Merge pull request #29 from simular-ai/1124-macos

Add more helpers to macOS and an example for web crawler on chrome

Author: simularlyon

README.md

@@ -42,6 +42,17 @@ python openaci/cli_app.py
 
 Please remember to review and modify the code according to your specific use case and requirements. It is highly recommended to thoroughly test the code in a controlled environment before using it in any production or critical systems.
 
+### Mac OS
+To run OpenACI on macOS, we need xcode developer package.
+```
+xcode-select --install
+```
+
+Also install
+```
+pip install pyobjc-framework-ApplicationServices
+```
+
 ## Contributing
 Contributions to this experimental codebase are welcome. If you discover any issues, have suggestions, or would like to add new features, please open an issue or submit a pull request.
 

examples/crawler/main.py

@@ -0,0 +1,86 @@
+from openaci.macos.Grounding import GroundingAgent
+import openaci.macos.system as system
+import openaci.macos.web as web
+import time
+import json
+import os
+
+def crawl(main_url, root_dir=None, browser="Chrome"):
+    # get domain from url in general way
+    domain = web.get_domain(main_url)
+    print('domain', domain)
+    content = system.fetch_content_from_url(main_url, where=browser, wait_time=2, retry=3)
+
+    # pretty print content
+    # print(json.dumps(content, indent=4))
+
+    # get all the links from content and open them in chrome
+    # do not open the same link twice
+    # do not open links that are not in the same domain
+    # create a directory to save the json content
+    visited_links = []
+    # default goes to home directory downloads folder
+    root_dir = root_dir or os.path.expanduser("~/Downloads/simular/")
+    root_dir = f"{root_dir}/{domain}"
+    os.makedirs(root_dir, exist_ok=True)
+
+    def get_file_path(link):
+        # handle / inside the link and remove https://
+        link = link.replace('/', '_').replace('https://', '')
+        return f"{root_dir}/{link}.json"
+    
+    # save json content to a file
+    def save_json_content(content, link):
+        with open(get_file_path(link), "w") as f:
+            # pretty print json
+            json.dump(content, f, indent=4)
+
+    def load_json_content(link):
+        with open(get_file_path(link), "r") as f:
+            return json.load(f)
+        return None
+
+    save_json_content(content, main_url)
+
+    # breadth first search visit all the links
+    queue = []
+
+    def add_to_queue(links):
+        for link in links:
+            # skip if link is already visited
+            if link in visited_links:
+                continue
+            # skip if link is not in the same domain
+            print(web.get_domain(link), domain)
+            if web.get_domain(link) != domain:
+                continue
+            queue.append(link)
+    
+    try:
+        add_to_queue([x['url'] for x in content['links'] if x['url']])
+    except Exception as e:
+        print(f"Error adding to queue: {e}")
+
+    print(content['links'])
+    print(f"Queue size: {len(queue)}")
+        
+    while queue:
+        link = queue.pop(0)
+        print(f"Visiting {link}")
+
+        visited_links.append(link)
+        # skip if file already exists
+        if os.path.exists(get_file_path(link)):
+            # load the json content
+            content = load_json_content(link)
+        else:
+            content = system.fetch_content_from_url(link, where=browser, wait_time=2)
+            save_json_content(content, link)
+
+        try:
+            add_to_queue([x['url'] for x in content['links'] if x['url']])
+        except Exception as e:
+            print(f"Error adding to queue: {e}")
+
+if __name__ == "__main__":
+    crawl("smarthistory.org")

openaci/macos/Grounding.py

@@ -21,6 +21,7 @@
 
 
 from openaci.macos.UIElement import UIElement
+from openaci.macos.system import open_running_app
 
 from AppKit import NSWorkspace, NSRunningApplication
 
@@ -289,6 +290,9 @@ def open_app(self, app_name):
         if app_name in self.all_apps:
             print(f"{app_name} has been opened successfully.")
             return f"""import subprocess; subprocess.run(["open", "-a", "{app_name}"], check=True)"""
+        elif open_running_app(app_name):
+            print(f"{app_name} has been opened successfully.")
+            return """WAIT"""
         else:
             self.execution_feedback = "There is no application " + app_name + " installed on the system. Please replan and avoid this action."
             print(self.execution_feedback)