Allow escaped strings in json_schema.py

Author: lapp0

Diff for: outlines/fsm/json_schema.py

@@ -10,14 +10,20 @@
 from referencing._core import Resolver
 from referencing.jsonschema import DRAFT202012
 
-STRING_INNER = r'([^"\\\x00-\x1f\x7f-\x9f]|\\\\)'
-STRING = f'"{STRING_INNER}*"'
 INTEGER = r"(-)?(0|[1-9][0-9]*)"
 NUMBER = rf"({INTEGER})(\.[0-9]+)?([eE][+-][0-9]+)?"
 BOOLEAN = r"(true|false)"
 NULL = r"null"
 WHITESPACE = r"[ ]?"
 
+# Disallow control sequences, e.g. "\n"
+_ANY_CHAR_NO_CTRL_SEQ = r'[^"\\\x00-\x1F\x7F-\x9F]'
+# Allow valid escapes, e.g. "\\n"
+# Disallow invalid escapes, e.g. "\escape"
+_VALID_ESCAPES = r'\\["\\/bfnrt]|\\u[0-9a-fA-F]{4}'
+STRING_INNER = rf"({_ANY_CHAR_NO_CTRL_SEQ}|{_VALID_ESCAPES})"
+STRING = f'"{STRING_INNER}*"'
+
 type_to_regex = {
     "string": STRING,
     "integer": INTEGER,

Diff for: tests/fsm/test_json_schema.py

@@ -124,6 +124,10 @@ def test_match_number(pattern, does_match):
                 ('"quoted_string"', True),
                 (r'"escape_\character"', False),
                 (r'"double_\\escape"', True),
+                ('"\n"', False),
+                ('"\\n"', True),
+                (r'"unescaped " quote"', False),
+                (r'"escaped \" quote"', True),
             ],
         ),
         # String with maximum length