Unescaped html data and tidied up test html files

snejus · Jan 19, 2021 · 14be671 · 14be671
1 parent 41310d7
commit 14be671
Show file tree

Hide file tree

Showing 7 changed files with 190 additions and 6,808 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,3 +1,10 @@
+# [0.5.3] 2021-01-19
+
+### Fixed
+- For data that is parsed directly from the html, ampersands are now correctly
+  unescaped.
+
+
 # [0.5.2] 2021-01-18
 
 ### Fixed

diff --git a/beetsplug/bandcamp/__init__.py b/beetsplug/bandcamp/__init__.py
@@ -19,6 +19,7 @@
 
 import logging
 import re
+from html import unescape
 from itertools import chain
 from operator import truth
 from typing import Any, Dict, Iterator, List, Optional, Sequence, Set
@@ -73,7 +74,7 @@ def _get(self, url: str) -> str:
         except requests.exceptions.RequestException:
             self._info("Error while fetching URL: {}", url)
             return ""
-        return response.text
+        return unescape(response.text)
 
 
 class BandcampAlbumArt(BandcampRequestsHandler, fetchart.RemoteArtSource):

diff --git a/pyproject.toml b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "beetcamp"
-version = "0.5.2"
+version = "0.5.3"
 description = "Bandcamp autotagger source for beets (http://beets.io)."
 authors = ["Šarūnas Nejus <[email protected]>"]
 readme = "README.md"

diff --git a/tests/album.html b/tests/album.html
diff --git a/tests/compilation.html b/tests/compilation.html
diff --git a/tests/single.html b/tests/single.html
diff --git a/tests/track_alt.html b/tests/track_alt.html