parsagon 0.16.0__py3-none-any.whl → 0.17.1__py3-none-any.whl

parsagon/api.py

@@ -247,7 +247,7 @@ def get_pipelines():
     return _api_call(httpx.get, f"/pipelines/")
 
 
-def get_pipeline_code(pipeline_name, variables, headless, use_uc, optimize, use_proxy, pipeline_id=None):
+def get_pipeline_code(pipeline_name, variables, headless, use_cdp, use_uc, optimize, use_proxy, pipeline_id=None):
     if pipeline_id:
         return _api_call(
             httpx.post,
@@ -255,6 +255,7 @@ def get_pipeline_code(pipeline_name, variables, headless, use_uc, optimize, use_
             json={
                 "variables": variables,
                 "headless": headless,
+                "use_cdp": use_cdp,
                 "use_uc": use_uc,
                 "optimize": optimize,
                 "use_proxy": use_proxy,
@@ -268,6 +269,7 @@ def get_pipeline_code(pipeline_name, variables, headless, use_uc, optimize, use_
             json={
                 "variables": variables,
                 "headless": headless,
+                "use_cdp": use_cdp,
                 "use_uc": use_uc,
                 "optimize": optimize,
                 "use_proxy": use_proxy,

parsagon/executor.py

@@ -17,11 +17,17 @@ from lxml import etree
 import lxml.html
 from lxml.html.clean import Cleaner
 from pypdf import PdfReader
-from seleniumbase import Driver
+from pyvirtualdisplay import Display
+import undetected_chromedriver as uc
+from selenium import webdriver
+from selenium.webdriver.chrome.service import Service as ChromeService
+from selenium.webdriver.chrome.options import Options
 from selenium.webdriver.common.action_chains import ActionChains
 from selenium.webdriver.common.by import By
 from selenium.webdriver.common.keys import Keys
 from selenium.webdriver.support.select import Select
+from webdriver_manager.chrome import ChromeDriverManager
+from webdriver_manager.core.driver_cache import DriverCacheManager
 
 from parsagon import settings
 from parsagon.api import (
@@ -76,6 +82,7 @@ class Executor:
         infer=False,
         use_uc=False,
         driver_path=None,
+        options=[],
         page_load_timeout=None,
         script_timeout=None,
         function_bank={},
@@ -83,8 +90,43 @@ class Executor:
         self.task = task
         self.headless = headless
         self.function_bank = function_bank
-        self.driver = Driver(external_pdf=True, undetectable=use_uc)
-        self.driver.maximize_window()
+        if self.headless:
+            self.display = Display(visible=False, size=(1280, 1050)).start()
+        if driver_path:
+            cache_manager = DriverCacheManager(root_dir=driver_path)
+            driver_executable_path = ChromeDriverManager(cache_manager=cache_manager).install()
+        else:
+            driver_executable_path = ChromeDriverManager().install()
+        if use_uc:
+            chrome_options = uc.ChromeOptions()
+            chrome_options.add_argument("--start-maximized")
+            for option in options:
+                chrome_options.add_argument(option)
+            chrome_options.add_experimental_option(
+                "prefs",
+                {
+                    "download.default_directory": os.getcwd(),
+                    "download.prompt_for_download": False,
+                    "download.directory_upgrade": True,
+                    "plugins.always_open_pdf_externally": True,
+                },
+            )
+            self.driver = uc.Chrome(driver_executable_path=driver_executable_path, options=chrome_options)
+        else:
+            chrome_options = webdriver.ChromeOptions()
+            chrome_options.add_argument("--start-maximized")
+            for option in options:
+                chrome_options.add_argument(option)
+            chrome_options.add_experimental_option(
+                "prefs",
+                {
+                    "download.default_directory": os.getcwd(),
+                    "download.prompt_for_download": False,
+                    "download.directory_upgrade": True,
+                    "plugins.always_open_pdf_externally": True,
+                },
+            )
+            self.driver = webdriver.Chrome(service=ChromeService(driver_executable_path), options=chrome_options)
         if page_load_timeout:
             self.driver.set_page_load_timeout(page_load_timeout)
         if script_timeout:
@@ -115,6 +157,7 @@ class Executor:
             "get_serp_data": get_serp_data,
             "get_pdf_text": self.get_pdf_text,
             "str_to_iso8601": self.str_to_iso8601,
+            "get_network_requests": self.get_network_requests,
         }
         self.custom_functions = {}
         self.infer = infer
@@ -671,6 +714,17 @@ class Executor:
                 s = " ".join(s.split()[:-1])
         return ""
 
