kreuzberg 4.9.0

High-performance document intelligence library for Rust. Extract text, metadata, and structured data from PDFs, Office documents, images, and 91+ formats and 248 programming languages via tree-sitter code intelligence 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
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232

//! Page structure types for documents.
//!
//! This module defines types for representing paginated document structures.

use serde::{Deserialize, Serialize};
use std::sync::Arc;

// Import serde helper and types from sibling modules
use super::extraction::{BoundingBox, ExtractedImage};
use super::serde_helpers::serde_vec_arc;
use super::tables::Table;

/// Unified page structure for documents.
///
/// Supports different page types (PDF pages, PPTX slides, Excel sheets)
/// with character offset boundaries for chunk-to-page mapping.
#[derive(Debug, Clone, Serialize, Deserialize)]
#[cfg_attr(feature = "api", derive(utoipa::ToSchema))]
pub struct PageStructure {
    /// Total number of pages/slides/sheets
    pub total_count: usize,

    /// Type of paginated unit
    pub unit_type: PageUnitType,

    /// Character offset boundaries for each page
    ///
    /// Maps character ranges in the extracted content to page numbers.
    /// Used for chunk page range calculation.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub boundaries: Option<Vec<PageBoundary>>,

    /// Detailed per-page metadata (optional, only when needed)
    #[serde(skip_serializing_if = "Option::is_none")]
    pub pages: Option<Vec<PageInfo>>,
}

/// Type of paginated unit in a document.
///
/// Distinguishes between different types of "pages" (PDF pages, presentation slides, spreadsheet sheets).
#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
#[serde(rename_all = "snake_case")]
#[cfg_attr(feature = "api", derive(utoipa::ToSchema))]
pub enum PageUnitType {
    /// Standard document pages (PDF, DOCX, images)
    Page,
    /// Presentation slides (PPTX, ODP)
    Slide,
    /// Spreadsheet sheets (XLSX, ODS)
    Sheet,
}

/// Byte offset boundary for a page.
///
/// Tracks where a specific page's content starts and ends in the main content string,
/// enabling mapping from byte positions to page numbers. Offsets are guaranteed to be
/// at valid UTF-8 character boundaries when using standard String methods (push_str, push, etc.).
#[derive(Debug, Clone, Serialize, Deserialize)]
#[cfg_attr(feature = "api", derive(utoipa::ToSchema))]
pub struct PageBoundary {
    /// Byte offset where this page starts in the content string (UTF-8 valid boundary, inclusive)
    pub byte_start: usize,
    /// Byte offset where this page ends in the content string (UTF-8 valid boundary, exclusive)
    pub byte_end: usize,
    /// Page number (1-indexed)
    pub page_number: usize,
}

/// Metadata for individual page/slide/sheet.
///
/// Captures per-page information including dimensions, content counts,
/// and visibility state (for presentations).
#[derive(Debug, Clone, Serialize, Deserialize)]
#[cfg_attr(feature = "api", derive(utoipa::ToSchema))]
pub struct PageInfo {
    /// Page number (1-indexed)
    pub number: usize,

    /// Page title (usually for presentations)
    #[serde(skip_serializing_if = "Option::is_none")]
    pub title: Option<String>,

    /// Dimensions in points (PDF) or pixels (images): (width, height)
    #[serde(skip_serializing_if = "Option::is_none")]
    pub dimensions: Option<(f64, f64)>,

    /// Number of images on this page
    #[serde(skip_serializing_if = "Option::is_none")]
    pub image_count: Option<usize>,

    /// Number of tables on this page
    #[serde(skip_serializing_if = "Option::is_none")]
    pub table_count: Option<usize>,

    /// Whether this page is hidden (e.g., in presentations)
    #[serde(skip_serializing_if = "Option::is_none")]
    pub hidden: Option<bool>,

    /// Whether this page is blank (no meaningful text, no images, no tables)
    ///
    /// A page is considered blank if it has fewer than 3 non-whitespace characters
    /// and contains no tables or images. This is useful for filtering out empty pages
    /// in scanned documents or PDFs with blank separator pages.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub is_blank: Option<bool>,
}

/// Content for a single page/slide.
///
/// When page extraction is enabled, documents are split into per-page content
/// with associated tables and images mapped to each page.
///
/// # Performance
///
/// Uses Arc-wrapped tables and images for memory efficiency:
/// - `Vec<Arc<Table>>` enables zero-copy sharing of table data
/// - `Vec<Arc<ExtractedImage>>` enables zero-copy sharing of image data
/// - Maintains exact JSON compatibility via custom Serialize/Deserialize
///
/// This reduces memory overhead for documents with shared tables/images
/// by avoiding redundant copies during serialization.
#[derive(Debug, Clone, Serialize, Deserialize)]
#[cfg_attr(feature = "api", derive(utoipa::ToSchema))]
pub struct PageContent {
    /// Page number (1-indexed)
    pub page_number: usize,

