Leveraging Coveo Atomic Framework for Request and Response Manipulation
PreProcessRequestMiddleware: The PreProcessRequestMiddleware provides direct access to requests before responses are received and pages are loaded. This allows for various manipulations, such as modifying request parameters or intercepting actions before execution. For instance, you can log query information for debugging purposes and perform conditional actions based on the query content. Example: preprocessRequest: (request, clientOrigin, metadata) => { const body = JSON.parse(request.body); request.body = JSON.stringify(body); // Reassign the modified body back to request // Execute the first search after preprocessing searchInterface.executeFirstSearch(); return request; // Return the modified request } preprocessSearchResponseMiddleware: The preprocessSearchResponseMiddleware operates after pages have loaded, providing access to search responses. Here, you can manipulate aspects such as total count and relevant data to meet specific requirements. This middleware allows for additional processing of search responses if necessary. Example: preprocessSearchResponseMiddleware: (response) => { // Additional processing of search response if needed return response; // Return the modified response } searchInterface.executeFirstSearch(); }(); Reference Link:https://docs.coveo.com/en/atomic/latest/usage/atomic-modify-requests-responses/#modify-responses By leveraging these middleware options, developers can effectively customize search experiences within the Coveo Atomic framework to align with unique business needs and user expectations.Switching Between List and Grid View in Coveo Atomic Search
This article describes how to create a custom component to switch between list and grid view for your Coveo Atomic search results. Coveo doesn't offer a built-in toggle component to switch between list and grid layouts. However, you can achieve this functionality using JavaScript (JS) and Cascading Style Sheets (CSS). Here's a step-by-step guide:Define Result Templates: Create two separate result list templates, one for grid view and another for list view. Assign Class Names: Assign appropriate class names to your <atomic-result-list> components. For example, use "list-view" for the list view template and "grid-view" for the grid view template. Create a Custom Button Component: Develop a custom component that displays buttons for switching between list and grid views. Implement Button Functionality: Use the onclick attribute for each button to trigger a JavaScript function when clicked. Inside the function, target the relevant <atomic-result-list> element using document.querySelector with the class name you assigned earlier (e.g., list_view = document.querySelector(".list-view")). Show/Hide Views with CSS:When the "List" button is clicked, set the display style of the list_view element to "block" (visible) and grid_view element to "none" (hidden) using: JavaScript list_view.style.display = "block"; grid_view.style.display = "none"; Implement similar logic for the grid button, showing the grid view and hiding the list view. Set Default View:Include logic to display the desired default view (list or grid) on initial page load. Benefits: Users can switch between list and grid layouts based on their preference. Improved user experience by providing more control over the search results display.Filter results on a specific date range with dates converted to Unix timestamps.
Unfortunately the Vertex Search filtering functionality is pretty limited at the moment. Here is the documentation for how to implement filtering using the API. You'll notice there is no built-in support for date filtering currently, but there is support for filtering on numerical fields. Workaround: One workaround I can think of right now is to ingest documents with their dates converted to Unix timestamps (integers). In the front-end of the application you could provide a normal date picker to provide a date range, and then in the backend convert this to unix time and perform numerical filtering. Example: "Jan 12 2023 to Jan 12 2024" would be equivalent to a unix time filter of between 1673481600 and 1705017600. In Vertex Search this would be applied as a filter in the syntax: date_field_name: IN (1673481600, 1705017600)Impersonating Tokens with Their Respective Service Accounts in GCP
What is Impersonation? Impersonation allows one service account to assume the identity of another, enabling it to perform actions on behalf of the impersonated account. This is particularly useful when you need to segregate duties and manage permissions more granularly. How to Print an Access Token Using Impersonation 1. Login to your google account using this command: google auth login 2. To generate an access token for a service account using impersonation, you can use the gcloud command-line tool(google cloud sdk shell). Here’s the command you need: gcloud auth print-access-token --impersonate-service-account=YOUR_SERVICE_ACCOUNT Replace YOUR_SERVICE_ACCOUNT with the email address of the service account you wish to impersonate. Prerequisites for Generating the Token Before you can generate the access token through impersonation, certain prerequisites must be met: 1. Principals and Roles To enable impersonation, you need to assign specific roles to the service account. Here’s a step-by-step guide: Enable the Service Account Token Creator Role: At the owner level, you must assign the Service Account Token Creator role to the service account that will perform the impersonation. This role allows the service account to generate access tokens. You can do this through the GCP Console or using the gcloud command-line tool. Using GCP Console: Navigate to the IAM & Admin section. Select the appropriate service account. Assign the Service Account Token Creator role. Here is the documentation for better understanding : Use service account impersonation Conclusion By following these steps, you can generate access tokens for specific service accounts using impersonation. Properly configuring roles and permissions ensures that only authorized accounts can perform actions, maintaining the integrity and security of your cloud resources.
Knowledge Base
Learn how leveraging Enterprise Search to suggest the most relevant content can help your business and your users!4 Posts
