Release of Version 4.8.3

Also simplify {Post,StoryItem}.date_local.

Discussed in #1316.

.github/ISSUE_TEMPLATE/feature_request.md

@@ -6,13 +6,15 @@ labels: "feature suggestion"
 ---
 
 **Provide us a use case of the feature**
-How could the user invoke the new function? Which problem would it solve? If new information is obtained, how would it be further processed?
+How could the user invoke the new function? Which problem would it solve?
+If new information is obtained, how would it be further processed?
 
 **Describe the solution you'd like**
 A clear and concise description of what you want to happen.
 
 **Describe alternatives you've considered**
 A clear and concise description of any alternative solutions or features you've considered.
+How do you achieve the here-described goal without the feature you are requesting?
 
 **If the feature request is accepted, would you be willing to submit a pull request?**
 Yes / No
@@ -21,4 +23,9 @@ Yes / No
 **Additional context**
 Add any other context about the feature request here.
 
-<!-- please also see https://instaloader.github.io/contributing.html for how to suggest a feature -->
+<!-- 
+Also see https://instaloader.github.io/contributing.html for how to suggest a feature.
+
+Please understand that a feature suggestion issue that lacks the required information cannot be
+processed.
+-->

.github/workflows/lint.yml

@@ -8,7 +8,7 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [3.8, 3.9]
+        python-version: ["3.8", "3.9", "3.10"]
     steps:
       - name: Checkout Instaloader Repository
         uses: actions/checkout@v2

.github/workflows/windows_exe.yml

@@ -10,9 +10,9 @@ jobs:
       - name: Checkout Instaloader repository
         uses: actions/checkout@v2
       - name: Setup Python
-        uses: actions/setup-python@v1
+        uses: actions/setup-python@v2
         with:
-          python-version: 3.9
+          python-version: "3.9"
           architecture: x64
       - name: Get the tagged version
         id: get_version

deploy/windows/create_exe.py

@@ -45,7 +45,7 @@ with open('__main__.py', 'w+') as f:
 
 # install dependencies and invoke PyInstaller
 commands = ["pip install pipenv==2020.11.15",
-            "pipenv sync --dev",
+            f"pipenv --python {sys.version_info.major}.{sys.version_info.minor} sync --dev",
             "pipenv run pyinstaller --log-level=DEBUG instaloader.spec"]
 
 for command in commands:

instaloader/__init__.py

@@ -1,7 +1,7 @@
 """Download pictures (or videos) along with their captions and other metadata from Instagram."""
 
 
-__version__ = '4.8rc4'
+__version__ = '4.8.3'
 
 
 try:

instaloader/__main__.py

@@ -93,6 +93,7 @@ def _main(instaloader: Instaloader, targetlist: List[str],
     latest_stamps = None
     if latest_stamps_file is not None:
         latest_stamps = LatestStamps(latest_stamps_file)
+        instaloader.context.log(f"Using latest stamps from {latest_stamps_file}.")
     # Login, if desired
     if username is not None:
         if not re.match(r"^[A-Za-z0-9._]+$", username):

instaloader/instaloader.py

@@ -453,9 +453,12 @@ class Instaloader:
     def save_location(self, filename: str, location: PostLocation, mtime: datetime) -> None:
         """Save post location name and Google Maps link."""
         filename += '_location.txt'
-        location_string = (location.name + "\n" +
-                           "https://maps.google.com/maps?q={0},{1}&ll={0},{1}\n".format(location.lat,
-                                                                                        location.lng))
+        if location.lat is not None and location.lng is not None:
+            location_string = (location.name + "\n" +
+                               "https://maps.google.com/maps?q={0},{1}&ll={0},{1}\n".format(location.lat,
+                                                                                            location.lng))
+        else:
+            location_string = location.name
         with open(filename, 'wb') as text_file:
             with BytesIO(location_string.encode()) as bio:
                 shutil.copyfileobj(cast(IO, bio), text_file)
@@ -788,6 +791,7 @@ class Instaloader:
 
