Module text 

Minimal text indexing for LSP position math (replaces ropey). Minimal text indexing for LSP position math.

Replaces our previous use of [ropey]. We never actually used ropey for what ropey is for — incremental edits and cheap snapshots — because the server uses full-document sync and rebuilds the buffer from a String on every change. The only thing we used was index conversion, and ropey hid the one decision that actually matters for correctness: which encoding a column is measured in.

LineIndex makes that explicit. It owns the source text plus a table of line-start byte offsets, and every conversion takes a PositionEncoding so the byte/UTF-16 choice is visible at the call site instead of buried in a dependency.

Encodings

An LSP Position.character is a count from the start of its line. The unit depends on the negotiated PositionEncoding:

• PositionEncoding::Utf16 — UTF-16 code units. This is the LSP default and what clients assume unless the server negotiates otherwise.
• PositionEncoding::Utf8 — bytes. Cheapest, and what this codebase used to emit unconditionally (correct only for ASCII).

Internally everything is a UTF-8 byte offset; encodings only ever affect the in-line column number.

Structs

LineIndex

Source text plus a line-start index, supporting byte ↔ Position conversion.

Enums

PositionEncoding

The unit in which an LSP Position.character column is measured.