graph-crawler 5.4.2__tar.gz → 5.4.4__tar.gz

graph_crawler-5.4.4/PKG-INFO

@@ -0,0 +1,130 @@
+Metadata-Version: 2.4
+Name: graph-crawler
+Version: 5.4.4
+Summary: Web crawler that builds a graph of a site. HTTP → TLS → browser escalation with pluggable browser engines.
+Author: 0-EternalJunior-0
+Maintainer: 0-EternalJunior-0
+License-Expression: MIT
+Project-URL: Homepage, https://gitlab.com/demoprogrammer/web_graf
+Project-URL: Documentation, https://gitlab.com/demoprogrammer/web_graf/-/blob/main/web_graf/README.md
+Project-URL: Repository, https://gitlab.com/demoprogrammer/web_graf
+Project-URL: Changelog, https://gitlab.com/demoprogrammer/web_graf/-/blob/main/web_graf/CHANGELOG.md
+Project-URL: Bug Tracker, https://gitlab.com/demoprogrammer/web_graf/-/issues
+Keywords: web,crawler,scraper,graph,spider,playwright,anti-bot,tls-fingerprint
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Topic :: Internet :: WWW/HTTP :: Indexing/Search
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: aiohttp>=3.13.0
+Requires-Dist: backports.zstd>=1.6.0; python_version < "3.14"
+Requires-Dist: Brotli
+Requires-Dist: selectolax<1.0.0,>=0.4.0
+Requires-Dist: defusedxml>=0.7.0
+Requires-Dist: pydantic>=2.12.0
+Requires-Dist: orjson>=3.11.0
+Requires-Dist: aiofiles>=24.1.0
+Requires-Dist: aiosqlite>=0.20.0
+Requires-Dist: typing_extensions>=4.5; python_version < "3.12"
+Requires-Dist: xxhash>=3.0.0
+Requires-Dist: resilix<2,>=1.1
+Provides-Extra: native
+Requires-Dist: cython>=3.0.0; extra == "native"
+Requires-Dist: mmh3>=5.0.0; extra == "native"
+Provides-Extra: browser
+Requires-Dist: playwright>=1.58.0; extra == "browser"
+Provides-Extra: playwright
+Requires-Dist: graph-crawler[browser]; extra == "playwright"
+Provides-Extra: tls
+Requires-Dist: curl_cffi>=0.7; extra == "tls"
+Provides-Extra: stealth
+Requires-Dist: patchright>=1.52; extra == "stealth"
+Requires-Dist: playwright-stealth>=1.0; extra == "stealth"
+Provides-Extra: cloak
+Requires-Dist: cloakbrowser==0.6.3; extra == "cloak"
+Requires-Dist: playwright>=1.58.0; extra == "cloak"
+Provides-Extra: undetected
+Requires-Dist: nodriver>=0.38; extra == "undetected"
+Provides-Extra: antidetect
+Requires-Dist: graph-crawler[browser,cloak,stealth,tls,undetected]; extra == "antidetect"
+Provides-Extra: server
+Requires-Dist: fastapi>=0.135.0; extra == "server"
+Requires-Dist: uvicorn>=0.30; extra == "server"
+Requires-Dist: websockets>=13; extra == "server"
+Provides-Extra: text
+Requires-Dist: tiktoken>=0.13.0; extra == "text"
+Requires-Dist: stop-words>=2018.7.23; extra == "text"
+Provides-Extra: captcha
+Provides-Extra: mongodb
+Requires-Dist: pymongo[srv]>=4.9.0; extra == "mongodb"
+Provides-Extra: postgresql
+Requires-Dist: asyncpg>=0.30.0; extra == "postgresql"
+Provides-Extra: newspaper
+Requires-Dist: newspaper3k>=0.2.8; extra == "newspaper"
+Provides-Extra: goose
+Requires-Dist: goose3>=3.1.0; extra == "goose"
+Provides-Extra: readability
+Requires-Dist: readability-lxml>=0.8.0; extra == "readability"
+Provides-Extra: markdown
+Requires-Dist: html2md-clean>=0.5.0; extra == "markdown"
+Requires-Dist: markdown-it-py>=4.0.0; extra == "markdown"
+Provides-Extra: sitemap-fast
+Requires-Dist: lxml>=5.1.0; extra == "sitemap-fast"
+Provides-Extra: articles
+Requires-Dist: newspaper3k>=0.2.8; extra == "articles"
+Requires-Dist: goose3>=3.1.0; extra == "articles"
+Requires-Dist: readability-lxml>=0.8.0; extra == "articles"
+Provides-Extra: viz
+Requires-Dist: pyvis>=0.3.0; extra == "viz"
+Requires-Dist: networkx>=3.6; extra == "viz"
+Provides-Extra: celery
+Requires-Dist: celery>=5.5.0; extra == "celery"
+Requires-Dist: redis>=5.2.0; extra == "celery"
+Provides-Extra: performance
+Requires-Dist: aiodns>=3.6.0; extra == "performance"
+Requires-Dist: uvloop>=0.22.0; platform_system != "Windows" and extra == "performance"
+Provides-Extra: fuzzy
+Requires-Dist: rapidfuzz>=3.6.0; extra == "fuzzy"
+Provides-Extra: dev
+Requires-Dist: pytest>=9.0.0; extra == "dev"
+Requires-Dist: pytest-asyncio>=0.24.0; extra == "dev"
+Requires-Dist: pytest-cov>=7.0.0; extra == "dev"
+Requires-Dist: black>=24.0.0; extra == "dev"
+Requires-Dist: ruff>=0.8.0; extra == "dev"
+Requires-Dist: mypy>=1.13.0; extra == "dev"
+Requires-Dist: psutil; extra == "dev"
+Requires-Dist: rapidfuzz>=3.6.0; extra == "dev"
+Requires-Dist: pytest-timeout>=2.3.0; extra == "dev"
+Requires-Dist: pytest-xdist>=3.5.0; extra == "dev"
+Requires-Dist: build>=1.2.0; extra == "dev"
+Requires-Dist: twine>=5.0.0; extra == "dev"
+Requires-Dist: playwright>=1.58.0; extra == "dev"
+Requires-Dist: curl_cffi>=0.7; extra == "dev"
+Requires-Dist: html2md-clean>=0.5.0; extra == "dev"
+Requires-Dist: markdown-it-py>=4.0.0; extra == "dev"
+Requires-Dist: tiktoken>=0.13.0; extra == "dev"
+Requires-Dist: stop-words>=2018.7.23; extra == "dev"
+Requires-Dist: fastapi>=0.135.0; extra == "dev"
+Requires-Dist: httpx>=0.28; extra == "dev"
+Provides-Extra: system
+Requires-Dist: psutil>=5.9.0; extra == "system"
+Provides-Extra: all
+Requires-Dist: graph-crawler[antidetect,articles,celery,fuzzy,markdown,mongodb,performance,postgresql,server,sitemap-fast,system,text,viz]; extra == "all"
+Dynamic: license-file
+
+# graph-crawler
+
+Веб-краулер, що будує граф сайту (сторінки — ноди, посилання — ребра) з ескалацією HTTP → TLS-fingerprint → браузер.
+
+```bash
+pip install graph-crawler            # лише HTTP
+pip install "graph-crawler[browser]"  # + Playwright
+```

