Merge pull request #66 from ToadKing/misskey

add support for misskey and calckey/firefish

README.md

@@ -136,6 +136,7 @@ Option | Required? | Notes |
 |`backfill-with-context` | No | Set to `0` to disable fetching remote replies while backfilling profiles. This is enabled by default, but you can disable it, if it's too slow for you.
 |`backfill-mentioned-users` | No | Set to `0` to disable backfilling any mentioned users when fetching the home timeline. This is enabled by default, but you can disable it, if it's too slow for you.
 | `remember-users-for-hours` | No | How long between back-filling attempts for non-followed accounts? Defaults to `168`, i.e. one week.
+| `remember-hosts-for-days` | No | How long should FediFetcher cache host info for? Defaults to `30`.
 | `http-timeout` | No | The timeout for any HTTP requests to the Mastodon API in seconds. Defaults to `5`.
 | `lock-hours` | No | Determines after how many hours a lock file should be discarded. Not relevant when running the script as GitHub Action, as concurrency is prevented using a different mechanism. Recommended value: `24`.
 | `lock-file` | No | Location for the lock file. If not specified, will use `lock.lock` under the state directory. Not relevant when running the script as GitHub Action.

find_posts.py

@@ -13,6 +13,7 @@ import time
 import argparse
 import uuid
 import git
+import defusedxml.ElementTree as ET
 
 argparser=argparse.ArgumentParser()
 
@@ -29,6 +30,7 @@ argparser.add_argument('--max-bookmarks', required = False, type=int, default=0,
 argparser.add_argument('--max-favourites', required = False, type=int, default=0, help="Fetch remote replies to the API key owners Favourites. We'll fetch replies to at most this many favourites")
 argparser.add_argument('--from-notifications', required = False, type=int, default=0, help="Backfill accounts of anyone appearing in your notifications, during the last hours")
 argparser.add_argument('--remember-users-for-hours', required=False, type=int, default=24*7, help="How long to remember users that you aren't following for, before trying to backfill them again.")
+argparser.add_argument('--remember-hosts-for-days', required=False, type=int, default=30, help="How long to remember host info for, before checking again.")
 argparser.add_argument('--http-timeout', required = False, type=int, default=5, help="The timeout for any HTTP requests to your own, or other instances.")
 argparser.add_argument('--backfill-with-context', required = False, type=int, default=1, help="If enabled, we'll fetch remote replies when backfilling profiles. Set to `0` to disable.")
 argparser.add_argument('--backfill-mentioned-users', required = False, type=int, default=1, help="If enabled, we'll backfill any mentioned users when fetching remote replies to timeline posts. Set to `0` to disable.")
@@ -66,17 +68,17 @@ def get_favourites(server, access_token, max):
         "Authorization": f"Bearer {access_token}",
     })
 
-def add_user_posts(server, access_token, followings, know_followings, all_known_users, seen_urls):
+def add_user_posts(server, access_token, followings, known_followings, all_known_users, seen_urls, seen_hosts):
     for user in followings:
         if user['acct'] not in all_known_users and not user['url'].startswith(f"https://{server}/"):
-            posts = get_user_posts(user, know_followings, server)
+            posts = get_user_posts(user, known_followings, server, seen_hosts)
 
             if(posts != None):
                 count = 0
                 failed = 0
                 for post in posts:
-                    if post.get('reblog') is None and post.get('url') is not None and post.get('url') not in seen_urls:
+                    if post.get('reblog') is None and post.get('renoteId') is None and post.get('url') is not None and post.get('url') not in seen_urls:
-                        added = add_post_with_context(post, server, access_token, seen_urls)
+                        added = add_post_with_context(post, server, access_token, seen_urls, seen_hosts)
                         if added is True:
                             seen_urls.add(post['url'])
                             count += 1
@@ -84,10 +86,10 @@ def add_user_posts(server, access_token, followings, know_followings, all_known_
                             failed += 1
                 log(f"Added {count} posts for user {user['acct']} with {failed} errors")
                 if failed == 0:
-                    know_followings.add(user['acct'])
+                    known_followings.add(user['acct'])
                     all_known_users.add(user['acct'])
 
