expr_str
, expr
= self
._ast
_expr
_from
_str
(m_expr
.group(1), begin_text_loc
)
929 m_attr
= self
._try
_parse
_pat
(self
._fl
_num
_len
_attr
_pat
)
933 m_attr
= self
._try
_parse
_pat
(self
._leb
128_int
_attr
_pat
)
936 # At this point it's invalid
938 "Expecting a length (multiple of eight bits), `uleb128`, or `sleb128`"
941 # Return LEB128 integer item
942 cls
= _ULeb128Int
if m_attr
.group(1) == "u" else _SLeb128Int
943 return cls(expr_str
, expr
, begin_text_loc
)
945 # Return fixed-length number item
949 int(m_attr
.group(0)),
953 # Patterns for _try_parse_var_assign()
954 _var_assign_name_equal_pat
= re
.compile(r
"({})\s*=".format(_py_name_pat
.pattern
))
955 _var_assign_expr_pat
= re
.compile(r
"[^}]+")
957 # Tries to parse a variable assignment, returning a variable
958 # assignment item on success.
959 def _try_parse_var_assign(self
):
960 begin_text_loc
= self
._text
_loc
963 m
= self
._try
_parse
_pat
(self
._var
_assign
_name
_equal
_pat
)
972 if name
== _icitte_name
:
974 "`{}` is a reserved variable name".format(_icitte_name
), begin_text_loc
977 if name
in self
._label
_names
:
978 _raise_error("Existing label named `{}`".format(name
), begin_text_loc
)
980 # Expect an expression
982 m
= self
._expect
_pat
(self
._var
_assign
_expr
_pat
, "Expecting an expression")
984 # Create an expression node from the expression string
985 expr_str
, expr
= self
._ast
_expr
_from
_str
(m
.group(0), begin_text_loc
)
987 # Add to known variable names
988 self
._var
_names
.add(name
)
998 # Pattern for _try_parse_set_bo()
999 _bo_pat
= re
.compile(r
"[bl]e")
1001 # Tries to parse a byte order name, returning a byte order setting
1003 def _try_parse_set_bo(self
):
1004 begin_text_loc
= self
._text
_loc
1007 m
= self
._try
_parse
_pat
(self
._bo
_pat
)
1013 # Return corresponding item
1014 if m
.group(0) == "be":
1015 return _SetBo(ByteOrder
.BE
, begin_text_loc
)
1017 assert m
.group(0) == "le"
1018 return _SetBo(ByteOrder
.LE
, begin_text_loc
)
1020 # Patterns for _try_parse_val_or_bo()
1021 _val_var_assign_set_bo_prefix_pat
= re
.compile(r
"\{")
1022 _val_var_assign_set_bo_suffix_pat
= re
.compile(r
"\}")
1024 # Tries to parse a value, a variable assignment, or a byte order
1025 # setting, returning an item on success.
1026 def _try_parse_val_or_var_assign_or_set_bo(self
):
1028 if self
._try
_parse
_pat
(self
._val
_var
_assign
_set
_bo
_prefix
_pat
) is None:
1034 # Variable assignment item?
1035 item
= self
._try
_parse
_var
_assign
()
1039 item
= self
._try
_parse
_num
_and
_attr
()
1042 # Byte order setting item?
1043 item
= self
._try
_parse
_set
_bo
()
1046 # At this point it's invalid
1048 "Expecting a fixed-length number, a variable assignment, or a byte order setting"
1053 self
._expect
_pat
(self
._val
_var
_assign
_set
_bo
_suffix
_pat
, "Expecting `}`")
1056 # Returns a normalized version (so as to be parseable by int()) of
1057 # the constant integer string `s`, possibly negative, dealing with
1060 def _norm_const_int(s
: str):
1064 if s
.startswith("-"):
1069 if pos
.startswith("0" + r
):
1085 for suf
in asm_suf_base
:
1087 s
= "{}0{}{}".format(neg
, asm_suf_base
[suf
], pos
.rstrip(suf
))
1091 # Common constant integer patterns
1092 _pos_const_int_pat
= re
.compile(
1093 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+"
1095 _const_int_pat
= re
.compile(r
"(?P<neg>-)?(?:{})".format(_pos_const_int_pat
.pattern
))
1097 # Tries to parse an offset setting value (after the initial `<`),
1098 # returning an offset item on success.
1099 def _try_parse_set_offset_val(self
):
1100 begin_text_loc
= self
._text
_loc
1103 m
= self
._try
_parse
_pat
(self
._pos
_const
_int
_pat
)
1110 return _SetOffset(int(self
._norm
_const
_int
(m
.group(0)), 0), begin_text_loc
)
1112 # Tries to parse a label name (after the initial `<`), returning a
1113 # label item on success.
1114 def _try_parse_label_name(self
):
1115 begin_text_loc
= self
._text
_loc
1118 m
= self
._try
_parse
_pat
(_py_name_pat
)
1127 if name
== _icitte_name
:
1129 "`{}` is a reserved label name".format(_icitte_name
), begin_text_loc
1132 if name
in self
._label
_names
:
1133 _raise_error("Duplicate label name `{}`".format(name
), begin_text_loc
)
1135 if name
in self
._var
_names
:
1136 _raise_error("Existing variable named `{}`".format(name
), begin_text_loc
)
1138 # Add to known label names
1139 self
._label
_names
.add(name
)
1142 return _Label(name
, begin_text_loc
)
1144 # Patterns for _try_parse_label_or_set_offset()
1145 _label_set_offset_prefix_pat
= re
.compile(r
"<")
1146 _label_set_offset_suffix_pat
= re
.compile(r
">")
1148 # Tries to parse a label or an offset setting, returning an item on
1150 def _try_parse_label_or_set_offset(self
):
1152 if self
._try
_parse
_pat
(self
._label
_set
_offset
_prefix
_pat
) is None:
1156 # Offset setting item?
