Implement DatasetContextManager and useDatasetManager

[nozomione]

Context

This is a rewrite of the original issue: https://github.com/AlexsLemonade/refinebio-web/issues/101

As with the existing refine.bio, we want to be able to manage users' datasets when they use the application.

Problem or idea

Using the context API, we'll create a dataset manager context DatasetManagerContext and implement its hook useDatasetManager which manages a user's dataset and keeps it in sync with the API.

The general strategy should be to have a manager that;

• allow a user to create a dataset and customize it by adding and removing samples, or moving and sharing it with others.
• track a user's dataset ID by storing it in the local storage. Any data associated with the ID must be re-fetched from the API when the user returns (*unless the user clears cache or uses a different web browser/device to access our application).
• keeps the dataset state in sync with the API and queues changes made to an existing dataset

❚ State Management

Context State

The DatasetManagerContext inherits the following states from the global context RefinebioContext that are shared across the application.

dataset: (object) the latest user-created dataset state retrieved from the API setDataset: an async method for setting dataset

datasetId: (uuid) a user-created dataset ID retrieved from the API at creation setDatasetId: an async method for setting datasetId

downloadOptions: (object) a user-selected download options for a file download setDownloadOptions: an async method for setting downloadOptions

email: (string) a user's provided email address entered in the email form for a file download

token: (uuid) a user's token required for a file download

Hook State

error: (object) an object that stores an error object (e.g., network) setError: an async method for setting error

loading: (boolean) the loading indicator for the dataset data setLoading: an async method for setting loading

❚ Method

The useDatasetManager hook contains the helper methods that are primarily used in the UI and each one corresponds to a specific application behavior.

Following are the implementation details and pseudocode (parameters marked with * are required):

Dataset

async createDataset(setCurrentDatasetId = false) return dataset Id

• makes API call to create a new dataset
• returns dataset Id

async getDataset(id) return dataset

• makes API call to fetch the latest state of the dataset with details
• returns dataset

async clearDataset(id) return dataset

• makes API call to clear all the samples in the dataset
• returns dataset

async updateDataset(id*, params*) return dataset

• makes API call to updates the dataset with the given dataset Id and params (e.g, sample data, or download options)
• returns dataset

Experiment

getTotalExperiments(data*) return number

• returns the total number of experiments in the dataset or 0 if none

formatExperiments() return object

• formats the sample/experiment arrays from the API response for UI
• returns formatted response

async removeExperiment(experimentAccessionCode*) return dataset

• makes API call to remove an experiment with the given accession code
• returns dataset

Samples

async addSamples(samplesToAdd*) return dataset

• makes API call to add the given samples to the dataset
• returns dataset

formatSampleMetadata(metadata*) return array

• formats the sample metadata names for UI
• returns the formatted sample metadata names

getTotalSamples(data*) return number

• returns the total number of samples in the dataset or 0 if none

async removeSamples(samplesToRemove*) return dataset

• makes API call to remove the given samples from the dataset
• returns dataset

async replaceSamples(samplesToReplace*) return dataset

• makes API call to replace the samples in a dataset with the given samples
• returns dataset

Example: To fetch the latest dataset state:

getDataset() // fetches the latest dataset associated with datasetId stored in the localStorage
getDataset(id) // fetches the latest dataset state with the given dataset Id

Solution or next step

Based on the above requirements, implement the dataset context manager and its hook.

[nozomione]

🗒️ Based on the 1:1 @davidsmejia, we should fetch datasets with the query string details: true by default (reason: future BE API revision is being taken into account). For this reason, we deleted getDatasetDetails and we also removed the details parameter from the rest of the methods.