-def add_post_with_context(post, server, access_token, seen_urls):
+def add_post_with_context(post, server, access_token, seen_urls, seen_hosts):
     added = add_context_url(post['url'], server, access_token)
     if added is True:
         seen_urls.add(post['url'])
@@ -96,27 +98,72 @@ def add_post_with_context(post, server, access_token, seen_urls):
             parsed = parse_url(post['url'], parsed_urls)
             if parsed == None:
                 return True
-            known_context_urls = get_all_known_context_urls(server, [post],parsed_urls)
+            known_context_urls = get_all_known_context_urls(server, [post],parsed_urls, seen_hosts)
             add_context_urls(server, access_token, known_context_urls, seen_urls)
         return True
     
     return False
 
-def get_user_posts(user, know_followings, server):
+def get_user_posts(user, known_followings, server, seen_hosts):
     parsed_url = parse_user_url(user['url'])
 
     if parsed_url == None:
         # We are adding it as 'known' anyway, because we won't be able to fix this.
-        know_followings.add(user['acct'])
+        known_followings.add(user['acct'])
         return None
     
     if(parsed_url[0] == server):
         log(f"{user['acct']} is a local user. Skip")
-        know_followings.add(user['acct'])
+        known_followings.add(user['acct'])
         return None
-    if re.match(r"^https:\/\/[^\/]+\/c\/", user['url']):
+
+    post_server = get_server_info(parsed_url[0], seen_hosts)
+    if post_server is None:
+        log(f'server {parsed_url[0]} not found for post')
+        return None
+
+    if post_server['mastodonApiSupport']:
+        return get_user_posts_mastodon(parsed_url[1], post_server['webserver'])
+
+    if post_server['lemmyApiSupport']:
+        return get_user_posts_lemmy(parsed_url[1], user['url'], post_server['webserver'])
+
+    if post_server['misskeyApiSupport']:
+        return get_user_posts_misskey(parsed_url[1], post_server['webserver'])
+
+    log(f'server api unknown for {post_server["webserver"]}, cannot fetch user posts')
+    return None
+
+def get_user_posts_mastodon(userName, webserver):
     try:
-            url = f"https://{parsed_url[0]}/api/v3/post/list?community_name={parsed_url[1]}&sort=New&limit=50"
+        user_id = get_user_id(webserver, userName)
+    except Exception as ex:
+        log(f"Error getting user ID for user {user['acct']}: {ex}")
+        return None
+
+    try:
+        url = f"https://{webserver}/api/v1/accounts/{user_id}/statuses?limit=40"
+        response = get(url)
+
+        if(response.status_code == 200):
+            return response.json()
+        elif response.status_code == 404:
+            raise Exception(
+                f"User {user['acct']} was not found on server {webserver}"
+            )
+        else:
+            raise Exception(
+                f"Error getting URL {url}. Status code: {response.status_code}"
+            )
+    except Exception as ex:
+        log(f"Error getting posts for user {user['acct']}: {ex}")
+        return None
+
+def get_user_posts_lemmy(userName, userUrl, webserver):
+    # community
+    if re.match(r"^https:\/\/[^\/]+\/c\/", userUrl):
+        try:
+            url = f"https://{webserver}/api/v3/post/list?community_name={userName}&sort=New&limit=50"
             response = get(url)
 
             if(response.status_code == 200):
@@ -126,12 +173,13 @@ def get_user_posts(user, know_followings, server):
                 return posts
 
         except Exception as ex:
-            log(f"Error getting community posts for community {parsed_url[1]}: {ex}")
+            log(f"Error getting community posts for community {userName}: {ex}")
         return None
 
-    if re.match(r"^https:\/\/[^\/]+\/u\/", user['url']):
+    # user
+    if re.match(r"^https:\/\/[^\/]+\/u\/", userUrl):
         try:
-            url = f"https://{parsed_url[0]}/api/v3/user?username={parsed_url[1]}&sort=New&limit=50"
+            url = f"https://{webserver}/api/v3/user?username={userName}&sort=New&limit=50"
             response = get(url)
 
             if(response.status_code == 200):
@@ -143,32 +191,53 @@ def get_user_posts(user, know_followings, server):
                 return all_posts
             
         except Exception as ex:
-            log(f"Error getting user posts for user {parsed_url[1]}: {ex}")
+            log(f"Error getting user posts for user {userName}: {ex}")
         return None
 
+def get_user_posts_misskey(userName, webserver):
+    # query user info via search api
+    # we could filter by host but there's no way to limit that to just the main host on firefish currently
+    # on misskey it works if you supply '.' as the host but firefish does not
+    userId = None
     try:
-        user_id = get_user_id(parsed_url[0], parsed_url[1])
+        url = f'https://{webserver}/api/users/search-by-username-and-host'
-    except Exception as ex:
+        resp = post(url, { 'username': userName })
-        log(f"Error getting user ID for user {user['acct']}: {ex}")
-        return None
 