         if not userids:
             self.context.log("Retrieving all visible stories...")
+            profile_count = None
         else:
             userids = [p if isinstance(p, int) else p.userid for p in userids]
             profile_count = len(userids)
@@ -807,7 +811,7 @@ class Instaloader:
                 last_scraped = latest_stamps.get_last_story_timestamp(name)
                 scraped_timestamp = datetime.now().astimezone()
             for item in user_story.get_items():
-                if latest_stamps is not None and item.date_utc.replace(tzinfo=timezone.utc) <= last_scraped:
+                if latest_stamps is not None and item.date_local <= last_scraped:
                     break
                 if storyitem_filter is not None and not storyitem_filter(item):
                     self.context.log("<{} skipped>".format(item), flush=True)
@@ -1202,7 +1206,7 @@ class Instaloader:
         posts_takewhile: Optional[Callable[[Post], bool]] = None
         if latest_stamps is not None:
             last_scraped = latest_stamps.get_last_tagged_timestamp(profile.username)
-            posts_takewhile = lambda p: p.date_utc.replace(tzinfo=timezone.utc) > last_scraped
+            posts_takewhile = lambda p: p.date_local > last_scraped
         tagged_posts = profile.get_tagged_posts()
         self.posts_download_loop(tagged_posts,
                                  target if target
@@ -1210,7 +1214,7 @@ class Instaloader:
                                        _PostPathFormatter.sanitize_path(':tagged')),
                                  fast_update, post_filter, takewhile=posts_takewhile)
         if latest_stamps is not None and tagged_posts.first_item is not None:
-            latest_stamps.set_last_tagged_timestamp(profile.username, tagged_posts.first_item.date_local.astimezone())
+            latest_stamps.set_last_tagged_timestamp(profile.username, tagged_posts.first_item.date_local)
 
     def download_igtv(self, profile: Profile, fast_update: bool = False,
                       post_filter: Optional[Callable[[Post], bool]] = None,
@@ -1225,12 +1229,12 @@ class Instaloader:
         posts_takewhile: Optional[Callable[[Post], bool]] = None
         if latest_stamps is not None:
             last_scraped = latest_stamps.get_last_igtv_timestamp(profile.username)
-            posts_takewhile = lambda p: p.date_utc.replace(tzinfo=timezone.utc) > last_scraped
+            posts_takewhile = lambda p: p.date_local > last_scraped
         igtv_posts = profile.get_igtv_posts()
         self.posts_download_loop(igtv_posts, profile.username, fast_update, post_filter,
                                  total_count=profile.igtvcount, owner_profile=profile, takewhile=posts_takewhile)
         if latest_stamps is not None and igtv_posts.first_item is not None:
-            latest_stamps.set_last_igtv_timestamp(profile.username, igtv_posts.first_item.date_local.astimezone())
+            latest_stamps.set_last_igtv_timestamp(profile.username, igtv_posts.first_item.date_local)
 
     def _get_id_filename(self, profile_name: str) -> str:
         if ((format_string_contains_key(self.dirname_pattern, 'profile') or
@@ -1423,14 +1427,14 @@ class Instaloader:
                     if latest_stamps is not None:
                         # pylint:disable=cell-var-from-loop
                         last_scraped = latest_stamps.get_last_post_timestamp(profile_name)
-                        posts_takewhile = lambda p: p.date_utc.replace(tzinfo=timezone.utc) > last_scraped
+                        posts_takewhile = lambda p: p.date_local > last_scraped
                     posts_to_download = profile.get_posts()
                     self.posts_download_loop(posts_to_download, profile_name, fast_update, post_filter,
                                              total_count=profile.mediacount, owner_profile=profile,
                                              takewhile=posts_takewhile)
                     if latest_stamps is not None and posts_to_download.first_item is not None:
                         latest_stamps.set_last_post_timestamp(profile_name,
-                                                              posts_to_download.first_item.date_local.astimezone())
+                                                              posts_to_download.first_item.date_local)
 
         if stories and profiles:
             with self.context.error_catcher("Download stories"):

instaloader/instaloadercontext.py

@@ -529,6 +529,28 @@ class InstaloaderContext:
         :raises ConnectionException: When download repeatedly failed."""
         self.write_raw(self.get_raw(url), filename)
 
+    def head(self, url: str, allow_redirects: bool = False) -> requests.Response:
+        """HEAD a URL anonymously.
+
+        :raises QueryReturnedNotFoundException: When the server responds with a 404.
+        :raises QueryReturnedForbiddenException: When the server responds with a 403.
+        :raises ConnectionException: When request failed.
+
+        .. versionadded:: 4.7.6
+        """
+        with self.get_anonymous_session() as anonymous_session:
+            resp = anonymous_session.head(url, allow_redirects=allow_redirects)
+        if resp.status_code == 200:
+            return resp
+        else:
+            if resp.status_code == 403:
+                # suspected invalid URL signature
+                raise QueryReturnedForbiddenException("403 when accessing {}.".format(url))
+            if resp.status_code == 404:
+                # 404 not worth retrying.
+                raise QueryReturnedNotFoundException("404 when accessing {}.".format(url))
+            raise ConnectionException("HTTP error code {}.".format(resp.status_code))
+
     @property
     def root_rhx_gis(self) -> Optional[str]:
         """rhx_gis string returned in the / query."""

