model-compose 0.4.67__tar.gz → 0.4.68__tar.gz

{model_compose-0.4.67 → model_compose-0.4.68}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: model-compose
-Version: 0.4.67
+Version: 0.4.68
 Summary: model-compose: Declarative AI Workflow Orchestrator
 Author-email: Hanyeol Cho <hanyeol.cho@gmail.com>
 License-Expression: MIT

{model_compose-0.4.67 → model_compose-0.4.68}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "model-compose"
-version = "0.4.67"
+version = "0.4.68"
 description = "model-compose: Declarative AI Workflow Orchestrator"
 authors = [
     { name = "Hanyeol Cho", email = "hanyeol.cho@gmail.com" }

{model_compose-0.4.67 → model_compose-0.4.68}/src/mindor/core/component/services/web_browser/drivers/chrome.py

@@ -107,10 +107,10 @@ class ChromeBrowserSession(WebBrowserSession):
         result = await self.client.send_command("Runtime.evaluate", {
             "expression": script, "returnByValue": True
         })
- 
+
         if "exceptionDetails" in result:
             raise RuntimeError(f"Extract JavaScript error: { result['exceptionDetails'] }")
- 
+
         return result.get("result", {}).get("value")
 
     # ---- Interaction ----

{model_compose-0.4.67 → model_compose-0.4.68}/src/mindor/core/component/services/web_browser/web_browser.py

@@ -92,7 +92,21 @@ class WebBrowserAction:
         attribute    = await context.render_variable(self.config.attribute) if self.config.attribute else None
         multiple     = await context.render_variable(self.config.multiple)
 
-        return await session.extract(selector, xpath, extract_mode, attribute, bool(multiple))
+        if selector:
+            if isinstance(selector, dict):
+                return { key: await session.extract(expr, None, extract_mode, attribute, bool(multiple)) for key, expr in selector.items() }
+            if isinstance(selector, list):
+                return [ await session.extract(expr, None, extract_mode, attribute, bool(multiple)) for expr in selector ]
+            return await session.extract(selector, None, extract_mode, attribute, bool(multiple))
+        
+        if xpath:
+            if isinstance(xpath, dict):
+                return { key: await session.extract(None, expr, extract_mode, attribute, bool(multiple)) for key, expr in xpath.items() }
+            if isinstance(xpath, list):
+                return [ await session.extract(None, expr, extract_mode, attribute, bool(multiple)) for expr in xpath ]
+            return await session.extract(None, xpath, extract_mode, attribute, bool(multiple))
+        
+        return await session.extract(None, None, extract_mode, attribute, bool(multiple))
 
     # ---- Interaction ----
 

{model_compose-0.4.67 → model_compose-0.4.68}/src/mindor/core/component/services/web_scraper.py

@@ -157,48 +157,56 @@ class WebScraperAction:
     def _extract_with_selector(
         self,
         html: str,
-        selector: str,
+        selector: Union[str, List[str], Dict[str, str]],
         extract_mode: str,
         attribute: Optional[str],
         multiple: bool
-    ) -> Union[str, List[str], None]:
-        """Extract content using CSS selector."""
+    ) -> Any:
+        """Extract content using CSS selector. Parses HTML once and reuses the parsed soup."""
         soup = BeautifulSoup(html, 'lxml')
 
-        if multiple:
-            elements = soup.select(selector)
-
-            if not elements:
-                return []
-
-            return [ self._extract_from_element(element, extract_mode, attribute) for element in elements ]
-        else:
-            element = soup.select_one(selector)
+        def _extract(expr: str) -> Union[str, List[str], None]:
+            if multiple:
+                elements = soup.select(expr)
+                if not elements:
+                    return []
+                return [ self._extract_from_element(element, extract_mode, attribute) for element in elements ]
 
+            element = soup.select_one(expr)
             if not element:
                 return None
-
             return self._extract_from_element(element, extract_mode, attribute)
 
