Merge pull request #3350 from Minnozz/custom-port

Correctly handle serving BookWyrm on custom port

.env.example

@@ -16,6 +16,11 @@ DEFAULT_LANGUAGE="English"
 ## Leave unset to allow all hosts
 # ALLOWED_HOSTS="localhost,127.0.0.1,[::1]"
 
+# Specify when the site is served from a port that is not the default
+# for the protocol (80 for HTTP or 443 for HTTPS).
+# Probably only necessary in development.
+# PORT=1333
+
 MEDIA_ROOT=images/
 
 # Database configuration
@@ -139,9 +144,9 @@ HTTP_X_FORWARDED_PROTO=false
 TWO_FACTOR_LOGIN_VALIDITY_WINDOW=2
 TWO_FACTOR_LOGIN_MAX_SECONDS=60
 
-# Additional hosts to allow in the Content-Security-Policy, "self" (should be DOMAIN)
-# and AWS_S3_CUSTOM_DOMAIN (if used) are added by default.
-# Value should be a comma-separated list of host names.
+# Additional hosts to allow in the Content-Security-Policy, "self" (should be
+# DOMAIN with optionally ":" + PORT) and AWS_S3_CUSTOM_DOMAIN (if used) are
+# added by default.  Value should be a comma-separated list of host names.
 CSP_ADDITIONAL_HOSTS=
 
 # Time before being logged out (in seconds)

bookwyrm/connectors/connector_manager.py

@@ -118,20 +118,22 @@ def get_connectors() -> Iterator[abstract_connector.AbstractConnector]:
 def get_or_create_connector(remote_id: str) -> abstract_connector.AbstractConnector:
     """get the connector related to the object's server"""
     url = urlparse(remote_id)
-    identifier = url.netloc
+    identifier = url.hostname
     if not identifier:
-        raise ValueError("Invalid remote id")
+        raise ValueError(f"Invalid remote id: {remote_id}")
+
+    base_url = f"{url.scheme}://{url.netloc}"
 
     try:
         connector_info = models.Connector.objects.get(identifier=identifier)
     except models.Connector.DoesNotExist:
         connector_info = models.Connector.objects.create(
             identifier=identifier,
             connector_file="bookwyrm_connector",
-            base_url=f"https://{identifier}",
-            books_url=f"https://{identifier}/book",
-            covers_url=f"https://{identifier}/images/covers",
-            search_url=f"https://{identifier}/search?q=",
+            base_url=base_url,
+            books_url=f"{base_url}/book",
+            covers_url=f"{base_url}/images/covers",
+            search_url=f"{base_url}/search?q=",
             priority=2,
         )
 
@@ -188,8 +190,11 @@ def raise_not_valid_url(url: str) -> None:
     if not parsed.scheme in ["http", "https"]:
         raise ConnectorException("Invalid scheme: ", url)
 
+    if not parsed.hostname:
+        raise ConnectorException("Hostname missing: ", url)
+
     try:
-        ipaddress.ip_address(parsed.netloc)
+        ipaddress.ip_address(parsed.hostname)
         raise ConnectorException("Provided url is an IP address: ", url)
     except ValueError:
         # it's not an IP address, which is good

bookwyrm/emailing.py

@@ -4,7 +4,7 @@
 
 from bookwyrm import models, settings
 from bookwyrm.tasks import app, EMAIL
-from bookwyrm.settings import DOMAIN
+from bookwyrm.settings import DOMAIN, BASE_URL
 
 
 def email_data():
@@ -14,6 +14,7 @@ def email_data():
         "site_name": site.name,
         "logo": site.logo_small_url,
         "domain": DOMAIN,
+        "base_url": BASE_URL,
         "user": None,
     }
 

bookwyrm/forms/links.py

@@ -26,7 +26,7 @@ def clean(self):
         url = cleaned_data.get("url")
         filetype = cleaned_data.get("filetype")
         book = cleaned_data.get("book")
-        domain = urlparse(url).netloc
+        domain = urlparse(url).hostname
         if models.LinkDomain.objects.filter(domain=domain).exists():
             status = models.LinkDomain.objects.get(domain=domain).status
             if status == "blocked":

bookwyrm/models/author.py

@@ -8,7 +8,7 @@
 import pgtrigger
 
 from bookwyrm import activitypub
-from bookwyrm.settings import DOMAIN
+from bookwyrm.settings import BASE_URL
 from bookwyrm.utils.db import format_trigger
 
 from .book import BookDataModel, MergedAuthor
@@ -70,7 +70,7 @@ def isfdb_link(self):
 
     def get_remote_id(self):
         """editions and works both use "book" instead of model_name"""
-        return f"https://{DOMAIN}/author/{self.id}"
+        return f"{BASE_URL}/author/{self.id}"
 
     class Meta:
         """sets up indexes and triggers"""

bookwyrm/models/base_model.py

@@ -10,7 +10,7 @@
 from django.utils.translation import gettext_lazy as _
 from django.utils.text import slugify
 
-from bookwyrm.settings import DOMAIN
+from bookwyrm.settings import BASE_URL
 from .fields import RemoteIdField
 
