Release of version 4.15.4

- `InstaloaderContext.get_page_data()` loads a page as a logged-out visitor and returns the query results that the page embeds. Instagram only embeds them when the request looks like a page navigation, so the request sends the `Sec-Fetch-*` headers of one.
- `Profile.from_username()` reads the profile from its public page. Anonymously, `_normalize_profile_data()` completes the node.
- Logged in, `Profile.get_posts()` uses query `28975909992013618` (`PolarisProfilePostsTabContentQuery_connection`) with the variables that the web app sends. The response has the same shape as before.

When logged in, `from_username()` does not normalize the page's node. The page shows the profile to a logged-out visitor, so normalizing would set `followed_by_viewer` to `False`, and `download_profiles()` would refuse private profiles that the session follows. Instead, the GraphQL profile query in `_obtain_metadata()` fills in the missing fields when they are first read.

.github/dependabot.yml

@@ -0,0 +1,15 @@
+version: 2
+updates:
+  - package-ecosystem: pip
+    directory: "/"
+    schedule:
+      interval: monthly
+    cooldown:
+      default-days: 7
+
+  - package-ecosystem: github-actions
+    directory: "/"
+    schedule:
+      interval: monthly
+    cooldown:
+      default-days: 7

.github/pull_request_template.md

@@ -2,17 +2,23 @@
 
 <!--
 Thanks for your willingness to contribute to Instaloader! Please briefly
-describe
+describe:
 
 - A motivation for this change, e.g.
   - Fixes # .
-  - More general: What problem does the pull request solve?
+  - What problem does the pull request solve?
 
 - The changes proposed in this pull request
+  - Keep the PR focused: avoid bundling independent fixes or refactorings.
 
 - The completeness of this change
-  - Is it just a proof of concept?
   - Is the documentation updated (if appropriate)?
-  - Do you consider it ready to be merged or is it a draft?
+  - How has it been tested?
+  - Do you consider it ready to be merged or is it a draft / proof of concept?
   - Can we help you at some point?
+
+- Further considerations
+  - Is it based on master (bugfix), upcoming/v4.X (new feature), or another pull request?
+  - Does it preserve backward compatibility and existing behavior?
+
 -->

.github/workflows/documentation.yml

@@ -8,17 +8,18 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: "Checkout Instaloader Repository"
-        uses: actions/checkout@v4
+        uses: actions/checkout@v7
         with:
           fetch-depth: 0  # needed for building docs
+          persist-credentials: false
       - name: "Setup Python"
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v7
         with:
-          python-version: 3.12
+          python-version: 3.15
           cache: 'pipenv'
       - name: "Install Dependencies"
         run: |
-          python -m pip install pipenv==2025.0.4
+          python -m pip install pipenv==2026.8.0
           pipenv --python `python --version | grep -Eo '3\.[0-9]+'` sync --dev
       - name: "Build Documentation"
         run: pipenv --python `python --version | grep -Eo '3\.[0-9]+'` run make -C docs html SPHINXOPTS="-W -n"

.github/workflows/lint.yml

@@ -9,18 +9,20 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.9", "3.10", "3.11", "3.12", "3.13"]
+        python-version: ["3.10", "3.11", "3.12", "3.13", "3.14", "3.15"]
     steps:
       - name: Checkout Instaloader Repository
-        uses: actions/checkout@v4
+        uses: actions/checkout@v7
+        with:
+          persist-credentials: false
       - name: Setup Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v7
         with:
           python-version: ${{ matrix.python-version }}
           cache: 'pipenv'
       - name: Install Dependencies
         run: |
-          python -m pip install pipenv==2025.0.4
+          python -m pip install pipenv==2026.8.0
           pipenv --python `python --version | grep -Eo '3\.[0-9]+'` sync --dev
       - name: PyLint
         run: pipenv --python `python --version | grep -Eo '3\.[0-9]+'` run pylint instaloader

.github/workflows/packages.yml

@@ -11,11 +11,13 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: "Checkout Repository"
-        uses: actions/checkout@v4
-      - name: "Setup Python"
-        uses: actions/setup-python@v5
+        uses: actions/checkout@v7
         with:
-          python-version: 3.12
+          persist-credentials: false
+      - name: "Setup Python"
+        uses: actions/setup-python@v7
+        with:
+          python-version: 3.15
       - name: "Get the tagged version"
         id: get_version
         env:

.github/workflows/stale.yml

@@ -11,7 +11,7 @@ jobs:
     runs-on: ubuntu-latest
     
     steps:
-    - uses: actions/stale@v9
+    - uses: actions/stale@v11
       with:
         repo-token: ${{ secrets.GITHUB_TOKEN }}
         only-labels: 'question'
@@ -21,7 +21,7 @@ jobs:
         days-before-stale: 21
         days-before-close: -1
         remove-stale-when-updated: false
-    - uses: actions/stale@v9
+    - uses: actions/stale@v11
       with:
         repo-token: ${{ secrets.GITHUB_TOKEN }}
         stale-issue-message: 'There has been no activity on this issue for an extended period of time. This issue will be closed after further 14 days of inactivity.'

.github/workflows/windows_exe.yml

@@ -8,11 +8,13 @@ jobs:
     runs-on: windows-latest
     steps:
       - name: Checkout Instaloader repository
-        uses: actions/checkout@v4
-      - name: Setup Python
-        uses: actions/setup-python@v5
+        uses: actions/checkout@v7
         with:
-          python-version: "3.13"
+          persist-credentials: false
+      - name: Setup Python
+        uses: actions/setup-python@v7
+        with:
+          python-version: 3.15
           cache: 'pipenv'
           architecture: x64
       - name: Get the tagged version
@@ -28,7 +30,7 @@ jobs:
       - name: Create draft release and upload EXE as release asset
         if: github.event_name == 'push' && startsWith(github.ref, 'refs/tags/v')
         id: create_release
-        uses: softprops/action-gh-release@v2
+        uses: softprops/action-gh-release@v3
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         with:

README.rst

@@ -124,11 +124,30 @@ Supporters
 .. current-sponsors-start
 
 | Instaloader is proudly sponsored by
-|  `@rocketapi-io <https://github.com/rocketapi-io>`__
+
+.. list-table::
+  :align: center
+
+  * - |socialapis|
+  * - `SocialAPIs <https://socialapis.io/?utm_source=github&utm_medium=sponsor&utm_campaign=instaloader>`__
+
+|
+
+.. list-table::
+  :align: center
+
+  * - `@rocketapi-io <https://github.com/rocketapi-io>`__
+
+|
 
 See `Alex' GitHub Sponsors <https://github.com/sponsors/aandergr>`__ page for
 how you can sponsor the development of Instaloader!
 
