Documentation is

• Missing
• Outdated
• Confusing
• Not sure?

Explain in Detail

On the Backend Integration page, parts 3 and 4 deal with the manifest.json file.

However, there are several parts that are not clearly explained, and I would appreciate clarification on, and may be helpful to others implementing something to convert it to HTML.

For example, asset chunks are not entry or dynamic entry chunks, but do not have a key starting with _.

Questions relating to Section 4:

• Does order matter for the steps in the bullet point list?
• The 2nd bullet point mentions recursive imports of css files.
  • Should that be just the css section, or does that include if file is a css file as well?
  • If it does include file, what is the impact of loading the files after loading all the css sections?
  • Can a parser assume that there will never be any cycles when recursively importing chunks?

Edit:

• For the 3rd bullet point, Can a parser assume that an entry point will always be a JS or CSS file?

Your Suggestion for Changes

Consider breaking manifest.json documentation to its own file.

Section 3

• Include a link to the TypeScript interface definition of a ManifestChunk.
• Include an example of an asset chunk, and a bullet point about it.
• Consider changing "main.js" to "index.html" or adding an example of that as well.
• Note in the 2nd / 3rd bullet point that entry/dynamic entry chunks are determined by the "isEntry"/"isDynamicEntry" fields.
• For each chunk type in the bullet points, link to the appropriate configuration option for the files produced:
  • Entry Point Selection
  • Asset Chunks.
  • Non entry Chunks.
  • Entry Chunks.

Section 4

• If order is important, change the bullet points to numbers.
  *Add information to the 2nd bullet point to answer the questions above.

Reproduction

https://vitejs.dev/guide/backend-integration.html

Steps to reproduce

No response