RealDevSquad / RealDevSquad/website-backend

[RFC] Transformation Layer

Open
#1,493 2 comments 0 reactions 1 assignee View on GitHub

@pallabez is already working on this.

Since Sep 7, 2023.

documentation
Dominant language
JavaScript
Stars
74
Forks
276
Avg merge
1d 26m
Merged PRs (30d)
14

Description

Problem

Currently we are fetching data from many external services (firestore, github, discord service). We mostly change a few fields of the data according to our needs and send it to controller and eventually as API response.

A few problem arises with this -

Transformation of data happening at different places
/controller/auth.js
      userData = {
        github_id: user.username,
        github_display_name: user.displayName,
        github_created_at: Number(new Date(user._json.created_at).getTime()),
        created_at: Date.now(),
        updated_at: Date.now(),
      };

      const { userId, incompleteUserDetails } = await users.addOrUpdate(userData);
/models/discordactions.js
    return groups.map((group) => {
      const groupCreator = groupCreatorsDetails[group.createdBy];
      return {
        ...group,
        firstName: groupCreator?.first_name,
        lastName: groupCreator?.last_name,
        image: groupCreator?.picture?.url,
        memberCount: roleIdToCountMap[group.roleid] || 0, // Number of users joined this group
        isMember: subscribedGroupIds.has(group.roleid), // Is current loggedIn user is a member of this group
      };
    });
No reliable way to ensure no sensitive data is sent.

We mostly return the data object to the frontend as we are given by firestore. Private fields need to be handled with more conditions throughout the codebase.

The below solution helps but is not reliable as - New private fields can be added to user as well as at other models.

/services/dataAccessLayer.js
const removeSensitiveInfo = function (obj, level = ACCESS_LEVEL.PUBLIC) {
  for (let i = 0; i < KEYS_NOT_ALLOWED[level].length; i++) {
    if (Object.prototype.hasOwnProperty.call(obj, KEYS_NOT_ALLOWED[level][i])) {
      delete obj[KEYS_NOT_ALLOWED[level][i]];
    }
  }
};
No fixed place to do date parsing or other parsing
/models/extensionRequest.js
    const request = {
      timestamp: Number((new Date().getTime() / 1000).toFixed(0)),
      ...extensionRequestData,
    };
    return await extensionRequestsModel.add(request);

Proposed Solution

  • A top level folder called transformers with files according to the models we are transforming
    Eg - /transformers/user.js, /transformers/goal.js
  • It would contain functions for sole of transforming data
    Eg - transformUserFromFirestore(), transformGoalsFromAPI()
  • A transformation function would look something like this
export const transformGoalFromApi = (goal: any): Goal => {
  const attributes = goal.attributes

  return {
    id: goal.id,
    title: attributes.title,
    description: attributes.description,
    createdAt: new Date(attributes.created_at),
    createdBy: new Date(attributes.created_by),
    startsOn: attributes.starts_on,
    endsOn: attributes.ends_on,
    percentageCompleted: attributes.percentage_completed,
    assignedBy: attributes.assigned_by,
    status: attributes.status,
    assignedTo: attributes.assigned_to
  }
}
  • This function is used inside models after edge cases has been handled

Architecture drawio(1)

Benifits

  • Standard data object for passing around.
  • Reliable way to ensure no private fields sent via API response.
  • Change in API response structure/field name would only require change in this file.
  • Central location for all data transformation/parsing.
  • Removes response ambiguity from the developer. (Developer can check what fields we are getting from API via this function)
  • Typescript integration is made much simpler.

Demerits

  • Changing might break features.
  • Lots of work.
    Thankfully, this can done in a step by step process as show in the migration plan

Migration Plan

Alternate Solution

It can be called serializer instead.
Ember data serializer does this functionality as mentioned here
(To be discussed more)

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.