Btabaska/74 pally ci api setup #2017
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
### Time to review: __1 mins__ ## Changes proposed Needed to upgrade dependencies for the API for grype issue: https://github.com/navapbc/simpler-grants-gov/actions/runs/9180615894/job/25245519194?pr=47 ## Additional information As usual, just ran `poetry update`
## Summary Fixes #9 ### Time to review: __10 mins__ ## Changes proposed Setup a search index to run locally via Docker Updated makefile to automatically initialize the index + added a script to wait for the index to start up before proceeding. Setup a very basic client for connecting to the search index (will be expanded more in subsequent PRs) Basic test / test utils to verify it is working (also will be expanded) ## Context for reviewers This is the first step in getting the search index working locally. This actually gets it running, and the client works, we just aren't doing anything meaningful with it yet besides tests. ## Additional information This doesn't yet create an index that we can use, except in the test. However, if you want to test out a search index, you can go to http://localhost:5601/app/dev_tools#/console (after running `make init`) to run some queries against the (one node) cluster. https://opensearch.org/docs/latest/getting-started/communicate/#sending-requests-in-dev-tools provides some examples of how to create + use indexes that you can follow.
… the index (#44) ## Summary Fixes #12 ### Time to review: __5 mins__ ## Changes proposed Made a new set of v1 endpoints that are basically copy-pastes of the v0.1 opportunity endpoints ## Context for reviewers Some changes I want to make to the schemas wouldn't make sense without the search index (eg. adding the filter counts to the response). As we have no idea what the actual launch of the v0.1 endpoint is going to look like, I don't want to mess with any of that code or try to make a weird hacky approach that needs to account for both the DB implementation and the search index one. Also, I think we've heard that with the launch of the search index, we'll be "officially" launched, so might as well call in v1 at the same time. Other than adjusting the names of a few schemas in v0.1, I left that implementation alone and just copied the boilerplate that I'll fill out in subsequent tickets. ## Additional information The endpoint appears locally:  --------- Co-authored-by: nava-platform-bot <[email protected]>
## Summary Fixes #10 ### Time to review: __10 mins__ ## Changes proposed Setup a script to populate the search index by loading opportunities from the DB, jsonify'ing them, loading them into a new index, and then aliasing that index. Several utilities were created for simplifying working with the OpenSearch client (a wrapper for setting up configuration / patterns) ## Context for reviewers Iterating over the opportunities and doing something with them is a common pattern in several of our scripts, so nothing is really different there. The meaningful implementation is how we handle creating and aliasing the index. In OpenSearch you can give any index an alias (including putting multiple indexes behind the same alias). The approach is pretty simple: * Create an index * Load opportunities into the index * Atomically swap the index backing the `opportunity-index-alias` * Delete the old index if they exist This approach means that our search endpoint just needs to query the alias, and we can keep making new indexes and swapping them out behind the scenes. Because we could remake the index every few minutes, if we ever need to re-configure things like the number of shards, or any other index-creation configuration, we just update that in this script and wait for it to run again. ## Additional information I ran this locally after loading `83250` records, and it took about 61s. You can run this locally yourself by doing: ```sh make init make db-seed-local poetry run flask load-search-data load-opportunity-data ``` If you'd like to see the data, you can test it out on http://localhost:5601/app/dev_tools#/console - here is an example query that filters by the word `research` across a few fields and filters to just forecasted/posted. ```json GET opportunity-index-alias/_search { "size": 25, "from": 0, "query": { "bool": { "must": [ { "simple_query_string": { "query": "research", "default_operator": "AND", "fields": ["agency.keyword^16", "opportunity_title^2", "opportunity_number^12", "summary.summary_description", "opportunity_assistance_listings.assistance_listing_number^10", "opportunity_assistance_listings.program_title^4"] } } ], "filter": [ { "terms": { "opportunity_status": [ "forecasted", "posted" ] } } ] } } } ```
## Summary Fixes #6 ### Time to review: __60 mins__ ## Changes proposed ### Move pages from page to app router: 1. Move all pages to [`[locale]`](https://next-intl-docs.vercel.app/docs/getting-started/app-router/with-i18n-routing#getting-started) folder 2. Add [`generateMetata()`](https://nextjs.org/docs/app/api-reference/functions/generate-metadata#generatemetadata-function) function and [next-intl `getTranslations()`](https://next-intl-docs.vercel.app/docs/environments/metadata-route-handlers#metadata-api) implementation * @rylew1 commented we