1158 item
= self
._try
_parse
_set
_offset
_val
()
1162 item
= self
._try
_parse
_label
_name
()
1165 # At this point it's invalid
1166 self
._raise
_error
("Expecting a label name or an offset setting value")
1170 self
._expect
_pat
(self
._label
_set
_offset
_suffix
_pat
, "Expecting `>`")
1173 # Pattern for _parse_pad_val()
1174 _pad_val_prefix_pat
= re
.compile(r
"~")
1176 # Tries to parse a padding value, returning the padding value, or 0
1178 def _parse_pad_val(self
):
1183 if self
._try
_parse
_pat
(self
._pad
_val
_prefix
_pat
) is not None:
1185 pad_val_text_loc
= self
._text
_loc
1186 m
= self
._expect
_pat
(
1187 self
._pos
_const
_int
_pat
,
1188 "Expecting a positive constant integer (byte value)",
1192 pad_val
= int(self
._norm
_const
_int
(m
.group(0)), 0)
1196 "Invalid padding byte value {}".format(pad_val
),
1202 # Patterns for _try_parse_align_offset()
1203 _align_offset_prefix_pat
= re
.compile(r
"@")
1204 _align_offset_val_pat
= re
.compile(r
"\d+")
1206 # Tries to parse an offset alignment, returning an offset alignment
1208 def _try_parse_align_offset(self
):
1209 begin_text_loc
= self
._text
_loc
1212 if self
._try
_parse
_pat
(self
._align
_offset
_prefix
_pat
) is None:
1216 # Expect an alignment
1218 align_text_loc
= self
._text
_loc
1219 m
= self
._expect
_pat
(
1220 self
._align
_offset
_val
_pat
,
1221 "Expecting an alignment (positive multiple of eight bits)",
1224 # Validate alignment
1225 val
= int(m
.group(0))
1227 if val
<= 0 or (val
% 8) != 0:
1229 "Invalid alignment value {} (not a positive multiple of eight)".format(
1236 pad_val
= self
._parse
_pad
_val
()
1239 return _AlignOffset(val
, pad_val
, begin_text_loc
)
1241 # Patterns for _try_parse_fill_until()
1242 _fill_until_prefix_pat
= re
.compile(r
"\+")
1243 _fill_until_pad_val_prefix_pat
= re
.compile(r
"~")
1245 # Tries to parse a filling, returning a filling item on success.
1246 def _try_parse_fill_until(self
):
1247 begin_text_loc
= self
._text
_loc
1250 if self
._try
_parse
_pat
(self
._fill
_until
_prefix
_pat
) is None:
1256 expr_str
, expr
= self
._expect
_const
_int
_name
_expr
(True)
1259 pad_val
= self
._parse
_pad
_val
()
1262 return _FillUntil(expr_str
, expr
, pad_val
, begin_text_loc
)
1264 # Patterns for _expect_rep_mul_expr()
1265 _inner_expr_prefix_pat
= re
.compile(r
"\{")
1266 _inner_expr_pat
= re
.compile(r
"[^}]+")
1267 _inner_expr_suffix_pat
= re
.compile(r
"\}")
1269 # Parses a constant integer if `accept_const_int` is `True`
1270 # (possibly negative if `allow_neg` is `True`), a name, or an
1271 # expression within `{` and `}`.
1272 def _expect_const_int_name_expr(
1273 self
, accept_const_int
: bool, allow_neg
: bool = False
1275 expr_text_loc
= self
._text
_loc
1280 if accept_const_int
:
1281 m
= self
._try
_parse
_pat
(self
._const
_int
_pat
)
1285 m
= self
._try
_parse
_pat
(_py_name_pat
)
1289 if self
._try
_parse
_pat
(self
._inner
_expr
_prefix
_pat
) is None:
1290 pos_msg
= "" if allow_neg
else "positive "
1292 if accept_const_int
:
1293 mid_msg
= "a {}constant integer, a name, or `{{`".format(
1297 mid_msg
= "a name or `{`"
1299 # At this point it's invalid
1300 self
._raise
_error
("Expecting {}".format(mid_msg
))
1302 # Expect an expression
1304 expr_text_loc
= self
._text
_loc
1305 m
= self
._expect
_pat
(self
._inner
_expr
_pat
, "Expecting an expression")
1306 expr_str
= m
.group(0)
1310 self
._expect
_pat
(self
._inner
_expr
_suffix
_pat
, "Expecting `}`")
1312 expr_str
= m
.group(0)
1314 if m
.group("neg") == "-" and not allow_neg
:
1315 _raise_error("Expecting a positive constant integer", expr_text_loc
)
1317 expr_str
= self
._norm
_const
_int
(m
.group(0))
1319 return self
._ast
_expr
_from
_str
(expr_str
, expr_text_loc
)
1321 # Parses the multiplier expression of a repetition (block or
1322 # post-item) and returns the expression string and AST node.
