Allow escaped strings in json_schema.py

outlines-dev · Jun 20, 2024 · 6d35ecd · 6d35ecd
1 parent 371313e
commit 6d35ecd
Show file tree

Hide file tree

Showing 2 changed files with 7 additions and 1 deletion.
diff --git a/outlines/fsm/json_schema.py b/outlines/fsm/json_schema.py
@@ -10,8 +10,10 @@
 from referencing._core import Resolver
 from referencing.jsonschema import DRAFT202012
 
-STRING_INNER = r'([^"\\\x00-\x1f\x7f-\x9f]|\\\\)'
+# allow any character which isn't a control sequence OR \" OR \\
+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)"

diff --git a/tests/fsm/test_json_schema.py b/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),
+                (r'"\n"', False),
+                (r'"\\n"', True),
+                (r'"unescaped " quote"', False),
+                (r'"escaped \" quote"', True),
             ],
         ),
         # String with maximum length