 
@@ -38,7 +38,7 @@ class BookWyrmModel(models.Model):
 
     def get_remote_id(self):
         """generate the url that resolves to the local object, without a slug"""
-        base_path = f"https://{DOMAIN}"
+        base_path = BASE_URL
         if hasattr(self, "user"):
             base_path = f"{base_path}{self.user.local_path}"
 
@@ -53,7 +53,7 @@ class Meta:
     @property
     def local_path(self):
         """how to link to this object in the local app, with a slug"""
-        local = self.get_remote_id().replace(f"https://{DOMAIN}", "")
+        local = self.get_remote_id().replace(BASE_URL, "")
 
         name = None
         if hasattr(self, "name_field"):

bookwyrm/models/book.py

@@ -21,7 +21,7 @@
 from bookwyrm.isbn.isbn import hyphenator_singleton as hyphenator
 from bookwyrm.preview_images import generate_edition_preview_image_task
 from bookwyrm.settings import (
-    DOMAIN,
+    BASE_URL,
     DEFAULT_LANGUAGE,
     LANGUAGE_ARTICLES,
     ENABLE_PREVIEW_IMAGES,
@@ -327,7 +327,7 @@ def save(self, *args: Any, **kwargs: Any) -> None:
 
     def get_remote_id(self):
         """editions and works both use "book" instead of model_name"""
-        return f"https://{DOMAIN}/book/{self.id}"
+        return f"{BASE_URL}/book/{self.id}"
 
     def guess_sort_title(self):
         """Get a best-guess sort title for the current book"""

bookwyrm/models/connector.py

@@ -11,7 +11,7 @@
 class Connector(BookWyrmModel):
     """book data source connectors"""
 
-    identifier = models.CharField(max_length=255, unique=True)
+    identifier = models.CharField(max_length=255, unique=True)  # domain
     priority = models.IntegerField(default=2)
     name = models.CharField(max_length=255, null=True, blank=True)
     connector_file = models.CharField(max_length=255, choices=ConnectorFiles.choices)

bookwyrm/models/federated_server.py

@@ -16,7 +16,7 @@
 class FederatedServer(BookWyrmModel):
     """store which servers we federate with"""
 
-    server_name = models.CharField(max_length=255, unique=True)
+    server_name = models.CharField(max_length=255, unique=True)  # domain
     status = models.CharField(
         max_length=255, default="federated", choices=FederationStatus
     )
@@ -64,5 +64,4 @@ def unblock(self):
     def is_blocked(cls, url: str) -> bool:
         """look up if a domain is blocked"""
         url = urlparse(url)
-        domain = url.netloc
-        return cls.objects.filter(server_name=domain, status="blocked").exists()
+        return cls.objects.filter(server_name=url.hostname, status="blocked").exists()

bookwyrm/models/group.py

@@ -1,7 +1,7 @@
 """ do book related things with other users """
 from django.db import models, IntegrityError, transaction
 from django.db.models import Q
-from bookwyrm.settings import DOMAIN
+from bookwyrm.settings import BASE_URL
 from .base_model import BookWyrmModel
 from . import fields
 from .relationship import UserBlocks
@@ -17,7 +17,7 @@ class Group(BookWyrmModel):
 
     def get_remote_id(self):
         """don't want the user to be in there in this case"""
-        return f"https://{DOMAIN}/group/{self.id}"
+        return f"{BASE_URL}/group/{self.id}"
 
     @classmethod
     def followers_filter(cls, queryset, viewer):

bookwyrm/models/link.py

@@ -38,7 +38,7 @@ def save(self, *args, **kwargs):
         """create a link"""
         # get or create the associated domain
         if not self.domain:
-            domain = urlparse(self.url).netloc
+            domain = urlparse(self.url).hostname
             self.domain, _ = LinkDomain.objects.get_or_create(domain=domain)
 
         # this is never broadcast, the owning model broadcasts an update

bookwyrm/models/list.py

@@ -7,7 +7,7 @@
 from django.utils import timezone
 
 from bookwyrm import activitypub
-from bookwyrm.settings import DOMAIN
+from bookwyrm.settings import BASE_URL
 
 from .activitypub_mixin import CollectionItemMixin, OrderedCollectionMixin
 from .base_model import BookWyrmModel
@@ -50,7 +50,7 @@ class List(OrderedCollectionMixin, BookWyrmModel):
 
     def get_remote_id(self):
         """don't want the user to be in there in this case"""
-        return f"https://{DOMAIN}/list/{self.id}"
+        return f"{BASE_URL}/list/{self.id}"
 
     @property
     def collection_queryset(self):

bookwyrm/models/report.py

@@ -3,7 +3,7 @@
 from django.db import models
 from django.utils.translation import gettext_lazy as _
 
-from bookwyrm.settings import DOMAIN
+from bookwyrm.settings import BASE_URL
 from .base_model import BookWyrmModel
 
