Handle UTF-16 surrogate pairs in quoted property selectors.

Author: jg-rp

CHANGELOG.md

@@ -2,12 +2,15 @@
 
 ## Version 0.3.0 (unreleased)
 
+**IETF JSONPath Draft compliance**
+
 - Added support for function extensions.
 - Added the built-in `length()` function.
 - Added the built-in `count()` function. `count()` is an alias for `length()`
-- Added the built-in `keys()` function.
+- Added the built-in `keys()` function (non-standard).
 - Support filters without parentheses.
 - Adhere to IETF JSONPath draft escaping in quoted property selectors.
+- Handle UTF-16 surrogate pairs in quoted property selectors.
 
 ## Version 0.2.0
 

jsonpath/parse.py

@@ -305,7 +305,9 @@ def parse_selector_list(self, stream: TokenStream) -> ListSelector:
                         token=stream.current,
                         name=codecs.decode(
                             stream.current.value.replace("\\/", "/"), "unicode-escape"
-                        ),
+                        )
+                        .encode("utf-16", "surrogatepass")
+                        .decode("utf-16"),
                     ),
                 )
             elif stream.current.kind == TOKEN_SLICE_START:

tests/compliance.py

@@ -33,10 +33,6 @@ class Case:
 
 
 SKIP = {
-    "name selector, double quotes, surrogate pair 𝄞": "TODO",
-    "name selector, double quotes, surrogate pair 😀": "TODO",
-    "name selector, single quotes, surrogate pair 𝄞": "TODO",
-    "name selector, single quotes, surrogate pair 😀": "TODO",
     "filter, nested": "not supported",
 }
 
@@ -59,7 +55,7 @@ def mangle_filter(case: Case) -> Case:
             case.selector = case.selector.replace("$[?", "$.*[?")
         return case
 
-    # XXX: skipping filter functions. Not supported.
+    # TODO: skipping filter functions. Not supported.
     return [
         mangle_filter(case)
         for case in cases()