graph_crawler-5.4.4/README.md

@@ -0,0 +1,8 @@
+# graph-crawler
+
+Веб-краулер, що будує граф сайту (сторінки — ноди, посилання — ребра) з ескалацією HTTP → TLS-fingerprint → браузер.
+
+```bash
+pip install graph-crawler            # лише HTTP
+pip install "graph-crawler[browser]"  # + Playwright
+```

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/data/backends/memory.py

@@ -182,6 +182,8 @@ class MemoryBackend:
         async with self._lock:
             count = 0
             for node in nodes:
+                if getattr(node, "_removed_from_graph", False):
+                    continue
                 normalized_url = self._normalize_url(node.url)
                 node.url = normalized_url
                 new_scanned = bool(getattr(node, "scanned", False))

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/data/backends/sqlite.py

@@ -235,6 +235,11 @@ class SQLiteBackend:
         """
         self.db_path = db_path
         self.journal_mode = journal_mode
+        # Скільки відкладених insert-ів пропущено через tombstone (діагностика).
+        self.tombstone_skips = 0
+        # URL нод, видалених з графа (remove_node): відкладений insert_nodes_batch
+        # їх не воскрешає; явний insert_node/insert_node_sync tombstone знімає.
+        self._tombstoned_urls: set[str] = set()
         self._conn = None
         self._is_open = False
 
@@ -895,6 +900,7 @@ class SQLiteBackend:
         # Normalize URL
         original_url = node.url
         normalized_url = self._normalize_url(original_url)
+        self._tombstoned_urls.discard(normalized_url)
 
         if original_url != normalized_url:
             node.url = normalized_url
@@ -1002,7 +1008,16 @@ class SQLiteBackend:
             for node in nodes:
                 from graph_crawler.shared.utils.id_generator import generate_node_id
 
