Struct SentimentAnalyzer 

pub struct SentimentAnalyzer {
    pub config: SentimentConfig,
    pub lexicon: HashMap<String, LexiconEntry>,
}

Lexicon-based sentiment analysis engine with aspect-level detection.

The analyser operates entirely in-process; no external model or network call is required. Accuracy is limited by the quality of the built-in (or user-supplied) lexicon — for high-stakes applications consider augmenting the lexicon via with_lexicon_entry.

Fields

config: SentimentConfig

Runtime configuration.

lexicon: HashMap<String, LexiconEntry>

Word-to-entry lookup table.

Implementations

impl SentimentAnalyzer

pub fn new(config: SentimentConfig) -> Self

Build a new analyser seeded with the built-in lexicon.

pub fn with_lexicon_entry(self, entry: LexiconEntry) -> Self

Add or replace a single lexicon entry (builder pattern).

pub fn analyze(&self, text_id: String, text: &str) -> SentimentResult

Analyse a single document and return a SentimentResult.

pub fn batch_analyze(&self, texts: &[(String, String)]) -> Vec<SentimentResult>

Analyse a batch of (text_id, text) pairs.

pub fn top_positive<'a>( &self, results: &'a [SentimentResult], n: usize, ) -> Vec<&'a SentimentResult>

Return the top n results ordered by compound score (most positive first).

pub fn top_negative<'a>( &self, results: &'a [SentimentResult], n: usize, ) -> Vec<&'a SentimentResult>

Return the n most negative results (lowest compound score first).

pub fn aggregate_sentiment(&self, results: &[SentimentResult]) -> SentimentScore

Compute the mean SentimentScore across a slice of results.

Returns SentimentScore::zero() for an empty slice.

pub fn stats(&self, results: &[SentimentResult]) -> SentimentAnalyzerStats

Compute aggregate statistics for a collection of results.