Merge pull request #602 from carmenbianca/special-parse-endings

Strip special non-comment-style endings from regex matches

CHANGELOG.md

@@ -51,6 +51,12 @@ The versions follow [semantic versioning](https://semver.org).
 
   - Apache Velocity Template (Extensions: `.vm`, `.vtl`)
 
+- Some special endings are always stripped from copyright and licensing
+  statements (#602):
+
+  - `">` (and variations such as `'>`, `" >`, and `"/>`)
+  - `] ::`
+
 ### Changed
 
 - Updated PyPI development status to 'production/stable' (#381)

src/reuse/_util.py

@@ -21,6 +21,7 @@
 from difflib import SequenceMatcher
 from gettext import gettext as _
 from hashlib import sha1
+from itertools import chain
 from os import PathLike
 from pathlib import Path
 from typing import BinaryIO, Iterator, List, Optional, Set
@@ -42,14 +43,32 @@
 _LOGGER = logging.getLogger(__name__)
 _LICENSING = Licensing()
 
-_END_PATTERN = r"{}$".format(  # pylint: disable=consider-using-f-string
+# REUSE-IgnoreStart
+
+_END_PATTERN = r"{}$".format(
     "".join(
         {
-            r"(?:{})*".format(  # pylint: disable=consider-using-f-string
-                re.escape(style.MULTI_LINE.end)
+            r"(?:{})*".format(item)  # pylint: disable=consider-using-f-string
+            for item in chain(
+                (
+                    re.escape(style.MULTI_LINE.end)
+                    for style in _all_style_classes()
+                    if style.MULTI_LINE.end
+                ),
+                # These are special endings which do not belong to specific
+                # comment styles, but which we want to nonetheless strip away
+                # while parsing.
+                (
+                    ending
+                    for ending in [
+                        # ex: <tag value="Copyright Jane Doe">
+                        r'"\s*/*>',
+                        r"'\s*/*>",
+                        # ex: [SPDX-License-Identifier: GPL-3.0-or-later] ::
+                        r"\]\s*::",
+                    ]
+                ),
             )
-            for style in _all_style_classes()
-            if style.MULTI_LINE.end
         }
     )
 )
@@ -75,14 +94,12 @@
 ]
 
 _COPYRIGHT_STYLES = {
-    # REUSE-IgnoreStart
     "spdx": "SPDX-FileCopyrightText:",
     "spdx-symbol": "SPDX-FileCopyrightText: ©",
     "string": "Copyright",
     "string-c": "Copyright (C)",
     "string-symbol": "Copyright ©",
     "symbol": "©",
-    # REUSE-IgnoreEnd
 }
 
 # Amount of bytes that we assume will be big enough to contain the entire
@@ -492,3 +509,6 @@ def detect_line_endings(text: str) -> str:
         if line_ending in text:
             return line_ending
     return os.linesep
+
+
+# REUSE-IgnoreEnd

tests/test_util.py

@@ -186,6 +186,27 @@ def test_extract_sameline_multiline():
     assert result.copyright_lines == {"SPDX-FileCopyrightText: Jane Doe"}
 
 
+def test_extract_special_endings():
+    """Strip some non-comment-style endings from the end of copyright and
+    licensing information.
+    """
+    text = cleandoc(
+        """
+        <tag value="Copyright 2019 Jane Doe">
+        <tag value="Copyright 2019 John Doe" >
+        <tag value="Copyright 2019 Joe Somebody" />
+        <tag value='Copyright 2019 Alice'>
+        <tag value='Copyright 2019 Bob' >
+        <tag value='Copyright 2019 Eve' />
+        [Copyright 2019 Ajnulo] ::
+        """
+    )
+    result = _util.extract_spdx_info(text)
+    for item in result.copyright_lines:
+        assert ">" not in item
+        assert "] ::" not in item
+
+
 def test_filter_ignore_block_with_comment_style():
     """Test that the ignore block is properly removed if start and end markers
     are in comment style.