+                # Tombstone: нода вже видалена з графа (редірект на існуючу) —
+                # відкладений write-through не повинен її воскресити.
+                if getattr(node, "_removed_from_graph", False):
+                    self.tombstone_skips += 1
+                    continue
+
                 normalized_url = self._normalize_url(node.url)
+                if normalized_url in self._tombstoned_urls:
+                    self.tombstone_skips += 1
+                    continue
                 node.url = normalized_url
 
                 # BUGFIX /: пере-обчислити node_id після
@@ -1241,6 +1256,7 @@ class SQLiteBackend:
             else:
                 url = node.url
 
+            self._tombstoned_urls.add(self._normalize_url(url))
             # Delete from DB
             await self._conn.execute("DELETE FROM nodes WHERE node_id = ?", (node_id,))
             await self._conn.execute(
@@ -1647,6 +1663,7 @@ class SQLiteBackend:
         from graph_crawler.shared.utils.id_generator import generate_node_id
 
         normalized_url = self._normalize_url(node.url)
+        self._tombstoned_urls.discard(normalized_url)
         cached = self._cache_get_by_url(normalized_url)
         if cached is not None:
             return cached
@@ -1690,6 +1707,7 @@ class SQLiteBackend:
             if not rows:
                 return False
             url = rows[0][0]
+        self._tombstoned_urls.add(self._normalize_url(url))
         deleted = self._sync_execute(
             [
                 ("DELETE FROM nodes WHERE node_id = ?", (node_id,)),

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/domain/entities/graph.py

@@ -124,6 +124,21 @@ from graph_crawler.shared.exceptions import GraphIntegrityError
 logger = logging.getLogger(__name__)
 
 
+
+def _set_removed_flag(node: Any, value: bool) -> None:
+    try:
+        node._removed_from_graph = value
+    except (AttributeError, ValueError, TypeError):
+        pass
+
+
+def is_removed_from_graph(node: Any) -> bool:
+    """Чи нода видалена з графа (tombstone від ``Graph.remove_node``)."""
+    try:
+        return bool(getattr(node, "_removed_from_graph", False))
+    except Exception:  # noqa: BLE001
+        return False
+
 class Graph:
     """Менеджер графу: CRUD над вузлами/ребрами, URL-дедуплікація, evict у RAM/на диск.
 
@@ -759,6 +774,8 @@ class Graph:
             self._stale_node_id_map[node.node_id] = expected_id
             node.node_id = expected_id
 
+        # Явне повторне додавання знімає tombstone.
+        _set_removed_flag(node, False)
         return node
 
     def add_node(self, node: Node, overwrite: bool = False) -> Node:
@@ -1339,6 +1356,10 @@ class Graph:
         if not node:
             return False
 
+        # Tombstone до видалення з backend: drain write-through, що тримає посилання
+        # на цей об'єкт, після DELETE не має зробити INSERT (зомбі-нода).
+        _set_removed_flag(node, True)
+
         # BUGFIX: у backend-режимі backend — джерело правди для
         # edges/adjacency. Делегуємо йому видалення (він почистить БД + власний
         # кеш edges + adjacency), інакше лишаються "мертві" edges у backend →

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/domain/entities/node.py

@@ -358,6 +358,8 @@ class Node(BaseModel):
     _lifecycle_lock: threading.Lock = PrivateAttr(default_factory=threading.Lock)
     # Лінки з ON_HTML_PARSED поточного enrich_node: переживають його переривання (per-node timeout).
     _discovered_links: list[str] = PrivateAttr(default_factory=list)
+    # Tombstone: Graph.remove_node видалив ноду — відкладений write-through не має її воскресити.
+    _removed_from_graph: bool = PrivateAttr(default=False)
 
     # Контент і обхід: хеші рахує enrich_node (дедуплікація, near-duplicates);
     # priority — з URL-правил для черги; lifecycle_stage змінює лише transition_state.

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/infrastructure/transport/playwright/browser_profile.py

@@ -1,6 +1,6 @@
 """BrowserProfile — єдине джерело істини для fingerprint.
 