    /// Text content for this page
    pub content: String,

    /// Tables found on this page (uses Arc for memory efficiency)
    ///
    /// Serializes as Vec<Table> for JSON compatibility while maintaining
    /// Arc semantics in-memory for zero-copy sharing.
    #[serde(skip_serializing_if = "Vec::is_empty", default, with = "serde_vec_arc")]
    #[cfg_attr(feature = "api", schema(value_type = Vec<Table>))]
    pub tables: Vec<Arc<Table>>,

    /// Images found on this page (uses Arc for memory efficiency)
    ///
    /// Serializes as Vec<ExtractedImage> for JSON compatibility while maintaining
    /// Arc semantics in-memory for zero-copy sharing.
    #[serde(skip_serializing_if = "Vec::is_empty", default, with = "serde_vec_arc")]
    #[cfg_attr(feature = "api", schema(value_type = Vec<ExtractedImage>))]
    pub images: Vec<Arc<ExtractedImage>>,

    /// Hierarchy information for the page (when hierarchy extraction is enabled)
    ///
    /// Contains text hierarchy levels (H1-H6) extracted from the page content.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub hierarchy: Option<PageHierarchy>,

    /// Whether this page is blank (no meaningful text content)
    ///
    /// Determined during extraction based on text content analysis.
    /// A page is blank if it has fewer than 3 non-whitespace characters
    /// and contains no tables or images.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub is_blank: Option<bool>,

    /// Layout detection regions for this page (when layout detection is enabled).
    ///
    /// Contains detected layout regions with class, confidence, bounding box,
    /// and area fraction. Only populated when layout detection is configured.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub layout_regions: Option<Vec<LayoutRegion>>,
}

/// A detected layout region on a page.
///
/// When layout detection is enabled, each page may have layout regions
/// identifying different content types (text, pictures, tables, etc.)
/// with confidence scores and spatial positions.
#[derive(Debug, Clone, Default, Serialize, Deserialize)]
#[cfg_attr(feature = "api", derive(utoipa::ToSchema))]
pub struct LayoutRegion {
    /// Layout class name (e.g. "picture", "table", "text", "section_header").
    pub class: String,
    /// Confidence score from the layout detection model (0.0 to 1.0).
    pub confidence: f64,
    /// Bounding box in document coordinate space.
    pub bounding_box: BoundingBox,
    /// Fraction of the page area covered by this region (0.0 to 1.0).
    pub area_fraction: f64,
}

/// Page hierarchy structure containing heading levels and block information.
///
/// Used when PDF text hierarchy extraction is enabled. Contains hierarchical
/// blocks with heading levels (H1-H6) for semantic document structure.
#[derive(Debug, Clone, Serialize, Deserialize)]
#[cfg_attr(feature = "api", derive(utoipa::ToSchema))]
pub struct PageHierarchy {
    /// Number of hierarchy blocks on this page
    pub block_count: usize,

    /// Hierarchical blocks with heading levels
    #[serde(skip_serializing_if = "Vec::is_empty", default)]
    pub blocks: Vec<HierarchicalBlock>,
}

/// A text block with hierarchy level assignment.
///
/// Represents a block of text with semantic heading information extracted from
/// font size clustering and hierarchical analysis.
#[derive(Debug, Clone, Serialize, Deserialize)]
#[cfg_attr(feature = "api", derive(utoipa::ToSchema))]
pub struct HierarchicalBlock {
    /// The text content of this block
    pub text: String,

    /// The font size of the text in this block
    pub font_size: f32,

    /// The hierarchy level of this block (H1-H6 or Body)
    ///
    /// Levels correspond to HTML heading tags:
    /// - "h1": Top-level heading
    /// - "h2": Secondary heading
    /// - "h3": Tertiary heading
    /// - "h4": Quaternary heading
    /// - "h5": Quinary heading
    /// - "h6": Senary heading
    /// - "body": Body text (no heading level)
    pub level: String,

    /// Bounding box information for the block
    ///
    /// Contains coordinates as (left, top, right, bottom) in PDF units.
    #[serde(skip_serializing_if = "Option::is_none")]
    pub bbox: Option<(f32, f32, f32, f32)>,
}