instaloader/nodeiterator.py

@@ -133,14 +133,15 @@ class NodeIterator(Iterator[T]):
             return item
         if self._data['page_info']['has_next_page']:
             query_response = self._query(self._data['page_info']['end_cursor'])
-            page_index, data = self._page_index, self._data
-            try:
-                self._page_index = 0
-                self._data = query_response
-            except KeyboardInterrupt:
-                self._page_index, self._data = page_index, data
-                raise
-            return self.__next__()
+            if self._data['edges'] != query_response['edges']:
+                page_index, data = self._page_index, self._data
+                try:
+                    self._page_index = 0
+                    self._data = query_response
+                except KeyboardInterrupt:
+                    self._page_index, self._data = page_index, data
+                    raise
+                return self.__next__()
         raise StopIteration()
 
     @property

instaloader/structures.py

@@ -3,9 +3,9 @@ import lzma
 import re
 from base64 import b64decode, b64encode
 from collections import namedtuple
-from datetime import datetime
+from datetime import datetime, timezone
 from pathlib import Path
-from typing import Any, Dict, Iterable, Iterator, List, Optional, Union
+from typing import Any, Dict, Iterable, Iterator, List, Optional, Tuple, Union
 
 from . import __version__
 from .exceptions import *
@@ -35,8 +35,8 @@ PostLocation.id.__doc__ = "ID number of location."
 PostLocation.name.__doc__ = "Location name."
 PostLocation.slug.__doc__ = "URL friendly variant of location name."
 PostLocation.has_public_page.__doc__ = "Whether location has a public page."
-PostLocation.lat.__doc__ = "Latitude (:class:`float`)."
-PostLocation.lng.__doc__ = "Longitude (:class:`float`)."
+PostLocation.lat.__doc__ = "Latitude (:class:`float` or None)."
+PostLocation.lng.__doc__ = "Longitude (:class:`float` or None)."
 
 
 class Post:
@@ -225,16 +225,12 @@ class Post:
     @property
     def date_local(self) -> datetime:
         """Timestamp when the post was created (local time zone)."""
-        return datetime.fromtimestamp(self._node["date"]
-                                      if "date" in self._node
-                                      else self._node["taken_at_timestamp"])
+        return datetime.fromtimestamp(self.get_timestamp_date_created()).astimezone()
 
     @property
     def date_utc(self) -> datetime:
         """Timestamp when the post was created (UTC)."""
-        return datetime.utcfromtimestamp(self._node["date"]
-                                         if "date" in self._node
-                                         else self._node["taken_at_timestamp"])
+        return datetime.utcfromtimestamp(self.get_timestamp_date_created()).replace(tzinfo=timezone.utc)
 
     @property
     def date(self) -> datetime:
@@ -275,6 +271,12 @@ class Post:
             return len(edges)
         return 1
 
+    def get_timestamp_date_created(self) -> float:
+        """Timestamp when the post was created"""
+        return (self._node["date"]
+                if "date" in self._node
+                else self._node["taken_at_timestamp"])
+
     def get_is_videos(self) -> List[bool]:
         """
         Return a list containing the ``is_video`` property for each media in the post.
@@ -377,13 +379,35 @@ class Post:
     def video_url(self) -> Optional[str]:
         """URL of the video, or None."""
         if self.is_video:
+            version_urls = []
+            try:
+                version_urls.append(self._field('video_url'))
+            except (InstaloaderException, KeyError, IndexError) as err:
+                self._context.error(f"Warning: Unable to fetch video from graphql of {self}: {err}")
             if self._context.iphone_support and self._context.is_logged_in:
                 try:
-                    url = self._iphone_struct['video_versions'][0]['url']
-                    return url
+                    version_urls.extend(version['url'] for version in self._iphone_struct['video_versions'])
                 except (InstaloaderException, KeyError, IndexError) as err:
-                    self._context.error('{} Unable to fetch high quality video version of {}.'.format(err, self))
-            return self._field('video_url')
+                    self._context.error(f"Unable to fetch high-quality video version of {self}: {err}")
+            version_urls = list(dict.fromkeys(version_urls))
+            if len(version_urls) == 0:
+                return None
+            if len(version_urls) == 1:
+                return version_urls[0]
+            url_candidates: List[Tuple[int, str]] = []
+            for idx, version_url in enumerate(version_urls):
+                try:
+                    url_candidates.append((
+                        int(self._context.head(version_url, allow_redirects=True).headers.get('Content-Length', 0)),
+                        version_url
+                    ))
+                except (InstaloaderException, KeyError, IndexError) as err:
+                    self._context.error(f"Video URL candidate {idx+1}/{len(version_urls)} for {self}: {err}")
+            if not url_candidates:
+                # All candidates fail: Fallback to default URL and handle errors later at the actual download attempt
+                return version_urls[0]
+            url_candidates.sort()
+            return url_candidates[-1][1]
         return None
 