could remove this from each page. To do that we could use [prop arguments](https://nextjs.org/docs/app/api-reference/functions/generate-metadata#with-segment-props) and update the based on the param. There is also more we can do with the metadata to properly add [app links and twitter cards](https://nextjs.org/docs/app/api-reference/functions/generate-metadata#applinks). TODO: create ticket 4. Replace i18n's `useTranslation` with next-intl's `useTranslations` 5. Remove hard-coded strings that were present b/c we were still b/w i18next and next-intl #### Changes * [Move process page to app](32ba4ee) * [Move research page to app](5b5ad1a) * [Move health page to app](a3e6255) * [Move feature flag page to app](395baed) * [Move search page to app router](1e261e3) * [Move newsletter pages to app router](b509ef8) * [Move home page to app router](de1be98) * [Move home page to app router](74077ae) * [Move 404 page to app router](ccbc956) ### Misc 1. [Delete hello api](5bad6ea) * This was left from the project creation 2. [Add USWDS icon component](0120c7b) * as noted in a slack discussion, when trying to access [one of the icons](https://github.com/trussworks/react-uswds/blob/main/src/components/Icon/Icons.ts) using `<Icon.Search/>` next errors: `You cannot dot into a client module from a server component. You can only pass the imported name through`. I'm not sure why it thinks the Icon component is a client module. [Dan A. suggests](vercel/next.js#51593 (comment)) trussworks should re-export as named exports. I tried importing the SVGs directly from the trussworks library but [svgr requires a custom webpack config](https://react-svgr.com/docs/next/) which is a road I didn't want to go down and [react svg](https://www.npmjs.com/package/react-svg) through an error in the app router 😥 . * I implemented @sawyerh 's [suggestion](0120c7b#diff-dadb35bd2f3f61f2c179f033cd0a2874fc343974236f2fb8613664703c751429), which did not work initially b/c next reported the USWDS icon was corrupt, which was fixed by adding a `viewBox` to the svg element 😮💨 . * [Remove unused WtGIContent](75490f7) ### Layout and component updates * [Move layout and update for app router](af112fd) * [Update global components for the app router](40119e6) ### Remaining next-intl config and removal of * [Move i18n strings for app router](eb3c07c) * [Adds next-intl config and removes i18n](c546571) * [Update tests for app router](3b9b193) * [Removes i18next and next-i18n packages](9d2e08a) * [Update storybook settings for app router](39f115d)
## Summary Fixes #56 ## Changes proposed - Update the date creation to the parsed year/month/day so it creates the `Date` object using the local time - this may prevent any type of rounding - Add some tests to ensure the right dates are showing up on search
…freshes (#67) ## Summary Fixes #58 ### Time to review: __3 mins__ ## Changes proposed Set the `persistAuthorization` OpenAPI config locally to True ## Context for reviewers For local development, we frequently need to go to http://localhost:8080/docs - enter the auth token, and then repeat this process every time we reopen this page or refresh. Having to either copy paste or retype in the auth token is tedious. This flag makes it so it gets preserved in your browsers local storage. We are only enabling this for the local endpoint at the moment as there are possibly security implications we would need to consider non-locally (eg. what if someone is using a public computer).
## Summary Fixes #64 ## Changes proposed - update to next 14.2.3 (https://nextjs.org/blog/next-14-2)
## Summary Fixes #69 ### Time to review: __5 mins__ ## Changes proposed Remove the `BASE_RESPONSE_SCHEMA` configuration. Adjust the healthcheck endpoint to be consistent in defining the schema / responses with the other endpoints. ## Context for reviewers APIFlask allows you to set a `BASE_RESPONSE_SCHEMA` - the idea is that its the shared schema that every API endpoint will return, and they will only differ in the `data` object within. This sounds great on paper as it prevents you from needing to define most of the response schema for many endpoints, but in practice, its clunky to use. If we want to modify anything in the response schema outside of the `data` object, it will affect every single endpoint. This means when we add something like the pagination info to our search endpoint, a pagination object appears on the healthcheck response. As we intend to make these docs something the public will use, this is going to be confusing. There is also a "bug" (unsure if it is as it was an intended change a few months ago in APIFlask/apispec) that the error response objects end up nested within themselves in the examples. For example, currently the error response for the healthcheck endpoint (which can literally only return a 5xx error) has an example response of: ```json { "data": { "data": "string", "errors": [ { "field": "string", "message": "string", "type": "string" } ], "message": "string", "status_code": 0 }, "message": "string", "pagination_info": { "order_by": "id", "page_offset": 1, "page_size": 25, "sort_direction": "ascending", "total_pages": 2, "total_records": 42 }, "status_code": 0, "warnings": [ { "field": "string", "message": "string", "type": "string" } ] } ``` When