+    def get_network_requests(window_id, delay=30):
+        requests = []
+
+        def handle_request(event):
+            requests.append(event)
+
+        driver.switch_to.window(window_id)
+        driver.add_cdp_listener("Network.requestWillBeSent", handle_request)
+        wait(delay)
+        return requests
+
     def execute(self, code):
         loc = {}
         try:

parsagon/main.py

@@ -144,6 +144,11 @@ def get_args(argv):
         type=str,
         help="write the data to the given file path",
     )
+    parser_run.add_argument(
+        "--use_cdp",
+        action="store_true",
+        help="enable CDP (only in undetected mode)",
+    )
     parser_run.add_argument(
         "--undetected",
         action="store_true",

parsagon/runs.py

@@ -28,6 +28,7 @@ def run(
     remote=False,
     output_log=False,
     output_file=None,
+    use_cdp=False,
     undetected=False,
     optimize=False,
     use_proxy=False,
@@ -83,7 +84,7 @@ def run(
 
     run = create_pipeline_run(pipeline_id, variables, True)
     code = get_pipeline_code(
-        program_name, variables, headless, undetected, optimize, use_proxy, pipeline_id=program_id
+        program_name, variables, headless, use_cdp, undetected, optimize, use_proxy, pipeline_id=program_id
     )["code"]
     start_time = datetime.datetime.now(datetime.timezone.utc).isoformat()
     run_data = {"start_time": start_time}

{parsagon-0.16.0.dist-info → parsagon-0.17.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: parsagon
-Version: 0.16.0
+Version: 0.17.1
 Summary: Allows you to create browser automations with natural language
 Author-email: Sandy Suh <sandy@parsagon.io>
 Project-URL: Homepage, https://parsagon.io
@@ -8,20 +8,19 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Operating System :: OS Independent
 Requires-Python: >=3.8
 Description-Content-Type: text/markdown
-Requires-Dist: seleniumbase==4.38.0
-Requires-Dist: selenium>=4.9.1
+Requires-Dist: selenium==4.9.1
 Requires-Dist: lxml[html_clean]==5.3.2
 Requires-Dist: httpx>=0.27.0
 Requires-Dist: psutil==7.0.0
-Requires-Dist: rich>=13.6.0
+Requires-Dist: rich==13.6.0
 Requires-Dist: PyVirtualDisplay==3.0
 Requires-Dist: selenium-wire==5.1.0
-Requires-Dist: cssselect>=1.2.0
+Requires-Dist: cssselect==1.2.0
 Requires-Dist: undetected-chromedriver==3.5.5
 Requires-Dist: webdriver-manager==4.0.2
 Requires-Dist: jsonpath-ng==1.5.3
 Requires-Dist: simplejson==3.19.1
-Requires-Dist: pypdf==4.2.0
+Requires-Dist: pypdf==5.9.0
 Requires-Dist: python-dateutil==2.9.0.post0
 Requires-Dist: html2text==2024.2.26
 Requires-Dist: feedparser==6.0.11