1323 def _expect_rep_mul_expr(self
):
1324 return self
._expect
_const
_int
_name
_expr
(True)
1326 # Common block end pattern
1327 _block_end_pat
= re
.compile(r
"!end\b")
1329 # Pattern for _try_parse_rep_block()
1330 _rep_block_prefix_pat
= re
.compile(r
"!r(?:epeat)?\b")
1332 # Tries to parse a repetition block, returning a repetition item on
1334 def _try_parse_rep_block(self
):
1335 begin_text_loc
= self
._text
_loc
1338 if self
._try
_parse
_pat
(self
._rep
_block
_prefix
_pat
) is None:
1343 self
._skip
_ws
_and
_comments
()
1344 expr_str
, expr
= self
._expect
_rep
_mul
_expr
()
1347 self
._skip
_ws
_and
_comments
()
1348 items_text_loc
= self
._text
_loc
1349 items
= self
._parse
_items
()
1351 # Expect end of block
1352 self
._skip
_ws
_and
_comments
()
1354 self
._block
_end
_pat
, "Expecting an item or `!end` (end of repetition block)"
1358 return _Rep(_Group(items
, items_text_loc
), expr_str
, expr
, begin_text_loc
)
1360 # Pattern for _try_parse_cond_block()
1361 _cond_block_prefix_pat
= re
.compile(r
"!if\b")
1362 _cond_block_else_pat
= re
.compile(r
"!else\b")
1364 # Tries to parse a conditional block, returning a conditional item
1366 def _try_parse_cond_block(self
):
1367 begin_text_loc
= self
._text
_loc
1370 if self
._try
_parse
_pat
(self
._cond
_block
_prefix
_pat
) is None:
1375 self
._skip
_ws
_and
_comments
()
1376 expr_str
, expr
= self
._expect
_const
_int
_name
_expr
(False)
1378 # Parse "true" items
1379 self
._skip
_ws
_and
_comments
()
1380 true_items_text_loc
= self
._text
_loc
1381 true_items
= self
._parse
_items
()
1382 false_items
= [] # type: List[_Item]
1383 false_items_text_loc
= begin_text_loc
1386 self
._skip
_ws
_and
_comments
()
1388 if self
._try
_parse
_pat
(self
._cond
_block
_else
_pat
) is not None:
1389 # Parse "false" items
1390 self
._skip
_ws
_and
_comments
()
1391 false_items_text_loc
= self
._text
_loc
1392 false_items
= self
._parse
_items
()
1394 # Expect end of block
1396 self
._block
_end
_pat
,
1397 "Expecting an item, `!else`, or `!end` (end of conditional block)",
1402 _Group(true_items
, true_items_text_loc
),
1403 _Group(false_items
, false_items_text_loc
),
1409 # Common left parenthesis pattern
1410 _left_paren_pat
= re
.compile(r
"\(")
1412 # Patterns for _try_parse_macro_def() and _try_parse_macro_exp()
1413 _macro_params_comma_pat
= re
.compile(",")
1415 # Patterns for _try_parse_macro_def()
1416 _macro_def_prefix_pat
= re
.compile(r
"!m(?:acro)?\b")
1418 # Tries to parse a macro definition, adding it to `self._macro_defs`
1419 # and returning `True` on success.
1420 def _try_parse_macro_def(self
):
1421 begin_text_loc
= self
._text
_loc
1424 if self
._try
_parse
_pat
(self
._macro
_def
_prefix
_pat
) is None:
1430 name_text_loc
= self
._text
_loc
1431 m
= self
._expect
_pat
(_py_name_pat
, "Expecting a valid macro name")
1436 if name
in self
._macro
_defs
:
1437 _raise_error("Duplicate macro named `{}`".format(name
), name_text_loc
)
1441 self
._expect
_pat
(self
._left
_paren
_pat
, "Expecting `(`")
1443 # Try to parse comma-separated parameter names
1444 param_names
= [] # type: List[str]
1445 expect_comma
= False
1451 if self
._try
_parse
_pat
(self
._right
_paren
_pat
) is not None:
1457 self
._expect
_pat
(self
._macro
_params
_comma
_pat
, "Expecting `,`")
1459 # Expect parameter name
1461 param_text_loc
= self
._text
_loc
1462 m
= self
._expect
_pat
(_py_name_pat
, "Expecting valid parameter name")
1464 if m
.group(0) in param_names
:
1466 "Duplicate macro parameter named `{}`".format(m
.group(0)),
1470 param_names
.append(m
.group(0))
1474 self
._skip
_ws
_and
_comments
()
1475 items_text_loc
= self
._text
_loc
1476 old_var_names
= self
._var
_names
.copy()
1477 old_label_names
= self
._label
_names
.copy()
1478 self
._var
_names
= set() # type: Set[str]
1479 self
._label
_names
= set() # type: Set[str]
1480 items
= self
._parse
_items
()
1481 self
._var
_names
= old_var_names
1482 self
._label
_names
= old_label_names
1486 self
._block
_end
_pat
, "Expecting an item or `!end` (end of macro block)"
1490 self
._macro
_defs
[name
] = _MacroDef(
1491 name
, param_names
, _Group(items
, items_text_loc
), begin_text_loc
1496 # Patterns for _try_parse_macro_exp()
1497 _macro_exp_prefix_pat
= re
.compile(r
"m\b")
1498 _macro_exp_colon_pat
= re
.compile(r
":")
1500 # Tries to parse a macro expansion, returning a macro expansion item
1502 def _try_parse_macro_exp(self
):
1503 begin_text_loc
= self
._text
_loc
1506 if self
._try
_parse
_pat
(self
._macro
_exp
_prefix
_pat
) is None:
1512 self
._expect
_pat
(self
._macro
_exp
_colon
_pat
, "Expecting `:`")
1514 # Expect a macro name
1516 name_text_loc
= self
._text
_loc
1517 m
= self
._expect
_pat
(_py_name_pat
, "Expecting a valid macro name")
1521 macro_def
= self
._macro
_defs
.get(name
)
1523 if macro_def
is None:
1524 _raise_error("Unknown macro name `{}`".format(name
), name_text_loc
)
1528 self
._expect
_pat
(self
._left
_paren
_pat
, "Expecting `(`")
1530 # Try to parse comma-separated parameter values
1531 params_text_loc
= self
._text
_loc
1532 params
= [] # type: List[_MacroExpParam]
1533 expect_comma
= False
1539 if self
._try
_parse
_pat
(self
._right
_paren
_pat
) is not None:
1545 self
._expect
_pat
(self
._macro
_params
_comma
_pat
, "Expecting `,`")
1548 param_text_loc
= self
._text
_loc
1551 *self
._expect
_const
_int
_name
_expr
(True, True), param_text_loc
1556 # Validate parameter values
1557 if len(params
) != len(macro_def
.param_names
):
1558 sing_plur
= "" if len(params
) == 1 else "s"
1560 "Macro expansion passes {} parameter{} while the definition expects {}".format(
1561 len(params
), sing_plur
, len(macro_def
.param_names
)
1567 return _MacroExp(name
, params
, begin_text_loc
)
1569 # Tries to parse a base item (anything except a repetition),
1570 # returning it on success.