-    try:
+        if resp.status_code == 200:
-        url = f"https://{parsed_url[0]}/api/v1/accounts/{user_id}/statuses?limit=40"
+            res = resp.json()
-        response = get(url)
+            for user in res:
-
+                if user['host'] is None:
-        if(response.status_code == 200):
+                    userId = user['id']
-            return response.json()
+                    break
-        elif response.status_code == 404:
-            raise Exception(
-                f"User {user['acct']} was not found on server {parsed_url[0]}"
-            )
         else:
-            raise Exception(
+            log(f"Error finding user {userName} from {webserver}. Status Code: {resp.status_code}")
-                f"Error getting URL {url}. Status code: {response.status_code}"
-            )
-    except Exception as ex:
-        log(f"Error getting posts for user {user['acct']}: {ex}")
             return None
+    except Exception as ex:
+        log(f"Error finding user {userName} from {webserver}. Exception: {ex}")
+        return None
+
+    if userId is None:
+        log(f'Error finding user {userName} from {webserver}: user not found on server in search')
+        return None
+
+    try:
+        url = f'https://{webserver}/api/users/notes'
+        resp = post(url, { 'userId': userId, 'limit': 40 })
+
+        if resp.status_code == 200:
+            notes = resp.json()
+            for note in notes:
+                if note.get('url') is None:
+                    # add this to make it look like Mastodon status objects
+                    note.update({ 'url': f"https://{webserver}/notes/{note['id']}" })
+            return notes
+        else:
+            log(f"Error getting posts by user {userName} from {webserver}. Status Code: {resp.status_code}")
+            return None
+    except Exception as ex:
+        log(f"Error getting posts by user {userName} from {webserver}. Exception: {ex}")
+        return None
+    
 
 def get_new_follow_requests(server, access_token, max, known_followings):
     """Get any new follow requests for the specified user, up to the max number provided"""
@@ -388,7 +457,7 @@ def get_reply_toots(user_id, server, access_token, seen_urls, reply_since):
     )
 
 
-def get_all_known_context_urls(server, reply_toots, parsed_urls):
+def get_all_known_context_urls(server, reply_toots, parsed_urls, seen_hosts):
     """get the context toots of the given toots from their original server"""
     known_context_urls = set()
     
@@ -396,7 +465,7 @@ def get_all_known_context_urls(server, reply_toots, parsed_urls):
         if toot_has_parseable_url(toot, parsed_urls):
             url = toot["url"] if toot["reblog"] is None else toot["reblog"]["url"]
             parsed_url = parse_url(url, parsed_urls)
-            context = get_toot_context(parsed_url[0], parsed_url[1], url)
+            context = get_toot_context(parsed_url[0], parsed_url[1], url, seen_hosts)
             if context is not None:
                 for item in context:
                     known_context_urls.add(item)
@@ -504,6 +573,11 @@ def parse_url(url, parsed_urls):
         if match is not None:
             parsed_urls[url] = match
 
+    if url not in parsed_urls:
+        match = parse_misskey_url(url)
+        if match is not None:
+            parsed_urls[url] = match
+
     if url not in parsed_urls:
         log(f"Error parsing toot URL {url}")
         parsed_urls[url] = None
@@ -560,6 +634,15 @@ def parse_pixelfed_url(url):
         return (match.group("server"), match.group("toot_id"))
     return None
 
+def parse_misskey_url(url):
+    """parse a Misskey URL and return the server and ID"""
+    match = re.match(
+        r"https://(?P<server>[^/]+)/notes/(?P<toot_id>[^/]+)", url
+    )
+    if match is not None:
+        return (match.group("server"), match.group("toot_id"))
+    return None
+
 def parse_pixelfed_profile_url(url):
     """parse a Pixelfed Profile URL and return the server and username"""
     match = re.match(r"https://(?P<server>[^/]+)/(?P<username>[^/]+)", url)
@@ -606,24 +689,37 @@ def get_redirect_url(url):
         return None
 
 
-def get_all_context_urls(server, replied_toot_ids):
+def get_all_context_urls(server, replied_toot_ids, seen_hosts):
     """get the URLs of the context toots of the given toots"""
     return filter(
         lambda url: not url.startswith(f"https://{server}/"),
         itertools.chain.from_iterable(
-            get_toot_context(server, toot_id, url)
+            get_toot_context(server, toot_id, url, seen_hosts)
             for (url, (server, toot_id)) in replied_toot_ids
         ),
     )
 
