fix: handle single closing bracket

tests/analyzers/test_template_tag.py

@@ -6,7 +6,7 @@
 
 from tests.dirty_equals import HasSubstring
 from tugboat.analyzers.template_tag import (
-    check_simple_tag_reference,
+    _check_simple_tag_reference,
     check_template_tags,
     parse_argo_template_tags,
     split_expr_membership,
@@ -16,7 +16,7 @@
 
 class TestParseArgoTemplateTags:
 
-    def test_pass(self):
+    def test_1(self):
         tree = parse_argo_template_tags(
             """
             Hello {{ inputs.parameters.name }}
@@ -34,6 +34,24 @@ def test_pass(self):
         (bot_name,) = bot_name_tag.find_token("REF")
         assert bot_name == "inputs.parameters['bot']"
 
+    def test_2(self):
+        tree = parse_argo_template_tags(
+            """
+            This is a {sample} document
+            that have some {{= expression{}}} inside.
+            """
+        )
+
+        simple_tags = list(tree.find_data("simple_tag"))
+        assert not simple_tags
+
+        expr_tags = list(tree.find_data("expression_tag"))
+        assert len(expr_tags) == 1
+
+    def test_empty(self):
+        tree = parse_argo_template_tags("")
+        assert tree.children == []
+
     def test_error(self):
         with pytest.raises(lark.exceptions.UnexpectedCharacters):
             parse_argo_template_tags("{{ inputs. parameters.name }}")
@@ -44,8 +62,7 @@ def test_error(self):
 class TestCheckTemplateTags:
 
     def test_pass(self):
-        references = ReferenceCollection()
-        references.add(("inputs", "parameters", "name"))
+        references = ReferenceCollection([("inputs", "parameters", "name")])
 
         diagnoses = list(
             check_template_tags("{{ inputs.parameters.name }}", references)
@@ -103,17 +120,15 @@ def test_unexpected_eof(self):
 class TestCheckSimpleTagReference:
 
     def test_pass(self):
-        references = ReferenceCollection()
-        references.add(("inputs", "parameters", "name"))
+        references = ReferenceCollection([("inputs", "parameters", "name")])
 
-        diagnosis = check_simple_tag_reference("inputs.parameters.name", references)
+        diagnosis = _check_simple_tag_reference("inputs.parameters.name", references)
         assert diagnosis is None
 
     def test_incorrect_format(self):
-        references = ReferenceCollection()
-        references.add(("inputs", "parameters", "name"))
+        references = ReferenceCollection([("inputs", "parameters", "name")])
 
-        diagnosis = check_simple_tag_reference("inputs.parameters['name']", references)
+        diagnosis = _check_simple_tag_reference("inputs.parameters['name']", references)
 
         assert diagnosis == IsPartialDict(
             code="VAR102",
@@ -123,19 +138,18 @@ def test_incorrect_format(self):
 
     def test_syntax_error(self):
         references = ReferenceCollection()
-        diagnosis = check_simple_tag_reference("inputs.parameters['name", references)
+        diagnosis = _check_simple_tag_reference("inputs.parameters['name", references)
         assert diagnosis == IsPartialDict(code="VAR101")
 
     def test_not_a_argo_variable_1(self):
         references = ReferenceCollection()
-        diagnosis = check_simple_tag_reference("inputs", references)
+        diagnosis = _check_simple_tag_reference("inputs", references)
         assert diagnosis == IsPartialDict(code="VAR202")
 
     def test_unknown_variable(self):
-        references = ReferenceCollection()
-        references.add(("demo",))
+        references = ReferenceCollection([("demo",)])
 
-        diagnosis = check_simple_tag_reference("inputs", references)
+        diagnosis = _check_simple_tag_reference("inputs", references)
         assert diagnosis == IsPartialDict(
             code="VAR201",
             fix="demo",

tugboat/analyzers/template_tag.py

@@ -27,6 +27,25 @@
     from tugboat.types import Diagnosis
 
 
+@functools.lru_cache(32)
+def parse_argo_template_tags(source: str) -> lark.Tree:
+    """
+    Parse Argo template tags in the given source string.
+
+    Parameters
+    ----------
+    source : str
+        The source string containing Argo template tags.
+
+    Returns
+    -------
+    lark.Tree
+        The parse tree representing the structure of the template tags.
+    """
+    parser = _argo_template_tag_parser()
+    return parser.parse(source)
+
+
 @functools.cache
 def _argo_template_tag_parser():
     return lark.Lark(
@@ -35,9 +54,7 @@ def _argo_template_tag_parser():
         %import common.LETTER
         %import common.WS
 
-        ?start: template
-
-        template: (_TEXT | expression_tag | simple_tag)+
+        ?start: (_TEXT | expression_tag | simple_tag)*
 
         # ignore other text
         _TEXT: /[^{]+/
@@ -48,31 +65,12 @@ def _argo_template_tag_parser():
         REF: (LETTER | DIGIT | "_" | "-" | "." | "\"" | "'" | "[" | "]")+
 
         # expression template tag
-        expression_tag: "{{=" WS? _ANY WS? "}}"
-        _ANY: /[^}]+/
+        expression_tag: "{{=" _ANY "}}"
+        _ANY: /([^}]|}(?!}))+/
         """
     )
 
 
-@functools.lru_cache(32)
-def parse_argo_template_tags(source: str) -> lark.Tree:
-    """
-    Parse Argo template tags in the given source string.
-
-    Parameters
-    ----------
-    source : str
-        The source string containing Argo template tags.
-
-    Returns
-    -------
-    lark.Tree
-        The parse tree representing the structure of the template tags.
-    """
-    parser = _argo_template_tag_parser()
-    return parser.parse(source)
-
-
 def check_template_tags(
     source: str, references: ReferenceCollection
 ) -> Iterator[Diagnosis]:
@@ -101,7 +99,7 @@ def check_template_tags(
     for tag in tree.find_data("simple_tag"):
         (ref_repr,) = tag.find_token("REF")
         ref_repr = typing.cast("Token", ref_repr)
-        if diagnosis := check_simple_tag_reference(ref_repr, references):
+        if diagnosis := _check_simple_tag_reference(ref_repr, references):
             yield diagnosis
 
 
@@ -137,7 +135,7 @@ def transform_lark_error(source: str, e: UnexpectedInput) -> Diagnosis:
     }
 
 
-def check_simple_tag_reference(
+def _check_simple_tag_reference(
     ref_repr: str, references: ReferenceCollection
 ) -> Diagnosis | None:
     # early exit if the reference is known

tugboat/references/context.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import typing
 from collections.abc import MutableSet
 
 from pydantic import BaseModel, Field, InstanceOf
@@ -10,6 +11,9 @@
     normalized_distance as dameau_levenshtein_normalized_distance,
 )
 
+if typing.TYPE_CHECKING:
+    from collections.abc import Iterable
+
 type _TR = tuple[str, ...]
 """Type alias for reference, which is a tuple of strings."""
 
@@ -39,10 +43,15 @@ class ReferenceCollection(MutableSet[_TR]):
     of :py:data:`AnyStr`.
     """
 
-    def __init__(self):
+    def __init__(self, iterable: Iterable[_TR] = ()):
+        super().__init__()
+
         self._static = set()
         self._dynamic = []
 
+        for item in iterable:
+            self.add(item)
+
     def __iter__(self):
         yield from self._static
         yield from self._dynamic