public: move page resolve code to utils (#57672)

combo/data/utils.py

@@ -27,7 +27,11 @@ from django.utils.translation import ugettext_lazy as _
 from combo.apps.assets.models import Asset
 from combo.apps.assets.utils import add_tar_content, clean_assets_files, tar_assets_files, untar_assets_files
 
-from .models import Page
+from .models import Page, extract_context_from_sub_slug
+
+
+class MissingSubSlug(Exception):
+    pass
 
 
 class ImportSiteError(Exception):
@@ -196,3 +200,56 @@ def import_site_tar(fd, if_empty=False, clean=False, overwrite=False, request=No
         data.update(untar_assets_files(tar, overwrite=overwrite))
         pages = import_site(data, if_empty=if_empty, clean=clean, request=request)
     return pages
+
+
+def get_page_from_url_parts(parts, request=None):
+    pages = {}
+    for page in Page.objects.filter(slug__in=parts):
+        if not page.slug in pages:
+            pages[page.slug] = []
+        pages[page.slug].append(page)
+
+    if not pages:
+        return
+
+    i = 0
+    hierarchy_ids = [None]
+    while i < len(parts):
+        slug_pages = pages.get(parts[i])
+        if slug_pages is None or len(slug_pages) == 0:
+            page = None
+            break
+        if len(slug_pages) == 1:
+            page = slug_pages[0]
+        else:
+            # multiple pages with same slugs
+            try:
+                page = [x for x in slug_pages if x.parent_id == hierarchy_ids[-1]][0]
+            except IndexError:
+                page = None
+                break
+        if page.parent_id != hierarchy_ids[-1]:
+            if i == 0:
+                # root page should be at root but maybe the page is a child of
+                # /index/, and as /index/ is silent the page would appear
+                # directly under /; this is not a suggested practice.
+                if page.parent.slug != 'index' and page.parent.parent_id is not None:
+                    page = None
+                    break
+            else:
+                page = None
+                break
+        if page.sub_slug:
+            if parts[i + 1 :] == []:
+                raise MissingSubSlug
+            extra = extract_context_from_sub_slug(page.sub_slug, parts[i + 1])
+            if extra is None:
+                page = None
+                break
+            if request:
+                request.extra_context_data.update(extra)
+            parts = parts[: i + 1] + parts[i + 2 :]  # skip variable component
+        i += 1
+        hierarchy_ids.append(page.id)
+
+    return page

combo/public/views.py

@@ -55,8 +55,8 @@ from combo.data.models import (
     PostException,
     Redirect,
     TextCell,
-    extract_context_from_sub_slug,
 )
+from combo.data.utils import MissingSubSlug, get_page_from_url_parts
 from combo.profile.models import Profile
 from combo.profile.utils import get_user_from_name_id
 
@@ -481,56 +481,16 @@ def page(request):
             request.session['visited'] = True
             return HttpResponseRedirect(settings.COMBO_WELCOME_PAGE_PATH)
 
-    pages = {}
-    for page in Page.objects.filter(slug__in=parts):
-        if not page.slug in pages:
-            pages[page.slug] = []
-        pages[page.slug].append(page)
+    try:
+        page = get_page_from_url_parts(parts, request)
+    except MissingSubSlug:
+        # a sub slug is expected but was not found; redirect to parent
+        # page as a mitigation.
+        return HttpResponseRedirect('..')
 
-    if pages == {} and parts == ['index'] and Page.objects.count() == 0:
+    if page is None and parts == ['index'] and Page.objects.count() == 0:
         return empty_site(request)
 
-    i = 0
-    hierarchy_ids = [None]
-    while i < len(parts):
-        slug_pages = pages.get(parts[i])
-        if slug_pages is None or len(slug_pages) == 0:
-            page = None
-            break
-        if len(slug_pages) == 1:
-            page = slug_pages[0]
-        else:
-            # multiple pages with same slugs
-            try:
-                page = [x for x in slug_pages if x.parent_id == hierarchy_ids[-1]][0]
-            except IndexError:
-                page = None
-                break
-        if page.parent_id != hierarchy_ids[-1]:
-            if i == 0:
-                # root page should be at root but maybe the page is a child of
-                # /index/, and as /index/ is silent the page would appear
-                # directly under /; this is not a suggested practice.
-                if page.parent.slug != 'index' and page.parent.parent_id is not None:
-                    page = None
-                    break
-            else:
-                page = None
-                break
-        if page.sub_slug:
-            if parts[i + 1 :] == []:
-                # a sub slug is expected but was not found; redirect to parent
-                # page as a mitigation.
-                return HttpResponseRedirect('..')
-            extra = extract_context_from_sub_slug(page.sub_slug, parts[i + 1])
-            if extra is None:
-                page = None
-                break
-            request.extra_context_data.update(extra)
-            parts = parts[: i + 1] + parts[i + 2 :]  # skip variable component
-        i += 1
-        hierarchy_ids.append(page.id)
-
     if not url.endswith('/') and settings.APPEND_SLASH:
         # this is useful to allow /login, /manage, and other non-page URLs to
         # work. re.sub is used to replace repeated slashes by single ones,