 
-def get_toot_context(server, toot_id, toot_url):
+def get_toot_context(server, toot_id, toot_url, seen_hosts):
     """get the URLs of the context toots of the given toot"""
-    if toot_url.find("/comment/") != -1:
+
-        return get_comment_context(server, toot_id, toot_url)
+    post_server = get_server_info(server, seen_hosts)
-    if toot_url.find("/post/") != -1:
+    if post_server is None:
-        return get_comments_urls(server, toot_id, toot_url)
+        log(f'server {server} not found for post')
-    url = f"https://{server}/api/v1/statuses/{toot_id}/context"
+        return []
+
+    if post_server['mastodonApiSupport']:
+        return get_mastodon_urls(post_server['webserver'], toot_id, toot_url)
+    if post_server['lemmyApiSupport']:
+        return get_lemmy_urls(post_server['webserver'], toot_id, toot_url)
+    if post_server['misskeyApiSupport']:
+        return get_misskey_urls(post_server['webserver'], toot_id, toot_url)
+
+    log(f'unknown server api for {server}')
+    return []
+
+def get_mastodon_urls(webserver, toot_id, toot_url):
+    url = f"https://{webserver}/api/v1/statuses/{toot_id}/context"
     try:
         resp = get(url)
     except Exception as ex:
@@ -642,16 +738,25 @@ def get_toot_context(server, toot_id, toot_url):
         reset = datetime.strptime(resp.headers['x-ratelimit-reset'], '%Y-%m-%dT%H:%M:%S.%fZ')
         log(f"Rate Limit hit when getting context for {toot_url}. Waiting to retry at {resp.headers['x-ratelimit-reset']}")
         time.sleep((reset - datetime.now()).total_seconds() + 1)
-        return get_toot_context(server, toot_id, toot_url)
+        return get_mastodon_urls(webserver, toot_id, toot_url)
 
     log(
         f"Error getting context for toot {toot_url}. Status code: {resp.status_code}"
     )
     return []
 
-def get_comment_context(server, toot_id, toot_url):
+def get_lemmy_urls(webserver, toot_id, toot_url):
+    if toot_url.find("/comment/") != -1:
+        return get_lemmy_comment_context(webserver, toot_id, toot_url)
+    if toot_url.find("/post/") != -1:
+        return get_lemmy_comments_urls(webserver, toot_id, toot_url)
+    else:
+        log(f'unknown lemmy url type {toot_url}')
+        return []
+
+def get_lemmy_comment_context(webserver, toot_id, toot_url):
     """get the URLs of the context toots of the given toot"""
-    comment = f"https://{server}/api/v3/comment?id={toot_id}"
+    comment = f"https://{webserver}/api/v3/comment?id={toot_id}"
     try:
         resp = get(comment)
     except Exception as ex:
@@ -662,7 +767,7 @@ def get_comment_context(server, toot_id, toot_url):
         try:
             res = resp.json()
             post_id = res['comment_view']['comment']['post_id']
-            return get_comments_urls(server, post_id, toot_url)
+            return get_lemmy_comments_urls(webserver, post_id, toot_url)
         except Exception as ex:
             log(f"Error parsing context for comment {toot_url}. Exception: {ex}")
         return []
@@ -670,12 +775,12 @@ def get_comment_context(server, toot_id, toot_url):
         reset = datetime.strptime(resp.headers['x-ratelimit-reset'], '%Y-%m-%dT%H:%M:%S.%fZ')
         log(f"Rate Limit hit when getting context for {toot_url}. Waiting to retry at {resp.headers['x-ratelimit-reset']}")
         time.sleep((reset - datetime.now()).total_seconds() + 1)
-        return get_comment_context(server, toot_id, toot_url)
+        return get_lemmy_comment_context(webserver, toot_id, toot_url)
 
-def get_comments_urls(server, post_id, toot_url):
+def get_lemmy_comments_urls(webserver, post_id, toot_url):
     """get the URLs of the comments of the given post"""
     urls = []
-    url = f"https://{server}/api/v3/post?id={post_id}"
+    url = f"https://{webserver}/api/v3/post?id={post_id}"
     try:
         resp = get(url)
     except Exception as ex:
@@ -691,7 +796,7 @@ def get_comments_urls(server, post_id, toot_url):
         except Exception as ex:
             log(f"Error parsing post {post_id} from {toot_url}. Exception: {ex}")
 