+.. |socialapis| image:: ./.sponsor-logos/socialapis.png
+   :alt: SocialAPIs
+   :target: https://socialapis.io/?utm_source=github&utm_medium=sponsor&utm_campaign=instaloader
+   :height: 96px
+
 .. current-sponsors-end
 
 It is a pleasure for us to share our Instaloader to the world, and we are proud

deploy/windows/create_exe.py

@@ -44,7 +44,7 @@ with open('__main__.py', 'w+') as f:
     f.writelines(lines)
 
 # install dependencies and invoke PyInstaller
-commands = ["pip install pipenv==2025.0.4",
+commands = ["pip install pipenv==2026.8.0",
             f"pipenv --python {sys.version_info.major}.{sys.version_info.minor} sync --dev",
             "pipenv run pyinstaller --log-level=DEBUG instaloader.spec"]
 

docs/.sponsor-logos

@@ -0,0 +1 @@
+../.sponsor-logos

docs/README.md

@@ -7,9 +7,10 @@ https://instaloader.github.io/
 
 ## Build
 
-The documentation is created with [Sphinx](http://www.sphinx-doc.org/). To build it, use
+The documentation is created with [Sphinx](http://www.sphinx-doc.org/). To build it, use (in the top-level directory
+of the Instaloader source tree):
 
 ```
-pip3 install -r requirements.txt
-make html
+pipenv sync --dev
+make -C docs html
 ```

docs/cli-options.rst

@@ -312,7 +312,7 @@ How to Download
 .. option:: --user-agent USER_AGENT
 
    User Agent to use for HTTP requests. Per default, Instaloader pretends being
-   Chrome/138 on Linux.
+   Chrome/142 on Linux.
 
 .. option:: --max-connection-attempts N
 

docs/codesnippets/113_only_one_per_user.py

@@ -2,7 +2,7 @@ import instaloader
 
 L = instaloader.Instaloader()
 
-posts = instaloader.Hashtag.from_name(L.context, 'urbanphotography').get_posts()
+posts = instaloader.Hashtag.from_name(L.context, 'urbanphotography').get_posts_resumable()
 
 users = set()
 

docs/codesnippets/615_import_firefox_session.py

@@ -14,8 +14,13 @@ def get_cookiefile():
     default_cookiefile = {
         "Windows": "~/AppData/Roaming/Mozilla/Firefox/Profiles/*/cookies.sqlite",
         "Darwin": "~/Library/Application Support/Firefox/Profiles/*/cookies.sqlite",
-    }.get(system(), "~/.mozilla/firefox/*/cookies.sqlite")
-    cookiefiles = glob(expanduser(default_cookiefile))
+    }.get(system(), None)
+    if default_cookiefile:
+        cookiefiles = glob(expanduser(default_cookiefile))
+    else:
+        # Check Firefox 147+ XDG path, then legacy
+        cookiefiles = glob(expanduser("~/.config/mozilla/firefox/*/cookies.sqlite")) or \
+                      glob(expanduser("~/.mozilla/firefox/*/cookies.sqlite"))
     if not cookiefiles:
         raise SystemExit("No Firefox cookies.sqlite file found. Use -c COOKIEFILE.")
     return cookiefiles[0]

docs/codesnippets/666_historical_hashtag_data.py

@@ -3,7 +3,7 @@ import instaloader
 
 L = instaloader.Instaloader()
 
-posts = instaloader.Hashtag.from_name(L.context, "urbanphotography").get_posts()
+posts = instaloader.Hashtag.from_name(L.context, "urbanphotography").get_posts_resumable()
 
 SINCE = datetime(2020, 5, 10)  # further from today, inclusive
 UNTIL = datetime(2020, 5, 11)  # closer to today, not inclusive

docs/contributing.rst

@@ -74,6 +74,14 @@ Improvements of the Instaloader source or its documentation can be proposed as a
     - a bug fix that does require thorough testing before being released to a
       final version.
 
+- Keep Pull Requests focused: address a single problem or feature in a
+  self-contained change. Avoid bundling multiple independent fixes, refactorings,
+  or architectural changes into one pull request, as this makes it hard to
+  review, validate, or selectively merge them.
+
+- Be mindful of backward compatibility: changes should preserve existing intended
+  behavior whenever possible.
+
 - All Pull Requests are analyzed by `Pylint <https://www.pylint.org/>`__ for
   error and syntax checking of the source and
   `Mypy <https://github.com/python/mypy>`__ for type checking. You can run them

docs/requirements.txt

@@ -1,2 +0,0 @@
-requests
-sphinx~=2.3

instaloader/__init__.py

@@ -1,7 +1,7 @@
 """Download pictures (or videos) along with their captions and other metadata from Instagram."""
 
 
-__version__ = '4.15a1'
+__version__ = '4.15.4'
 
 
 try:

instaloader/__main__.py

@@ -82,7 +82,7 @@ def filterstr_to_filterfunc(filter_str: str, item_type: type):
     return filterfunc
 
 
-def get_cookies_from_instagram(domain, browser, cookie_file='', cookie_name=''):
+def get_cookies_from_instagram(domain: str, browser: str, cookie_file: Optional[str] = None):
     supported_browsers = {
         "brave": browser_cookie3.brave,
         "chrome": browser_cookie3.chrome,
@@ -114,13 +114,10 @@ def get_cookies_from_instagram(domain, browser, cookie_file='', cookie_name=''):
         raise LoginException(f"No cookies found for Instagram in {browser}, "
                              f"Are you logged in successfully in {browser}?")
 
