[FIX] website_slides: get video id from youtube url

The regexp used to get the youtube video id from its url is broken in
multiple ways.

If the video ID starts with the letter 'v' in the short (youtu.be) url
then the url is considered invalid. i.e. https://youtu.be/vmhB-pt7EfA

If the query option `v=<video id>` is not the first option of the query
in the regular url (youtube.com/watch) then the url is also considered
invalid. i.e. https://www.youtube.com/watch?feature=youtu.be&v=vmhB-pt7EfA

The problem has been solved by replacing the over-engineered regexp by
the dedicated parsing tools of `urllib.parse`.

opw-2006330

closes odoo/odoo#34036

Signed-off-by: Julien Castiaux <Julien00859@users.noreply.github.com>

3 files changed, 41 insertions, 5 deletions

diff --git a/addons/website_slides/models/slides.py b/addons/website_slides/models/slides.py
index 0261d04725d..11d1f50253d 100644
--- a/addons/website_slides/models/slides.py
+++ b/addons/website_slides/models/slides.py
@@ -531,11 +531,16 @@ class Slide(models.Model):
         return result
 
     def _find_document_data_from_url(self, url):
-        expr = re.compile(r'^.*((youtu.be/)|(v/)|(\/u\/\w\/)|(embed\/)|(watch\?))\??v?=?([^#\&\?]*).*')
-        arg = expr.match(url)
-        document_id = arg and arg.group(7) or False
-        if document_id:
-            return ('youtube', document_id)
+        url_obj = urls.url_parse(url)
+        if url_obj.ascii_host == 'youtu.be':
+            return ('youtube', url_obj.path[1:] if url_obj.path else False)
+        elif url_obj.ascii_host in ('youtube.com', 'www.youtube.com', 'm.youtube.com'):
+            v_query_value = url_obj.decode_query().get('v')
+            if v_query_value:
+                return ('youtube', v_query_value)
+            split_path = url_obj.path.split('/')
+            if len(split_path) >= 3 and split_path[1] in ('v', 'embed'):
+                return ('youtube', split_path[2])
 
         expr = re.compile(r'(^https:\/\/docs.google.com|^https:\/\/drive.google.com).*\/d\/([^\/]*)')
         arg = expr.match(url)
diff --git a/addons/website_slides/tests/__init__.py b/addons/website_slides/tests/__init__.py
new file mode 100644
index 00000000000..801e9234160
--- /dev/null
+++ b/addons/website_slides/tests/__init__.py
@@ -0,0 +1 @@
+from . import test_from_url
\ No newline at end of file
diff --git a/addons/website_slides/tests/test_from_url.py b/addons/website_slides/tests/test_from_url.py
new file mode 100644
index 00000000000..dc9a1d6768e
--- /dev/null
+++ b/addons/website_slides/tests/test_from_url.py
@@ -0,0 +1,30 @@
+import odoo.tests
+
+
+class TestFromURL(odoo.tests.TransactionCase):
+    def test_youtube_urls(self):
+        urls = {
+            'W0JQcpGLSFw': [
+                'https://youtu.be/W0JQcpGLSFw',
+                'https://www.youtube.com/watch?v=W0JQcpGLSFw',
+                'https://www.youtube.com/watch?v=W0JQcpGLSFw&list=PL1-aSABtP6ACZuppkBqXFgzpNb2nVctZx',
+            ],
+            'vmhB-pt7EfA': [  # id starts with v, it is important
+                'https://youtu.be/vmhB-pt7EfA',
+                'https://www.youtube.com/watch?feature=youtu.be&v=vmhB-pt7EfA',
+                'https://www.youtube.com/watch?v=vmhB-pt7EfA&list=PL1-aSABtP6ACZuppkBqXFgzpNb2nVctZx&index=7',
+            ],
+            'hlhLv0GN1hA': [
+                'https://www.youtube.com/v/hlhLv0GN1hA',
+                'https://www.youtube.com/embed/hlhLv0GN1hA',
+                'https://m.youtube.com/watch?v=hlhLv0GN1hA'
+            ],
+        }
+
+        model = self.env['slide.slide']
+        for id, urls in urls.items():
+            for url in urls:
+                with self.subTest(url=url, id=id):
+                    document = model._find_document_data_from_url(url)
+                    self.assertEqual(document[0], 'youtube')
+                    self.assertEqual(document[1], id)