in reality, the error response it actually returns looks like: ```json { "data": {}, "errors": [], "message": "Service Unavailable", "status_code": 503 } ``` This set of changes works around all of these confusing issues and just requires us to define specific response schemas for each endpoint with some small set of details. I've kept some base schema classes to derive from that we can use in most cases. ## Additional information Before & After Examples in OpenAPI <table> <tr><th>Endpoint</th><th>Before</th><th>After</th><th>Actual Response (before change)</th></tr> <tr><td>GET /health (200) </td> <td> ```json { "data": { "message": "string" }, "message": "string", "pagination_info": { "order_by": "id", "page_offset": 1, "page_size": 25, "sort_direction": "ascending", "total_pages": 2, "total_records": 42 }, "status_code": 0, "warnings": [ { "field": "string", "message": "string", "type": "string" } ] } ``` </td> <td> ```json { "data": null, "message": "Success", "status_code": 200 } ``` </td> <td> ```json { "data": {}, "message": "Service healthy", "pagination_info": null, "status_code": 200, "warnings": [] } ``` </td> </tr> <tr><td>GET /health (503)</td> <td> ```json { "data": { "data": "string", "errors": [ { "field": "string", "message": "string", "type": "string" } ], "message": "string", "status_code": 0 }, "message": "string", "pagination_info": { "order_by": "id", "page_offset": 1, "page_size": 25, "sort_direction": "ascending", "total_pages": 2, "total_records": 42 }, "status_code": 0, "warnings": [ { "field": "string", "message": "string", "type": "string" } ] } ``` </td> <td> ```json { "data": {}, "errors": [], "message": "Error", "status_code": 0 } ``` </td> <td> ```json { "data": {}, "message": "Service unavailable", "pagination_info": null, "status_code": 200, "warnings": [] } ``` </td> </tr> <tr><td>POST /v0.1/opportunities/search (200)</td> <td> ```json { "data": [ {.. Excluding for brevity } ], "message": "string", "pagination_info": { "order_by": "id", "page_offset": 1, "page_size": 25, "sort_direction": "ascending", "total_pages": 2, "total_records": 42 }, "status_code": 0, "warnings": [ { "field": "string", "message": "string", "type": "string" } ] } ``` </td> <td> ```json { "data": [ {.. Excluding for brevity } ], "message": "Success", "pagination_info": { "order_by": "id", "page_offset": 1, "page_size": 25, "sort_direction": "ascending", "total_pages": 2, "total_records": 42 }, "status_code": 200 } ``` </td> <td> ```json { "data": [ {}, {}, {} // excluded for brevity ], "message": "Success", "pagination_info": { "order_by": "opportunity_id", "page_offset": 1, "page_size": 3, "sort_direction": "ascending", "total_pages": 1010, "total_records": 3030 }, "status_code": 200, "warnings": [] } ``` </td> </tr> <tr><td>POST /v0.1/opportunities/search (401 or 422)</td> <td> ```json { "data": { "data": "string", "errors": [ { "field": "string", "message": "string", "type": "string" } ], "message": "string", "status_code": 0 }, "message": "string", "pagination_info": { "order_by": "id", "page_offset": 1, "page_size": 25, "sort_direction": "ascending", "total_pages": 2, "total_records": 42 }, "status_code": 0, "warnings": [ { "field": "string", "message": "string", "type": "string" } ] } ``` </td> <td> ```json { "data": {}, "errors": [], "message": "Error", "status_code": 0 } ``` </td> <td> ```json { "data": {}, "errors": [], "message": "The server could not verify that you are authorized to access the URL requested", "status_code": 401 } ``` </td> </tr> </table> --------- Co-authored-by: nava-platform-bot <[email protected]>
## Summary Fixes #45 ### Time to review: __3 mins__ ## Changes proposed * update `config.py` database url * add function in `cli.py` * updated packages in `poetry.lock` ## Context for reviewers N/A ## Additional information Row created manually in the database alongside a row created via `test_connection` 
## Summary Fixes #88 ### Time to review: __5 mins__ ## Changes proposed Remove the version field from the docker-compose files Adjust the docker commands to use `docker compose` instead of `docker-compose` ## Context for reviewers A recent version of docker removed the need for specifying a version in the compose files - so the field is now obsolete and just gives a warning whenever you run a command: https://docs.docker.com/compose/compose-file/04-version-and-name/ The change in the command we run is actually from 2021 and makes sure we use docker compose v2. Depending on how you've setup your local instance of docker, `docker-compose` may have been aliased to `docker compose` anyways (I actually think if it wasn't, certain things break anyways). This is just using the proper format for certain. ## Additional information I went through running several docker commands and noticed no difference, as this change shouldn't meaningfully change anything, that is to be expected.
btabaska
requested review from
aplybeah,
coilysiren,
ebuwa-evbuoma-fike,
widal001,
AlexanderStephensonUSDS,
jamesbursa,
chouinar and
andycochran
as code owners
github-actions
bot
added
documentation
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary
Fixes #{ISSUE}
Time to review: x mins
Changes proposed
Context for reviewers
Additional information