1571 def _try_parse_base_item(self
):
1573 item
= self
._try
_parse
_byte
()
1575 if item
is not None:
1579 item
= self
._try
_parse
_str
()
1581 if item
is not None:
1584 # Value, variable assignment, or byte order setting item?
1585 item
= self
._try
_parse
_val
_or
_var
_assign
_or
_set
_bo
()
1587 if item
is not None:
1590 # Label or offset setting item?
1591 item
= self
._try
_parse
_label
_or
_set
_offset
()
1593 if item
is not None:
1596 # Offset alignment item?
1597 item
= self
._try
_parse
_align
_offset
()
1599 if item
is not None:
1603 item
= self
._try
_parse
_fill
_until
()
1605 if item
is not None:
1609 item
= self
._try
_parse
_group
()
1611 if item
is not None:
1614 # Repetition block item?
1615 item
= self
._try
_parse
_rep
_block
()
1617 if item
is not None:
1620 # Conditional block item?
1621 item
= self
._try
_parse
_cond
_block
()
1623 if item
is not None:
1627 item
= self
._try
_parse
_macro
_exp
()
1629 if item
is not None:
1632 # Pattern for _try_parse_rep_post()
1633 _rep_post_prefix_pat
= re
.compile(r
"\*")
1635 # Tries to parse a post-item repetition, returning the expression
1636 # string and AST expression node on success.
1637 def _try_parse_rep_post(self
):
1639 if self
._try
_parse
_pat
(self
._rep
_post
_prefix
_pat
) is None:
1643 # Return expression string and AST expression
1644 self
._skip
_ws
_and
_comments
()
1645 return self
._expect
_rep
_mul
_expr
()
1647 # Tries to parse an item, possibly followed by a repetition,
1648 # returning `True` on success.
1650 # Appends any parsed item to `items`.
1651 def _try_append_item(self
, items
: List
[_Item
]):
1652 self
._skip
_ws
_and
_comments
()
1655 item
= self
._try
_parse
_base
_item
()
1660 # Parse repetition if the base item is repeatable
1661 if isinstance(item
, _RepableItem
):
1662 self
._skip
_ws
_and
_comments
()
1663 rep_text_loc
= self
._text
_loc
1664 rep_ret
= self
._try
_parse
_rep
_post
()
1666 if rep_ret
is not None:
1667 item
= _Rep(item
, *rep_ret
, rep_text_loc
)
1672 # Parses and returns items, skipping whitespaces, insignificant
1673 # symbols, and comments when allowed, and stopping at the first
1674 # unknown character.
1676 # Accepts and registers macro definitions if `accept_macro_defs`
1678 def _parse_items(self
, accept_macro_defs
: bool = False) -> List
[_Item
]:
1679 items
= [] # type: List[_Item]
1681 while self
._isnt
_done
():
1682 # Try to append item
1683 if not self
._try
_append
_item
(items
):
1684 if accept_macro_defs
and self
._try
_parse
_macro
_def
():
1687 # Unknown at this point
1692 # Parses the whole Normand input, setting `self._res` to the main
1693 # group item on success.
1695 if len(self
._normand
.strip()) == 0:
1696 # Special case to make sure there's something to consume
1697 self
._res
= _Group([], self
._text
_loc
)
1700 # Parse first level items
1701 items
= self
._parse
_items
(True)
1703 # Make sure there's nothing left
1704 self
._skip
_ws
_and
_comments
()
1706 if self
._isnt
_done
():
1708 "Unexpected character `{}`".format(self
._normand
[self
._at
])
1711 # Set main group item
1712 self
._res
= _Group(items
, self
._text
_loc
)
1715 # The return type of parse().
1721 variables
: VariablesT
,
1724 bo
: Optional
[ByteOrder
],
1726 self
= cls
.__new
__(cls
)
1727 self
._init
(data
, variables
, labels
, offset
, bo
)
1730 def __init__(self
, *args
, **kwargs
): # type: ignore
1731 raise NotImplementedError
1736 variables
: VariablesT
,
1739 bo
: Optional
[ByteOrder
],
1742 self
._vars
= variables
1743 self
._labels
= labels
1744 self
._offset
= offset
1752 # Dictionary of updated variable names to their last computed value.
1754 def variables(self
):
1757 # Dictionary of updated main group label names to their computed
1768 # Updated byte order.
1770 def byte_order(self
):
1774 # Raises a parse error for the item `item`, creating it using the
1776 def _raise_error_for_item(msg
: str, item
: _Item
) -> NoReturn
:
1777 _raise_error(msg
, item
.text_loc
)
1780 # The `ICITTE` reserved name.
1781 _icitte_name
= "ICITTE"
1784 # Base node visitor.
1786 # Calls the _visit_name() method for each name node which isn't the name
1788 class _NodeVisitor(ast
.NodeVisitor
):
1790 self
._parent
_is
_call
= False
1792 def generic_visit(self
, node
: ast
.AST
):
1793 if type(node
) is ast
.Call
:
1794 self
._parent
_is
_call
= True
1795 elif type(node
) is ast
.Name
and not self
._parent
_is
_call
:
1796 self
._visit
_name
(node
.id)
1798 super().generic_visit(node
)
1799 self
._parent
_is
_call
= False
1802 def _visit_name(self
, name
: str):
1806 # Expression validator: validates that all the names within the
1807 # expression are allowed.
