healthtree / firestore-join Goto Github PK

Is your feature request related to a problem? Please describe.
Firestore stores dates as their own timestamp type. To use dates on our application and with third party libraries we need to convert them to JS dates.

We have a transform function that checks for timestamps at the first level of data and if the value has the .toDate function defined, we call it and store that value instead.

The problem is that now we are storing dates on nested properties, either on objects or arrays of objects. Our transform function is not picking this cases.

Describe the solution you'd like
Modify the transform function:

Refactor the code so we can recursively transform the data inside serializedDocument.

Is your feature request related to a problem? Please describe.
When we use firebase join to include reference relations, we do this on an inefficient way.

Example:

We get all the news for the homepage, and for each new piece we include the categories.

This means that if we get 50 news, we hit the db to get the categories. Even tho we only have 10 categories.

Same thing happens on the category's articles page, we hit the db to get same category N (number of articles) times.

This problem will repeat across the app because most of the relations we have are belongs to.

Describe the solution you'd like
Create a cache layer that we can manage from the application to set cached collections.

I imagine a function exported from 'firestore-join' with this signature:

setCachedCollection('apps/community/articleCategories', serializedDocumentArray);

serializedDocumentArray could come from a snapshot listener, or from a simple .get() that was serialized.

Then I would store a map for the cache with {'collectionName1': array1, 'collectionName2: array2, ... }

if we wanted to clear the cache we could just do:

setCachedCollection('apps/community/articleCategories',[]);

Then from inside firestore-join we need to modify the parts of the code that usedocumentReference.get() (line 99 and 111) a reference to check on the cache first based on the documentReference.collection and documentReference.path .