check if expression contains a perspective snapshot regardless of language and make viewable as graph

[weswalla]

a few questions:

1. do we assume the expression (regardless of language) is of the same structure? for example, the following is the structure of the perspective snapshot language:

{
  author: 'did:key:zQ3shYfQkxSgLDqG3VuQVnYvGsp23kRHL6btm1YzN8Xyw1mCV',
  timestamp: '2022-01-25T18:59:28.004Z',
  data: '{"links":[{"author":"did:key:zQ3shYfQkxSgLDqG3VuQVnYvGsp23kRHL6btm1YzN8Xyw1mCV","timestamp":"2022-01-25T17:43:10.798Z","data":{"predicate":"w","source":"q","target":"e"},"proof":{"key":"#zQ3shYfQkxSgLDqG3VuQVnYvGsp23kRHL6btm1YzN8Xyw1mCV","signature":"ce66f6e3e20dad8d7bc8b3b88a92f6721648ea854095b42da45542baa77106e457edf5ecd9dba23fc60e678285ecc18269a3c81409096dd7708369b026fb2de2","valid":true,"invalid":false}}]}',
  language: { address: 'perspective' },
  proof: { valid: true, invalid: null }
}

Do we assume that the expression data contains the ad4m-related metadata, particularily the "data" field which contains the relevant "link" field, and then each link nesting the link data in a "data" field?

2. are we assuming that any language could contain such expression types, or only http/https languages? As in, do we need to check each expression in a perspective to be of a valid type, or only http expressions (in addition to the perspective language)?

3. if we are going to check each expression in a perspective, would this not introduce performance issues as a perspective gets bigger?