Bootstrap webmention endpoint

app/main.py

@@ -38,6 +38,7 @@ from app import httpsig
 from app import indieauth
 from app import models
 from app import templates
+from app import webmentions
 from app.actor import LOCAL_ACTOR
 from app.actor import get_actors_metadata
 from app.boxes import public_outbox_objects_count
@@ -82,6 +83,7 @@ app.mount("/static", StaticFiles(directory="app/static"), name="static")
 app.include_router(admin.router, prefix="/admin")
 app.include_router(admin.unauthenticated_router, prefix="/admin")
 app.include_router(indieauth.router)
+app.include_router(webmentions.router)
 
 logger.configure(extra={"request_id": "no_req_id"})
 logger.remove()

app/templates/object.html

@@ -3,6 +3,7 @@
 
 {% block head %}
 {% if outbox_object %}
+<link rel="webmention" href="{{ url_for("webmention_endpoint") }}">
 <link rel="alternate" href="{{ request.url }}" type="application/activity+json">
 <meta name="description" content="{{ outbox_object.content | html2text | trim | truncate(50) }}">
 <meta content="article" property="og:type" />

app/utils/indieauth.py

@@ -1,11 +1,7 @@
 from dataclasses import dataclass
 from typing import Any
 
-import httpx
-import mf2py  # type: ignore
-from loguru import logger
-
-from app import config
+from app.utils import microformats
 from app.utils.url import make_abs
 
 
@@ -26,21 +22,10 @@ def _get_prop(props: dict[str, Any], name: str, default=None) -> Any:
 
 
 async def get_client_id_data(url: str) -> IndieAuthClient | None:
-    async with httpx.AsyncClient() as client:
-        try:
-            resp = await client.get(
-                url,
-                headers={
-                    "User-Agent": config.USER_AGENT,
-                },
-                follow_redirects=True,
-            )
-            resp.raise_for_status()
-        except (httpx.HTTPError, httpx.HTTPStatusError):
-            logger.exception(f"Failed to discover webmention endpoint for {url}")
-            return None
+    maybe_data_and_html = await microformats.fetch_and_parse(url)
+    if maybe_data_and_html is not None:
+        data: dict[str, Any] = maybe_data_and_html[0]
 
-    data = mf2py.parse(doc=resp.text)
         for item in data["items"]:
             if "h-x-app" in item["type"] or "h-app" in item["type"]:
                 props = item.get("properties", {})

app/utils/microformats.py

@@ -0,0 +1,25 @@
+from typing import Any
+
+import httpx
+import mf2py  # type: ignore
+from loguru import logger
+
+from app import config
+
+
+async def fetch_and_parse(url: str) -> tuple[dict[str, Any], str] | None:
+    async with httpx.AsyncClient() as client:
+        try:
+            resp = await client.get(
+                url,
+                headers={
+                    "User-Agent": config.USER_AGENT,
+                },
+                follow_redirects=True,
+            )
+            resp.raise_for_status()
+        except (httpx.HTTPError, httpx.HTTPStatusError):
+            logger.exception(f"Failed to discover webmention endpoint for {url}")
+            return None
+
+    return mf2py.parse(doc=resp.text), resp.text

app/webmentions.py

@@ -0,0 +1,62 @@
+from bs4 import BeautifulSoup  # type: ignore
+from fastapi import APIRouter
+from fastapi import HTTPException
+from fastapi import Request
+from fastapi.responses import JSONResponse
+from loguru import logger
+
+from app.utils import microformats
+from app.utils.url import check_url
+from app.utils.url import is_url_valid
+
+router = APIRouter()
+
+
+def is_source_containing_target(source_html: str, target_url: str) -> bool:
+    soup = BeautifulSoup(source_html, "html5lib")
+    for link in soup.find_all("a"):
+        h = link.get("href")
+        if not is_url_valid(h):
+            continue
+
+        if h == target_url:
+            return True
+
+    return False
+
+
+@router.post("/webmentions")
+async def webmention_endpoint(
+    request: Request,
+) -> JSONResponse:
+    form_data = await request.form()
+    try:
+        source = form_data["source"]
+        target = form_data["target"]
+
+        if source == target:
+            raise ValueError("source URL is the same as target")
+
+        check_url(source)
+        check_url(target)
+    except Exception:
+        logger.exception("Invalid webmention request")
+        raise HTTPException(status_code=400, detail="Invalid payload")
+
+    logger.info(f"Received webmention {source=} {target=}")
+
+    # TODO: get outbox via ap_id (URL is the same as ap_id)
+    maybe_data_and_html = await microformats.fetch_and_parse(source)
+    if not maybe_data_and_html:
+        logger.info("failed to fetch source")
+        raise HTTPException(status_code=400, detail="failed to fetch source")
+
+    data, html = maybe_data_and_html
+
+    if not is_source_containing_target(html, target):
+        logger.warning("target not found in source")
+        raise HTTPException(status_code=400, detail="target not found in source")
+
+    logger.info(f"{data=}")
+
+    return JSONResponse(content={}, status_code=200)