added [playwright] extra

aider/commands.py

@@ -9,7 +9,7 @@ import git
 from aider import models, prompts, voice
 from aider.help import Help, install_help_extra
 from aider.llm import litellm
-from aider.scrape import Scraper
+from aider.scrape import Scraper, install_playwright
 from aider.utils import is_image_file
 
 from .dump import dump  # noqa: F401
@@ -65,17 +65,17 @@ class Commands:
             return
 
         if not self.scraper:
-            self.scraper = Scraper(print_error=self.io.tool_error)
+            res = install_playwright(self.io)
+            if not res:
+                self.io.tool_error("Unable to initialize playwright.")
+
+            self.scraper = Scraper(print_error=self.io.tool_error, playwright_available=res)
 
         content = self.scraper.scrape(url) or ""
         # if content:
         #    self.io.tool_output(content)
 
-        instructions = self.scraper.get_playwright_instructions()
-        if instructions:
-            self.io.tool_error(instructions)
-
-        content = f"{url}:\n\n" + content  # noqa: E231
+        content = f"{url}:\n\n" + content
 
         return content
 

aider/scrape.py

@@ -12,14 +12,59 @@ aider_user_agent = f"Aider/{__version__} +{urls.website}"
 
 # Playwright is nice because it has a simple way to install dependencies on most
 # platforms.
-PLAYWRIGHT_INFO = """
-For better web scraping, install Playwright chromium:
+
+
+def install_playwright(io):
+    try:
+        from playwright.sync_api import sync_playwright
+
+        has_pip = True
+    except ImportError:
+        has_pip = False
+
+    try:
+        with sync_playwright() as p:
+            p.chromium.launch()
+            has_chromium = True
+    except Exception as err:
+        dump(err)
+        has_chromium = False
+
+    if has_pip and has_chromium:
+        return True
+
+    pip_cmd = utils.get_pip_install(["aider-chat[playwright]"])
+    chromium_cmd = "playwright install --with-deps chromium".split()
+
+    cmds = ""
+    if not has_pip:
+        cmds += " ".join(pip_cmd) + "\n"
+    if not has_chromium:
+        cmds += " ".join(chromium_cmd) + "\n"
+
+    text = f"""For the best web scraping, install Playwright:
 
 {cmds}
-
 See {urls.enable_playwright} for more info.
 """
 
+    io.tool_error(text)
+    if not io.confirm_ask("Install playwright?", default="y"):
+        return
+
+    if not has_pip:
+        success, output = utils.run_install(pip_cmd)
+        if not success:
+            io.tool_error(output)
+            return
+
+    success, output = utils.run_install(chromium_cmd)
+    if not success:
+        io.tool_error(output)
+        return
+
+    return True
+
 
 class Scraper:
     pandoc_available = None
@@ -27,7 +72,7 @@ class Scraper:
     playwright_instructions_shown = False
 
     # Public API...
-    def __init__(self, print_error=None):
+    def __init__(self, print_error=None, playwright_available=None):
         """
         `print_error` - a function to call to print error/debug info.
         """
@@ -36,13 +81,14 @@ class Scraper:
         else:
             self.print_error = print
 
+        self.playwright_available = playwright_available
+
     def scrape(self, url):
         """
         Scrape a url and turn it into readable markdown.
 
         `url` - the URLto scrape.
         """
-        self.try_playwright()
 
         if self.playwright_available:
             content = self.scrape_with_playwright(url)
@@ -88,46 +134,8 @@ class Scraper:
 
         return content
 
-    def try_playwright(self):
-        if self.playwright_available is not None:
-            return
-
-        try:
-            from playwright.sync_api import sync_playwright
-
-            has_pip = True
-        except ImportError:
-            has_pip = False
-
-        try:
-            with sync_playwright() as p:
-                p.chromium.launch()
-                has_chromium = True
-        except Exception:
-            has_chromium = False
-
-        if has_pip and has_chromium:
-            self.playwright_available = True
-
-        pip_cmd = utils.get_pip_cmd("playwright")
-        chromium_cmd = "playwright install --with-deps chromium".split()
-
-        cmds = ""
-        if not has_pip:
-            cmds += " ".join(pip_cmd) + "\n"
-        if not has_chromium:
-            cmds += " ".join(chromium_cmd) + "\n"
-
-        text = PLAYWRIGHT_INFO.format(cmds=cmds)
-
     def get_playwright_instructions(self):
-        if self.playwright_available in (True, None):
-            return
-        if self.playwright_instructions_shown:
-            return
-
-        self.playwright_instructions_shown = True
-        return PLAYWRIGHT_INFO
+        return
 
     def scrape_with_httpx(self, url):
         import httpx

aider/utils.py

@@ -224,18 +224,21 @@ def run_install(cmd):
                 last_update = current_time
 
         return_code = process.wait()
+        output = "".join(output)
+
+        dump(output)
 
         if return_code == 0:
             print("\rInstallation complete.")
             print()
-            return True, ''.join(output)
+            return True, output
 
     except subprocess.CalledProcessError as e:
         print(f"\nError running pip install: {e}")
 
     print("\nInstallation failed.\n")
 
-    return False, ''.join(output)
+    return False, output
 
 
 def check_pip_install_extra(io, module, prompt, pip_install_cmd):
@@ -263,8 +266,7 @@ def check_pip_install_extra(io, module, prompt, pip_install_cmd):
     except (ImportError, ModuleNotFoundError):
         pass
 
-    for line in output:
-        print(line)
+    io.tool_error(output)
 
     print()
     print(f"Failed to install {pip_install_cmd[0]}")