erjemin/instaloader
erjemin/instaloader
1
0
Fork 0
Code Issues Wiki Activity

Revert date_utc and date_local being timezone aware

Browse Source 
Fixes #1379.

Revert "Make {Post,StoryItem}.date_utc timezone aware"
d864ce08ff43a2d86ac6123459bb9fa32b924657.

Revert "Remove unnecessary conversion to timezone aware timestamp (#1372)"
555c86633c61b106c5c3201e1d5f030a04a4f801.

Revert "Timezone option for datetime style formatting (#1316)"
21365ec6688b55c91803040693c3b67cc8820cb8.
This commit is contained in:
Alexander Graf 2022-01-15 17:05:41 +01:00
parent d6e5e31005
commit ca78fee307
2 changed files with 16 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
instaloader/instaloader.py
View File

@ -811,7 +811,7 @@ class Instaloader:
last_scraped = latest_stamps.get_last_story_timestamp(name) last_scraped = latest_stamps.get_last_story_timestamp(name)
scraped_timestamp = datetime.now().astimezone() scraped_timestamp = datetime.now().astimezone()
for item in user_story.get_items(): for item in user_story.get_items():
if latest_stamps is not None and item.date_local <= last_scraped: if latest_stamps is not None and item.date_utc.replace(tzinfo=timezone.utc) <= last_scraped:
break break
if storyitem_filter is not None and not storyitem_filter(item): if storyitem_filter is not None and not storyitem_filter(item):
self.context.log("<{} skipped>".format(item), flush=True) self.context.log("<{} skipped>".format(item), flush=True)
@ -1206,7 +1206,7 @@ class Instaloader:
posts_takewhile: Optional[Callable[[Post], bool]] = None posts_takewhile: Optional[Callable[[Post], bool]] = None
if latest_stamps is not None: if latest_stamps is not None:
last_scraped = latest_stamps.get_last_tagged_timestamp(profile.username) last_scraped = latest_stamps.get_last_tagged_timestamp(profile.username)
posts_takewhile = lambda p: p.date_local > last_scraped posts_takewhile = lambda p: p.date_utc.replace(tzinfo=timezone.utc) > last_scraped
tagged_posts = profile.get_tagged_posts() tagged_posts = profile.get_tagged_posts()
self.posts_download_loop(tagged_posts, self.posts_download_loop(tagged_posts,
target if target target if target
@ -1214,7 +1214,7 @@ class Instaloader:
_PostPathFormatter.sanitize_path(':tagged')), _PostPathFormatter.sanitize_path(':tagged')),
fast_update, post_filter, takewhile=posts_takewhile) fast_update, post_filter, takewhile=posts_takewhile)
if latest_stamps is not None and tagged_posts.first_item is not None: 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) latest_stamps.set_last_tagged_timestamp(profile.username, tagged_posts.first_item.date_local.astimezone())
def download_igtv(self, profile: Profile, fast_update: bool = False, def download_igtv(self, profile: Profile, fast_update: bool = False,
post_filter: Optional[Callable[[Post], bool]] = None, post_filter: Optional[Callable[[Post], bool]] = None,
@ -1229,12 +1229,12 @@ class Instaloader:
posts_takewhile: Optional[Callable[[Post], bool]] = None posts_takewhile: Optional[Callable[[Post], bool]] = None
if latest_stamps is not None: if latest_stamps is not None:
last_scraped = latest_stamps.get_last_igtv_timestamp(profile.username) last_scraped = latest_stamps.get_last_igtv_timestamp(profile.username)
posts_takewhile = lambda p: p.date_local > last_scraped posts_takewhile = lambda p: p.date_utc.replace(tzinfo=timezone.utc) > last_scraped
igtv_posts = profile.get_igtv_posts() igtv_posts = profile.get_igtv_posts()
self.posts_download_loop(igtv_posts, profile.username, fast_update, post_filter, self.posts_download_loop(igtv_posts, profile.username, fast_update, post_filter,
total_count=profile.igtvcount, owner_profile=profile, takewhile=posts_takewhile) total_count=profile.igtvcount, owner_profile=profile, takewhile=posts_takewhile)
if latest_stamps is not None and igtv_posts.first_item is not None: 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) latest_stamps.set_last_igtv_timestamp(profile.username, igtv_posts.first_item.date_local.astimezone())
def _get_id_filename(self, profile_name: str) -> str: def _get_id_filename(self, profile_name: str) -> str:
if ((format_string_contains_key(self.dirname_pattern, 'profile') or if ((format_string_contains_key(self.dirname_pattern, 'profile') or
@ -1427,14 +1427,14 @@ class Instaloader:
if latest_stamps is not None: if latest_stamps is not None:
# pylint:disable=cell-var-from-loop # pylint:disable=cell-var-from-loop
last_scraped = latest_stamps.get_last_post_timestamp(profile_name) last_scraped = latest_stamps.get_last_post_timestamp(profile_name)
posts_takewhile = lambda p: p.date_local > last_scraped posts_takewhile = lambda p: p.date_utc.replace(tzinfo=timezone.utc) > last_scraped
posts_to_download = profile.get_posts() posts_to_download = profile.get_posts()
self.posts_download_loop(posts_to_download, profile_name, fast_update, post_filter, self.posts_download_loop(posts_to_download, profile_name, fast_update, post_filter,
total_count=profile.mediacount, owner_profile=profile, total_count=profile.mediacount, owner_profile=profile,
takewhile=posts_takewhile) takewhile=posts_takewhile)
if latest_stamps is not None and posts_to_download.first_item is not None: if latest_stamps is not None and posts_to_download.first_item is not None:
latest_stamps.set_last_post_timestamp(profile_name, latest_stamps.set_last_post_timestamp(profile_name,
posts_to_download.first_item.date_local) posts_to_download.first_item.date_local.astimezone())
if stories and profiles: if stories and profiles:
with self.context.error_catcher("Download stories"): with self.context.error_catcher("Download stories"):

