]> git.mxchange.org Git - fba.git/blobdiff - fetch_blocks.py
Continued:
[fba.git] / fetch_blocks.py
index 4b4aba4dc1c19a09323e49dffadea5aa8b1ec6e4..bbccb931f121fc76052c7b01af8a09aef603f17d 100644 (file)
@@ -1,3 +1,19 @@
+# Fedi API Block - An aggregator for fetching blocking data from fediverse nodes
+# Copyright (C) 2023 Free Software Foundation
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as published
+# by the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU Affero General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 import reqto
 import time
 import bs4
@@ -6,11 +22,13 @@ import re
 import fba
 
 fba.cursor.execute(
-    "SELECT domain, software FROM instances WHERE software IN ('pleroma', 'mastodon', 'friendica', 'misskey', 'gotosocial', 'bookwyrm', 'takahe') AND (last_blocked IS NULL OR last_blocked < ?) ORDER BY rowid DESC", [time.time() - fba.config["recheck_block"]]
+    "SELECT domain, software, origin, nodeinfo_url FROM instances WHERE software IN ('pleroma', 'mastodon', 'friendica', 'misskey', 'gotosocial', 'bookwyrm', 'takahe') AND (last_blocked IS NULL OR last_blocked < ?) ORDER BY rowid DESC", [time.time() - fba.config["recheck_block"]]
 )
 
-for blocker, software in fba.cursor.fetchall():
-    # NOISY-DEBUG: print("DEBUG: BEFORE blocker,software:", blocker, software)
+rows = fba.cursor.fetchall()
+print(f"INFO: Checking {len(rows)} entries ...")
+for blocker, software, origin, nodeinfo_url in rows:
+    # NOISY-DEBUG: print("DEBUG: BEFORE blocker,software,origin,nodeinfo_url:", blocker, software, origin, nodeinfo_url)
     blockdict = []
     blocker = fba.tidyup(blocker)
     # NOISY-DEBUG: print("DEBUG: AFTER blocker,software:", blocker, software)
@@ -29,7 +47,7 @@ for blocker, software in fba.cursor.fetchall():
         print("INFO: blocker:", blocker)
         try:
             # Blocks
-            json = fba.fetch_nodeinfo(blocker)
+            json = fba.fetch_nodeinfo(blocker, nodeinfo_url)
             if json is None:
                 print("WARNING: Could not fetch nodeinfo from blocker:", blocker)
                 continue
@@ -68,25 +86,22 @@ for blocker, software in fba.cursor.fetchall():
                         if blocked.count("*") > 1:
                             # -ACK!-oma also started obscuring domains without hash
                             fba.cursor.execute(
-                                "SELECT domain FROM instances WHERE domain LIKE ? ORDER BY rowid LIMIT 1", [blocked.replace("*", "_")]
+                                "SELECT domain, nodeinfo_url FROM instances WHERE domain LIKE ? ORDER BY rowid LIMIT 1", [blocked.replace("*", "_")]
                             )
                             searchres = fba.cursor.fetchone()
                             # NOISY-DEBUG: print("DEBUG: searchres[]:", type(searchres))
                             if searchres != None:
                                 blocked = searchres[0]
+                                nodeinfo_url = searchres[1]
                                 # NOISY-DEBUG: print("DEBUG: Looked up domain:", blocked)
 
                         # NOISY-DEBUG: print("DEBUG: Looking up instance by domain:", blocked)
-                        fba.cursor.execute(
-                            "SELECT domain FROM instances WHERE domain = ? LIMIT 1", [blocked]
-                        )
-
-                        if fba.cursor.fetchone() == None:
+                        if not fba.is_instance_registered(blocked):
                             # NOISY-DEBUG: print("DEBUG: Domain wasn't found, adding:", blocked, blocker)
-                            fba.add_instance(blocked, blocker, argv[0])
+                            fba.add_instance(blocked, blocker, origin, nodeinfo_url)
 
