Struct PromptCompressor 

pub struct PromptCompressor { /* private fields */ }

Prompt compressor for optimizing token usage

Implementations

impl PromptCompressor

pub fn new() -> Self

Create a new prompt compressor with default settings

pub fn with_whitespace_removal(self, remove: bool) -> Self

Configure whitespace removal

pub fn with_empty_line_removal(self, remove: bool) -> Self

Configure empty line removal

pub fn with_line_trimming(self, trim: bool) -> Self

Configure line trimming

pub fn estimate_tokens(text: &str) -> u32

Estimate token count for a text

Uses a simple heuristic: ~4 characters per token for English text. This is approximate and may vary by language and tokenizer.

pub fn compress(&self, text: &str) -> (String, CompressionStats)

Compress a prompt to reduce token usage

Arguments

• text - The text to compress

Returns

A tuple of (compressed_text, compression_stats)

pub fn check_limit(text: &str, model: &str) -> Option<u32>

Check if a prompt exceeds the model’s token limit

Arguments

• text - The text to check
• model - The model name

Returns

Some(tokens) if the limit is exceeded, None otherwise

pub fn get_limit_warning(text: &str, model: &str) -> Option<String>

Get a warning message if the prompt exceeds the model limit

Arguments

• text - The text to check
• model - The model name

Returns

A warning message if the limit is exceeded, None otherwise

Trait Implementations

impl Default for PromptCompressor

fn default() -> Self

Returns the “default value” for a type.