Fix parsing of scientific notation in JSONPath queries

jg-rp · Sep 28, 2023 · f2bc762 · f2bc762
1 parent 65c8920
commit f2bc762
Show file tree

Hide file tree

Showing 4 changed files with 17 additions and 5 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -10,6 +10,7 @@
 **Fixes**
 
 - We no longer silently ignore invalid escape sequences in JSONPath string literals. For example, `$['\"']` used to be OK, it now raises a `JSONPathSyntaxError`.
+- Fixed parsing of JSONPath integer literals that use scientific notation.
 
 ## Version 0.9.0
 

diff --git a/jsonpath/lex.py b/jsonpath/lex.py
@@ -139,7 +139,7 @@ def compile_rules(self) -> Pattern[str]:
             (TOKEN_BRACKET_PROPERTY, self.bracketed_property_pattern),
             (TOKEN_DOT_PROPERTY, self.dot_property_pattern),
             (TOKEN_FLOAT, r"-?\d+\.\d*(?:e[+-]?\d+)?"),
-            (TOKEN_INT, r"-?\d+(?:e[+\-]?\d+)?\b"),
+            (TOKEN_INT, r"-?\d+(?P<G_EXP>e[+\-]?\d+)?\b"),
             (TOKEN_DDOT, r"\.\."),
             (TOKEN_AND, self.bool_and_pattern),
             (TOKEN_OR, self.bool_or_pattern),
@@ -265,6 +265,19 @@ def tokenize(self, path: str) -> Iterator[Token]:  # noqa PLR0912
                     value=match.group("G_SQUOTE"),
                     index=match.start("G_SQUOTE"),
                 )
+            elif kind == TOKEN_INT:
+                if match.group("G_EXP") and match.group("G_EXP")[1] == "-":
+                    yield _token(
+                        kind=TOKEN_FLOAT,
+                        value=match.group(),
+                        index=match.start(),
+                    )
+                else:
+                    yield _token(
+                        kind=TOKEN_INT,
+                        value=match.group(),
+                        index=match.start(),
+                    )
             elif kind == TOKEN_RE_PATTERN:
                 yield _token(
                     kind=TOKEN_RE_PATTERN,

diff --git a/jsonpath/parse.py b/jsonpath/parse.py
@@ -454,7 +454,8 @@ def parse_string_literal(self, stream: TokenStream) -> FilterExpression:
         return StringLiteral(value=self._decode_string_literal(stream.current))
 
     def parse_integer_literal(self, stream: TokenStream) -> FilterExpression:
-        return IntegerLiteral(value=int(stream.current.value))
+        # Convert to float first to handle scientific notation.
+        return IntegerLiteral(value=int(float(stream.current.value)))
 
     def parse_float_literal(self, stream: TokenStream) -> FilterExpression:
         return FloatLiteral(value=float(stream.current.value))

diff --git a/tests/test_compliance.py b/tests/test_compliance.py
@@ -78,9 +78,6 @@ class Case:
     "filter, multiple selectors, filter and wildcard": "TODO",
     "filter, multiple selectors, filter and slice": "TODO",
     "filter, multiple selectors, comparison filter, index and slice": "TODO",
-    "filter, equals number, exponent": "TODO",
-    "filter, equals number, positive exponent": "TODO",
-    "filter, equals number, negative exponent": "TODO",
     "filter, equals number, decimal fraction, no fractional digit": "TODO",
     "functions, length, result must be compared": "ignore",
     "functions, count, result must be compared": "ignore",