Preparatory refactor for split schema support

[SevInf]

Introduces abstractions over schema files that in future would allow us to support multiple schemas files. At the moment, this does not yet means Language tools support split schema - we are still working with a single files in there.

Abstractions I am talking about:

• SchemaDocument - single .prisma document, part of the schema. Can be created either from TextDocument supplied by language tools or file read from the disk.
• Line - slightly nicer abstraction over single line of text, main point of it's existence - pointer back to SchemaDocument that contains it.
• PrismaSchema - collection of schema documents comprising a single DB schema. Allows to iterate over lines of all of the contained files. For the most parts, this replaces lines: string[] array as an input of various language tools operations.

Block got a pointer back to the containing documents. Renames and jumps to definition are aware that their target might not be the same file they started in and adjusted accordingly.

Close prisma/team-orm#1126

[CLAassistant]

All committers have signed the CLA.

[github-actions[bot]]

🤖 Pull request artifacts

empty string

file	commit
prisma.vsix	6a126af715ac560ab81f1eaeb18dd8f947e05f78

[github-actions[bot]]

Try this extension build via GitPod