1808 class _ExprValidator(_NodeVisitor
):
1809 def __init__(self
, expr_str
: str, text_loc
: TextLocation
, allowed_names
: Set
[str]):
1811 self
._expr
_str
= expr_str
1812 self
._text
_loc
= text_loc
1813 self
._allowed
_names
= allowed_names
1815 def _visit_name(self
, name
: str):
1816 # Make sure the name refers to a known and reachable
1817 # variable/label name.
1818 if name
!= _icitte_name
and name
not in self
._allowed
_names
:
1819 msg
= "Illegal (unknown or unreachable) variable/label name `{}` in expression `{}`".format(
1820 name
, self
._expr
_str
1823 allowed_names
= self
._allowed
_names
.copy()
1824 allowed_names
.add(_icitte_name
)
1826 if len(allowed_names
) > 0:
1827 allowed_names_str
= ", ".join(
1828 sorted(["`{}`".format(name
) for name
in allowed_names
])
1830 msg
+= "; the legal names are {{{}}}".format(allowed_names_str
)
1842 variables
: VariablesT
,
1845 bo
: Optional
[ByteOrder
],
1847 self
.variables
= variables
.copy()
1848 self
.labels
= labels
.copy()
1849 self
.offset
= offset
1853 return "_GenState({}, {}, {}, {})".format(
1854 repr(self
.variables
), repr(self
.labels
), repr(self
.offset
), repr(self
.bo
)
1858 # Fixed-length number item instance.
1859 class _FlNumItemInst
:
1863 offset_in_data
: int,
1865 parse_error_msgs
: List
[ParseErrorMessage
],
1868 self
._offset
_in
_data
= offset_in_data
1870 self
._parse
_error
_msgs
= parse_error_msgs
1877 def offset_in_data(self
):
1878 return self
._offset
_in
_data
1885 def parse_error_msgs(self
):
1886 return self
._parse
_error
_msgs
1889 # Generator of data and final state from a group item.
1891 # Generation happens in memory at construction time. After building, use
1892 # the `data`, `variables`, `labels`, `offset`, and `bo` properties to
1893 # get the resulting context.
1895 # The steps of generation are:
1897 # 1. Handle each item in prefix order.
1899 # The handlers append bytes to `self._data` and update some current
1900 # state object (`_GenState` instance).
1902 # When handling a fixed-length number item, try to evaluate its
1903 # expression using the current state. If this fails, then it might be
1904 # because the expression refers to a "future" label: save the current
1905 # offset in `self._data` (generated data) and a snapshot of the
1906 # current state within `self._fl_num_item_insts` (`_FlNumItemInst`
1907 # object). _gen_fl_num_item_insts() will deal with this later. A
1908 # `_FlNumItemInst` instance also contains a snapshot of the current
1909 # parsing error messages (`self._parse_error_msgs`) which need to be
1910 # taken into account when handling the instance later.
1912 # When handling the items of a group, keep a map of immediate label
1913 # names to their offset. Then, after having processed all the items,
1914 # update the relevant saved state snapshots in
1915 # `self._fl_num_item_insts` with those immediate label values.
1916 # _gen_fl_num_item_insts() will deal with this later.
1918 # 2. Handle all the fixed-length number item instances of which the
1919 # expression evaluation failed before.
1921 # At this point, `self._fl_num_item_insts` contains everything that's
1922 # needed to evaluate the expressions, including the values of
1923 # "future" labels from the point of view of some fixed-length number
1926 # If an evaluation fails at this point, then it's a user error. Add
1927 # to the parsing error all the saved parsing error messages of the
1928 # instance. Those additional messages add precious context to the
1934 macro_defs
: _MacroDefsT
,
1935 variables
: VariablesT
,
1938 bo
: Optional
[ByteOrder
],
1940 self
._macro
_defs
= macro_defs
1941 self
._fl
_num
_item
_insts
= [] # type: List[_FlNumItemInst]
1942 self
._parse
_error
_msgs
= [] # type: List[ParseErrorMessage]
1943 self
._gen
(group
, _GenState(variables
, labels
, offset
, bo
))
1950 # Updated variables.
1952 def variables(self
):
1953 return self
._final
_state
.variables
1955 # Updated main group labels.
1958 return self
._final
_state
.labels
1963 return self
._final
_state
.offset
1965 # Updated byte order.
1968 return self
._final
_state
.bo
1970 # Evaluates the expression `expr` of which the original string is
1971 # `expr_str` at the location `text_loc` considering the current
1972 # generation state `state`.
1974 # If `allow_float` is `True`, then the type of the result may be
1979 expr
: ast
.Expression
,
1980 text_loc
: TextLocation
,
1982 allow_float
: bool = False,
1984 syms
= {} # type: VariablesT
1985 syms
.update(state
.labels
)
1987 # Set the `ICITTE` name to the current offset
1988 syms
[_icitte_name
] = state
.offset
1990 # Add the current variables
1991 syms
.update(state
.variables
)
1993 # Validate the node and its children
1994 _ExprValidator(expr_str
, text_loc
, set(syms
.keys())).visit(expr
)
1996 # Compile and evaluate expression node
1998 val
= eval(compile(expr
, "", "eval"), None, syms
)
1999 except Exception as exc
:
2001 "Failed to evaluate expression `{}`: {}".format(expr_str
, exc
),
2005 # Convert `bool` result type to `int` to normalize
2006 if type(val
) is bool:
2009 # Validate result type
2010 expected_types
= {int}
# type: Set[type]
2014 expected_types
.add(float)
2015 type_msg
+= " or `float`"
2017 if type(val
) not in expected_types
:
2019 "Invalid expression `{}`: expecting result type {}, not `{}`".format(
2020 expr_str
, type_msg
, type(val
).__name
__
2027 # Evaluates the expression of `item` considering the current
2028 # generation state `state`.