-    if cookie_name:
-        return cookies.get(cookie_name, {})
-    else:
-        return cookies
+    return cookies
 
 
-def import_session(browser, instaloader, cookiefile):
+def import_session(browser: str, instaloader: Instaloader, cookiefile: Optional[str]):
     cookie = get_cookies_from_instagram('instagram', browser, cookiefile)
     if cookie is not None:
         instaloader.context.update_cookies(cookie)

instaloader/instaloader.py

@@ -330,6 +330,9 @@ class Instaloader:
     def __exit__(self, *args):
         self.close()
 
+    def __repr__(self):
+        return f"<Instaloader context={self.context}>"
+
     @_retry_on_connection_error
     def download_pic(self, filename: str, url: str, mtime: datetime,
                      filename_suffix: Optional[str] = None, _attempt: int = 1) -> bool:

instaloader/instaloadercontext.py

@@ -2,6 +2,7 @@ import json
 import os
 import pickle
 import random
+import re
 import shutil
 import sys
 import textwrap
@@ -30,9 +31,22 @@ def copy_session(session: requests.Session, request_timeout: Optional[float] = N
     return new
 
 
+def _embedded_query_data(obj: Any) -> Iterator[Dict[str, Any]]:
+    """Yield the data of each GraphQL query result in JSON that a page of www.instagram.com embeds."""
+    if isinstance(obj, dict):
+        result = obj['__bbox'].get('result') if isinstance(obj.get('__bbox'), dict) else None
+        if isinstance(result, dict) and isinstance(result.get('data'), dict):
+            yield result['data']
+        for value in obj.values():
+            yield from _embedded_query_data(value)
+    elif isinstance(obj, list):
+        for value in obj:
+            yield from _embedded_query_data(value)
+
+
 def default_user_agent() -> str:
     return ('Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 '
-            '(KHTML, like Gecko) Chrome/138.0.0.0 Safari/537.36')
+            '(KHTML, like Gecko) Chrome/142.0.0.0 Safari/537.36')
 
 
 def default_iphone_headers() -> Dict[str, Any]:
@@ -149,6 +163,9 @@ class InstaloaderContext:
         if repeat_at_end:
             self.error_log.append(msg)
 
+    def __repr__(self):
+        return f"<InstaloaderContext is_logged_in={self.is_logged_in}>"
+
     @property
     def has_stored_errors(self) -> bool:
         """Returns whether any error has been reported and stored to be repeated at program termination.
@@ -191,7 +208,8 @@ class InstaloaderContext:
                   'Referer': 'https://www.instagram.com/',
                   'User-Agent': self.user_agent,
                   'X-Instagram-AJAX': '1',
-                  'X-Requested-With': 'XMLHttpRequest'}
+                  'X-Requested-With': 'XMLHttpRequest',
+                  'x-ig-app-id': '936619743392459'}
         if empty_session_only:
             del header['Host']
             del header['Origin']
@@ -395,7 +413,10 @@ class InstaloaderContext:
         :param session: Session to use, or None to use self.session
         :param use_post: Use POST instead of GET to make the request
         :return: Decoded response dictionary
-        :raises QueryReturnedBadRequestException: When the server responds with a 400.
+        :raises AbortDownloadException: When the server responds with
+            'feedback_required'/'checkpoint_required'/'challenge_required'
+        :raises QueryReturnedBadRequestException: When the server responds with a 400 (and not
+            'feedback_required'/'checkpoint_required'/'challenge_required').
         :raises QueryReturnedNotFoundException: When the server responds with a 404.
         :raises ConnectionException: When query repeatedly failed.
 
@@ -447,6 +468,15 @@ class InstaloaderContext:
                 response_headers.clear()
                 response_headers.update(resp.headers)
             if resp.status_code == 400:
+                with suppress(json.decoder.JSONDecodeError):
+                    if resp.json().get("message") in [
+                        "feedback_required",
+                        "checkpoint_required",
+                        "challenge_required",
+                    ]:
+                        # Raise AbortDownloadException in case of substantial Instagram
+                        # requirements to stop producing more requests
+                        raise AbortDownloadException(self._response_error(resp))
                 raise QueryReturnedBadRequestException(self._response_error(resp))
             if resp.status_code == 404:
                 raise QueryReturnedNotFoundException(self._response_error(resp))
@@ -528,6 +558,13 @@ class InstaloaderContext:
         :param referer: HTTP Referer, or None.
         :return: The server's response dictionary.
         """
+        csrf = next((c.value for c in self._session.cookies
+                     if c.name == 'csrftoken' and c.value), None)
+        if not csrf:
+            self._session.get('https://www.instagram.com/', timeout=self.request_timeout)
+            csrf = next((c.value for c in self._session.cookies
+                         if c.name == 'csrftoken' and c.value), '')
+
         with copy_session(self._session, self.request_timeout) as tmpsession:
             tmpsession.headers.update(self._default_http_header(empty_session_only=True))
             del tmpsession.headers['Connection']
@@ -535,6 +572,7 @@ class InstaloaderContext:
             tmpsession.headers['authority'] = 'www.instagram.com'
             tmpsession.headers['scheme'] = 'https'
             tmpsession.headers['accept'] = '*/*'
+            tmpsession.headers['x-csrftoken'] = csrf
             if referer is not None:
                 tmpsession.headers['referer'] = urllib.parse.quote(referer)
 
@@ -650,6 +688,38 @@ class InstaloaderContext:
 