 
@@ -46,7 +46,7 @@ def raise_not_editable(self, viewer):
         raise PermissionDenied()
 
     def get_remote_id(self):
-        return f"https://{DOMAIN}/settings/reports/{self.id}"
+        return f"{BASE_URL}/settings/reports/{self.id}"
 
     def comment(self, user, note):
         """comment on a report"""

bookwyrm/models/shelf.py

@@ -6,7 +6,7 @@
 from django.utils import timezone
 
 from bookwyrm import activitypub
-from bookwyrm.settings import DOMAIN
+from bookwyrm.settings import BASE_URL
 from bookwyrm.tasks import BROADCAST
 from .activitypub_mixin import CollectionItemMixin, OrderedCollectionMixin
 from .base_model import BookWyrmModel
@@ -71,7 +71,7 @@ def get_remote_id(self):
     @property
     def local_path(self):
         """No slugs"""
-        return self.get_remote_id().replace(f"https://{DOMAIN}", "")
+        return self.get_remote_id().replace(BASE_URL, "")
 
     def raise_not_deletable(self, viewer):
         """don't let anyone delete a default shelf"""

bookwyrm/models/site.py

@@ -12,7 +12,7 @@
 
 from bookwyrm.connectors.abstract_connector import get_data
 from bookwyrm.preview_images import generate_site_preview_image_task
-from bookwyrm.settings import DOMAIN, ENABLE_PREVIEW_IMAGES, STATIC_FULL_URL
+from bookwyrm.settings import BASE_URL, ENABLE_PREVIEW_IMAGES, STATIC_FULL_URL
 from bookwyrm.settings import RELEASE_API
 from bookwyrm.tasks import app, MISC
 from .base_model import BookWyrmModel, new_access_code
@@ -188,7 +188,7 @@ def valid(self):
     @property
     def link(self):
         """formats the invite link"""
-        return f"https://{DOMAIN}/invite/{self.code}"
+        return f"{BASE_URL}/invite/{self.code}"
 
 
 class InviteRequest(BookWyrmModel):
@@ -235,7 +235,7 @@ def valid(self):
     @property
     def link(self):
         """formats the invite link"""
-        return f"https://{DOMAIN}/password-reset/{self.code}"
+        return f"{BASE_URL}/password-reset/{self.code}"
 
 
 # pylint: disable=unused-argument

bookwyrm/models/user.py

@@ -19,7 +19,7 @@
 from bookwyrm.models.shelf import Shelf
 from bookwyrm.models.status import Status
 from bookwyrm.preview_images import generate_user_preview_image_task
-from bookwyrm.settings import DOMAIN, ENABLE_PREVIEW_IMAGES, USE_HTTPS, LANGUAGES
+from bookwyrm.settings import BASE_URL, ENABLE_PREVIEW_IMAGES, LANGUAGES
 from bookwyrm.signatures import create_key_pair
 from bookwyrm.tasks import app, MISC
 from bookwyrm.utils import regex
@@ -42,12 +42,6 @@ def get_feed_filter_choices():
     return [f[0] for f in FeedFilterChoices]
 
 
-def site_link():
-    """helper for generating links to the site"""
-    protocol = "https" if USE_HTTPS else "http"
-    return f"{protocol}://{DOMAIN}"
-
-
 # pylint: disable=too-many-public-methods
 class User(OrderedCollectionPageMixin, AbstractUser):
     """a user who wants to read books"""
@@ -214,8 +208,7 @@ def active_follower_requests(self):
     @property
     def confirmation_link(self):
         """helper for generating confirmation links"""
-        link = site_link()
-        return f"{link}/confirm-email/{self.confirmation_code}"
+        return f"{BASE_URL}/confirm-email/{self.confirmation_code}"
 
     @property
     def following_link(self):
@@ -349,7 +342,7 @@ def save(self, *args, **kwargs):
         if not self.local and not re.match(regex.FULL_USERNAME, self.username):
             # generate a username that uses the domain (webfinger format)
             actor_parts = urlparse(self.remote_id)
-            self.username = f"{self.username}@{actor_parts.netloc}"
+            self.username = f"{self.username}@{actor_parts.hostname}"
 
         # this user already exists, no need to populate fields
         if not created:
@@ -369,11 +362,10 @@ def save(self, *args, **kwargs):
 
         with transaction.atomic():
             # populate fields for local users
-            link = site_link()
-            self.remote_id = f"{link}/user/{self.localname}"
+            self.remote_id = f"{BASE_URL}/user/{self.localname}"
             self.followers_url = f"{self.remote_id}/followers"
             self.inbox = f"{self.remote_id}/inbox"
-            self.shared_inbox = f"{link}/inbox"
+            self.shared_inbox = f"{BASE_URL}/inbox"
             self.outbox = f"{self.remote_id}/outbox"
 
             # an id needs to be set before we can proceed with related models
@@ -558,7 +550,7 @@ def set_remote_server(user_id, allow_external_connections=False):
     user = User.objects.get(id=user_id)
     actor_parts = urlparse(user.remote_id)
     federated_server = get_or_create_remote_server(
-        actor_parts.netloc, allow_external_connections=allow_external_connections
+        actor_parts.hostname, allow_external_connections=allow_external_connections
     )
     # if we were unable to find the server, we need to create a new entry for it
     if not federated_server: