[deviantart] add /view URL support

[the-blank-x]

URLs I found when searching for deviantart.com/view on web.archive.org:

• https://www.deviantart.com/view/<id>

  • http:// redirects you to https://
  • deviantart.com redirects you to www.deviantart.com
  • One trailing slash in the URL is ignored
  • If the deviation exists, it redirects you to https://www.deviantart.com/<username>/<art|journal>/<id>
  • If no deviation exists, it sends a 404
  • There are some URLs with an offset parameter, e.g. https://www.deviantart.com/view/14864502/?offset=80 (wayback machine), don't know what it means

• https://www.deviantart.com/view.php?id=<id>
https://www.deviantart.com/view-full.php?id=<id>

  • http:// redirects you to https://
  • deviantart.com redirects you to www.deviantart.com
  • Trailing slashes in the URL redirect you to the URL but with one trailing slash removed
  • Currently sends a 404, regardless if there's a deviation or not

[ClosedPort22]

The /view URL comes in very handy when only the index of the deviation (or its base36 equivalent) is known.

[rautamiekka]

Are ()() in-dev placeholders ?

[the-blank-x]

No, the first two groups are treated as an optional username, so you need to have two empty groups for URLs without a username, like DeviantartWatchPostsExtractor

[mikf]

It would also be possible to slightly extend the existing DeviationExtractor to have the same result without an extra HEAD request and without much extra code.

Deviantart does not seem to care about username, type, or slug in deviation URLs, so we can just use default names if none are given.

diff --git a/gallery_dl/extractor/deviantart.py b/gallery_dl/extractor/deviantart.py
index df59be4a..597abccf 100644
--- a/gallery_dl/extractor/deviantart.py
+++ b/gallery_dl/extractor/deviantart.py
@@ -854,7 +854,9 @@ class DeviantartDeviationExtractor(DeviantartExtractor):
     """Extractor for single deviations"""
     subcategory = "deviation"
     archive_fmt = "g_{_username}_{index}.{extension}"
-    pattern = BASE_PATTERN + r"/(art|journal)/(?:[^/?#]+-)?(\d+)"
+    pattern = (BASE_PATTERN + r"/(art|journal)/(?:[^/?#]+-)?(\d+)"
+               r"|(?:https?://)?(?:www\.)?deviantart\.com/"
+               r"(?:view/|view(?:-full)?\.php/*\?(?:[^#]+&)?id=)(\d+)")
     test = (
         (("https://www.deviantart.com/shimoda7/art/For-the-sake-10073852"), {
             "options": (("original", 0),),
@@ -919,11 +921,12 @@ class DeviantartDeviationExtractor(DeviantartExtractor):
     def __init__(self, match):
         DeviantartExtractor.__init__(self, match)
         self.type = match.group(3)
-        self.deviation_id = match.group(4)
+        self.deviation_id = match.group(4) or match.group(5)
 
     def deviations(self):
         url = "{}/{}/{}/{}".format(
-            self.root, self.user, self.type, self.deviation_id)
+            self.root, self.user or "u", self.type or "art", self.deviation_id)
+
         uuid = text.extract(self._limited_request(url).text,
                             '"deviationUuid\\":\\"', '\\')[0]
         if not uuid:

[the-blank-x]

Was a bit on the fence about guessing the full URL since there's no guarantee that it'll continue to work, but I suppose it works