[WIP] 10390/fix/sanitize single double quote in solr query

openlibrary/plugins/worksearch/schemes/__init__.py

@@ -8,6 +8,7 @@
     escape_unknown_fields,
     fully_escape_query,
     luqum_parser,
+    sanitize_solr_symbols,
 )
 
 logger = logging.getLogger("openlibrary.worksearch")
@@ -82,6 +83,7 @@ def process_user_query(self, q_param: str) -> str:
             return q_param
 
         try:
+            q_param = sanitize_solr_symbols(q_param, ['"'])
             q_param = escape_unknown_fields(
                 (
                     # Solr 4+ has support for regexes (eg `key:/foo.*/`)! But for now,
@@ -104,6 +106,7 @@ def process_user_query(self, q_param: str) -> str:
             q_tree = luqum_parser(fully_escape_query(q_param))
 
         q_tree = self.transform_user_query(q_param, q_tree)
+
         return str(q_tree)
 
     def transform_user_query(

openlibrary/plugins/worksearch/schemes/tests/test_works.py

@@ -35,6 +35,14 @@
         'title:"food rules" author:pollan',
         'alternative_title:"food rules" author_name:pollan',
     ),
+    'Unmatched double-quote': (
+        'title:Compilation Group for the "History of Modern China',
+        'alternative_title:(Compilation Group for the) "History of Modern China"',
+    ),
+    'Unmatched double-quotes': (
+        'title:"Compilation Group for the "History of Modern China"',
+        'alternative_title:"Compilation Group for the "History of Modern China""',
+    ),
     'Leading text': (
         'query here title:food rules author:pollan',
         'query here alternative_title:(food rules) author_name:pollan',

openlibrary/solr/query_utils.py

@@ -146,6 +146,15 @@ def fully_escape_query(query: str) -> str:
     return escaped
 
 
+# Ensure solr special symbols are valid in query
+def sanitize_solr_symbols(query: str, chars: list[str]) -> str:
+    for c in chars:
+        # Handles unclosed symbols (i.e. adding closing '"' for Solr Terms)
+        if c in ('"') and query.count(c) % 2 == 1:
+            query = query + f'{c}'
+    return query
+
+
 def luqum_parser(query: str) -> Item:
     """
     Parses a lucene-like query, with the special binding rules of Open Library.