kreuzberg 4.5.0

High-performance document intelligence library for Rust. Extract text, metadata, and structured data from PDFs, Office documents, images, and 88+ formats with async/sync APIs.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41

//! Layout detection configuration.

use serde::{Deserialize, Serialize};

/// Layout detection configuration.
///
/// Controls layout detection behavior in the extraction pipeline.
/// When set on [`ExtractionConfig`](super::ExtractionConfig), layout detection
/// is enabled for PDF extraction.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct LayoutDetectionConfig {
    /// Preset for model selection. Currently only `"accurate"` (RT-DETR) is supported.
    #[serde(default = "default_preset")]
    pub preset: String,

    /// Confidence threshold override (None = use model default).
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub confidence_threshold: Option<f32>,

    /// Whether to apply postprocessing heuristics (default: true).
    #[serde(default = "default_true")]
    pub apply_heuristics: bool,
}

impl Default for LayoutDetectionConfig {
    fn default() -> Self {
        Self {
            preset: default_preset(),
            confidence_threshold: None,
            apply_heuristics: true,
        }
    }
}

fn default_preset() -> String {
    "accurate".to_string()
}

fn default_true() -> bool {
    true
}