     @property
@@ -556,10 +580,10 @@ class Post:
             return None
         location_id = int(loc['id'])
         if any(k not in loc for k in ('name', 'slug', 'has_public_page', 'lat', 'lng')):
-            loc = self._context.get_json("explore/locations/{0}/".format(location_id),
-                                         params={'__a': 1})['graphql']['location']
+            loc.update(self._context.get_json("explore/locations/{0}/".format(location_id),
+                                              params={'__a': 1})['native_location_data']['location_info'])
         self._location = PostLocation(location_id, loc['name'], loc['slug'], loc['has_public_page'],
-                                      loc['lat'], loc['lng'])
+                                      loc.get('lat'), loc.get('lng'))
         return self._location
 
 
@@ -1057,12 +1081,12 @@ class StoryItem:
     @property
     def date_local(self) -> datetime:
         """Timestamp when the StoryItem was created (local time zone)."""
-        return datetime.fromtimestamp(self._node['taken_at_timestamp'])
+        return datetime.fromtimestamp(self._node['taken_at_timestamp']).astimezone()
 
     @property
     def date_utc(self) -> datetime:
         """Timestamp when the StoryItem was created (UTC)."""
-        return datetime.utcfromtimestamp(self._node['taken_at_timestamp'])
+        return datetime.utcfromtimestamp(self._node['taken_at_timestamp']).replace(tzinfo=timezone.utc)
 
     @property
     def date(self) -> datetime:
@@ -1110,7 +1134,35 @@ class StoryItem:
     def video_url(self) -> Optional[str]:
         """URL of the video, or None."""
         if self.is_video:
-            return self._node['video_resources'][-1]['src']
+            version_urls = []
+            try:
+                version_urls.append(self._node['video_resources'][-1]['src'])
+            except (InstaloaderException, KeyError, IndexError) as err:
+                self._context.error(f"Warning: Unable to fetch video from graphql of {self}: {err}")
+            if self._context.iphone_support and self._context.is_logged_in:
+                try:
+                    version_urls.extend(version['url'] for version in self._iphone_struct['video_versions'])
+                except (InstaloaderException, KeyError, IndexError) as err:
+                    self._context.error(f"Unable to fetch high-quality video version of {self}: {err}")
+            version_urls = list(dict.fromkeys(version_urls))
+            if len(version_urls) == 0:
+                return None
+            if len(version_urls) == 1:
+                return version_urls[0]
+            url_candidates: List[Tuple[int, str]] = []
+            for idx, version_url in enumerate(version_urls):
+                try:
+                    url_candidates.append((
+                        int(self._context.head(version_url, allow_redirects=True).headers.get('Content-Length', 0)),
+                        version_url
+                    ))
+                except (InstaloaderException, KeyError, IndexError) as err:
+                    self._context.error(f"Video URL candidate {idx+1}/{len(version_urls)} for {self}: {err}")
+            if not url_candidates:
+                # All candidates fail: Fallback to default URL and handle errors later at the actual download attempt
+                return version_urls[0]
+            url_candidates.sort()
+            return url_candidates[-1][1]
         return None
 
 
@@ -1509,7 +1561,7 @@ class TopSearchResults:
             place = location.get('place', {})
             slug = place.get('slug')
             loc = place.get('location', {})
-            yield PostLocation(int(loc['pk']), loc['name'], slug, None, loc['lat'], loc['lng'])
+            yield PostLocation(int(loc['pk']), loc['name'], slug, None, loc.get('lat'), loc.get('lng'))
 
     def get_hashtag_strings(self) -> Iterator[str]:
         """