-ANTIBOT_AUDIT_pooled_driver.md §1.6/§1.7/§2: раніше UA, platform, screen,
+UA, platform, screen,
 viewport, WebGL, hardwareConcurrency, languages, locale/timezone задавались у
 різних місцях (constants, StealthPlugin, EnhancedStealthPlugin, UserProfilePlugin)
 і між собою НЕ узгоджувались — це головна причина детекту. Тепер усе виводиться

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/infrastructure/transport/playwright/net/__init__.py

@@ -1,6 +1,6 @@
 """Мережевий шар для Playwright-гілки.
 
-ANTIBOT_AUDIT_pooled_driver.md §4.4: proxy не підтримується в
+proxy не підтримується в
 `PooledPlaywrightDriver` жодним чином → увесь флот сидить на одному IP,
 а locale/timezone/Accept-Language не пов'язані з GeoIP.
 

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/infrastructure/transport/playwright/net/geoip.py

@@ -1,6 +1,5 @@
 """GeoResolver — виведення locale/timezone з проксі-запису (Крок 3).
 
-ANTIBOT_AUDIT_pooled_driver.md §2.6:
 IP з Німеччини + `timeZone: "UTC"` + `Accept-Language: en-US` +
 Windows-UA на Linux — чотири незалежні суперечності в ОДНОМУ запиті.
 Для Cloudflare/DataDome достатньо challenge без JS-аналізу.

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/infrastructure/transport/playwright/net/proxy_pool.py

@@ -1,6 +1,5 @@
 """ProxyPool — sticky-прив'язка «домен ↔ proxy».
 
-ANTIBOT_AUDIT_pooled_driver.md §4.4 / §5.4:
 Правильна архітектура:
 
     ProxyPool → обраний proxy → GeoIP → BrowserProfile → new_context()

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/infrastructure/transport/playwright/net/rate_limiter.py

@@ -1,6 +1,5 @@
 """DomainRateLimiter — per-domain concurrency + мінімальний інтервал (Крок 3).
 
-ANTIBOT_AUDIT_pooled_driver.md §3.2:
 Наявний `calculate_optimal_jitter` збільшує затримку між стартами,
 але **не обмежує concurrency per-domain**: 25 URL одного домену
 підуть у 25 паралельних вкладок з одного IP → 429/challenge раніше,

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/infrastructure/transport/playwright/plugins/human_bezier.py

@@ -1,6 +1,5 @@
 """BezierHumanBehaviorPlugin.
 
-ANTIBOT_AUDIT_pooled_driver.md §3.3:
 Існуючий `HumanBehaviorPlugin` виконує рухи миші прямолінійно
 (steps=3-5), а `_scroll_page()` у драйвері викликає
 `page.evaluate("window.scrollTo(...)")` — але події, ініційовані з

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/infrastructure/transport/playwright/pooled_driver.py

@@ -570,7 +570,7 @@ class PooledPlaywrightDriver(PageFetcher, BrowserPool, BatchScheduler, BaseDrive
     ) -> list[FetchResponse]:
         """Async паралельне завантаження багатьох сторінок.
 
-        Аудит §1.9 (ANTIBOT_AUDIT_pooled_driver.md): раніше ці параметри
+        Ці параметри
         мовчки ігнорувались — тепер `referer` передається у `page.goto`
         per-URL, а `cookies` / `extra_headers` мержаться і застосовуються
         на рівні контексту (Playwright не дає per-URL headers/cookies при

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler/shared/constants.py

@@ -660,7 +660,6 @@ DEFAULT_POST_NAVIGATION_WAIT_NETWORKIDLE_MS = 5000
 """
 
 # Playwright Stealth Mode Arguments
-# Аудит ANTIBOT_AUDIT_pooled_driver.md §1.3 / §1.14 / Крок 0:
 # прибрано `--disable-web-security` та `--disable-features=IsolateOrigins,site-per-process`
 # (вимкнення Site Isolation детектується як anti-stealth сигнал у реальному Chrome
 # такої конфігурації не буває) та `--disable-background-networking`
@@ -747,7 +746,6 @@ F15 (§11.2): 600 → 500 за реальним заміром — PSS дере
 Перевизначається env ``GRAPH_CRAWLER_BROWSER_RAM_BASELINE_MB``."""
 