-    url = f"https://{server}/api/v3/comment/list?post_id={post_id}&sort=New&limit=50"
+    url = f"https://{webserver}/api/v3/comment/list?post_id={post_id}&sort=New&limit=50"
     try:
         resp = get(url)
     except Exception as ex:
@@ -711,11 +816,53 @@ def get_comments_urls(server, post_id, toot_url):
         reset = datetime.strptime(resp.headers['x-ratelimit-reset'], '%Y-%m-%dT%H:%M:%S.%fZ')
         log(f"Rate Limit hit when getting comments for {toot_url}. Waiting to retry at {resp.headers['x-ratelimit-reset']}")
         time.sleep((reset - datetime.now()).total_seconds() + 1)
-        return get_comments_urls(server, post_id, toot_url)
+        return get_lemmy_comments_urls(webserver, post_id, toot_url)
 
     log(f"Error getting comments for post {toot_url}. Status code: {resp.status_code}")
     return []
 
+def get_misskey_urls(webserver, post_id, toot_url):
+    """get the URLs of the comments of a given misskey post"""
+
+    urls = []
+    url = f"https://{webserver}/api/notes/children"
+    try:
+        resp = post(url, { 'noteId': post_id, 'limit': 100, 'depth': 12 })
+    except Exception as ex:
+        log(f"Error getting post {post_id} from {toot_url}. Exception: {ex}")
+        return []
+
+    if resp.status_code == 200:
+        try:
+            res = resp.json()
+            log(f"Got children for misskey post {toot_url}")
+            list_of_urls = [f'https://{webserver}/notes/{comment_info["id"]}' for comment_info in res]
+            urls.extend(list_of_urls)
+        except Exception as ex:
+            log(f"Error parsing post {post_id} from {toot_url}. Exception: {ex}")
+    else:
+        log(f"Error getting post {post_id} from {toot_url}. Status Code: {resp.status_code}")
+
+    url = f"https://{webserver}/api/notes/conversation"
+    try:
+        resp = post(url, { 'noteId': post_id, 'limit': 100 })
+    except Exception as ex:
+        log(f"Error getting post {post_id} from {toot_url}. Exception: {ex}")
+        return []
+
+    if resp.status_code == 200:
+        try:
+            res = resp.json()
+            log(f"Got conversation for misskey post {toot_url}")
+            list_of_urls = [f'https://{webserver}/notes/{comment_info["id"]}' for comment_info in res]
+            urls.extend(list_of_urls)
+        except Exception as ex:
+            log(f"Error parsing post {post_id} from {toot_url}. Exception: {ex}")
+    else:
+        log(f"Error getting post {post_id} from {toot_url}. Status Code: {resp.status_code}")
+
+    return urls
+
 def add_context_urls(server, access_token, context_urls, seen_urls):
     """add the given toot URLs to the server"""
     count = 0
@@ -845,9 +992,61 @@ def get(url, headers = {}, timeout = 0, max_tries = 5):
         raise Exception(f"Maximum number of retries exceeded for rate limited request {url}")
     return response
 
+def post(url, json, headers = {}, timeout = 0, max_tries = 5):
+    """A simple wrapper to make a post request while providing our user agent, and respecting rate limits"""
+    h = headers.copy()
+    if 'User-Agent' not in h:
+        h['User-Agent'] = 'FediFetcher (https://go.thms.uk/mgr)'
+
+    if timeout == 0:
+        timeout = arguments.http_timeout
+
+    response = requests.post( url, json=json, headers= h, timeout=timeout)
+    if response.status_code == 429:
+        if max_tries > 0:
+            reset = parser.parse(response.headers['x-ratelimit-reset'])
+            now = datetime.now(datetime.now().astimezone().tzinfo)
+            wait = (reset - now).total_seconds() + 1
+            log(f"Rate Limit hit requesting {url}. Waiting {wait} sec to retry at {response.headers['x-ratelimit-reset']}")
+            time.sleep(wait)
+            return post(url, json, headers, timeout, max_tries - 1)
+
+        raise Exception(f"Maximum number of retries exceeded for rate limited request {url}")
+    return response
+
 def log(text):
     print(f"{datetime.now()} {datetime.now().astimezone().tzinfo}: {text}")
 
+class ServerList:
+    def __init__(self, iterable):
+        self._dict = {}
+        for item in iterable:
+            if('last_checked' in iterable[item]):
+                iterable[item]['last_checked'] = parser.parse(iterable[item]['last_checked'])
+            self.add(item, iterable[item])
+
+    def add(self, key, item):
+        self._dict[key] = item
+
+    def get(self, key):
+        return self._dict[key]
+    
+    def pop(self,key):
+        return self._dict.pop(key)
+    
+    def __contains__(self, item):
+        return item in self._dict
+
+    def __iter__(self):
+        return iter(self._dict)
+
+    def __len__(self):
+        return len(self._dict)
+    
+    def toJSON(self):
+        return json.dumps(self._dict,default=str)
+
+
 class OrderedSet:
     """An ordered set implementation over a dict"""
 