             return response
 
+    def get_page_data(self, path: str) -> List[Dict[str, Any]]:
+        """Load a page of ``www.instagram.com`` as a logged-out visitor and return the data it embeds.
+
+        Instagram renders public profile and post pages with the results of their GraphQL queries
+        embedded in the HTML, even when it refuses anonymous API requests for the same data.
+
+        :param path: URL, relative to ``www.instagram.com/``
+        :return: The ``data`` object of each query result that the page embeds, in page order.
+        :raises QueryReturnedNotFoundException: When the server responds with a 404.
+        :raises ConnectionException: When the request failed.
+        """
+        self.do_sleep()
+        self._rate_controller.wait_before_query('other')
+        url = 'https://www.instagram.com/{0}'.format(path)
+        # Instagram only embeds the data when the request looks like a browser's page navigation.
+        headers = {'Accept': 'text/html', 'Sec-Fetch-Dest': 'document', 'Sec-Fetch-Mode': 'navigate',
+                   'Sec-Fetch-Site': 'none'}
+        try:
+            with self.get_anonymous_session() as anonymous_session:
+                resp = anonymous_session.get(url, headers=headers, allow_redirects=False)
+        except requests.exceptions.RequestException as err:
+            raise ConnectionException("Page request to {}: {}".format(url, err)) from err
+        if resp.status_code == 404:
+            raise QueryReturnedNotFoundException(self._response_error(resp))
+        if resp.status_code != 200:
+            raise ConnectionException(self._response_error(resp))
+        data: List[Dict[str, Any]] = []
+        for script in re.finditer(r'<script type="application/json"[^>]*>(.*?)</script>', resp.text, re.DOTALL):
+            with suppress(json.decoder.JSONDecodeError):
+                data.extend(_embedded_query_data(json.loads(script.group(1))))
+        return data
+
     def write_raw(self, resp: Union[bytes, requests.Response], filename: str) -> None:
         """Write raw response data into a file.
 

instaloader/nodeiterator.py

@@ -7,7 +7,7 @@ from datetime import datetime, timedelta
 from lzma import LZMAError
 from typing import Any, Callable, Dict, Iterable, Iterator, NamedTuple, Optional, Tuple, TypeVar
 
-from .exceptions import AbortDownloadException, InvalidArgumentException
+from .exceptions import InvalidArgumentException
 from .instaloadercontext import InstaloaderContext
 
 class FrozenNodeIterator(NamedTuple):
@@ -318,7 +318,7 @@ def resumable_iteration(context: InstaloaderContext,
             context.error("Warning: Not resuming from {}: {}".format(resume_file_path, exc))
     try:
         yield is_resuming, start_index
-    except (KeyboardInterrupt, AbortDownloadException):
+    except (Exception, KeyboardInterrupt):
         if os.path.dirname(resume_file_path):
             os.makedirs(os.path.dirname(resume_file_path), exist_ok=True)
         save(iterator.freeze(), resume_file_path)

instaloader/structures.py

@@ -152,7 +152,7 @@ _hashtag_regex = re.compile(r"(?:#)((?:\w){1,150})")
 # support Unicode and a word/beginning of string delimiter at the beginning to ensure
 # that no email addresses join the list of mentions.
 # http://blog.jstassen.com/2016/03/code-regex-for-instagram-username-and-hashtags/
-_mention_regex = re.compile(r"(?:^|[^\w\n]|_)(?:@)(\w(?:(?:\w|(?:\.(?!\.))){0,28}(?:\w))?)", re.ASCII)
+_mention_regex = re.compile(r"(?:^|\W|_)(?:@)(\w(?:(?:\w|(?:\.(?!\.))){0,28}(?:\w))?)", re.ASCII)
 
 
 def _optional_normalize(string: Optional[str]) -> Optional[str]:
@@ -258,6 +258,100 @@ class Post:
             fake_node["video_url"] = iphone_node["video_versions"][0]["url"]
         return fake_node
 
+    @staticmethod
+    def _fetch_play_count_from_clips(context: 'InstaloaderContext', user_id: str,
+                                     shortcode: str) -> Optional[int]:
+        """Fetch play_count for a reel via the clips connection endpoint as a fallback."""
+        try:
+            resp = context.doc_id_graphql_query(
+                "27234427476213202",
+                {"data": {"include_feed_video": True, "page_size": 12,
+                           "target_user_id": str(user_id)}},
+            )
+            edges = ((resp.get("data") or {})
+                     .get("xdt_api__v1__clips__user__connection_v2") or {})
+            for edge in edges.get("edges") or []:
+                media = (edge.get("node") or {}).get("media") or {}
+                if media.get("code") == shortcode:
+                    return media.get("play_count")
+        except (ConnectionException, BadResponseException):
+            pass
+        return None
+
+    @staticmethod
+    def _normalize_post_data(media: Dict[str, Any], context: 'InstaloaderContext') -> Dict[str, Any]:
+        """Normalize a Polaris media item to a legacy-compatible node, preserving the source structure."""
+        media_types = {1: "GraphImage", 2: "GraphVideo", 8: "GraphSidecar"}
+        media_type = media.get("media_type")
+        typename = None
+        if isinstance(media_type, int):
+            typename = media_types.get(media_type)
+        if not typename:
+            raise BadResponseException(f"Unknown media_type in metadata: {media_type}.")
+        pic_json: Dict[str, Any] = media.copy()
+        pic_json["shortcode"] = media["code"]
+        pic_json["id"] = media["pk"]
+        pic_json["__typename"] = typename
+        pic_json["is_video"] = media_type == 2
+        pic_json["taken_at_timestamp"] = media["taken_at"]
+        pic_json["owner"] = {
+            "id": media["user"]["pk"],
+            "username": media["user"].get("username", ""),
+            "full_name": media["user"].get("full_name", ""),
+        }
+        candidates = (media.get("image_versions2") or {}).get("candidates") or []
+        pic_json["display_url"] = candidates[0]["url"] if candidates else None
+        video_versions = media.get("video_versions") or []
+        pic_json["video_url"] = video_versions[0]["url"] if video_versions else None
+        pic_json["video_duration"] = media.get("video_duration")
+        pic_json["video_view_count"] = media.get("view_count")
+        pic_json["video_play_count"] = media.get("play_count")
+        if media_type == 2 and pic_json["video_view_count"] is None:
+            pic_json["video_play_count"] = Post._fetch_play_count_from_clips(
+                context, media["user"]["pk"], media["code"]
+            )
+        caption = media.get("caption")
+        caption_text = caption.get("text") if isinstance(caption, dict) else None
+        pic_json["edge_media_to_caption"] = (
+            {"edges": [{"node": {"text": caption_text}}]} if caption_text is not None
+            else {"edges": []}
+        )
+        pic_json["edge_media_preview_like"] = {"count": media.get("like_count") or 0}
+        pic_json["edge_media_to_parent_comment"] = {
+            "count": media.get("comment_count") or 0,
+            "edges": [],
+        }
+        if media.get("has_liked") is not None:
+            pic_json["viewer_has_liked"] = media["has_liked"]
+        carousel = media.get("carousel_media") or []
+        if carousel:
+            carousel_nodes = []
+            for item in carousel:
+                item_type = item.get("media_type", 1)
+                node: Dict[str, Any] = {
+                    "shortcode": item.get("code", ""),
+                    "__typename": media_types.get(item_type, "GraphImage"),
+                    "is_video": item_type == 2,
+                }
+                item_candidates = (item.get("image_versions2") or {}).get("candidates") or []
+                node["display_url"] = item_candidates[0]["url"] if item_candidates else ""
+                item_videos = item.get("video_versions") or []
+                node["video_url"] = item_videos[0]["url"] if item_videos else None
+                if item.get("accessibility_caption") is not None:
+                    node["accessibility_caption"] = item["accessibility_caption"]
+                carousel_nodes.append({"node": node})
+            pic_json["edge_sidecar_to_children"] = {"edges": carousel_nodes}
+        tagged = (media.get("usertags") or {}).get("in") or []
+        if tagged:
+            pic_json["edge_media_to_tagged_user"] = {
+                "edges": [
+                    {"node": {"user": {"username": t["user"]["username"].lower()}}}
+                    for t in tagged
+                    if (t.get("user") or {}).get("username")
+                ]
+            }
+        return pic_json
+
     @staticmethod
     def shortcode_to_mediaid(code: str) -> int:
         if len(code) > 11:
@@ -319,23 +413,18 @@ class Post:
 
