- Fixed #1301: Import headings in nested sections leads to check errors

CHANGELOG.md

@@ -3,6 +3,9 @@ Changelog
 
 NOTE: isort follows the [semver](https://semver.org/) versioning standard.
 
+### 5.0.9 July 11, 2020
+  - Fixed #1301: Import headings in nested sections leads to check errors
+
 ### 5.0.8 July 11, 2020
   - Fixed #1277 & #1278: New line detection issues on Windows.
   - Fixed #1294: Fix bundled git hook.

isort/api.py

@@ -8,18 +8,10 @@
 from warnings import warn
 
 from . import io, output, parse
-from .exceptions import (
-    ExistingSyntaxErrors,
-    FileSkipComment,
-    FileSkipSetting,
-    IntroducedSyntaxErrors,
-)
-from .format import (
-    ask_whether_to_apply_changes_to_file,
-    format_natural,
-    remove_whitespace,
-    show_unified_diff,
-)
+from .exceptions import (ExistingSyntaxErrors, FileSkipComment,
+                         FileSkipSetting, IntroducedSyntaxErrors)
+from .format import (ask_whether_to_apply_changes_to_file, format_natural,
+                     remove_whitespace, show_unified_diff)
 from .io import Empty
 from .place import module as place_module  # skipcq: PYL-W0611 (intended export of public API)
 from .place import module_with_reason as place_module_with_reason  # skipcq: PYL-W0611 (^)
@@ -398,7 +390,7 @@ def _sort_imports(
         else:
             stripped_line = line.strip()
             if stripped_line and not line_separator:
-                line_separator = line[len(line.rstrip()) :]
+                line_separator = line[len(line.rstrip()) :].replace(" ", "").replace("\t", "")
 
             for file_skip_comment in FILE_SKIP_COMMENTS:
                 if file_skip_comment in line:
@@ -451,6 +443,9 @@ def _sort_imports(
                     isort_off = True
                 elif stripped_line == "# isort: split":
                     not_imports = True
+                elif stripped_line in config.section_comments and not import_section:
+                    import_section += line
+                    indent = line[: -len(line.lstrip())]
                 elif not (stripped_line or contains_imports):
                     if add_imports and not indent:
                         import_section += line_separator.join(add_imports) + line_separator

isort/hooks.py

@@ -57,9 +57,11 @@ def git_hook(strict: bool = False, modify: bool = False) -> int:
             staged_cmd = ["git", "show", f":{filename}"]
             staged_contents = get_output(staged_cmd)
 
-            if not api.check_code_string(staged_contents, file_path=Path(filename)):
+            if not api.check_code_string(
+                staged_contents, file_path=Path(filename), settings_path=Path(filename).parent
+            ):
                 errors += 1
                 if modify:
-                    api.sort_file(filename)
+                    api.sort_file(filename, settings_path=Path(filename).parent)
 
     return errors if strict else 0

isort/output.py

@@ -185,7 +185,7 @@ def sorted_imports(
                 line,
                 in_quote="",
                 index=len(formatted_output),
-                section_comments=parsed.section_comments,
+                section_comments=config.section_comments,
             )
             if not should_skip and line.strip():
                 if (

isort/parse.py

@@ -80,7 +80,7 @@ def _strip_syntax(import_string: str) -> str:
 
 
 def skip_line(
-    line: str, in_quote: str, index: int, section_comments: List[str]
+    line: str, in_quote: str, index: int, section_comments: Tuple[str, ...]
 ) -> Tuple[bool, str]:
     """Determine if a given line should be skipped.
 
