Repository navigation
Administrator Content and Routing
This chapter explains how content becomes a public URL in Stream Engine. Use it after completing the initial configuration.
Make structural changes on a staging site first. A page-tree error can prevent normal requests from starting, so take a current database backup before changing an established site's routes.
Stream Engine separates content from the URL and navigation that expose it:
flowchart LR
F[Feed<br>content record] --> P[Page action<br>selects and renders content]
P --> R[Route<br>public URL]
M[Menu item<br>optional navigation] --> R
- A feed is a content record. It has a type, title, slug, body, optional parent, visibility, image, and ordering position.
- A page is a node in the public routing tree. Its pattern and parent determine a URL segment, while its action determines which module renders the request.
- A page action is supplied by an enabled module. It defines which feed or other fields are required. The page editor displays and validates that contract.
- A menu item is an optional link. It does not create a route and its access setting is not a security boundary.
Saving a feed alone does not guarantee that it has a public URL. Saving a menu item alone does not create a page. The route must exist, its page action must be configured correctly, and both the page and the selected content must allow the visitor to access them.
Sign in as an administrator and open Administration. The relevant areas are Feeds, Pages, and Menus.
Before a route change:
- record the current page ID, parent, pattern, action, feed fields, access, and menu items that reference it;
- save a database backup;
- test the proposed URL while signed in and in a private browser window;
- plan redirects for any public URL that will move; and
- make one structural change at a time.
The administration interface has no redirect manager. Configure permanent redirects in the reverse proxy or web server before moving a URL that visitors or search engines may already know.
Open Administration → Feeds to find content records. You can filter by type, ID, slug, owner ID, or parent ID. The default view is limited to 100 newest records; use a narrow filter instead of assuming that an absent record does not exist.
| Field | Meaning |
|---|---|
| Title | Human-readable content title. |
| Type | Content contract supplied by a module, such as article or article-section. Choose the type before connecting the feed to a page action. |
| Slug | URL-safe name used by dynamic routes. Keep it stable and unique within the relevant parent and type. Use short lowercase words separated by hyphens. |
| Parent ID | Optional feed hierarchy. This is a feed ID, not a page ID. Dynamic article routes use it to keep content in the same hierarchy as the page route. |
| Visibility | Content-level access: public, members, or private. See Access has three layers. |
| Position | Non-negative ordering value used by parent-based listings; lower values sort first where the listing supports position order. |
| Description | Summary used by supported views and metadata. |
| Image URL | Existing image URL. The upload button beside the content editor inserts an uploaded image into the body instead of filling this field. |
| Content | Sanitized HTML body edited in the source editor. Preview the result in the active theme after saving. |
After creation, the browser address changes to /admin/feeds/ID. Record that
ID, or read it from the Feeds list, when a page action needs a fixed Feed
ID or another feed needs this record as its parent.
Caution
The generic feed editor does not currently provide a delete operation. Do not delete feed rows directly from the database: related content and cached URLs may depend on them. Restrict access while deciding on a supported content-removal workflow.
Open Administration → Pages to see the routing table in hierarchy order.
Page ID 1 is the root page at /. Its pattern and parent cannot be changed,
and it cannot be deleted.
Every other page requires a parent. The public path is built from its ancestor patterns. For example:
| Parent path | Child pattern | Resulting path |
|---|---|---|
/ |
about |
/about/ |
/ |
articles |
/articles/ |
/articles/ |
{slug} |
/articles/{slug}/ |
A static pattern is literal. A dynamic pattern contains a placeholder such as
{slug}. Stream Engine also supports a constrained placeholder such as
{id:\d+}, but administrators should use only a pattern required by the
selected module's documented route. An invalid regular expression can break
request startup.
Static siblings take precedence over a dynamic sibling. If /profile/ and
/{slug}/ share the root parent, /profile/ reaches the static page and an
otherwise unmatched segment can reach the dynamic page.
- Page IDs must be unique.
- Every non-root page must reference an existing parent.
- A page cannot be its own parent or a child of one of its descendants.
- Static sibling patterns must be unique.
- Only one dynamic sibling pattern is allowed under a given parent.
- Page actions marked as singleton must remain unique in the complete runtime page tree. Other actions may be reused by multiple concrete pages.
The editor prevents invalid parent cycles and validates action-specific fields. The current release does not reject every duplicate pattern before writing it to the database. The runtime validates the complete tree on the next request and may fail to start if a conflict was saved. Always search the Pages list for the intended pattern and action before creating a page, and keep the database backup until the site has passed its post-change checks.
Select the action first. The editor then marks Feed ID, Feed type, List feed type, and Term vocabulary as required, optional, fixed, or unused for that action. Clear values reported as unsupported. Do not copy field combinations from a different action.
Common article actions are:
| Action | Use |
|---|---|
article.show-id |
Render one fixed article selected by Feed ID. The installed root page uses this action for the welcome article. |
article.show-slug |
Render an article selected from a {slug} route. Feed type is fixed to article. |
articles.list |
Resolve an article-section from the route and list its child article feeds. |
sections.list |
List article-section children under a fixed root Feed ID. |
Other enabled modules add their own actions. Use the descriptions and field validation shown by the editor; unavailable legacy actions may be preserved on an existing page but cannot be selected for a new one.
Page name supplies a fallback display name. Article content width is available for article display actions. Show comments and Show share buttons affect only views and themes that support those features.
This example creates one reusable dynamic article route, publishes an About article through it, and adds a navigation link. Perform it on a new or staging installation first.