     def _obtain_metadata(self):
         if not self._full_metadata_dict:
-            pic_json = self._context.doc_id_graphql_query(
-                "8845758582119845", {"shortcode": self.shortcode}
-            )["data"]["xdt_shortcode_media"]
-            if pic_json is None:
+            resp = self._context.doc_id_graphql_query(
+                "27128499623469141",
+                {
+                    "shortcode": self.shortcode,
+                    "__relay_internal__pv__PolarisAIGMMediaWebLabelEnabledrelayprovider": False,
+                },
+            )
+            web_info = (resp.get("data") or {}).get("xdt_api__v1__media__shortcode__web_info") or {}
+            items = web_info.get("items")
+            if not items:
                 raise BadResponseException("Fetching Post metadata failed.")
-            try:
-                xdt_types = {
-                    "XDTGraphImage": "GraphImage",
-                    "XDTGraphVideo": "GraphVideo",
-                    "XDTGraphSidecar": "GraphSidecar",
-                }
-                pic_json["__typename"] = xdt_types[pic_json["__typename"]]
-            except KeyError as exc:
-                raise BadResponseException(
-                    f"Unknown __typename in metadata: {pic_json['__typename']}."
-                ) from exc
-            self._full_metadata_dict = pic_json
+            self._full_metadata_dict = Post._normalize_post_data(items[0], self._context)
             if self.shortcode != self._full_metadata_dict['shortcode']:
                 self._node.update(self._full_metadata_dict)
                 raise PostChangedException
@@ -602,6 +691,15 @@ class Post:
             return self._field('video_view_count')
         return None
 
+    @property
+    def video_play_count(self) -> Optional[int]:
+        """Play count of the video, or None.
+
+        .. versionadded:: 4.14.3"""
+        if self.is_video:
+            return self._field('video_play_count')
+        return None
+
     @property
     def video_duration(self) -> Optional[float]:
         """Duration of the video in seconds, or None.
@@ -880,7 +978,8 @@ class Profile:
     Also, this class implements == and is hashable.
     """
     def __init__(self, context: InstaloaderContext, node: Dict[str, Any]):
-        assert 'username' in node
+        assert "username" in node
+        assert "id" in node or "pk" in node
         self._context = context
         self._has_public_story: Optional[bool] = None
         self._node = node
@@ -900,10 +999,30 @@ class Profile:
         :param username: Username
         :raises: :class:`ProfileNotExistsException`
         """
-        # pylint:disable=protected-access
-        profile = cls(context, {'username': username.lower()})
-        profile._obtain_metadata()  # to raise ProfileNotExistsException now in case username is invalid
-        return profile
+        # Resolve the profile through its public web page, which Instagram renders with the
+        # profile embedded. The web_profile_info endpoint previously used here is refused
+        # with HTTP 400 for some professional profiles, and with HTTP 401 or 429 for some
+        # sessions, both anonymously and when logged in.
+        try:
+            page = context.get_page_data("{0}/".format(username.lower()))
+        except QueryReturnedNotFoundException:
+            page = []
+        # The page embeds two queries for the profile: one for its details and one for its
+        # posts.
+        user: Dict[str, Any] = {}
+        for data in page:
+            user.update(data.get("xig_user_by_username") or {})
+        if not user.get("pk"):
+            raise ProfileNotExistsException("Profile {} does not exist.".format(username))
+        user.pop("polaris_ordered_timeline_connection", None)
+        # The page's "id" is not the user ID, which is "pk".
+        node = dict(user, id=user["pk"])
+        if context.is_logged_in:
+            # The page shows the profile as a logged-out visitor sees it. Leave the missing
+            # fields, notably the viewer's relationship to the profile, such as
+            # followed_by_viewer, to the GraphQL profile query in _obtain_metadata().
+            return cls(context, node)
+        return cls(context, cls._normalize_profile_data(node))
 