2030 # If `allow_float` is `True`, then the type of the result may be
2033 def _eval_item_expr(
2034 item
: Union
[_FlNum
, _Leb128Int
, _FillUntil
, _VarAssign
, _Rep
, _Cond
],
2036 allow_float
: bool = False,
2038 return _Gen
._eval
_expr
(
2039 item
.expr_str
, item
.expr
, item
.text_loc
, state
, allow_float
2042 # Handles the byte item `item`.
2043 def _handle_byte_item(self
, item
: _Byte
, state
: _GenState
):
2044 self
._data
.append(item
.val
)
2045 state
.offset
+= item
.size
2047 # Handles the string item `item`.
2048 def _handle_str_item(self
, item
: _Str
, state
: _GenState
):
2049 self
._data
+= item
.data
2050 state
.offset
+= item
.size
2052 # Handles the byte order setting item `item`.
2053 def _handle_set_bo_item(self
, item
: _SetBo
, state
: _GenState
):
2054 # Update current byte order
2057 # Handles the variable assignment item `item`.
2058 def _handle_var_assign_item(self
, item
: _VarAssign
, state
: _GenState
):
2060 state
.variables
[item
.name
] = self
._eval
_item
_expr
(item
, state
, True)
2062 # Handles the fixed-length number item `item`.
2063 def _handle_fl_num_item(self
, item
: _FlNum
, state
: _GenState
):
2064 # Validate current byte order
2065 if state
.bo
is None and item
.len > 8:
2066 _raise_error_for_item(
2067 "Current byte order isn't defined at first fixed-length number (`{}`) to encode on more than 8 bits".format(
2073 # Try an immediate evaluation. If it fails, then keep everything
2074 # needed to (try to) generate the bytes of this item later.
2076 data
= self
._gen
_fl
_num
_item
_inst
_data
(item
, state
)
2078 self
._fl
_num
_item
_insts
.append(
2082 copy
.deepcopy(state
),
2083 copy
.deepcopy(self
._parse
_error
_msgs
),
2087 # Reserve space in `self._data` for this instance
2088 data
= bytes([0] * (item
.len // 8))
2094 state
.offset
+= len(data
)
2096 # Returns the size, in bytes, required to encode the value `val`
2097 # with LEB128 (signed version if `is_signed` is `True`).
2099 def _leb128_size_for_val(val
: int, is_signed
: bool):
2101 # Equivalent upper bound.
2103 # For example, if `val` is -128, then the full integer for
2104 # this number of bits would be [-128, 127].
2107 # Number of bits (add one for the sign if needed)
2108 bits
= val
.bit_length() + int(is_signed
)
2113 # Seven bits per byte
2114 return math
.ceil(bits
/ 7)
2116 # Handles the LEB128 integer item `item`.
2117 def _handle_leb128_int_item(self
, item
: _Leb128Int
, state
: _GenState
):
2119 val
= self
._eval
_item
_expr
(item
, state
, False)
2122 size
= self
._leb
128_size
_for
_val
(val
, type(item
) is _SLeb128Int
)
2125 for _
in range(size
):
2126 # Seven LSBs, MSB of the byte set (continue)
2127 self
._data
.append((val
& 0x7F) |
0x80)
2130 # Clear MSB of last byte (stop)
2131 self
._data
[-1] &= ~
0x80
2134 state
.offset
+= size
2136 # Handles the group item `item`, removing the immediate labels from
2137 # `state` at the end if `remove_immediate_labels` is `True`.
2138 def _handle_group_item(
2139 self
, item
: _Group
, state
: _GenState
, remove_immediate_labels
: bool = True
2141 first_fl_num_item_inst_index
= len(self
._fl
_num
_item
_insts
)
2142 immediate_labels
= {} # type: LabelsT
2145 for subitem
in item
.items
:
2146 if type(subitem
) is _Label
:
2147 # Add to local immediate labels
2148 immediate_labels
[subitem
.name
] = state
.offset
2150 self
._handle
_item
(subitem
, state
)
2152 # Remove immediate labels from current state if needed
2153 if remove_immediate_labels
:
2154 for name
in immediate_labels
:
2155 del state
.labels
[name
]
2157 # Add all immediate labels to all state snapshots since
2158 # `first_fl_num_item_inst_index`.
2159 for inst
in self
._fl
_num
_item
_insts
[first_fl_num_item_inst_index
:]:
2160 inst
.state
.labels
.update(immediate_labels
)
2162 # Handles the repetition item `item`.
2163 def _handle_rep_item(self
, item
: _Rep
, state
: _GenState
):
2164 # Compute the repetition count
2165 mul
= _Gen
._eval
_item
_expr
(item
, state
)
2169 _raise_error_for_item(
2170 "Invalid expression `{}`: unexpected negative result {:,}".format(
2176 # Generate item data `mul` times
2177 for _
in range(mul
):
2178 self
._handle
_item
(item
.item
, state
)
2180 # Handles the conditional item `item`.
2181 def _handle_cond_item(self
, item
: _Cond
, state
: _GenState
):
2182 # Compute the conditional value
2183 val
= _Gen
._eval
_item
_expr
(item
, state
)
2185 # Generate item data if needed
2187 self
._handle
_item
(item
.true_item
, state
)
2189 self
._handle
_item
(item
.false_item
, state
)
2191 # Evaluates the parameters of the macro expansion item `item`
2192 # considering the initial state `init_state` and returns a new state
2193 # to handle the items of the macro.
2194 def _eval_macro_exp_params(self
, item
: _MacroExp
, init_state
: _GenState
):
2196 exp_state
= _GenState({}, {}, init_state
.offset
, init_state
.bo
)
2198 # Evaluate the parameter expressions
2199 macro_def
= self
._macro
_defs
[item
.name
]
2201 for param_name
, param
in zip(macro_def
.param_names
, item
.params
):
2202 exp_state
.variables
[param_name
] = _Gen
._eval
_expr
(
2203 param
.expr_str
, param
.expr
, param
.text_loc
, init_state
, True
2208 # Handles the macro expansion item `item`.
