@danhachuel/thunderbolt 0.6.59 → 0.6.61

package/integrations/instagram_public.py

@@ -185,23 +185,45 @@ def _metric_text_value(value: Any) -> int | None:
     return int(digits) * multiplier
 
 
+def _extract_bio_from_html(document: str) -> str:
+    """Extract only the JSON biography value; never use an SEO description as Bio."""
+    patterns = (
+        r'"biography"\s*:\s*"((?:\\.|[^"\\])*)"',
+        r"'biography'\s*:\s*'((?:\\.|[^'\\])*)'",
+    )
+    for pattern in patterns:
+        for match in re.finditer(pattern, document, flags=re.IGNORECASE | re.DOTALL):
+            raw_value = match.group(1)
+            try:
+                decoded = json.loads(f'"{raw_value}"')
+            except (TypeError, ValueError, json.JSONDecodeError):
+                decoded = bytes(raw_value, 'utf-8').decode('unicode_escape')
+            if str(decoded or '').strip():
+                return str(decoded).strip()
+    return ''
+
+
 def _embedded_profile_user(document: str, username: str) -> Mapping[str, Any]:
-    wanted = username.casefold()
+    wanted = str(username or '').strip().lstrip('@').casefold()
     for payload in _embedded_json_documents(document):
         for node in _walk_json(payload):
             if not isinstance(node, Mapping):
                 continue
-            node_username = str(node.get('username') or node.get('handle') or '').lstrip('@').casefold()
-            if node_username == wanted and any(key in node for key in ('biography', 'bio', 'edge_followed_by', 'edge_follow', 'followers', 'following')):
+            node_username = str(node.get('username') or node.get('handle') or '').strip().lstrip('@').casefold()
+            has_profile_fields = any(key in node for key in ('biography', 'bio', 'edge_followed_by', 'edge_follow', 'followers', 'following'))
+            if node_username == wanted and has_profile_fields:
                 return node
     return {}
 
 
 def _extract_profile_user_from_html(document: str, username: str) -> dict[str, Any]:
     embedded_user = _embedded_profile_user(document, username)
+    regex_biography = _extract_bio_from_html(document)
     seo_description = _meta(document, 'og:description')
-    biography = str(embedded_user.get('biography') or embedded_user.get('bio') or '').strip()
-    description = biography or seo_description
+    biography = str(embedded_user.get('biography') or embedded_user.get('bio') or regex_biography or '').strip()
+    print(f'embedded_user for @{username}: {embedded_user}')
+    print(f'biography extracted for @{username}: {biography!r}')
+    description = biography
     title = str(embedded_user.get('full_name') or embedded_user.get('name') or _meta(document, 'og:title') or username).split('(')[0].strip() or username
     avatar_url = str(embedded_user.get('profile_pic_url_hd') or embedded_user.get('profile_pic_url') or _meta(document, 'og:image')).strip()
 
@@ -214,9 +236,9 @@ def _extract_profile_user_from_html(document: str, username: str) -> dict[str, A
                     return parsed
         return None
 