 # Дефолтні ресурси для блокування (економія трафіку та RAM)
-# Аудит ANTIBOT_AUDIT_pooled_driver.md §2.5 / §3.5 / Крок 0:
 # прибрано "font" і "stylesheet" — жоден реальний браузер не вантажить HTML
 # без CSS/шрифтів; порожній network waterfall (нуль .css/.woff2) + зламаний
 # layout під getBoundingClientRect() = поведінковий маркер рівня «скрапер».

graph_crawler-5.4.4/graph_crawler.egg-info/PKG-INFO

@@ -0,0 +1,130 @@
+Metadata-Version: 2.4
+Name: graph-crawler
+Version: 5.4.4
+Summary: Web crawler that builds a graph of a site. HTTP → TLS → browser escalation with pluggable browser engines.
+Author: 0-EternalJunior-0
+Maintainer: 0-EternalJunior-0
+License-Expression: MIT
+Project-URL: Homepage, https://gitlab.com/demoprogrammer/web_graf
+Project-URL: Documentation, https://gitlab.com/demoprogrammer/web_graf/-/blob/main/web_graf/README.md
+Project-URL: Repository, https://gitlab.com/demoprogrammer/web_graf
+Project-URL: Changelog, https://gitlab.com/demoprogrammer/web_graf/-/blob/main/web_graf/CHANGELOG.md
+Project-URL: Bug Tracker, https://gitlab.com/demoprogrammer/web_graf/-/issues
+Keywords: web,crawler,scraper,graph,spider,playwright,anti-bot,tls-fingerprint
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Topic :: Internet :: WWW/HTTP :: Indexing/Search
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: aiohttp>=3.13.0
+Requires-Dist: backports.zstd>=1.6.0; python_version < "3.14"
+Requires-Dist: Brotli
+Requires-Dist: selectolax<1.0.0,>=0.4.0
+Requires-Dist: defusedxml>=0.7.0
+Requires-Dist: pydantic>=2.12.0
+Requires-Dist: orjson>=3.11.0
+Requires-Dist: aiofiles>=24.1.0
+Requires-Dist: aiosqlite>=0.20.0
+Requires-Dist: typing_extensions>=4.5; python_version < "3.12"
+Requires-Dist: xxhash>=3.0.0
+Requires-Dist: resilix<2,>=1.1
+Provides-Extra: native
+Requires-Dist: cython>=3.0.0; extra == "native"
+Requires-Dist: mmh3>=5.0.0; extra == "native"
+Provides-Extra: browser
+Requires-Dist: playwright>=1.58.0; extra == "browser"
+Provides-Extra: playwright
+Requires-Dist: graph-crawler[browser]; extra == "playwright"
+Provides-Extra: tls
+Requires-Dist: curl_cffi>=0.7; extra == "tls"
+Provides-Extra: stealth
+Requires-Dist: patchright>=1.52; extra == "stealth"
+Requires-Dist: playwright-stealth>=1.0; extra == "stealth"
+Provides-Extra: cloak
+Requires-Dist: cloakbrowser==0.6.3; extra == "cloak"
+Requires-Dist: playwright>=1.58.0; extra == "cloak"
+Provides-Extra: undetected
+Requires-Dist: nodriver>=0.38; extra == "undetected"
+Provides-Extra: antidetect
+Requires-Dist: graph-crawler[browser,cloak,stealth,tls,undetected]; extra == "antidetect"
+Provides-Extra: server
+Requires-Dist: fastapi>=0.135.0; extra == "server"
+Requires-Dist: uvicorn>=0.30; extra == "server"
+Requires-Dist: websockets>=13; extra == "server"
+Provides-Extra: text
+Requires-Dist: tiktoken>=0.13.0; extra == "text"
+Requires-Dist: stop-words>=2018.7.23; extra == "text"
+Provides-Extra: captcha
+Provides-Extra: mongodb
+Requires-Dist: pymongo[srv]>=4.9.0; extra == "mongodb"
+Provides-Extra: postgresql
+Requires-Dist: asyncpg>=0.30.0; extra == "postgresql"
+Provides-Extra: newspaper
+Requires-Dist: newspaper3k>=0.2.8; extra == "newspaper"
+Provides-Extra: goose
+Requires-Dist: goose3>=3.1.0; extra == "goose"
+Provides-Extra: readability
+Requires-Dist: readability-lxml>=0.8.0; extra == "readability"
+Provides-Extra: markdown
+Requires-Dist: html2md-clean>=0.5.0; extra == "markdown"
+Requires-Dist: markdown-it-py>=4.0.0; extra == "markdown"
+Provides-Extra: sitemap-fast
+Requires-Dist: lxml>=5.1.0; extra == "sitemap-fast"
+Provides-Extra: articles
+Requires-Dist: newspaper3k>=0.2.8; extra == "articles"
+Requires-Dist: goose3>=3.1.0; extra == "articles"
+Requires-Dist: readability-lxml>=0.8.0; extra == "articles"
+Provides-Extra: viz
+Requires-Dist: pyvis>=0.3.0; extra == "viz"
+Requires-Dist: networkx>=3.6; extra == "viz"
+Provides-Extra: celery
+Requires-Dist: celery>=5.5.0; extra == "celery"
+Requires-Dist: redis>=5.2.0; extra == "celery"
+Provides-Extra: performance
+Requires-Dist: aiodns>=3.6.0; extra == "performance"
+Requires-Dist: uvloop>=0.22.0; platform_system != "Windows" and extra == "performance"
+Provides-Extra: fuzzy
+Requires-Dist: rapidfuzz>=3.6.0; extra == "fuzzy"
+Provides-Extra: dev
+Requires-Dist: pytest>=9.0.0; extra == "dev"
+Requires-Dist: pytest-asyncio>=0.24.0; extra == "dev"
+Requires-Dist: pytest-cov>=7.0.0; extra == "dev"
+Requires-Dist: black>=24.0.0; extra == "dev"
+Requires-Dist: ruff>=0.8.0; extra == "dev"
+Requires-Dist: mypy>=1.13.0; extra == "dev"
+Requires-Dist: psutil; extra == "dev"
+Requires-Dist: rapidfuzz>=3.6.0; extra == "dev"
+Requires-Dist: pytest-timeout>=2.3.0; extra == "dev"
+Requires-Dist: pytest-xdist>=3.5.0; extra == "dev"
+Requires-Dist: build>=1.2.0; extra == "dev"
+Requires-Dist: twine>=5.0.0; extra == "dev"
+Requires-Dist: playwright>=1.58.0; extra == "dev"
+Requires-Dist: curl_cffi>=0.7; extra == "dev"
+Requires-Dist: html2md-clean>=0.5.0; extra == "dev"
+Requires-Dist: markdown-it-py>=4.0.0; extra == "dev"
+Requires-Dist: tiktoken>=0.13.0; extra == "dev"
+Requires-Dist: stop-words>=2018.7.23; extra == "dev"
+Requires-Dist: fastapi>=0.135.0; extra == "dev"
+Requires-Dist: httpx>=0.28; extra == "dev"
+Provides-Extra: system
+Requires-Dist: psutil>=5.9.0; extra == "system"
+Provides-Extra: all
+Requires-Dist: graph-crawler[antidetect,articles,celery,fuzzy,markdown,mongodb,performance,postgresql,server,sitemap-fast,system,text,viz]; extra == "all"
+Dynamic: license-file
+
+# graph-crawler
+
+Веб-краулер, що будує граф сайту (сторінки — ноди, посилання — ребра) з ескалацією HTTP → TLS-fingerprint → браузер.
+
+```bash
+pip install graph-crawler            # лише HTTP
+pip install "graph-crawler[browser]"  # + Playwright
+```