20
instaloader/structures.py
View File

@ -3,7 +3,7 @@ import lzma
import re import re
from base64 import b64decode, b64encode from base64 import b64decode, b64encode
from collections import namedtuple from collections import namedtuple
from datetime import datetime, timezone from datetime import datetime
from pathlib import Path from pathlib import Path
from typing import Any, Dict, Iterable, Iterator, List, Optional, Tuple, Union from typing import Any, Dict, Iterable, Iterator, List, Optional, Tuple, Union
@ -225,12 +225,16 @@ class Post:
@property @property
def date_local(self) -> datetime: def date_local(self) -> datetime:
"""Timestamp when the post was created (local time zone).""" """Timestamp when the post was created (local time zone)."""
return datetime.fromtimestamp(self.get_timestamp_date_created()).astimezone() return datetime.fromtimestamp(self._node["date"]
if "date" in self._node
else self._node["taken_at_timestamp"])
@property @property
def date_utc(self) -> datetime: def date_utc(self) -> datetime:
"""Timestamp when the post was created (UTC).""" """Timestamp when the post was created (UTC)."""
return datetime.utcfromtimestamp(self.get_timestamp_date_created()).replace(tzinfo=timezone.utc) return datetime.utcfromtimestamp(self._node["date"]
if "date" in self._node
else self._node["taken_at_timestamp"])
@property @property
def date(self) -> datetime: def date(self) -> datetime:
@ -271,12 +275,6 @@ class Post:
return len(edges) return len(edges)
return 1 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]: def get_is_videos(self) -> List[bool]:
""" """
Return a list containing the ``is_video`` property for each media in the post. Return a list containing the ``is_video`` property for each media in the post.
@ -1081,12 +1079,12 @@ class StoryItem:
@property @property
def date_local(self) -> datetime: def date_local(self) -> datetime:
"""Timestamp when the StoryItem was created (local time zone).""" """Timestamp when the StoryItem was created (local time zone)."""
return datetime.fromtimestamp(self._node['taken_at_timestamp']).astimezone() return datetime.fromtimestamp(self._node['taken_at_timestamp'])
@property @property
def date_utc(self) -> datetime: def date_utc(self) -> datetime:
"""Timestamp when the StoryItem was created (UTC).""" """Timestamp when the StoryItem was created (UTC)."""
return datetime.utcfromtimestamp(self._node['taken_at_timestamp']).replace(tzinfo=timezone.utc) return datetime.utcfromtimestamp(self._node['taken_at_timestamp'])
@property @property
def date(self) -> datetime: def date(self) -> datetime: