improving insta archiver

src/auto_archiver/archivers/instagram_api_archiver.py

@@ -73,9 +73,9 @@ class InstagramAPIArchiver(Archiver):
         if type(d) == list: return [self.cleanup_dict(v) for v in d]
         if type(d) != dict: return d
         return {
-                k: self.cleanup_dict(v) if type(v) in [dict, list] else v 
+                k: clean_v
                 for k, v in d.items() 
-                if v not in [0.0, 0, [], {}, "", None, "null"] and
+                if (clean_v := self.cleanup_dict(v)) not in [0.0, 0, [], {}, "", None, "null"] and
                 k not in ["x", "y", "width", "height"]
         }
 
@@ -93,9 +93,6 @@ class InstagramAPIArchiver(Archiver):
 
         if self.full_profile:
             user_id = user.get("pk")
-            # download all posts
-            self.download_all_posts(result, user_id)
-
             # download all stories
             try:
                 stories = self._download_stories_reusable(result, username)
@@ -104,6 +101,20 @@ class InstagramAPIArchiver(Archiver):
                 result.append("errors", f"Error downloading stories for {username}")
                 logger.error(f"Error downloading stories for {username}: {e}")
 
+            # download all posts
+            try:
+                self.download_all_posts(result, user_id)
+            except Exception as e:
+                result.append("errors", f"Error downloading posts for {username}")
+                logger.error(f"Error downloading posts for {username}: {e}")
+
+            # download all tagged
+            try:
+                self.download_all_tagged(result, user_id)
+            except Exception as e:
+                result.append("errors", f"Error downloading tagged posts for {username}")
+                logger.error(f"Error downloading tagged posts for {username}: {e}")
+
             # download all highlights
             try:
                 count_highlights = 0
@@ -120,6 +131,7 @@ class InstagramAPIArchiver(Archiver):
                 result.append("errors", f"Error downloading highlights for {username}")
                 logger.error(f"Error downloading highlights for {username}: {e}")
 
+
         result.set_url(url) # reset as scrape_item modifies it
         return result.success("insta profile")
 
@@ -200,6 +212,28 @@ class InstagramAPIArchiver(Archiver):
                 pbar.update(1)
                 post_count+=1
         result.set("#posts", post_count)
+        
+    def download_all_tagged(self, result: Metadata, user_id: str):
+        next_page_id = ""
+        pbar = tqdm(desc="downloading tagged posts")
+
+        tagged_count = 0
+        while next_page_id != None:
+            resp = self.call_api(f"v2/user/tag/medias", {"user_id": user_id, "page_id": next_page_id})
+            posts = resp.get("response", {}).get("items", [])
+            if not len(posts): break
+            next_page_id = resp.get("next_page_id")
+            
+            logger.info(f"parsing {len(posts)} tagged posts, next {next_page_id=}")
+
+            for p in posts:
+                try: self.scrape_item(result, p, "tagged")
+                except Exception as e:
+                    result.append("errors", f"Error downloading tagged post {p.get('id')}")
+                    logger.error(f"Error downloading tagged post, skipping {p.get('id')}: {e}")
+                pbar.update(1)
+                tagged_count+=1
+        result.set("#tagged", tagged_count)
 
 
 ### reusable parsing utils below
@@ -217,10 +251,10 @@ class InstagramAPIArchiver(Archiver):
             if self.minimize_json_output: 
                 del item["clips_metadata"]
 
-        if code := item.get("code"): 
-            result.set("url", f"https://www.instagram.com/p/{code}/")
+        if code := item.get("code") and not result.get("url"): 
+            result.set_url(f"https://www.instagram.com/p/{code}/")
             
-        resources = item.get("resources", [])
+        resources = item.get("resources", item.get("carousel_media", []))
         item, media, media_id = self.scrape_media(item, context)
         # if resources are present take the main media from the first resource
         if not media and len(resources):
@@ -242,7 +276,7 @@ class InstagramAPIArchiver(Archiver):
     def scrape_media(self, item: dict, context:str) -> tuple[dict, Media, str]:
         # remove unnecessary info
         if self.minimize_json_output: 
-            for k in ["image_versions", "video_versions", "video_dash_manifest"]:
+            for k in ["image_versions", "video_versions", "video_dash_manifest", "image_versions2", "video_versions2"]:
                 if k in item: del item[k]
         item = self.cleanup_dict(item)
 
@@ -253,19 +287,24 @@ class InstagramAPIArchiver(Archiver):
             
         # retrieve video info
         best_id = item.get('id', item.get('pk'))
-        taken_at = item.get("taken_at")
+        taken_at = item.get("taken_at", item.get("taken_at_ts"))
         code = item.get("code")
+        caption_text = item.get("caption_text")
+        if "carousel_media" in item: del item["carousel_media"]
+
         if video_url := item.get("video_url"):
             filename = self.download_from_url(video_url, verbose=False)
             video_media = Media(filename=filename)
             if taken_at: video_media.set("date", taken_at)
             if code: video_media.set("url", f"https://www.instagram.com/p/{code}")
+            if caption_text: video_media.set("text", caption_text)
             video_media.set("preview", [image_media])
             video_media.set("data", [item])
             return item, video_media, f"{context or 'video'} {best_id}"
         elif image_media:
             if taken_at: image_media.set("date", taken_at)
             if code: image_media.set("url", f"https://www.instagram.com/p/{code}")
+            if caption_text: image_media.set("text", caption_text)
             image_media.set("data", [item])
             return item, image_media, f"{context or 'image'} {best_id}"
         

src/auto_archiver/archivers/youtubedl_archiver.py

@@ -39,7 +39,7 @@ class YoutubeDLArchiver(Archiver):
         ydl = yt_dlp.YoutubeDL(ydl_options) # allsubtitles and subtitleslangs not working as expected, so default lang is always "en"
 
         try:
-            # don'd download since it can be a live stream
+            # don't download since it can be a live stream
             info = ydl.extract_info(url, download=False)
             if info.get('is_live', False) and not self.livestreams:
                 logger.warning("Livestream detected, skipping due to 'livestreams' configuration setting")
@@ -64,13 +64,17 @@ class YoutubeDLArchiver(Archiver):
 
         result = Metadata()
         result.set_title(info.get("title"))
+        if "description" in info: result.set_content(info["description"])
         for entry in entries:
             try:
                 filename = ydl.prepare_filename(entry)
                 if not os.path.exists(filename):
                     filename = filename.split('.')[0] + '.mkv'
-                new_media = Media(filename).set("duration", info.get("duration"))
-                
+
+                new_media = Media(filename)
+                for x in ["duration", "original_url", "fulltitle", "description", "upload_date"]:
+                    if x in entry: new_media.set(x, entry[x])
+
                 # read text from subtitles if enabled
                 if self.subtitles:
                     for lang, val in (info.get('requested_subtitles') or {}).items():

src/auto_archiver/enrichers/wacz_enricher.py

@@ -38,20 +38,19 @@ class WaczArchiverEnricher(Enricher, Archiver):
     
     def setup(self) -> None:
         self.use_docker = os.environ.get('WACZ_ENABLE_DOCKER') or not os.environ.get('RUNNING_IN_DOCKER')
+        self.docker_in_docker = os.environ.get('WACZ_ENABLE_DOCKER') and os.environ.get('RUNNING_IN_DOCKER')
+
+        self.cwd_dind = f"/crawls/crawls{random_str(8)}"
         self.browsertrix_home_host = os.environ.get('BROWSERTRIX_HOME_HOST')
         self.browsertrix_home_container = os.environ.get('BROWSERTRIX_HOME_CONTAINER') or self.browsertrix_home_host
         # create crawls folder if not exists, so it can be safely removed in cleanup
-        if self.use_docker:
-            if self.browsertrix_home_container:
-                os.makedirs(self.browsertrix_home_container, exist_ok=True)
+        if self.docker_in_docker:
+            os.makedirs(self.cwd_dind, exist_ok=True)
 
     def cleanup(self) -> None:
-        if self.use_docker:
-            if self.browsertrix_home_container:
-                logger.debug(f"Removing {self.browsertrix_home_container=}")
-                shutil.rmtree(self.browsertrix_home_container, ignore_errors=True)
-
-
+        if self.docker_in_docker:
+            logger.debug(f"Removing {self.cwd_dind=}")
+            shutil.rmtree(self.cwd_dind, ignore_errors=True)
 
     def download(self, item: Metadata) -> Metadata:
         # this new Metadata object is required to avoid duplication
@@ -84,6 +83,9 @@ class WaczArchiverEnricher(Enricher, Archiver):
             "--behaviors", "autoscroll,autoplay,autofetch,siteSpecific",
             "--behaviorTimeout", str(self.timeout),
             "--timeout", str(self.timeout)]
+        
+        if self.docker_in_docker:
+            cmd.extend(["--cwd", self.cwd_dind])
 
         # call docker if explicitly enabled or we are running on the host (not in docker)
         if self.use_docker:
@@ -118,7 +120,10 @@ class WaczArchiverEnricher(Enricher, Archiver):
             logger.error(f"WACZ generation failed: {e}")
             return False
 
-        if self.use_docker:
+        
+        if self.docker_in_docker:
+            wacz_fn = os.path.join(self.cwd_dind, "collections", collection, f"{collection}.wacz")
+        elif self.use_docker:
             wacz_fn = os.path.join(browsertrix_home_container, "collections", collection, f"{collection}.wacz")
         else:
             wacz_fn = os.path.join("collections", collection, f"{collection}.wacz")
@@ -131,7 +136,9 @@ class WaczArchiverEnricher(Enricher, Archiver):
         if self.extract_media or self.extract_screenshot:
             self.extract_media_from_wacz(to_enrich, wacz_fn)
 
-        if self.use_docker:
+        if self.docker_in_docker:
+            jsonl_fn = os.path.join(self.cwd_dind, "collections", collection, "pages", "pages.jsonl")
+        elif self.use_docker:
             jsonl_fn = os.path.join(browsertrix_home_container, "collections", collection, "pages", "pages.jsonl")
         else:
             jsonl_fn = os.path.join("collections", collection, "pages", "pages.jsonl")

src/auto_archiver/version.py

@@ -3,7 +3,7 @@ _MAJOR = "0"
 _MINOR = "9"
 # On main and in a nightly release the patch should be one ahead of the last
 # released build.
-_PATCH = "4"
+_PATCH = "9"
 # This is mainly for nightly builds which have the suffix ".dev$DATE". See
 # https://semver.org/#is-v123-a-semantic-version for the semantics.
 _SUFFIX = ""