     @classmethod
     def from_id(cls, context: InstaloaderContext, profile_id: int):
@@ -916,15 +1035,16 @@ class Profile:
         """
         if profile_id in context.profile_id_cache:
             return context.profile_id_cache[profile_id]
-        data = context.graphql_query('7c16654f22c819fb63d1183034a5162f',
-                                     {'user_id': str(profile_id),
-                                      'include_chaining': False,
-                                      'include_reel': True,
-                                      'include_suggested_users': False,
-                                      'include_logged_out_extras': False,
-                                      'include_highlight_reels': False})['data']['user']
-        if data:
-            profile = cls(context, data['reel']['owner'])
+        # Resolve the current username from the user id, then load the full profile.
+        # The GraphQL user query previously used here started responding with HTTP 400.
+        try:
+            user = context.get_json(
+                "api/v1/users/{0}/info/".format(profile_id), params={}
+            ).get('user')
+        except QueryReturnedNotFoundException:
+            user = None
+        if user and user.get('username'):
+            profile = cls.from_username(context, user['username'])
         else:
             raise ProfileNotExistsException("No profile found, the user may have blocked you (ID: " +
                                             str(profile_id) + ").")
@@ -970,11 +1090,41 @@ class Profile:
     def _obtain_metadata(self):
         try:
             if not self._has_full_metadata:
-                metadata = self._context.get_iphone_json(f'api/v1/users/web_profile_info/?username={self.username}',
-                                                         params={})
-                if metadata['data']['user'] is None:
+                if not self._context.is_logged_in:
+                    # Anonymous access: the web_profile_info endpoint returns the full
+                    # node in the legacy format and still works, unlike the GraphQL
+                    # profile query.
+                    try:
+                        data = self._context.get_json(
+                            "api/v1/users/web_profile_info/",
+                            params={"username": self.username},
+                        ).get('data')
+                    except QueryReturnedNotFoundException as err:
+                        raise ProfileNotExistsException(
+                            'Profile {} does not exist.'.format(self.username)) from err
+                    user_data = data.get('user') if data else None
+                    if user_data is None:
+                        raise ProfileNotExistsException('Profile {} does not exist.'.format(self.username))
+                    self._node = user_data
+                    self._has_full_metadata = True
+                    return
+                user_id = self._node.get('id') or self._node.get('pk')
+                variables = {
+                    "id": str(user_id),
+                    "render_surface": "PROFILE",
+                    "__relay_internal__pv__PolarisCannesGuardianExperienceEnabledrelayprovider": True,
+                    "__relay_internal__pv__PolarisCASB976ProfileEnabledrelayprovider": False,
+                    "__relay_internal__pv__PolarisRepostsConsumptionEnabledrelayprovider": False,
+                    "__relay_internal__pv__PolarisWebSchoolsEnabledrelayprovider": False,
+                    "enable_integrity_filters": True,
+                }
+                data = self._context.doc_id_graphql_query('27937681195819736', variables)
+                if data is None:
+                    raise QueryReturnedNotFoundException('GraphQL query returned None')
+                user_data = data.get('data', {}).get('user')
+                if user_data is None:
                     raise ProfileNotExistsException('Profile {} does not exist.'.format(self.username))
-                self._node = metadata['data']['user']
+                self._node = self._normalize_profile_data(user_data)
                 self._has_full_metadata = True
         except (QueryReturnedNotFoundException, KeyError) as err:
             top_search_results = TopSearchResults(self._context, self.username)
@@ -989,6 +1139,45 @@ class Profile:
                                                         ', '.join(similar_profiles[0:5]))) from err
             raise ProfileNotExistsException('Profile {} does not exist.'.format(self.username)) from err
 
+    @staticmethod
+    def _normalize_profile_data(user_data: Dict[str, Any]) -> Dict[str, Any]:
+        """Normalize PolarisProfilePageContentQuery response to match legacy format."""
+        normalized = user_data.copy()
+        if 'id' not in normalized and 'pk' in normalized:
+            normalized['id'] = normalized['pk']
+        if 'edge_owner_to_timeline_media' not in normalized and 'media_count' in normalized:
+            normalized['edge_owner_to_timeline_media'] = {'count': normalized['media_count']}
+        if 'edge_felix_video_timeline' not in normalized:
+            normalized['edge_felix_video_timeline'] = {'count': 0}
+        if 'edge_followed_by' not in normalized and 'follower_count' in normalized:
+            normalized['edge_followed_by'] = {'count': normalized['follower_count']}
+        if 'edge_follow' not in normalized and 'following_count' in normalized:
+            normalized['edge_follow'] = {'count': normalized['following_count']}
+        if 'is_business_account' not in normalized and 'is_business' in normalized:
+            normalized['is_business_account'] = normalized['is_business']
+        if 'business_category_name' not in normalized and 'category' in normalized:
+            normalized['business_category_name'] = normalized['category']
+        friendship = normalized.get('friendship_status', {}) or {}
+        if 'followed_by_viewer' not in normalized:
+            normalized['followed_by_viewer'] = friendship.get('following', False)
+        if 'follows_viewer' not in normalized:
+            normalized['follows_viewer'] = friendship.get('followed_by', False)
+        if 'blocked_by_viewer' not in normalized:
+            normalized['blocked_by_viewer'] = friendship.get('blocking', False)
+        if 'has_blocked_viewer' not in normalized:
+            normalized['has_blocked_viewer'] = False
+        if 'has_requested_viewer' not in normalized:
+            normalized['has_requested_viewer'] = friendship.get('incoming_request', False)
+        if 'requested_by_viewer' not in normalized:
+            normalized['requested_by_viewer'] = friendship.get('outgoing_request', False)
+        if 'profile_pic_url_hd' not in normalized:
+            hd_info = normalized.get('hd_profile_pic_url_info')
+            if hd_info and 'url' in hd_info:
+                normalized['profile_pic_url_hd'] = hd_info['url']
+            elif 'profile_pic_url' in normalized:
+                normalized['profile_pic_url_hd'] = normalized['profile_pic_url']
+        return normalized
+
     def _metadata(self, *keys) -> Any:
         try:
             d = self._node
@@ -1197,18 +1386,52 @@ class Profile:
 
