Removes leading whitespace when tailoring BUILD files without heade…

…r text (#13394)

Resolves #13351:

Additionally to allowing different names for `tailor`ed BUILD files, new behaviour was added to add copyright headers to BUILD files, but no tests were present to observe behaviour when the default (empty) header text was used. 

This adds such a test, and replaces the default header case with None. Things seem to work better now.

(Cherry-pick of 056c3fd / #13375)

# Rust tests and lints will be skipped. Delete if not intended.
[ci skip-rust]

# Building wheels and fs_util will be skipped. Delete if not intended.
[ci skip-build-wheels]

src/python/pants/core/goals/tailor.py

@@ -9,7 +9,7 @@
 from abc import ABCMeta
 from collections import defaultdict
 from dataclasses import dataclass
-from typing import Iterable, Mapping, cast
+from typing import Iterable, Mapping, Optional, cast
 
 from pants.base.specs import (
     AddressSpecs,
@@ -259,7 +259,7 @@ def register_options(cls, register):
             "--build-file-header",
             advanced=True,
             type=str,
-            default="",
+            default=None,
             help="A header, e.g., a copyright notice, to add to the content of created BUILD files.",
         )
 
@@ -285,8 +285,8 @@ def build_file_name(self) -> str:
         return cast(str, self.options.build_file_name)
 
     @property
-    def build_file_header(self) -> str:
-        return cast(str, self.options.build_file_header)
+    def build_file_header(self) -> str | None:
+        return cast(Optional[str], self.options.build_file_header)
 
     @property
     def build_file_indent(self) -> str:
@@ -416,7 +416,7 @@ async def restrict_conflicting_sources(ptgt: PutativeTarget) -> DisjointSourcePu
 class EditBuildFilesRequest:
     putative_targets: PutativeTargets
     name: str
-    header: str
+    header: str | None
     indent: str
 
 

src/python/pants/core/goals/tailor_test.py

@@ -1,8 +1,10 @@
 # Copyright 2021 Pants project contributors (see CONTRIBUTORS.md).
 # Licensed under the Apache License, Version 2.0 (see LICENSE).
+from __future__ import annotations
 
 import textwrap
 from dataclasses import dataclass
+from textwrap import dedent
 
 import pytest
 
@@ -123,19 +125,19 @@ def test_make_content_str() -> None:
         ],
     )
     assert (
-        textwrap.dedent(
-            """
-    fortran_library()
+        dedent(
+            """\
+            fortran_library()
 
-    fortran_tests(
-        name="tests",
-        sources=[
-            "test1.f90",
-            "test2.f90",
-        ],
-    )
-    """
-        ).lstrip()
+            fortran_tests(
+                name="tests",
+                sources=[
+                    "test1.f90",
+                    "test2.f90",
+                ],
+            )
+            """
+        )
         == content
     )
 
@@ -285,40 +287,75 @@ def test_edit_build_files(rule_runner: RuleRunner, name: str) -> None:
     expected = [
         FileContent(
             f"src/fortran/baz/{name}.pants",
-            textwrap.dedent(
-                """
-                Copyright © 2021 FooCorp.
+            dedent(
+                """\
+            Copyright © 2021 FooCorp.
 
-                fortran_library()
+            fortran_library()
             """
-            )
-            .lstrip()
-            .encode(),
+            ).encode(),
         ),
         FileContent(
             f"src/fortran/foo/{name}",
             textwrap.dedent(
+                """\
+                fortran_library(sources=["bar1.f90"])
+
+                # A comment spread
+                # over multiple lines.
+                fortran_library(
+                    name="foo0",
+                    sources=[
+                        "bar2.f90",
+                        "bar3.f90",
+                    ],
+                )
+
+                fortran_tests(
+                    name="tests",
+                    life_the_universe_and_everything=42,
+                )
                 """
-            fortran_library(sources=["bar1.f90"])
+            ).encode(),
+        ),
+    ]
+    actual = list(contents)
+    # We do these more laborious asserts instead of just comparing the lists so that
+    # on a text mismatch we see the actual string diff on the decoded strings.
+    assert len(expected) == len(actual)
+    for efc, afc in zip(expected, actual):
+        assert efc.path == afc.path
+        assert efc.content.decode() == afc.content.decode()
+        assert efc.is_executable == afc.is_executable
 
-            # A comment spread
-            # over multiple lines.
-            fortran_library(
-                name="foo0",
-                sources=[
-                    "bar2.f90",
-                    "bar3.f90",
-                ],
-            )
 
-            fortran_tests(
-                name="tests",
-                life_the_universe_and_everything=42,
-            )
-            """
-            )
-            .lstrip()
-            .encode(),
+def test_edit_build_files_without_header_text(rule_runner: RuleRunner) -> None:
+    rule_runner.create_dir("src/fortran/baz/BUILD")  # NB: A directory, not a file.
+    req = EditBuildFilesRequest(
+        PutativeTargets(
+            [
+                PutativeTarget.for_target_type(
+                    FortranLibrary, "src/fortran/baz", "baz", ["qux1.f90"]
+                ),
+            ]
+        ),
+        name="BUILD",
+        header=None,
+        indent="    ",
+    )
+    edited_build_files = rule_runner.request(EditedBuildFiles, [req])
+
+    assert edited_build_files.created_paths == ("src/fortran/baz/BUILD.pants",)
+
+    contents = rule_runner.request(DigestContents, [edited_build_files.digest])
+    expected = [
+        FileContent(
+            "src/fortran/baz/BUILD.pants",
+            dedent(
+                """\
+               fortran_library()
+               """
+            ).encode(),
         ),
     ]
     actual = list(contents)
@@ -331,6 +368,34 @@ def test_edit_build_files(rule_runner: RuleRunner, name: str) -> None:
         assert efc.is_executable == afc.is_executable
 
 
+@pytest.mark.parametrize("header", [None, "I am some header text"])
+def test_build_file_lacks_leading_whitespace(rule_runner: RuleRunner, header: str | None) -> None:
+    rule_runner.create_dir("src/fortran/baz/BUILD")  # NB: A directory, not a file.
+    req = EditBuildFilesRequest(
+        PutativeTargets(
+            [
+                PutativeTarget.for_target_type(
+                    FortranLibrary, "src/fortran/baz", "baz", ["qux1.f90"]
+                ),
+            ]
+        ),
+        name="BUILD",
+        header=header,
+        indent="    ",
+    )
+    edited_build_files = rule_runner.request(EditedBuildFiles, [req])
+
+    assert edited_build_files.created_paths == ("src/fortran/baz/BUILD.pants",)
+
+    contents = rule_runner.request(DigestContents, [edited_build_files.digest])
+    actual = list(contents)
+    # We do these more laborious asserts instead of just comparing the lists so that
+    # on a text mismatch we see the actual string diff on the decoded strings.
+    for afc in actual:
+        content = afc.content.decode()
+        assert content.lstrip() == content
+
+
 def test_group_by_dir() -> None:
     paths = {
         "foo/bar/baz1.ext",