2209 def _handle_macro_exp_item(self
, item
: _MacroExp
, state
: _GenState
):
2210 parse_error_msg_text
= "While expanding the macro `{}`:".format(item
.name
)
2214 exp_state
= self
._eval
_macro
_exp
_params
(item
, state
)
2216 # Process the contained group
2217 init_data_size
= len(self
._data
)
2219 ParseErrorMessage
._create
( # pyright: ignore[reportPrivateUsage]
2220 parse_error_msg_text
, item
.text_loc
2223 self
._parse
_error
_msgs
.append(parse_error_msg
)
2224 self
._handle
_item
(self
._macro
_defs
[item
.name
].group
, exp_state
)
2225 self
._parse
_error
_msgs
.pop()
2226 except ParseError
as exc
:
2227 _augment_error(exc
, parse_error_msg_text
, item
.text_loc
)
2229 # Update state offset and return
2230 state
.offset
+= len(self
._data
) - init_data_size
2232 # Handles the offset setting item `item`.
2233 def _handle_set_offset_item(self
, item
: _SetOffset
, state
: _GenState
):
2234 state
.offset
= item
.val
2236 # Handles the offset alignment item `item` (adds padding).
2237 def _handle_align_offset_item(self
, item
: _AlignOffset
, state
: _GenState
):
2238 init_offset
= state
.offset
2239 align_bytes
= item
.val
// 8
2240 state
.offset
= (state
.offset
+ align_bytes
- 1) // align_bytes
* align_bytes
2241 self
._data
+= bytes([item
.pad_val
] * (state
.offset
- init_offset
))
2243 # Handles the filling item `item` (adds padding).
2244 def _handle_fill_until_item(self
, item
: _FillUntil
, state
: _GenState
):
2245 # Compute the new offset
2246 new_offset
= _Gen
._eval
_item
_expr
(item
, state
)
2248 # Validate the new offset
2249 if new_offset
< state
.offset
:
2250 _raise_error_for_item(
2251 "Invalid expression `{}`: new offset {:,} is less than current offset {:,}".format(
2252 item
.expr_str
, new_offset
, state
.offset
2258 self
._data
+= bytes([item
.pad_val
] * (new_offset
- state
.offset
))
2261 state
.offset
= new_offset
2263 # Handles the label item `item`.
2264 def _handle_label_item(self
, item
: _Label
, state
: _GenState
):
2265 state
.labels
[item
.name
] = state
.offset
2267 # Handles the item `item`, returning the updated next repetition
2269 def _handle_item(self
, item
: _Item
, state
: _GenState
):
2270 return self
._item
_handlers
[type(item
)](item
, state
)
2272 # Generates the data for a fixed-length integer item instance having
2273 # the value `val` and returns it.
2274 def _gen_fl_int_item_inst_data(self
, val
: int, item
: _FlNum
, state
: _GenState
):
2276 if val
< -(2 ** (item
.len - 1)) or val
> 2**item
.len - 1:
2277 _raise_error_for_item(
2278 "Value {:,} is outside the {}-bit range when evaluating expression `{}`".format(
2279 val
, item
.len, item
.expr_str
2284 # Encode result on 64 bits (to extend the sign bit whatever the
2285 # value of `item.len`).
2288 ">" if state
.bo
in (None, ByteOrder
.BE
) else "<",
2289 "Q" if val
>= 0 else "q",
2294 # Keep only the requested length
2295 len_bytes
= item
.len // 8
2297 if state
.bo
in (None, ByteOrder
.BE
):
2298 # Big endian: keep last bytes
2299 data
= data
[-len_bytes
:]
2301 # Little endian: keep first bytes
2302 assert state
.bo
== ByteOrder
.LE
2303 data
= data
[:len_bytes
]
2308 # Generates the data for a fixed-length floating point number item
2309 # instance having the value `val` and returns it.
2310 def _gen_fl_float_item_inst_data(self
, val
: float, item
: _FlNum
, state
: _GenState
):
2312 if item
.len not in (32, 64):
2313 _raise_error_for_item(
2314 "Invalid {}-bit length for a fixed-length floating point number (value {:,})".format(
2320 # Encode and return result
2323 ">" if state
.bo
in (None, ByteOrder
.BE
) else "<",
2324 "f" if item
.len == 32 else "d",
2329 # Generates the data for a fixed-length number item instance and
2331 def _gen_fl_num_item_inst_data(self
, item
: _FlNum
, state
: _GenState
):
2333 val
= self
._eval
_item
_expr
(item
, state
, True)
2335 # Handle depending on type
2336 if type(val
) is int:
2337 return self
._gen
_fl
_int
_item
_inst
_data
(val
, item
, state
)
2339 assert type(val
) is float
2340 return self
._gen
_fl
_float
_item
_inst
_data
(val
, item
, state
)
2342 # Generates the data for all the fixed-length number item instances
2343 # and writes it at the correct offset within `self._data`.
2344 def _gen_fl_num_item_insts(self
):
2345 for inst
in self
._fl
_num
_item
_insts
:
2348 data
= self
._gen
_fl
_num
_item
_inst
_data
(inst
.item
, inst
.state
)
2349 except ParseError
as exc
:
2350 # Add all the saved parse error messages for this
2352 for msg
in reversed(inst
.parse_error_msgs
):
2353 _add_error_msg(exc
, msg
.text
, msg
.text_location
)
2357 # Insert bytes into `self._data`
2358 self
._data
[inst
.offset_in_data
: inst
.offset_in_data
+ len(data
)] = data
2360 # Generates the data (`self._data`) and final state
2361 # (`self._final_state`) from `group` and the initial state `state`.