         :rtype: NodeIterator[Post]"""
         self._obtain_metadata()
+        if not self._context.is_logged_in:
+            return NodeIterator(
+                context=self._context,
+                edge_extractor=lambda d: d["data"]["user"]["edge_owner_to_timeline_media"],
+                node_wrapper=lambda n: Post(self._context, n, self),
+                query_variables={
+                    "data": {
+                        "count": 12,
+                        "include_relationship_info": True,
+                        "latest_besties_reel_media": True,
+                        "latest_reel_media": True,
+                    },
+                    "id": self.userid,
+                },
+                query_referer="https://www.instagram.com/{0}/".format(self.username),
+                is_first=Profile._make_is_newest_checker(),
+                doc_id="7950326061742207",
+                query_hash=None,
+                first_data=self._metadata("edge_owner_to_timeline_media"),
+            )
         return NodeIterator(
-            context = self._context,
-            edge_extractor = lambda d: d['data']['xdt_api__v1__feed__user_timeline_graphql_connection'],
-            node_wrapper = lambda n: Post.from_iphone_struct(self._context, n),
-            query_variables = {'data': {
-                'count': 12, 'include_relationship_info': True,
-                'latest_besties_reel_media': True, 'latest_reel_media': True},
-             'username': self.username},
-            query_referer = 'https://www.instagram.com/{0}/'.format(self.username),
-            is_first = Profile._make_is_newest_checker(),
-            doc_id = '7898261790222653',
-            query_hash = None,
+            context=self._context,
+            edge_extractor=lambda d: d["data"]["xdt_api__v1__feed__user_timeline_graphql_connection"],
+            node_wrapper=lambda n: Post.from_iphone_struct(self._context, n),
+            # The query and variables that the web app sends for a profile's posts tab
+            # (PolarisProfilePostsTabContentQuery_connection). Instagram answers the
+            # previous query, 7898261790222653, with "execution error".
+            query_variables={
+                "data": {
+                    "count": 12,
+                    "include_reel_media_seen_timestamp": True,
+                    "include_relationship_info": True,
+                    "latest_besties_reel_media": True,
+                    "latest_reel_media": True,
+                },
+                "first": 12,
+                "include_multi_captions": True,
+                "username": self.username,
+                "__relay_internal__pv__PolarisMultiCaptionCarouselEnabledrelayprovider": True,
+                "__relay_internal__pv__PolarisShortDramaEnabledrelayprovider": False,
+                "__relay_internal__pv__PolarisReelsRecoDebugOverlayEnabledrelayprovider": False,
+            },
+            query_referer="https://www.instagram.com/{0}/".format(self.username),
+            is_first=Profile._make_is_newest_checker(),
+            doc_id="28975909992013618",
+            query_hash=None,
         )
 
     def get_saved_posts(self) -> NodeIterator[Post]:

setup.py

@@ -18,8 +18,8 @@ def get_version():
     raise SystemExit("Could not find version string.")
 
 
-if sys.version_info < (3, 9):
-    sys.exit('Instaloader requires Python >= 3.9.')
+if sys.version_info < (3, 10):
+    sys.exit('Instaloader requires Python >= 3.10.')
 
 requirements = ['requests>=2.25']
 optional_requirements = {
@@ -46,7 +46,7 @@ setup(
                 'from Instagram.',
     long_description=open(os.path.join(SRC, 'README.rst')).read(),
     install_requires=requirements,
-    python_requires='>=3.9',
+    python_requires='>=3.10',
     extras_require=optional_requirements,
     entry_points={'console_scripts': ['instaloader=instaloader.__main__:main']},
     zip_safe=False,
@@ -58,11 +58,12 @@ setup(
         'Intended Audience :: Developers',
         'License :: OSI Approved :: MIT License',
         'Operating System :: OS Independent',
-        'Programming Language :: Python :: 3.9',
         'Programming Language :: Python :: 3.10',
         'Programming Language :: Python :: 3.11',
         'Programming Language :: Python :: 3.12',
         'Programming Language :: Python :: 3.13',
+        'Programming Language :: Python :: 3.14',
+        'Programming Language :: Python :: 3.15',
         'Programming Language :: Python :: 3 :: Only',
         'Topic :: Internet',
         'Topic :: Multimedia :: Graphics'

test/README.md

@@ -0,0 +1,4 @@
+# Instaloader Tests
+
+Integration tests for Instaloader features that require Instagram access are maintained here:
+https://github.com/instaloader/instaloader-tests