+        if isinstance(selector, dict):
+            return { key: _extract(expr) for key, expr in selector.items() }
+        if isinstance(selector, list):
+            return [ _extract(expr) for expr in selector ]
+        return _extract(selector)
+
     def _extract_with_xpath(
         self,
         html: str,
-        xpath: str,
+        xpath: Union[str, List[str], Dict[str, str]],
         extract_mode: str,
         attribute: Optional[str],
         multiple: bool
-    ) -> Union[str, List[str], None]:
-        """Extract content using XPath."""
+    ) -> Any:
+        """Extract content using XPath. Parses HTML once and reuses the parsed tree."""
         tree = etree.HTML(html)
-        elements = tree.xpath(xpath)
-
-        if not elements:
-            return [] if multiple else None
 
-        if multiple:
-            return [ self._extract_from_xpath_element(elem, extract_mode, attribute) for elem in elements ]
-
-        return self._extract_from_xpath_element(elements[0], extract_mode, attribute)
+        def _extract(expr: str) -> Union[str, List[str], None]:
+            elements = tree.xpath(expr)
+            if not elements:
+                return [] if multiple else None
+            if multiple:
+                return [ self._extract_from_xpath_element(element, extract_mode, attribute) for element in elements ]
+            return self._extract_from_xpath_element(elements[0], extract_mode, attribute)
+
+        if isinstance(xpath, dict):
+            return { key: _extract(expr) for key, expr in xpath.items() }
+        if isinstance(xpath, list):
+            return [ _extract(expr) for expr in xpath ]
+        return _extract(xpath)
 
     def _extract_from_element(self, element, extract_mode: str, attribute: Optional[str]) -> Optional[str]:
         """Extract content from a BeautifulSoup element."""

{model_compose-0.4.67 → model_compose-0.4.68}/src/mindor/dsl/schema/action/impl/web_browser.py

@@ -48,8 +48,8 @@ class WebBrowserScreenshotActionConfig(CommonWebBrowserActionConfig):
 
 class WebBrowserExtractActionConfig(CommonWebBrowserActionConfig):
     method: Literal[WebBrowserActionMethod.EXTRACT]
-    selector: Optional[str] = Field(default=None, description="CSS selector.")
-    xpath: Optional[str] = Field(default=None, description="XPath expression.")
+    selector: Optional[Union[str, List[str], Dict[str, str]]] = Field(default=None, description="CSS selector(s) to extract elements.")
+    xpath: Optional[Union[str, List[str], Dict[str, str]]] = Field(default=None, description="XPath expression(s) to extract elements.")
     extract_mode: Union[Literal["text", "html", "attribute"], str] = Field(default="text", description="Extraction mode.")
     attribute: Optional[str] = Field(default=None, description="Attribute name when extract_mode='attribute'.")
     multiple: Union[bool, str] = Field(default=False, description="Return all matches as a list.")

{model_compose-0.4.67 → model_compose-0.4.68}/src/mindor/dsl/schema/action/impl/web_scraper.py

@@ -19,8 +19,8 @@ class WebScraperActionConfig(CommonActionConfig):
     url: str = Field(..., description="URL to scrape")
     headers: Dict[str, str] = Field(default_factory=dict, description="HTTP headers to include in the request")
     cookies: Dict[str, str] = Field(default_factory=dict, description="Cookies to include in the request")
-    selector: Optional[str] = Field(default=None, description="CSS selector to extract elements")
-    xpath: Optional[str] = Field(default=None, description="XPath expression to extract elements")
+    selector: Optional[Union[str, List[str], Dict[str, str]]] = Field(default=None, description="CSS selector(s) to extract elements.")
+    xpath: Optional[Union[str, List[str], Dict[str, str]]] = Field(default=None, description="XPath expression(s) to extract elements.")
     extract_mode: Union[Literal[ "text", "html", "attribute" ], str] = Field(default="text", description="Extraction mode")
     attribute: Optional[str] = Field(default=None, description="Attribute name to extract when extract_mode='attribute'")
     multiple: Union[bool, str] = Field(default=False, description="Extract multiple elements (returns list) or single element")

{model_compose-0.4.67 → model_compose-0.4.68}/src/model_compose.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: model-compose
-Version: 0.4.67
+Version: 0.4.68
 Summary: model-compose: Declarative AI Workflow Orchestrator
 Author-email: Hanyeol Cho <hanyeol.cho@gmail.com>
 License-Expression: MIT