2362 def _gen(self
, group
: _Group
, state
: _GenState
):
2364 self
._data
= bytearray()
2367 self
._item
_handlers
= {
2368 _AlignOffset
: self
._handle
_align
_offset
_item
,
2369 _Byte
: self
._handle
_byte
_item
,
2370 _Cond
: self
._handle
_cond
_item
,
2371 _FillUntil
: self
._handle
_fill
_until
_item
,
2372 _FlNum
: self
._handle
_fl
_num
_item
,
2373 _Group
: self
._handle
_group
_item
,
2374 _Label
: self
._handle
_label
_item
,
2375 _MacroExp
: self
._handle
_macro
_exp
_item
,
2376 _Rep
: self
._handle
_rep
_item
,
2377 _SetBo
: self
._handle
_set
_bo
_item
,
2378 _SetOffset
: self
._handle
_set
_offset
_item
,
2379 _SLeb128Int
: self
._handle
_leb
128_int
_item
,
2380 _Str
: self
._handle
_str
_item
,
2381 _ULeb128Int
: self
._handle
_leb
128_int
_item
,
2382 _VarAssign
: self
._handle
_var
_assign
_item
,
2383 } # type: Dict[type, Callable[[Any, _GenState], None]]
2385 # Handle the group item, _not_ removing the immediate labels
2386 # because the `labels` property offers them.
2387 self
._handle
_group
_item
(group
, state
, False)
2389 # This is actually the final state
2390 self
._final
_state
= state
2392 # Generate all the fixed-length number bytes now that we know
2394 self
._gen
_fl
_num
_item
_insts
()
2397 # Returns a `ParseResult` instance containing the bytes encoded by the
2398 # input string `normand`.
2400 # `init_variables` is a dictionary of initial variable names (valid
2401 # Python names) to integral values. A variable name must not be the
2402 # reserved name `ICITTE`.
2404 # `init_labels` is a dictionary of initial label names (valid Python
2405 # names) to integral values. A label name must not be the reserved name
2408 # `init_offset` is the initial offset.
2410 # `init_byte_order` is the initial byte order.
2412 # Raises `ParseError` on any parsing error.
2415 init_variables
: Optional
[VariablesT
] = None,
2416 init_labels
: Optional
[LabelsT
] = None,
2417 init_offset
: int = 0,
2418 init_byte_order
: Optional
[ByteOrder
] = None,
2420 if init_variables
is None:
2423 if init_labels
is None:
2426 parser
= _Parser(normand
, init_variables
, init_labels
)
2435 return ParseResult
._create
( # pyright: ignore[reportPrivateUsage]
2436 gen
.data
, gen
.variables
, gen
.labels
, gen
.offset
, gen
.bo
2440 # Raises a command-line error with the message `msg`.
2441 def _raise_cli_error(msg
: str) -> NoReturn
:
2442 raise RuntimeError("Command-line error: {}".format(msg
))
2445 # Returns a dictionary of string to integers from the list of strings
2446 # `args` containing `NAME=VAL` entries.
2447 def _dict_from_arg(args
: Optional
[List
[str]]):
2448 d
= {} # type: LabelsT
2454 m
= re
.match(r
"({})=(\d+)$".format(_py_name_pat
.pattern
), arg
)
2457 _raise_cli_error("Invalid assignment {}".format(arg
))
2459 d
[m
.group(1)] = int(m
.group(2))
2464 # Parses the command-line arguments and returns, in this order:
2466 # 1. The input file path, or `None` if none.
2467 # 2. The Normand input text.
2468 # 3. The initial offset.
2469 # 4. The initial byte order.
2470 # 5. The initial variables.
2471 # 6. The initial labels.
2472 def _parse_cli_args():
2476 ap
= argparse
.ArgumentParser()
2483 help="initial offset (positive)",
2489 choices
=["be", "le"],
2491 help="initial byte order (`be` or `le`)",
2497 help="add an initial variable (may be repeated)",
2504 help="add an initial label (may be repeated)",
2507 "--version", action
="version", version
="Normand {}".format(__version__
)
2514 help="input path (none means standard input)",
2518 args
= ap
.parse_args()
2521 if args
.path
is None:
2522 normand
= sys
.stdin
.read()
2524 with
open(args
.path
) as f
:
2527 # Variables and labels
2528 variables
= typing
.cast(VariablesT
, _dict_from_arg(args
.var
))
2529 labels
= _dict_from_arg(args
.label
)
2533 _raise_cli_error("Invalid negative offset {}")
2535 # Validate and set byte order
2536 bo
= None # type: Optional[ByteOrder]
2538 if args
.byte_order
is not None:
2539 if args
.byte_order
== "be":
2542 assert args
.byte_order
== "le"
2545 # Return input and initial state
2546 return args
.path
, normand
, args
.offset
, bo
, variables
, labels
2549 # CLI entry point without exception handling.
2550 def _run_cli_with_args(
2553 bo
: Optional
[ByteOrder
],
2554 variables
: VariablesT
,
2557 sys
.stdout
.buffer.write(parse(normand
, variables
, labels
, offset
, bo
).data
)
2560 # Prints the exception message `msg` and exits with status 1.
2561 def _fail(msg
: str) -> NoReturn
:
2562 if not msg
.endswith("."):
2565 print(msg
.strip(), file=sys
.stderr
)
2572 args
= _parse_cli_args()
2573 except Exception as exc
:
2577 _run_cli_with_args(*args
[1:])
2578 except ParseError
as exc
:
2581 prefix
= "" if args
[0] is None else "{}:".format(os
.path
.abspath(args
[0]))
2584 for msg
in reversed(exc
.messages
):
2585 fail_msg
+= "{}{}:{} - {}".format(
2587 msg
.text_location
.line_no
,
2588 msg
.text_location
.col_no
,
2592 if fail_msg
[-1] not in ".:;":
2597 _fail(fail_msg
.strip())
2598 except Exception as exc
:
2602 if __name__
== "__main__":