PubMed: RangeQuery

Author: Gerit Wagner

search_query/constants.py

@@ -31,6 +31,7 @@ class TokenTypes(Enum):
     SEARCH_TERM = "SEARCH_TERM"
     PARENTHESIS_OPEN = "PARENTHESIS_OPEN"
     PARENTHESIS_CLOSED = "PARENTHESIS_CLOSED"
+    RANGE_OPERATOR = "RANGE_OPERATOR"
     UNKNOWN = "UNKNOWN"
 
 
@@ -72,6 +73,8 @@ def get_operator_type(self) -> str:
             return Operators.OR
         if self.value.upper() == "NOT":
             return Operators.NOT
+        if self.value == ":":
+            return Operators.RANGE
         raise ValueError()  # pragma: no cover
 
 
@@ -115,6 +118,7 @@ class Operators:
     NOT = "NOT"
     NEAR = "NEAR"
     WITHIN = "WITHIN"
+    RANGE = "RANGE"
 
 
 class Fields:

search_query/pubmed/linter.py

@@ -44,11 +44,15 @@ class PubmedQueryStringLinter(QueryStringLinter):
         TokenTypes.FIELD: [
             TokenTypes.LOGIC_OPERATOR,
             TokenTypes.PARENTHESIS_CLOSED,
+            TokenTypes.RANGE_OPERATOR,
         ],
         TokenTypes.LOGIC_OPERATOR: [
             TokenTypes.SEARCH_TERM,
             TokenTypes.PARENTHESIS_OPEN,
         ],
+        TokenTypes.RANGE_OPERATOR: [
+            TokenTypes.SEARCH_TERM,
+        ],
     }
 
     def __init__(self, query_str: str = "") -> None:

search_query/pubmed/parser.py

@@ -64,6 +64,8 @@ def tokenize(self) -> None:
 
             if value.upper() in {"AND", "OR", "NOT", "|", "&"}:
                 token_type = TokenTypes.LOGIC_OPERATOR
+            elif value == ":":
+                token_type = TokenTypes.RANGE_OPERATOR
             elif value == "(":
                 token_type = TokenTypes.PARENTHESIS_OPEN
             elif value == ")":
@@ -129,7 +131,10 @@ def _get_operator_indices(self, tokens: list) -> list:
             elif token.type == TokenTypes.PARENTHESIS_CLOSED:
                 i = i - 1
 
-            if i == 0 and token.type == TokenTypes.LOGIC_OPERATOR:
+            if i == 0 and token.type in [
+                TokenTypes.LOGIC_OPERATOR,
+                TokenTypes.RANGE_OPERATOR,
+            ]:
                 operator = token.get_operator_type()
                 if not first_operator_found:
                     first_operator = operator

search_query/query.py

@@ -181,6 +181,7 @@ def value(self, v: str) -> None:
             Operators.NOT,
             Operators.NEAR,
             Operators.WITHIN,
+            Operators.RANGE,
         ]:
             raise ValueError(f"Invalid operator value: {v}")
         self._value = v

search_query/query_range.py

@@ -0,0 +1,39 @@
+#!/usr/bin/env python
+"""Range Query"""
+import typing
+
+from search_query.constants import Operators
+from search_query.query import Query
+from search_query.query import SearchField
+
+# pylint: disable=duplicate-code
+
+
+class RangeQuery(Query):
+    """Range Query"""
+
+    def __init__(
+        self,
+        children: typing.List[typing.Union[str, Query]],
+        *,
+        search_field: typing.Optional[typing.Union[SearchField, str]] = None,
+        position: typing.Optional[tuple] = None,
+        platform: str = "generic",
+    ) -> None:
+        """init method
+        search terms: strings which you want to include in the search query
+        nested queries: queries whose roots are appended to the query
+        search field: search field to which the query should be applied
+        """
+        assert len(children) == 2, "RangeQuery must have exactly two children"
+        super().__init__(
+            value=Operators.RANGE,
+            children=children,
+            search_field=search_field
+            if isinstance(search_field, SearchField)
+            else SearchField(search_field)
+            if search_field is not None
+            else None,
+            position=position,
+            platform=platform,
+        )

test/test_pubmed.py

@@ -879,6 +879,16 @@ def test_linter_with_general_search_field(
             "All Fields",
             'OR["eHealth"[[all]], AND["digital health"[[all]], NOT["bias"[[all]], "equity"[[all]]]], "policy"[[all]]]',
         ),
+        (
+            'eHealth[ti] AND ("2006/01/01"[Date - Create] : "2023/08/18"[Date - Create])',
+            "",
+            'AND[eHealth[[ti]], RANGE["2006/01/01"[[Date - Create]], "2023/08/18"[[Date - Create]]]]',
+        ),
+        (
+            '("1995/01/01"[pdat] : "3000"[pdat])',
+            "",
+            'RANGE["1995/01/01"[[pdat]], "3000"[[pdat]]]',
+        ),
     ],
 )
 def test_parser(