Merge pull request #723 from onekey-sec/hexstring-anchors

parser: support regex style begin-end anchors (`^`, `$`)

docs/development.md

@@ -443,6 +443,10 @@ PATTERNS = [
 ]
 ```
 
+In addition, start and end of input anchors (`^` and `$` like in regular
+expressions) can also be used to restrict a match to the beginning or the end of
+the input file.
+
 ### DirectoryPatterns
 
 The `DirectoryHandler` uses these patterns to identify the starting/main file of a given

tests/test_parser.py

@@ -44,6 +44,8 @@
             rb"\x00(\x01|\x02(\x03|\x04))\x05",
             id="nested-alternative",
         ),
+        pytest.param("^ 00", rb"^\x00", id="start-anchor"),
+        pytest.param("00 $", rb"\x00$", id="end-anchor"),
     ],
 )
 def test_simple_convert(hex_string, expected_regex):
@@ -72,6 +74,16 @@ def test_single_comment():
     assert regex == rb"\x01\x02"
 
 
-def test_invalid_hexstring():
+@pytest.mark.parametrize(
+    "pattern",
+    [
+        pytest.param("invalid hexstring", id="invalid"),
+        pytest.param("00 ^", id="start-anchor-at-end"),
+        pytest.param("00 ^ 01", id="start-anchor-at-middle"),
+        pytest.param("$ 00", id="end-anchor-at-start"),
+        pytest.param("00 $ 01", id="end-anchor-at-middle"),
+    ],
+)
+def test_invalid_hexstring(pattern):
     with pytest.raises(InvalidHexString):
-        hexstring2regex("invalid hexstring")
+        hexstring2regex(pattern)

unblob/parser.py

@@ -1,6 +1,6 @@
 import itertools
 
-from lark.exceptions import UnexpectedCharacters
+from lark.exceptions import UnexpectedInput
 from lark.lark import Lark
 from lark.visitors import Transformer
 
@@ -16,7 +16,7 @@
 
     %ignore COMMENT
 
-    start: item+
+    start: START_ANCHOR? item+ END_ANCHOR?
 
     item: LITERAL           -> literal
         | WILDCARD          -> wildcard
@@ -29,6 +29,8 @@
     alternative: "(" item+ (ALTERNATIVE_SEPARATOR item+)+ ")"
     ALTERNATIVE_SEPARATOR: "|"
     LITERAL: HEXDIGIT HEXDIGIT
+    START_ANCHOR: "^"
+    END_ANCHOR: "$"
     WILDCARD: "??"
     FIRSTNIBLE: "?" HEXDIGIT
     SECONDNIBLE: HEXDIGIT "?"
@@ -89,6 +91,6 @@ class InvalidHexString(ValueError):
 def hexstring2regex(hexastr):
     try:
         parsed = _hex_string_parser.parse(hexastr)
-    except UnexpectedCharacters as e:
+    except UnexpectedInput as e:
         raise InvalidHexString(str(e)) from e
     return _HexStringToRegex().transform(parsed)