@@ -134,7 +134,6 @@ class ParsedContent(NamedTuple):
     original_line_count: int
     line_separator: str
     sections: Any
-    section_comments: List[str]
 
 
 def file_contents(contents: str, config: Config = DEFAULT_CONFIG) -> ParsedContent:
@@ -143,7 +142,6 @@ def file_contents(contents: str, config: Config = DEFAULT_CONFIG) -> ParsedConte
     in_lines = contents.split(line_separator)
     out_lines = []
     original_line_count = len(in_lines)
-    section_comments = [f"# {heading}" for heading in config.import_headings.values()]
     if config.old_finders:
         finder = FindersManager(config=config).find
     else:
@@ -172,10 +170,10 @@ def file_contents(contents: str, config: Config = DEFAULT_CONFIG) -> ParsedConte
         index += 1
         statement_index = index
         (skipping_line, in_quote) = skip_line(
-            line, in_quote=in_quote, index=index, section_comments=section_comments
+            line, in_quote=in_quote, index=index, section_comments=config.section_comments
         )
 
-        if line in section_comments and not skipping_line:
+        if line in config.section_comments and not skipping_line:
             if import_index == -1:
                 import_index = index - 1
             continue
@@ -435,5 +433,4 @@ def file_contents(contents: str, config: Config = DEFAULT_CONFIG) -> ParsedConte
         original_line_count=original_line_count,
         line_separator=line_separator,
         sections=config.sections,
-        section_comments=section_comments,
     )

isort/settings.py

@@ -11,7 +11,8 @@
 from distutils.util import strtobool as _as_bool
 from functools import lru_cache
 from pathlib import Path
-from typing import Any, Callable, Dict, FrozenSet, Iterable, List, Optional, Pattern, Set, Tuple
+from typing import (Any, Callable, Dict, FrozenSet, Iterable, List, Optional,
+                    Pattern, Set, Tuple)
 from warnings import warn
 
 from . import stdlibs
@@ -206,11 +207,15 @@ def __init__(
         config: Optional[_Config] = None,
         **config_overrides,
     ):
+        self._known_patterns: Optional[List[Tuple[Pattern[str], str]]] = None
+        self._section_comments: Optional[Tuple[str, ...]] = None
+
         if config:
             config_vars = vars(config).copy()
             config_vars.update(config_overrides)
             config_vars["py_version"] = config_vars["py_version"].replace("py", "")
             config_vars.pop("_known_patterns")
+            config_vars.pop("_section_comments")
             super().__init__(**config_vars)  # type: ignore
             return
 
@@ -318,7 +323,6 @@ def __init__(
                 combined_config.pop(f"{IMPORT_HEADING_PREFIX}{import_heading_key}")
             combined_config["import_headings"] = import_headings
 
-        self._known_patterns: Optional[List[Tuple[Pattern[str], str]]] = None
         super().__init__(sources=tuple(sources), **combined_config)  # type: ignore
 
     def is_skipped(self, file_path: Path) -> bool:
@@ -378,6 +382,14 @@ def known_patterns(self):
 
         return self._known_patterns
 
+    @property
+    def section_comments(self) -> Tuple[str, ...]:
+        if self._section_comments is not None:
+            return self._section_comments
+
+        self._section_comments = tuple(f"# {heading}" for heading in self.import_headings.values())
+        return self._section_comments
+
     @staticmethod
     def _parse_known_pattern(pattern: str) -> List[str]:
         """Expand pattern if identified as a directory and return found sub packages"""

tests/test_parse.py

@@ -36,7 +36,6 @@ def test_file_contents():
         original_line_count,
         _,
         _,
-        _,
     ) = parse.file_contents(TEST_CONTENTS, config=Config(default_section=""))
     assert "\n".join(in_lines) == TEST_CONTENTS
     assert "import" not in "\n".join(out_lines)

tests/test_regressions.py

@@ -245,3 +245,20 @@ def test_windows_newline_issue_1278():
         "from ..log import CheckLoggingAdapter, init_logging\r\n\r\n    init_logging()\r\n"
         "except ImportError:\r\n    pass\r\n"
     )
+
+
+def test_check_never_passes_with_indented_headings_1301():
+    """Test to ensure that test can pass even when there are indented headings.
+    See: https://github.com/timothycrosley/isort/issues/1301
+    """
+    assert isort.check_code(
+        """
+try:
+    # stdlib
+    import logging
+    from os import abc, path
+except ImportError:
+    pass
+""",
+        import_heading_stdlib="stdlib",
+    )