Struct SplitResult 

pub struct SplitResult {
    pub to_compact: Vec<Message>,
    pub preserved: Vec<Message>,
    pub compact_tokens: u64,
    pub preserved_tokens: u64,
    pub split_index: usize,
}

Result of splitting a conversation into old and recent portions.

Fields

to_compact: Vec<Message>

Messages to compact or summarize (the old portion).

These are the messages before the split point. They are candidates for summarization, truncation, or removal by the compactor.

preserved: Vec<Message>

Messages to preserve as-is (the recent portion).

These are the messages after the split point. They are kept untouched in the conversation history.

compact_tokens: u64

Estimated token count of the old portion (to_compact).

Approximate token usage of the messages eligible for summarization or removal, computed from the pre-split history so callers can budget how much context compaction should reclaim.

preserved_tokens: u64

Estimated token count of the recent portion (preserved).

Approximate token usage of the messages kept intact after the split, giving callers the residual context footprint that will carry into the next model call.

split_index: usize

The index in the original message list where the split occurred.

Zero when no split was needed (the entire conversation was preserved).

Trait Implementations

impl Clone for SplitResult

fn clone(&self) -> SplitResult

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SplitResult

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.