refactor(ui): Loading schema dynamically for dataset profile #7558
Conversation
| }); | ||
| const isHideSiblingMode = useIsSeparateSiblingsMode(); | ||
| // Merge with sibling information as required. | ||
| const combinedData = rawData && !isHideSiblingMode ? combineEntityDataWithSiblings(cloneDeep(rawData)) : rawData; |
There was a problem hiding this comment.
does the cloneDeep here just ensure we don't mess with the original rawData object in this method? is that necessary in the other places we call combineEntityDataWithSiblings or only here because it's cache-first?
There was a problem hiding this comment.
it's because of cache! apollo hates it when we f with the objects it has in cache!
| urn, | ||
| }, | ||
| skip: entityType !== EntityType.Dataset, | ||
| fetchPolicy: 'cache-first', |
There was a problem hiding this comment.
is editing schema field documentation, tags, and terms affected here with cache-first?
There was a problem hiding this comment.
i've actually updated to address!
| export const useGetEntityWithSchema = () => { | ||
| const { urn, entityData, entityType } = useEntityData(); | ||
| // Load the dataset schema lazily. | ||
| const { data: rawData, loading } = useGetDatasetSchemaQuery({ |
There was a problem hiding this comment.
also i think this query got lost in the commit so it's not defined? or is it already defined?
There was a problem hiding this comment.
it should be inside dataset.graphql!
remove all ref to schema metadata Adding back test changes Adding schema context fix checkstyler Add new data to ensure test passes
…om/jjoyce0510/datahub-1 into jj-load-schema-dynamically-dataset
…et' into jj-load-schema-dynamically-dataset
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
…-project#7558) Co-authored-by: david-leifker <114954101+david-leifker@users.noreply.github.com>
…-project#7558) Co-authored-by: david-leifker <114954101+david-leifker@users.noreply.github.com>
Summary
We've found that large schemas can bottleneck the time to first meaningful render on the Dataset profile page. This PR speeds things up by separating the schema tab load from the main getDataset query (for datasets only).
Features include a new loading indicator on the schema tab while the fetch for the schema loads. We also use the apollo cache to ensure that switching between tabs remains smooth.
Status
Ready for review.
Checklist