Add scrap patent

app.py

@@ -4,11 +4,17 @@ from mcp.server.fastmcp import FastMCP
 
 server = FastMCP(name="streamable-http-mcp-server-test", json_response=False, stateless_http=False)
 
-async def make_request(url: str, data: Dict[str, Any]):
+async def make_request(url: str, method: Literal["GET", "POST"], data: Dict[str, Any] = {}):
     headers = {"Accept": "application/json"}
     async with httpx.AsyncClient(verify=False) as client:
         try:
-            response = await client.post(url, headers=headers, json=data)
+            if method == "GET":
+                response = await client.get(url, headers=headers)
+            elif method == "POST":
+                response = await client.post(url, headers=headers, json=data)
+            else:
+                print("Method not allowed !")
+                return None
             response.raise_for_status()
             return response.json()
         except:
@@ -46,7 +52,7 @@ async def get_document_url(doc_id: str) -> str:
     Returns the URL (also scope + version if doc is a specification [not all specifications have a version or scope])
     Arguments: doc_id -> string
     """
-    response = await make_request('https://organizedprogrammers-docfinder.hf.space/find/single', {"doc_id": doc_id})
+    response = await make_request('https://organizedprogrammers-docfinder.hf.space/find/single', "POST", {"doc_id": doc_id})
     if not response:
         return "Unable to find document/specification"
     version = response.get('version', 'unavailable')
@@ -64,7 +70,7 @@ async def search_specifications_with_keywords(keywords: str, threshold: int = 60
         - source -> string (either '3GPP', 'ETSI' or 'all', by default, set to 'all')
         - spec_type -> string (either 'TS' or 'TR' or None, by default, set to None)
     """
-    response = await make_request('https://organizedprogrammers-docfinder.hf.space/search/bm25', {"keywords": keywords, "threshold": threshold, "source": source, "spec_type": spec_type})
+    response = await make_request('https://organizedprogrammers-docfinder.hf.space/search/bm25', "POST", {"keywords": keywords, "threshold": threshold, "source": source, "spec_type": spec_type})
     if not response:
         return "Unable to search specifications | No specifications has been found"
     results = response["results"]
@@ -78,7 +84,7 @@ async def get_spec_text(spec_id: str) -> str:
     Returns a dictionary k:v where k is the section (1., 2.2.1, ...) and v, the content of k, or a string if failed
     Args: spec_id -> string
     """
-    response = await make_request('https://organizedprogrammers-specsplitter.hf.space/extract_text/structured', {"spec_id": spec_id})
+    response = await make_request('https://organizedprogrammers-specsplitter.hf.space/extract_text/structured', "POST", {"spec_id": spec_id})
     if not response:
         return "Unable to extract specification text"
     return "\n".join([f"{k}: {v}" for k, v in response.keys()])
@@ -93,8 +99,27 @@ async def search_google_patents(queries: List[str], n_results: int) -> str:
     Returns a list of patents from queries, for each query, {n_results} patents will be retrieved
     Args: queries -> list of string, n_results -> integer [by default: 10]
     """
-    response = await make_request("https://organizedprogrammers-serpent.hf.space/serp/search_patents", {"queries": queries, "n_results": n_results})
+    response = await make_request("https://organizedprogrammers-serpent.hf.space/serp/search_patents", "POST", {"queries": queries, "n_results": n_results})
     if not response:
         return "Unable to fetch patents"
     return "\n".join(f"[Patent ID: {patent['id']} | Title: {patent['title']} | Body: {patent['body']}]" for patent in response.results)
+
+@server.tool()
+async def scrap_google_patents(patent_ids: List[str]) -> str:
+    """
+    Scrap patents from one or many patents from Google Patents
+    Returns a list of patents with their title, abstract, description, claims, field of invention and background
+    Args: patent_ids -> list of strings corresponding to Google Patent ID [min. 1]
+    """
+    if len(patent_ids) > 1:
+        response = await make_request("https://organizedprogrammers-serpent.hf.space/scrap/scrap_patents_bulk", "POST", {"patent_ids": patent_ids})
+        if not response:
+            return "Unable to scrap patents"
+        return "\n---\n".join([f"Title: {pat['title']}\nAbstract: {pat['abstract']}\nDescription: {pat['description']}\nClaims: {pat['claims']}\nField of invention{pat['field_of_invention']}\nBackground: {pat['background']}" for pat in response['patents']])
+    elif len(patent_ids) == 1:
+        response = await make_request("https://organizedprogrammers-serpent.hf.space/scrap/scrap_patent/"+patent_ids[0], "GET")
+        if not response:
+            return "Unable to scrap patent"
+        return f"Title: {response['title']}\nAbstract: {response['abstract']}\nDescription: {response['description']}\nClaims: {response['claims']}\nField of invention{response['field_of_invention']}\nBackground: {response['background']}"
+
 app = server.streamable_http_app