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler.egg-info/SOURCES.txt

@@ -1,4 +1,3 @@
-CHANGELOG.md
 LICENSE
 MANIFEST.in
 README.md

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/graph_crawler.egg-info/requires.txt

@@ -36,7 +36,7 @@ celery>=5.5.0
 redis>=5.2.0
 
 [cloak]
-cloakbrowser>=0.5
+cloakbrowser==0.6.3
 playwright>=1.58.0
 
 [dev]

{graph_crawler-5.4.2 → graph_crawler-5.4.4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "graph-crawler"
-version = "5.4.2"
+version = "5.4.4"
 description = "Web crawler that builds a graph of a site. HTTP → TLS → browser escalation with pluggable browser engines."
 readme = "README.md"
 requires-python = ">=3.11"
@@ -60,7 +60,7 @@ playwright = ["graph-crawler[browser]"]                    # alias на один
 # Anti-detect tiers (see README "Anti-detect")
 tls = ["curl_cffi>=0.7"]                                   # tier "tls": TLS fingerprint (JA3/JA4)
 stealth = ["patchright>=1.52", "playwright-stealth>=1.0"]  # engine "patchright"
-cloak = ["cloakbrowser>=0.5", "playwright>=1.58.0"]        # engine "cloakbrowser" (default browser tier)
+cloak = ["cloakbrowser==0.6.3", "playwright>=1.58.0"]        # engine "cloakbrowser" (default browser tier)
 undetected = ["nodriver>=0.38"]                            # tier "undetected": driver="nodriver"
 antidetect = ["graph-crawler[tls,browser,stealth,cloak,undetected]"]