@@ -890,8 +1089,160 @@ class OrderedSet:
         return len(self._dict)
     
     def toJSON(self):
-        return json.dump(self._dict, f, default=str)
+        return json.dumps(self._dict,default=str)
 
+def get_server_from_host_meta(server):
+    url = f'https://{server}/.well-known/host-meta'
+    try:
+        resp = get(url, timeout = 30)
+    except Exception as ex:
+        log(f"Error getting host meta for {server}. Exception: {ex}")
+        return None
+
+    if resp.status_code == 200:
+        try:
+            hostMeta = ET.fromstring(resp.text)
+            lrdd = hostMeta.find('.//{http://docs.oasis-open.org/ns/xri/xrd-1.0}Link[@rel="lrdd"]')
+            url = lrdd.get('template')
+            match = re.match(
+                r"https://(?P<server>[^/]+)/", url
+            )
+            if match is not None:
+                return match.group("server")
+            else:
+                raise Exception(f'server not found in lrdd for {server}')
+                return None
+        except Exception as ex:
+            log(f'Error parsing host meta for {server}. Exception: {ex}')
+            return None
+    else:
+        log(f'Error getting host meta for {server}. Status Code: {resp.status_code}')
+        return None
+
+def get_nodeinfo(server, seen_hosts, host_meta_fallback = False):
+    url = f'https://{server}/.well-known/nodeinfo'
+    try:
+        resp = get(url, timeout = 30)
+    except Exception as ex:
+        log(f"Error getting host node info for {server}. Exception: {ex}")
+        return None
+
+    # if well-known nodeinfo isn't found, try to check host-meta for a webfinger URL
+    # needed on servers where the display domain is different than the web domain
+    if resp.status_code != 200 and not host_meta_fallback:
+        # not found, try to check host-meta as a fallback
+        log(f'nodeinfo for {server} not found, checking host-meta')
+        new_server = get_server_from_host_meta(server)
+        if new_server is not None:
+            if new_server == server:
+                log(f'host-meta for {server} did not get a new server.')
+                return None
+            else:
+                return get_nodeinfo(new_server, seen_hosts, True)
+        else:
+            return None
+
+    if resp.status_code == 200:
+        try:
+            nodeInfo = resp.json()
+            for link in nodeInfo['links']:
+                if link['rel'] in [
+                    'http://nodeinfo.diaspora.software/ns/schema/2.0',
+                    'http://nodeinfo.diaspora.software/ns/schema/2.1',
+                ]:
+                    nodeLoc = link['href']
+                    break
+        except Exception as ex:
+            log(f'error getting server {server} info from well-known node info. Exception: {ex}')
+            return None
+    else:
+        log(f'Error getting well-known host node info for {server}. Status Code: {resp.status_code}')
+        return None
+
+    if nodeLoc is None:
+        log(f'could not find link to node info in well-known nodeinfo of {server}')
+        return None
+
+    # regrab server from nodeLoc, again in the case of different display and web domains
+    match = re.match(
+        r"https://(?P<server>[^/]+)/", nodeLoc
+    )
+    if match is None:
+        log(f"Error getting web server name from {server}.")
+        return None
+
+    server = match.group('server')
+
+    # return early if the web domain has been seen previously (in cases with host-meta lookups)
+    if server in seen_hosts:
+        return seen_hosts[server]
+
+    try:
+        resp = get(nodeLoc, timeout = 30)
+    except Exception as ex:
+        log(f"Error getting host node info for {server}. Exception: {ex}")
+        return None
+
+    if resp.status_code == 200:
+        try:
+            nodeInfo = resp.json()
+            if 'activitypub' not in nodeInfo['protocols']:
+                log(f'server {server} does not support activitypub, skipping')
+                return None
+            return {
+                'webserver': server,
+                'software': nodeInfo['software']['name'],
+                'version': nodeInfo['software']['version'],
+                'rawnodeinfo': nodeInfo,
+            }
+        except Exception as ex:
+            log(f'error getting server {server} info from nodeinfo. Exception: {ex}')
+            return None
+    else:
+        log(f'Error getting host node info for {server}. Status Code: {resp.status_code}')
+        return None
+
+def get_server_info(server, seen_hosts):
+    if server in seen_hosts:
+        serverInfo = seen_hosts.get(server)
+        if('info' in serverInfo and serverInfo['info'] == None):
+            return None
+        return serverInfo
+
+    nodeinfo = get_nodeinfo(server, seen_hosts)
+    if nodeinfo is None:
+        seen_hosts.add(server, {
+            'info': None,
+            'last_checked': datetime.now()
+        })
+    else:
+        set_server_apis(nodeinfo)
+        seen_hosts.add(server, nodeinfo)
+        if server is not nodeinfo['webserver']:
+            seen_hosts.add(nodeinfo['webserver'], nodeinfo)
+    return nodeinfo
+
+def set_server_apis(server):
+    # support for new server software should be added here
+    software_apis = {
+        'mastodonApiSupport': ['mastodon', 'pleroma', 'akkoma', 'pixelfed', 'hometown'],
+        'misskeyApiSupport': ['misskey', 'calckey', 'firefish', 'foundkey'],
+        'lemmyApiSupport': ['lemmy']
+    }
+
+    # software that has specific API support but is not compatible with FediFetcher for various reasons:
+    # * gotosocial - All Mastodon APIs require access token (https://github.com/superseriousbusiness/gotosocial/issues/2038)
+
+    for api, softwareList in software_apis.items():
+        server[api] = server['software'] in softwareList
+
+    # search `features` list in metadata if available
+    if 'metadata' in server['rawnodeinfo'] and 'features' in server['rawnodeinfo']['metadata'] and type(server['rawnodeinfo']['metadata']['features']) is list:
+        features = server['rawnodeinfo']['metadata']['features']
+        if 'mastodon_api' in features:
+            server['mastodonApiSupport'] = True
+
+    server['last_checked'] = datetime.now()
 