- Open Administration → Pages.
- Find page
#1, the root page. - Record the Feed ID displayed for it. Call this value
ROOT_FEED_IDin the following steps. - Search the page list for action
article.show-slug.
If that action already exists, do not create another page with the same action. Inspect the existing page and use the parent feed and URL hierarchy it defines.
Open Administration → Feeds → New feed and enter:
| Field | Value |
|---|---|
| Title | About |
| Type | article |
| Slug | about |
| Parent ID | ROOT_FEED_ID |
| Visibility | public |
| Position |
0 or the desired order among sibling articles |
| Description | A short public summary |
| Content | The About page body as valid HTML |
Save the feed. Confirm that it now has an ID and remains attached to the
expected parent. The parent matters: article.show-slug resolves the article
inside the feed hierarchy anchored by the nearest ancestor page.
Skip this step if the required article.show-slug page already exists.
Otherwise, open Administration → Pages → New page and enter:
| Field | Value |
|---|---|
| Pattern | {slug} |
| Action | article.show-slug |
| Page name | Article |
| Parent ID |
#1, the root page |
| Feed type |
article (filled or restricted by the action contract) |
| List feed type | empty |
| Term vocabulary | empty |
| Feed ID | empty |
| Change frequency | monthly |
| Access | public |
Choose the desired content width, comments, and share buttons, then save. Do
not enter the About feed ID: this route selects articles by slug and can serve
all public article children of ROOT_FEED_ID.
Open https://example.com/about/, replacing the host with the site's
canonical host. Confirm all of the following:
- the request returns the About article rather than a 404 or another page;
- the title, description, body, and images render correctly;
- the canonical URL uses the expected HTTPS host and path;
- comments and share controls match the page settings; and
- the same URL works in a private browser window.
Also revisit existing top-level URLs. The static routes should still win over
{slug}, but this smoke test detects a misplaced parent or conflicting route.
Open Administration → Menus → New menu item and enter:
| Field | Value |
|---|---|
| Menu group | top |
| Type | external |
| Label | About |
| Parent item | empty for a top-level item, or the desired item in the same group |
| URL | /about/ |
| Access | public |
| Enabled | on |
The relative URL keeps the link on the canonical site. This example uses the
external menu type because the target page is dynamic and the menu editor
cannot bind the fixed about slug to an internal page reference. Save the
item, inspect the menu preview, then open the public site and test the link.
Further articles need only another article feed under ROOT_FEED_ID with a
unique slug. They reuse the same {slug} page. Add a menu item only when that
article belongs in persistent navigation.
Access must be considered separately at each layer:
-
Page access gates the route:
public,authenticated,moderator, oradmin. -
Feed visibility gates the selected content:
public,members, orprivate, with owner and administrator overrides. - Menu access controls whether a navigation item is displayed.
Hiding a menu item does not protect its URL. Set the page access rule and feed visibility first, then give the menu item the same or a stricter audience.
members and private visibility depend on membership in a content container.
For a generic global feed without a container, they do not mean “all signed-in
users” and may leave the item available only to its owner and administrators.
Use public for the worked example. User, moderator, ownership, and community
rules are covered in the next administrator chapter.
Change frequency is a search-engine hint, not a publishing schedule and
not a cache lifetime. Use a realistic value. never still permits a sitemap
entry; it tells crawlers that the URL is not expected to change.
In the current release:
- a public static HTML page is included in
sitemap-pages.xmlunless its value isnoindex; - a public dynamic page with a feed type enables feed sitemap entries for that type;
- dynamic feed sitemap entries use the feed update time and do not include the page's change-frequency value; and
-
noindexon a page controls generated sitemap inclusion only where the sitemap implementation checks it. It does not by itself guarantee arobotsmeta tag or HTTP header.
If exclusion from search engines is mandatory, verify the rendered response and apply an appropriate reverse-proxy header or theme-level robots directive in addition to checking the generated sitemap. Do not rely on the field label alone.
The generated sitemap index is available at /sitemap-index.xml and is
advertised through /robots.txt. After publishing, verify both endpoints and
search the relevant sitemap for the expected canonical URL.
Changing a page pattern or parent changes its URL and the paths of its descendants. Changing a feed slug changes a dynamic URL. Before either change:
- inventory inbound menu links and known external links;
- create the replacement route and web-server redirect plan;
- make the change during a maintenance window;
- test as every intended audience; and
- verify canonical links, menus, and sitemaps.
Feed canonical URLs are cached in Memcached for up to 24 hours. A route or slug change can therefore leave an old canonical URL in rendered content or a sitemap until that entry expires. Do not flush a shared Memcached service indiscriminately; it may affect other applications or Stream Engine sites.
To delete a page, remove or repoint menu items that reference it and delete its child pages first. The root page cannot be deleted. Page deletion is immediate and cannot be undone through the interface, so retain the database backup until verification is complete.
- The intended URL works with and without a trailing slash as expected by the web-server configuration.
- Static sibling URLs still resolve correctly.
- Guests and each required signed-in role receive the intended response.
- Content visibility agrees with page and menu access.
- Canonical URLs use
SITE_URLand HTTPS. - Menus show the correct label and target to the correct audience.
-
/sitemap-index.xmland the relevant child sitemap are valid. - Application and PHP-FPM logs contain no page-tree, route-regex, action, or missing-feed errors.
- The rollback backup and old route details are retained until monitoring shows normal traffic.
Back: Initial configuration · Administrator Guide · Next: Users and permissions