                         fba.cursor.execute(
-                            "SELECT * FROM blocks WHERE blocker = ? AND blocked = ? AND block_level = ?",
+                            "SELECT * FROM blocks WHERE blocker = ? AND blocked = ? AND block_level = ? LIMIT 1",
                             (
                                blocker,
                                blocked,
@@ -156,7 +171,7 @@ for blocker, software in fba.cursor.fetchall():
 
             fba.connection.commit()
         except Exception as e:
-            print("error:", e, blocker, software)
+            print(f"ERROR: blocker='{blocker}',software='{software}',exception='{str(e)}'")
     elif software == "mastodon":
         print("INFO: blocker:", blocker)
         try:
@@ -178,10 +193,10 @@ for blocker, software in fba.cursor.fetchall():
                 try:
                     csrf = meta.find("meta", attrs={"name": "csrf-token"})["content"]
                     # NOISY-DEBUG: print("DEBUG: Adding CSRF token:", blocker, csrf)
-                    reqheaders = {**fba.headers, **{"x-csrf-token": csrf}}
+                    reqheaders = {**fba.api_headers, **{"X-CSRF-Token": csrf}}
                 except:
                     # NOISY-DEBUG: print("DEBUG: No CSRF token found, using normal headers:", blocker)
-                    reqheaders = fba.headers
+                    reqheaders = fba.api_headers
 
                 # NOISY-DEBUG: print("DEBUG: Quering API domain_blocks:", blocker)
                 blocks = reqto.get(f"https://{blocker}/api/v1/instance/domain_blocks", headers=reqheaders, timeout=(fba.config["connection_timeout"], config["read_timeout"])).json()
@@ -235,7 +250,7 @@ for blocker, software in fba.cursor.fetchall():
 
                         if fba.cursor.fetchone() == None:
                             # NOISY-DEBUG: print("DEBUG: Hash wasn't found, adding:", blocked, blocker)
-                            fba.add_instance(blocked, blocker, argv[0])
+                            fba.add_instance(blocked, blocker, origin)
                     else:
                         # Doing the hash search for instance names as well to tidy up DB
                         fba.cursor.execute(
@@ -248,7 +263,7 @@ for blocker, software in fba.cursor.fetchall():
                             blocked = searchres[0]
 
                     fba.cursor.execute(
-                        "SELECT * FROM blocks WHERE blocker = ? AND blocked = ? AND block_level = ?",
+                        "SELECT * FROM blocks WHERE blocker = ? AND blocked = ? AND block_level = ? LIMIT 1",
                         (
                             blocker,
                             blocked if blocked.count("*") <= 1 else blocked_hash,
@@ -268,13 +283,13 @@ for blocker, software in fba.cursor.fetchall():
                     else:
                         fba.update_last_seen(blocker, blocked if blocked.count("*") <= 1 else blocked_hash, block_level)
 
-                    if reason != '':
+                    if reason != "":
                         # NOISY-DEBUG: print("DEBUG: Updating block reason:", blocker, blocked, reason)
                         fba.update_block_reason(reason, blocker, blocked if blocked.count("*") <= 1 else blocked_hash, block_level)
 
             fba.connection.commit()
         except Exception as e:
-            print("error:", e, blocker, software)
+            print(f"ERROR: blocker='{blocker}',software='{software}',exception='{str(e)}'")
     elif software == "friendica" or software == "misskey" or software == "bookwyrm" or software == "takahe":
         print("INFO: blocker:", blocker)
         try:
@@ -318,25 +333,24 @@ for blocker, software in fba.cursor.fetchall():
                     if blocked.count("?") > 0:
                         # Some obscure them with question marks, not sure if that's dependent on version or not
                         fba.cursor.execute(
-                            "SELECT domain FROM instances WHERE domain LIKE ? ORDER BY rowid LIMIT 1", [blocked.replace("?", "_")]
+                            "SELECT domain, origin, nodeinfo_url FROM instances WHERE domain LIKE ? ORDER BY rowid LIMIT 1", [blocked.replace("?", "_")]
                         )
                         searchres = fba.cursor.fetchone()
                         if searchres != None:
                             blocked = searchres[0]
+                            origin = searchres[1]
+                            nodeinfo_url = searchres[2]
 
                     # NOISY-DEBUG: print("DEBUG: AFTER-blocked:", blocked)
-                    fba.cursor.execute(
-                        "SELECT domain FROM instances WHERE domain = ? LIMIT 1", [blocked]
-                    )
-
-                    if fba.cursor.fetchone() == None:
+                    if not fba.is_instance_registered(blocked):
                         # NOISY-DEBUG: print("DEBUG: Hash wasn't found, adding:", blocked, blocker)
-                        fba.add_instance(blocked, blocker)
+                        fba.add_instance(blocked, blocker, origin, nodeinfo_url)
 
                     fba.cursor.execute(
                         "SELECT * FROM blocks WHERE blocker = ? AND blocked = ?",
                         (blocker, blocked),
                     )
+
                     if fba.cursor.fetchone() == None:
                         fba.block_instance(blocker, blocked, reason, block_level)
 
@@ -355,12 +369,12 @@ for blocker, software in fba.cursor.fetchall():
 
             fba.connection.commit()
         except Exception as e:
-            print("error:", e, blocker, software)
+            print(f"ERROR: blocker='{blocker}',software='{software}',exception='{str(e)}'")
     elif software == "gotosocial":
         print("INFO: blocker:", blocker)
         try:
             # Blocks
-            federation = reqto.get(f"https://{blocker}/api/v1/instance/peers?filter=suspended", headers=fba.headers, timeout=(fba.config["connection_timeout"], config["read_timeout"])).json()
+            federation = reqto.get(f"https://{blocker}{get_peers_url}?filter=suspended", headers=fba.api_headers, timeout=(fba.config["connection_timeout"], config["read_timeout"])).json()
 
             if (federation == None):
                 print("WARNING: No valid response:", blocker);
@@ -380,23 +394,21 @@ for blocker, software in fba.cursor.fetchall():
                     elif blocked.count("*") > 0:
                         # GTS does not have hashes for obscured domains, so we have to guess it
                         fba.cursor.execute(
-                            "SELECT domain FROM instances WHERE domain LIKE ? ORDER BY rowid LIMIT 1", [blocked.replace("*", "_")]
+                            "SELECT domain, origin, nodeinfo_url FROM instances WHERE domain LIKE ? ORDER BY rowid LIMIT 1", [blocked.replace("*", "_")]
                         )
                         searchres = fba.cursor.fetchone()
 
                         if searchres != None:
                             blocked = searchres[0]
+                            origin = searchres[1]
+                            nodeinfo_url = searchres[2]
 
-                    fba.cursor.execute(
-                        "SELECT domain FROM instances WHERE domain = ? LIMIT 1", [blocked]
-                    )
-
-                    if fba.cursor.fetchone() == None:
-                        # NOISY-DEBUG: print("DEBUG: Updating block reason:", blocker, blocked, reason)
-                        fba.update_block_reason(reason, blocker, blocked, block_level)
+                    if not fba.is_instance_registered(blocked):
+                        # NOISY-DEBUG: print("DEBUG: Domain wasn't found, adding:", blocked, blocker)
+                        fba.add_instance(blocked, blocker, origin, nodeinfo_url)
 
                     fba.cursor.execute(
-                        "SELECT * FROM blocks WHERE blocker = ? AND blocked = ? AND block_level = ?",
+                        "SELECT * FROM blocks WHERE blocker = ? AND blocked = ? AND block_level = ? LIMIT 1",
                         (
                             blocker,
                             blocked,
@@ -405,6 +417,7 @@ for blocker, software in fba.cursor.fetchall():
                     )
 
                     if fba.cursor.fetchone() == None:
+                        # NOISY-DEBUG: print(f"DEBUG: blocker='{blocker}' is blocking '{blocked}' for unknown reason at this point")
                         fba.block_instance(blocker, blocked, "unknown", "reject")
 
                         blockdict.append(
@@ -421,11 +434,12 @@ for blocker, software in fba.cursor.fetchall():
 
                         for entry in blockdict:
                             if entry["blocked"] == blocked:
+                                # NOISY-DEBUG: print(f"DEBUG: Setting block reason for blocked='{blocked}':'{peer['public_comment']}'")
                                 entry["reason"] = peer["public_comment"]
 
                 fba.connection.commit()
         except Exception as e:
-            print("error:", e, blocker, software)
+            print(f"ERROR: blocker='{blocker}',software='{software}',exception='{str(e)}'")
     else:
         print("WARNING: Unknown software:", blocker, software)