 if __name__ == "__main__":
     start = datetime.now()
@@ -978,6 +1329,7 @@ if __name__ == "__main__":
         REPLIED_TOOT_SERVER_IDS_FILE = os.path.join(arguments.state_dir, "replied_toot_server_ids")
         KNOWN_FOLLOWINGS_FILE = os.path.join(arguments.state_dir, "known_followings")
         RECENTLY_CHECKED_USERS_FILE = os.path.join(arguments.state_dir, "recently_checked_users")
+        SEEN_HOSTS_FILE = os.path.join(arguments.state_dir, "seen_hosts")
 
 
         seen_urls = OrderedSet([])
@@ -1011,6 +1363,22 @@ if __name__ == "__main__":
 
         all_known_users = OrderedSet(list(known_followings) + list(recently_checked_users))
 
+        if os.path.exists(SEEN_HOSTS_FILE):
+            with open(SEEN_HOSTS_FILE, "r", encoding="utf-8") as f:
+                seen_hosts = ServerList(json.load(f))
+
+            for host in list(seen_hosts):
+                serverInfo = seen_hosts.get(host)
+                if 'last_checked' in serverInfo:
+                    serverAge = datetime.now(serverInfo['last_checked'].tzinfo) - serverInfo['last_checked']
+                    if(serverAge.total_seconds() > arguments.remember_hosts_for_days * 24 * 60 * 60 ):
+                        seen_hosts.pop(host)
+                    elif('info' in serverInfo and serverInfo['info'] == None and serverAge.total_seconds() > 60 * 60 ):
+                        # Don't cache failures for more than 24 hours
+                        seen_hosts.pop(host)
+        else:
+            seen_hosts = ServerList({})
+
         if(isinstance(arguments.access_token, str)):
             setattr(arguments, 'access_token', [arguments.access_token])
 
@@ -1023,19 +1391,19 @@ if __name__ == "__main__":
                 reply_toots = get_all_reply_toots(
                     arguments.server, user_ids, token, seen_urls, arguments.reply_interval_in_hours
                 )
-                known_context_urls = get_all_known_context_urls(arguments.server, reply_toots,parsed_urls)
+                known_context_urls = get_all_known_context_urls(arguments.server, reply_toots,parsed_urls, seen_hosts)
                 seen_urls.update(known_context_urls)
                 replied_toot_ids = get_all_replied_toot_server_ids(
                     arguments.server, reply_toots, replied_toot_server_ids, parsed_urls
                 )
-                context_urls = get_all_context_urls(arguments.server, replied_toot_ids)
+                context_urls = get_all_context_urls(arguments.server, replied_toot_ids, seen_hosts)
                 add_context_urls(arguments.server, token, context_urls, seen_urls)
 
 
             if arguments.home_timeline_length > 0:
                 """Do the same with any toots on the key owner's home timeline """
                 timeline_toots = get_timeline(arguments.server, token, arguments.home_timeline_length)
-                known_context_urls = get_all_known_context_urls(arguments.server, timeline_toots,parsed_urls)
+                known_context_urls = get_all_known_context_urls(arguments.server, timeline_toots,parsed_urls, seen_hosts)
                 add_context_urls(arguments.server, token, known_context_urls, seen_urls)
 
