Search
How to find documents and files - what the search boxes look at, how words are matched, combining text with tags, jumping to an item by its ID, and the keyboard shortcuts. For editors.
On this page
A workspace grows quickly. Search finds a document by a word from its title, summary, tags, file name, folder or text, and finds any document or file by its ID. This page explains what the search boxes do, what they leave out, and how to paste an ID.
Where to search
| Box | Where | What it finds |
|---|---|---|
| Search documents | Top of the left column in the editor | Documents |
| Find a document | The dialog Link to document in the editor | Documents, to link to |
| Search files | The Media page | Images and files |
| Search name, alt text or tag | The Media tab in the right column of a document | Images and files, to put in the text |
There is no single box for everything. Search documents in the editor, and files in the media library.
Search documents
Type in Search documents. After a short pause the tree is replaced by a flat list of matching documents. Each row shows a dot (published or draft), the title, the folder, the summary and the tags. Clear the box, choose Clear search (the X), or press Escape in the list to get the tree back.
The heading of the list says how many documents match: “3 documents”, or “No documents match.” The list shows at most 200, and says “(first 200)” when there are more.
What is searched
A document matches when every word you typed is found in one of these:
- the title
- the summary
- the tags
- the file name and the folder name
- the body text
Search always looks at the draft, the text as you see it in the editor, not at what is published. A document you created a moment ago is found at once.
Important
Search does not cover the Trash. A document or file that you moved to the Trash does not appear in the results, even when its words match. To find a trashed item, search by its ID (see below) or open the Trash and look there (Trash).
How words are matched
- Every word must match.
fleet servicefinds documents that contain both words, in any order. The words do not have to be next to each other. - A word matches inside longer words.
markfindswordmark. This holds for the title, summary, tags, file name and folder. For the body, the match is at the start of a word first, and a match in the middle of a word is added when that finds only a few documents. - Case does not matter.
- Punctuation separates words. Anything that is not a letter or a digit splits the query.
case-studyiscaseandstudy. Accents and non-Latin letters count as letters. - Only the first 8 different words are used.
- There is no spell-checking and no stemming.
servicedoes not findserving. - A query with no letter or digit, such as
%, searches nothing, and the tree stays.
Matches that have the words in their title are listed first. Otherwise the list is in title order. Words that start a word in the title, summary or tag are highlighted. A match in the middle of a word is found but not highlighted.
Combine text and tags
Use Filter by tag under the search box. A document must have all the tags you choose, and match the text. See Folders and tags.
Lists of things to fix
The home page has links that open the document list as a work queue: documents with unpublished changes, without a summary, or with broken links. A banner above the list names the queue (for example “Only documents without a summary”) and shows your progress. A document that you fix stays in the list, marked with a check mark, so that you do not lose your place. Choose Show all documents (the X in the banner) to leave the queue.
A document you are editing also stays in the results when your edit makes it stop matching. It says “No longer matches the search” until you search again.
Search by ID
Every document and every file has an ID, a 26-character code such as 01M5YPMJRG2SN2C75GJWV9DAWF. Developers, and the AI tools that build sites, refer to content by ID. When they report a problem, they give you the ID. Paste it into the search box and you land on the item.
The search box understands these forms:
| You paste | Example |
|---|---|
| The whole ID | 01M5YPMJRG2SN2C75GJWV9DAWF |
| The start of an ID, 8 to 25 characters | 01M5YPMJ |
| A media link | asset://01M5YPMJRG2SN2C75GJWV9DAWF |
| A document link | lake://01M5YPMJRG2SN2C75GJWV9DAWF |
| A file path in the lake | assets/01M5YPMJRG2SN2C75GJWV9DAWF.json or docs/01M5YPMJRG2SN2C75GJWV9DAWF.md |
| Any of these in quotes or backticks | `01M5YPMJRG2SN2C75GJWV9DAWF` |
Upper or lower case both work. The query must be only the ID, with nothing else around it. If there is other text, the search treats it as ordinary words.
How it behaves:
- In Search documents, the document with that ID is listed first. For a partial ID, up to 20 documents whose ID starts that way come first, then the ordinary matches.
- Press Enter in the search box with a whole ID and the document opens. If the result has not arrived yet, it opens as soon as it does.
- On the Media page, Search files does the same for files. A file found by its ID is shown wherever it is in the workspace, whatever folder is open and whatever filters are on.
- If the ID belongs to a document in the trash, the list says “This document is in the trash.” and offers Open the trash. For a file, the Media page says “This file is in the trash” and offers Open it in the trash. Items from another workspace are never found.
The same rules are used by the CMS and by the tools, so an ID copied from a lake file path works.
Copy an ID
- In a document, the ID is shown below the body as
IDfollowed by the code. Choose the copy button (Copy ID) next to it. The button changes to a check mark and a screen reader says “ID copied”. - For a file, the ID is shown in the details pane with the same copy button.
- The code is also plain text, so you can select it if copying is not allowed in your browser.
Keyboard
| Keys | What they do |
|---|---|
| Ctrl+K, or Cmd+K on a Mac | Focus Search documents from anywhere in the editor and select its text. |
| Arrow down in the box | Move into the results. |
| Arrow keys in the list | Move between results. |
| Enter in the box | Open the first result. With a whole ID, wait for it and open that document. |
| Enter on a result | Open it. |
| Escape in the results | Clear the search and return to the box. |
If something goes wrong
| What you see | What to do |
|---|---|
| “No documents match.” | Check spelling, use fewer words, and remove tag filters. Every word must match, so one wrong word hides the document. |
| A document you know exists is missing | It may be in the Trash. Search by its ID to find out. Or it is in another workspace. |
| “Could not search the documents.” | Your connection or the CMS had a problem. Try again in a moment. |
| An ID does nothing | It has fewer than 8 characters, it has letters that cannot be in an ID (I, L, O, U), or it does not start with a digit from 0 to 7. It is then treated as an ordinary word. |
| The text you changed a minute ago is not found | Wait a second after you stop typing, so the draft is saved. Then search again. |
What’s next
- Keep content tidy so people can find it: Folders and tags.
- Let tools refer to your content by ID: Writing documents.
- Bring a lot of content in at once: Importing content.