-    followers = _metric_text_value(embedded_user.get('edge_followed_by')) or _metric_text_value(embedded_user.get('followers')) or metric((r'([\d.,]+)\s*(k|mil)?\s*(?:followers|seguidores)',)) or _structured_metric(document, 'edge_followed_by', 'followers', 'follower_count') or _structured_metric_from_json(document, 'edge_followed_by', 'followers', 'follower_count', 'followerCount')
-    following = _metric_text_value(embedded_user.get('edge_follow')) or _metric_text_value(embedded_user.get('following')) or metric((r'([\d.,]+)\s*(k|mil)?\s*(?:following|seguindo)',)) or _structured_metric(document, 'edge_follow', 'follows', 'following', 'following_count', 'followingCount') or _structured_metric_from_json(document, 'edge_follow', 'follows', 'following', 'following_count', 'followingCount')
-    post_count = _metric_text_value(embedded_user.get('edge_owner_to_timeline_media')) or _metric_text_value(embedded_user.get('posts')) or metric((r'([\d.,]+)\s*(k|mil)?\s*(?:posts|publicações|publications)',)) or _structured_metric(document, 'edge_owner_to_timeline_media', 'posts', 'post_count')
+    followers = _metric_text_value(embedded_user.get('edge_followed_by')) or _metric_text_value(embedded_user.get('followers')) or metric((r'([\d.,]+)\s*(k|mil)?\s*(?:followers|seguidores)', r'(?:followers|seguidores)\s*([\d.,]+)\s*(k|mil)?')) or _structured_metric(document, 'edge_followed_by', 'followers', 'follower_count') or _structured_metric_from_json(document, 'edge_followed_by', 'followers', 'follower_count', 'followerCount')
+    following = _metric_text_value(embedded_user.get('edge_follow')) or _metric_text_value(embedded_user.get('following')) or metric((r'([\d.,]+)\s*(k|mil)?\s*(?:following|seguindo)', r'(?:following|seguindo)\s*([\d.,]+)\s*(k|mil)?')) or _structured_metric(document, 'edge_follow', 'follows', 'following', 'following_count', 'followingCount') or _structured_metric_from_json(document, 'edge_follow', 'follows', 'following', 'following_count', 'followingCount')
+    post_count = _metric_text_value(embedded_user.get('edge_owner_to_timeline_media')) or _metric_text_value(embedded_user.get('posts')) or metric((r'([\d.,]+)\s*(k|mil)?\s*(?:posts|publicações|publications)', r'(?:posts|publicações|publications)\s*([\d.,]+)\s*(k|mil)?')) or _structured_metric(document, 'edge_owner_to_timeline_media', 'posts', 'post_count')
     country = _country_value(embedded_user.get('country') or embedded_user.get('country_code')) or extract_public_instagram_country(embedded_user)
     if not country:
         for payload in _embedded_json_documents(document):
@@ -227,7 +249,7 @@ def _extract_profile_user_from_html(document: str, username: str) -> dict[str, A
     user: dict[str, Any] = {
         'username': username,
         'full_name': title.split('(')[0].strip() or username,
-        'biography': description,
+        'biography': biography,
         'edge_followed_by': {'count': followers} if followers is not None else {},
         'edge_follow': {'count': following} if following is not None else {},
         'edge_owner_to_timeline_media': {'count': post_count, 'edges': [{'node': post} for post in posts]},
@@ -235,7 +257,7 @@ def _extract_profile_user_from_html(document: str, username: str) -> dict[str, A
         'profile_pic_url': avatar_url,
         '_html_posts': posts,
     }
-    return user if any((description, followers, following, post_count, country, posts)) else {}
+    return user if any((biography, followers, following, post_count, country, posts)) else {}
 
 
 def _fetch_web_profile_user(username: str) -> dict[str, Any] | None:
@@ -261,6 +283,15 @@ def _fetch_web_profile_user(username: str) -> dict[str, Any] | None:
                     merged[key] = value
         return merged
 
+    def profile_is_complete(user: Mapping[str, Any]) -> bool:
+        if profile_quality(user) < 4:
+            return False
+        if platform.system() != 'Windows':
+            return True
+        has_bio = bool(str(user.get('biography') or user.get('bio') or '').strip())
+        has_following = any(user.get(key) not in (None, '', {}, []) for key in ('edge_follow', 'follows', 'following', 'following_count', 'followingCount'))
+        return has_bio and has_following
+
     best_user: dict[str, Any] | None = None
     for endpoint in endpoints:
         try:
@@ -273,7 +304,7 @@ def _fetch_web_profile_user(username: str) -> dict[str, Any] | None:
         user = ((payload.get('data') or {}).get('user') if isinstance(payload, dict) else None)
         if isinstance(user, dict):
             best_user = merge_profile(best_user, user)
-            if profile_quality(best_user) >= 4:
+            if profile_is_complete(best_user):
                 return best_user
 
     curl = shutil.which('curl')
@@ -453,6 +484,8 @@ def fetch_public_instagram_profile(source: str) -> IntegrationResult:
     api_user = _fetch_web_profile_user(reference['username'])
     if api_user:
         return IntegrationResult(True, 'Perfil Instagram encontrado publicamente.', _profile_data_from_api(api_user, reference))
+    if platform.system() == 'Windows':
+        return IntegrationResult(False, 'Não foi possível encontrar o perfil público estruturado do Instagram.', reference)
     try:
         response = requests.get(reference['url'], headers={'User-Agent': 'Mozilla/5.0', 'Accept-Language': 'pt-BR,pt;q=0.9,en;q=0.8'}, timeout=12)
     except requests.RequestException as exc:

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@danhachuel/thunderbolt",
-  "version": "0.6.59",
+  "version": "0.6.61",
   "description": "Thunderbolt — interface local para operação de canais faceless e motor MoneyPrinterTurbo",
   "license": "MIT",
   "main": "scripts/cli.mjs",