feat #198: support psycopg placeholders (#233)

CHANGELOG.md

@@ -4,6 +4,10 @@ All notable changes to this project will be documented in this file.
 
 ## [Unreleased]
 
+### Features
+
+-   sqlfmt now supports the psycopg placeholders `%s` and `%(name)s` ([#198](https://github.com/tconbeer/sqlfmt/issues/198) - thank you [@snorkysnark](https://github.com/snorkysnark)!)
+
 ### Formatting Changes + Bug Fixes
 
 -   sqlfmt now standardizes whitespace inside word tokens ([#201](https://github.com/tconbeer/sqlfmt/issues/201))

src/sqlfmt/dialect.py

@@ -219,6 +219,18 @@ def __init__(self) -> None:
                         actions.add_node_to_buffer, token_type=TokenType.COLON
                     ),
                 ),
+                Rule(
+                    name="other_identifiers",
+                    priority=907,
+                    pattern=group(
+                        r"@\w+",  # stages
+                        r"\$\d+",  # pg placeholders
+                        r"%(\([^%()]+\))?s",  # psycopg placeholders
+                    ),
+                    action=partial(
+                        actions.add_node_to_buffer, token_type=TokenType.NAME
+                    ),
+                ),
                 Rule(
                     name="operator",
                     priority=910,
@@ -249,6 +261,7 @@ def __init__(self) -> None:
                         r"-\|-",
                         r"[*+?]?\?",  # regex greedy/non-greedy, also ?
                         r"!!",  # negate text match
+                        r"%%",  # psycopg escaped mod operator
                         r"[+\-*/%&|^=<>:#!]=?",  # singles
                     ),
                     action=partial(
@@ -410,14 +423,6 @@ def __init__(self) -> None:
                     + group(r"\W", r"$"),
                     action=actions.handle_set_operator,
                 ),
-                Rule(
-                    name="other_identifiers",
-                    priority=4000,
-                    pattern=group(r"@\w+", r"\$\d+"),
-                    action=partial(
-                        actions.add_node_to_buffer, token_type=TokenType.NAME
-                    ),
-                ),
                 Rule(
                     name="name",
                     priority=5000,

tests/data/unformatted/119_psycopg_placeholders.sql

@@ -0,0 +1,11 @@
+SELECT image_data FROM images WHERE id = %s;
+SELECT image_data, dividend %% divisor as escaped_mod_operator FROM images WHERE id in (%(one)s, %(two)s, %(three)s, %(f!o^u+r)s);
+)))))__SQLFMT_OUTPUT__(((((
+select image_data
+from images
+where id = %s
+;
+select image_data, dividend %% divisor as escaped_mod_operator
+from images
+where id in (%(one)s, %(two)s, %(three)s, %(f!o^u+r)s)
+;

tests/functional_tests/test_general_formatting.py

@@ -31,6 +31,7 @@
         "unformatted/116_chained_booleans.sql",
         "unformatted/117_whitespace_in_tokens.sql",
         "unformatted/118_within_group.sql",
+        "unformatted/119_psycopg_placeholders.sql",
         "unformatted/200_base_model.sql",
         "unformatted/201_basic_snapshot.sql",
         "unformatted/202_unpivot_macro.sql",

tests/unit_tests/test_dialect.py

@@ -88,6 +88,8 @@ def test_rule_props_are_unique(self, dialect: Polyglot) -> None:
             ("main", "operator", "+"),
             ("main", "operator", "-"),
             ("main", "operator", "/"),
+            ("main", "operator", "%"),
+            ("main", "operator", "%%"),
             ("main", "operator", "<>"),
             ("main", "operator", "||"),
             ("main", "operator", "=>"),
@@ -183,6 +185,9 @@ def test_rule_props_are_unique(self, dialect: Polyglot) -> None:
             ("main", "name", "replace"),
             ("main", "other_identifiers", "$2"),
             ("main", "other_identifiers", "@my_unquoted_stage"),
+            ("main", "other_identifiers", "%s"),
+            ("main", "other_identifiers", "%(name)s"),
+            ("main", "other_identifiers", "%(anything! else!)s"),
             ("main", "newline", "\n"),
             ("jinja", "jinja_comment", "{# my comment #}"),
             ("jinja", "jinja_comment", "{#-my comment -#}"),