test/instaloader_unittests.py

@@ -1,214 +0,0 @@
-"""Unit Tests for Instaloader"""
-
-import os
-import shutil
-import tempfile
-import unittest
-from itertools import islice
-from typing import Optional
-
-import instaloader
-
-PROFILE_WITH_HIGHLIGHTS = 325732271
-PUBLIC_PROFILE = "selenagomez"
-PUBLIC_PROFILE_ID = 460563723
-PUBLIC_PROFILE_WITH_IGTV = "natgeo"
-HASHTAG = "kitten"
-LOCATION = "362629379"
-OWN_USERNAME = "aandergr"
-NORMAL_MAX_COUNT = 2
-PAGING_MAX_COUNT = 15
-PRIVATE_PROFILE = "aandergr"
-PRIVATE_PROFILE_ID = 1706625676
-EMPTY_PROFILE = "not_public"
-EMPTY_PROFILE_ID = 1928659031
-
-ratecontroller: Optional[instaloader.RateController] = None
-
-
-class TestInstaloaderAnonymously(unittest.TestCase):
-
-    def setUp(self):
-        self.dir = tempfile.mkdtemp()
-        print("Testing in {}".format(self.dir))
-        os.chdir(self.dir)
-        self.L = instaloader.Instaloader(download_geotags=True,
-                                         download_comments=True,
-                                         save_metadata=True)
-        self.L.context.raise_all_errors = True
-        if ratecontroller is not None:
-            # pylint:disable=protected-access
-            ratecontroller._context = self.L.context
-            self.L.context._rate_controller = ratecontroller
-
-    def tearDown(self):
-        # pylint:disable=global-statement,protected-access
-        global ratecontroller
-        ratecontroller = self.L.context._rate_controller
-        self.L.close()
-        os.chdir('/')
-        print("Removing {}".format(self.dir))
-        shutil.rmtree(self.dir)
-
-    def post_paging_test(self, iterator):
-        previous_post = None
-        for post in islice(iterator, PAGING_MAX_COUNT):
-            print(post)
-            if previous_post:
-                self.assertTrue(post.date_utc < previous_post.date_utc)
-            previous_post = post
-
-    @unittest.skip('')
-    def test_public_profile_download(self):
-        profiles = {self.L.check_profile_id(PUBLIC_PROFILE)}
-        self.L.download_profiles(profiles, profile_pic=False, fast_update=True, raise_errors=True)
-        self.L.download_profiles(profiles, profile_pic=False, fast_update=True, raise_errors=True)
-
-    def test_public_profile_paging(self):
-        self.post_paging_test(instaloader.Profile.from_username(self.L.context, PUBLIC_PROFILE).get_posts())
-
-    def test_profile_pic_download(self):
-        self.L.download_profiles({self.L.check_profile_id(PUBLIC_PROFILE)}, posts=False, raise_errors=True)
-
-    def test_hashtag_download(self):
-        self.L.download_hashtag(HASHTAG, NORMAL_MAX_COUNT)
-
-    def test_hashtag_paging(self):
-        for count, post in enumerate(instaloader.Hashtag.from_name(self.L.context, HASHTAG).get_all_posts()):
-            print(post)
-            if count == PAGING_MAX_COUNT:
-                break
-
-    def test_get_id_by_username(self):
-        self.assertEqual(PUBLIC_PROFILE_ID,
-                         instaloader.Profile.from_username(self.L.context, PUBLIC_PROFILE).userid)
-
-    def test_get_username_by_id_private(self):
-        self.assertEqual(PRIVATE_PROFILE.lower(),
-                         instaloader.Profile.from_id(self.L.context, PRIVATE_PROFILE_ID).username)
-
-    def test_get_username_by_id_public(self):
-        self.assertEqual(PUBLIC_PROFILE.lower(),
-                         instaloader.Profile.from_id(self.L.context, PUBLIC_PROFILE_ID).username)
-
-    def test_get_username_by_id_empty(self):
-        self.assertEqual(EMPTY_PROFILE.lower(),
-                         instaloader.Profile.from_id(self.L.context, EMPTY_PROFILE_ID).username)
-
-    def test_get_username_by_name_empty(self):
-        self.assertEqual(EMPTY_PROFILE_ID,
-                         instaloader.Profile.from_username(self.L.context, EMPTY_PROFILE).userid)
-
-    def test_post_from_mediaid(self):
-        for post in instaloader.Profile.from_username(self.L.context, PUBLIC_PROFILE).get_posts():
-            post2 = instaloader.Post.from_mediaid(self.L.context, post.mediaid)
-            self.assertEqual(post, post2)
-            break
-
-    def test_public_profile_tagged_paging(self):
-        for post in islice(instaloader.Profile.from_username(self.L.context, PUBLIC_PROFILE).get_tagged_posts(),
-                           PAGING_MAX_COUNT):
-            print(post)
-
-    def test_public_profile_igtv(self):
-        for post in islice(instaloader.Profile.from_username(self.L.context, PUBLIC_PROFILE_WITH_IGTV).get_igtv_posts(),
-                           PAGING_MAX_COUNT):
-            print(post)
-
-
-class TestInstaloaderLoggedIn(TestInstaloaderAnonymously):
-
-    def setUp(self):
-        super().setUp()
-        self.L.load_session_from_file(OWN_USERNAME)
-
-    @unittest.skip('')
-    def test_stories_download(self):
-        self.L.download_stories()
-
-    @unittest.skip('')
-    def test_private_profile_download(self):
-        self.L.download_profiles({self.L.check_profile_id(PRIVATE_PROFILE)}, stories=True, raise_errors=True)
-
-    def test_stories_paging(self):
-        for user_story in self.L.get_stories():
-            print("profile {}.".format(user_story.owner_username))
-            for item in user_story.get_items():
-                print(item)
-
-    def test_highlights_paging(self):
-        for user_highlight in self.L.get_highlights(PROFILE_WITH_HIGHLIGHTS):
-            print("Retrieving {} highlights \"{}\" from profile {}".format(user_highlight.itemcount,
-                                                                           user_highlight.title,
-                                                                           user_highlight.owner_username))
-            for item in user_highlight.get_items():
-                print(item)
-
-    def test_private_profile_paging(self):
-        self.post_paging_test(instaloader.Profile.from_username(self.L.context, PRIVATE_PROFILE).get_posts())
-
-    def test_feed_download(self):
-        self.L.download_feed_posts(NORMAL_MAX_COUNT)
-
-    def test_feed_paging(self):
-        for count, post in enumerate(self.L.get_feed_posts()):
-            print(post)
-            if count == PAGING_MAX_COUNT:
-                break
-
-    def test_saved_download(self):
-        self.L.download_saved_posts(NORMAL_MAX_COUNT)
-
-    def test_saved_paging(self):
-        for count, post in enumerate(instaloader.Profile.from_username(self.L.context, OWN_USERNAME).get_saved_posts()):
-            print(post)
-            if count == PAGING_MAX_COUNT:
-                break
-
-    def test_test_login(self):
-        self.assertEqual(OWN_USERNAME, self.L.test_login())
-
-    def test_followees_and_stories(self):
-        profile = instaloader.Profile.from_username(self.L.context, OWN_USERNAME)
-        followees = set(islice(profile.get_followees(), PAGING_MAX_COUNT))
-        self.L.download_profiles(followees, profile_pic=False, stories=True, posts=False, raise_errors=True)
-
-    def test_get_followees(self):
-        profile = instaloader.Profile.from_username(self.L.context, OWN_USERNAME)
-        for f in profile.get_followees():
-            print(f.username)
-
-    def test_get_followers(self):
-        profile = instaloader.Profile.from_username(self.L.context, OWN_USERNAME)
-        for f in profile.get_followers():
-            print(f.username)
-
-    def test_get_similar_accounts(self):
-        profile = instaloader.Profile.from_username(self.L.context, OWN_USERNAME)
-        for f in profile.get_similar_accounts():
-            print(f.username)
-
-    def test_get_likes(self):
-        for post in instaloader.Profile.from_username(self.L.context, OWN_USERNAME).get_posts():
-            for like in post.get_likes():
-                print(like.username)
-            break
-
-    def test_explore_paging(self):
-        for count, post in enumerate(self.L.get_explore_posts()):
-            print(post)
-            if count == PAGING_MAX_COUNT:
-                break
-
-    def test_location_download(self):
-        self.L.download_location(LOCATION, NORMAL_MAX_COUNT)
-
-    def test_location_paging(self):
-        for count, post in enumerate(self.L.get_location_posts(LOCATION)):
-            print(post)
-            if count == PAGING_MAX_COUNT:
-                break
-
-
-if __name__ == '__main__':
-    unittest.main()