                 # Backfill any post authors, and any mentioned users
@@ -1057,40 +1425,40 @@ if __name__ == "__main__":
                             if user not in mentioned_users and user['acct'] not in all_known_users:
                                 mentioned_users.append(user)
 
-                    add_user_posts(arguments.server, token, filter_known_users(mentioned_users, all_known_users), recently_checked_users, all_known_users, seen_urls)
+                    add_user_posts(arguments.server, token, filter_known_users(mentioned_users, all_known_users), recently_checked_users, all_known_users, seen_urls, seen_hosts)
 
             if arguments.max_followings > 0:
                 log(f"Getting posts from last {arguments.max_followings} followings")
                 user_id = get_user_id(arguments.server, arguments.user, token)
                 followings = get_new_followings(arguments.server, user_id, arguments.max_followings, all_known_users)
-                add_user_posts(arguments.server, token, followings, known_followings, all_known_users, seen_urls)
+                add_user_posts(arguments.server, token, followings, known_followings, all_known_users, seen_urls, seen_hosts)
             
             if arguments.max_followers > 0:
                 log(f"Getting posts from last {arguments.max_followers} followers")
                 user_id = get_user_id(arguments.server, arguments.user, token)
                 followers = get_new_followers(arguments.server, user_id, arguments.max_followers, all_known_users)
-                add_user_posts(arguments.server, token, followers, recently_checked_users, all_known_users, seen_urls)
+                add_user_posts(arguments.server, token, followers, recently_checked_users, all_known_users, seen_urls, seen_hosts)
 
             if arguments.max_follow_requests > 0:
                 log(f"Getting posts from last {arguments.max_follow_requests} follow requests")
                 follow_requests = get_new_follow_requests(arguments.server, token, arguments.max_follow_requests, all_known_users)
-                add_user_posts(arguments.server, token, follow_requests, recently_checked_users, all_known_users, seen_urls)
+                add_user_posts(arguments.server, token, follow_requests, recently_checked_users, all_known_users, seen_urls, seen_hosts)
 
             if arguments.from_notifications > 0:
                 log(f"Getting notifications for last {arguments.from_notifications} hours")
                 notification_users = get_notification_users(arguments.server, token, all_known_users, arguments.from_notifications)
-                add_user_posts(arguments.server, token, notification_users, recently_checked_users, all_known_users, seen_urls)
+                add_user_posts(arguments.server, token, notification_users, recently_checked_users, all_known_users, seen_urls, seen_hosts)
 
             if arguments.max_bookmarks > 0:
                 log(f"Pulling replies to the last {arguments.max_bookmarks} bookmarks")
                 bookmarks = get_bookmarks(arguments.server, token, arguments.max_bookmarks)
-                known_context_urls = get_all_known_context_urls(arguments.server, bookmarks,parsed_urls)
+                known_context_urls = get_all_known_context_urls(arguments.server, bookmarks,parsed_urls, seen_hosts)
                 add_context_urls(arguments.server, token, known_context_urls, seen_urls)
 
             if arguments.max_favourites > 0:
                 log(f"Pulling replies to the last {arguments.max_favourites} favourites")
                 favourites = get_favourites(arguments.server, token, arguments.max_favourites)
-                known_context_urls = get_all_known_context_urls(arguments.server, favourites,parsed_urls)
+                known_context_urls = get_all_known_context_urls(arguments.server, favourites,parsed_urls, seen_hosts)
                 add_context_urls(arguments.server, token, known_context_urls, seen_urls)
 
         with open(KNOWN_FOLLOWINGS_FILE, "w", encoding="utf-8") as f:
@@ -1103,7 +1471,10 @@ if __name__ == "__main__":
             json.dump(dict(list(replied_toot_server_ids.items())[-10000:]), f)
 
         with open(RECENTLY_CHECKED_USERS_FILE, "w", encoding="utf-8") as f:
-            recently_checked_users.toJSON()
+            f.write(recently_checked_users.toJSON())
+
+        with open(SEEN_HOSTS_FILE, "w", encoding="utf-8") as f:
+            f.write(seen_hosts.toJSON())
 
         os.remove(LOCK_FILE)
 

requirements.txt

@@ -9,3 +9,4 @@ requests==2.28.2
 six==1.16.0
 smmap==5.0.0
 urllib3==1.26.14
+defusedxml==0.7.1