{parsagon-0.16.0.dist-info → parsagon-0.17.1.dist-info}/RECORD

@@ -1,16 +1,16 @@
 parsagon/__init__.py,sha256=_S5MlYHBViB4iY96_UbGo3mfGndE8MmFmb1EORJppK8,452
-parsagon/api.py,sha256=TCS1iGuzuMZI-MOtdDFA-Xwdb8ogtuOQVUJhKtTG0DY,10484
+parsagon/api.py,sha256=8ZaJfx1Jnf18rFWhydcdRPswnddgDsc8UqTLIctMR0g,10565
 parsagon/assistant.py,sha256=V3NL6UdDqe74W_X3wPQ1qwFuJRvHzitXtOt-XSCXvds,4065
 parsagon/create.py,sha256=BERrBviwMvifg5OwApqdanvULJHHk39fIvnTCZN3Xkk,4432
 parsagon/custom_function.py,sha256=oEj28qItaHUnsvLIHD7kg5QL3J3aO6rW6xKKP-H-Drs,770
 parsagon/edit.py,sha256=5gtnx0gNB7Gvz8ET00SczE-ZS0TomN1um6uObP-OObE,3120
 parsagon/exceptions.py,sha256=tG1vnpmUN1GdJ1GSpe1MaWH3zWmFLZCwtOfEGu8qPP0,910
-parsagon/executor.py,sha256=wmHkp21RfReVay9t80r_eof0rVc86wCCWwQvj0au_fQ,27551
+parsagon/executor.py,sha256=_jPdLxclmLekb03JCtxvslRpfUhVotL6rnHFw9EibXE,29927
 parsagon/gui_entry.py,sha256=bqG9K0CArXWWwDGoT8aV17YLNM8MfjSf6SJ_B3QbNeA,671
 parsagon/highlights.js,sha256=2UDfUApblU9xtGgTLCq4X7rHRV0wcqDSSFZPmJS6fJg,16643
-parsagon/main.py,sha256=LIUE3sfiU7cuz44jgTTfSG807EHJa8avCKtJS1fF18s,9863
+parsagon/main.py,sha256=_ww08Fpf1S8ePXSDg2GqtmnQvdfSgMK_AiNq5CuhWko,10001
 parsagon/print.py,sha256=-7iVKil0W9e8zX1EJMcdlqNdfpmfPxKTBtZfwzWpGYU,4106
-parsagon/runs.py,sha256=gi36oak3RxKo1BxB0Bc6GV8K06vrrBrEXu3NYkIVPwY,8633
+parsagon/runs.py,sha256=VDvWpg3x6D67jNSffimeL7rxQy8rycsys0kjV0fNWRU,8661
 parsagon/secrets.py,sha256=72dr-6q1q2ATBkE75fT18tcvwDM-4nymTb9NDVwjHTE,545
 parsagon/settings.py,sha256=ejd9wGCOEB68n8a0M2SyzlmlwPofnzMKaxpBqYhcs34,3552
 parsagon/gui/__init__.py,sha256=ZvKZfcchSnBh__UY_XeMnQJACd1EG99ix3_SfSHth8g,68
@@ -25,8 +25,8 @@ parsagon/tests/test_invalid_args.py,sha256=TAFdHGy92lUxjljPrtODOuEGVss6rn-F5GvEK
 parsagon/tests/test_pipeline_operations.py,sha256=aEwZNtIwOl9X7jdLDLB4YEdgMp7_x8PXCINAE7RT4NY,805
 parsagon/tests/test_print.py,sha256=BG7f55YDBoL0S7k291-so_Gje_hUAQOkB-jh-bEYsJY,198
 parsagon/tests/test_secrets.py,sha256=Ctsscl2tmMTZcFAy5dnyqUlgTov2UharZgLpbRCLdEg,2662
-parsagon-0.16.0.dist-info/METADATA,sha256=UoHsc13frMxoCOPYqdKX_-BIFEVQ88b3h2CQROq1fSk,2603
-parsagon-0.16.0.dist-info/WHEEL,sha256=DnLRTWE75wApRYVsjgc6wsVswC54sMSJhAEd4xhDpBk,91
-parsagon-0.16.0.dist-info/entry_points.txt,sha256=I1UlPUb4oY2k9idkI8kvdkEcrjKGRSOl5pMbA6uu6kw,48
-parsagon-0.16.0.dist-info/top_level.txt,sha256=ih5uYQzW4qjhRKppys-WiHLIbXVZ99YdqDcfAtlcQwk,9
-parsagon-0.16.0.dist-info/RECORD,,
+parsagon-0.17.1.dist-info/METADATA,sha256=-ylaryWc8rMx_nq9m3Oz_WUsQ7Ygw37pCQqnQn-1hkc,2567
+parsagon-0.17.1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+parsagon-0.17.1.dist-info/entry_points.txt,sha256=I1UlPUb4oY2k9idkI8kvdkEcrjKGRSOl5pMbA6uu6kw,48
+parsagon-0.17.1.dist-info/top_level.txt,sha256=ih5uYQzW4qjhRKppys-WiHLIbXVZ99YdqDcfAtlcQwk,9
+parsagon-0.17.1.dist-info/RECORD,,

{parsagon-0.16.0.dist-info → parsagon-0.17.1.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (80.4.0)
+Generator: setuptools (80.9.0)
 Root-Is-Purelib: true
 Tag: py3-none-any