xberg 1.1.2

High-performance document intelligence library for Rust. Extract text, metadata, and structured data from PDFs, Office documents, images, and 107 formats and 371 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

//! OCR functionality for PDF extraction.
//!
//! Handles text quality evaluation, OCR fallback decision logic, and OCR processing.
//!
//! Split into topical submodules: [`scoring`] (native-text/OCR-output quality evaluation and
//! the skip/fallback gate), [`rendering`] (PDF page rasterization and image-XObject recovery),
//! [`document`] (bbox/geometry transforms and assembling OCR output into page documents,
//! including merging OCR pages back into a natively-extracted document), and [`pipeline`]
//! (the top-level mixed-OCR and per-page pipeline orchestrators). Items used across submodule
//! boundaries are `pub(super)`; items already reachable from outside `ocr` keep their original
//! visibility and are re-exported here so external call sites are unaffected by the split.
//! Unit tests live in `tests.rs` and `recognition_noise_tests.rs`, not inline here.

mod document;
mod pipeline;
mod rendering;
mod scoring;

#[cfg(any(feature = "ocr", feature = "ocr-pipeline"))]
pub(crate) use scoring::{OcrGateOutcome, evaluate_ocr_skip_gate, evaluate_per_page_ocr};

// ~keep The standalone-image extractor builds the same `PageContent.ocr_confidence` summary
// from its own single OCR run (#1568), so these three leave `ocr` rather than staying
// `pub(super)`. They are pure readers/builders with no PDF dependency of their own.
#[cfg(feature = "ocr")]
pub(crate) use scoring::{mean_text_conf_of, page_ocr_confidence, word_count_of};

// ~keep These three are reachable only from `extractors::pdf`'s own `#[cfg(test)]` unit tests,
// so a non-test build of this crate never exercises the re-export itself (the underlying items
// are not dead code -- `scoring` uses `NativeTextStats` and `evaluate_native_text_for_ocr`
// internally too, and `OcrFallbackDecision` is `pub`). `#[allow(unused_imports)]` documents that
// on purpose rather than leaving an unexplained warning suppression.
#[cfg(any(feature = "ocr", feature = "ocr-pipeline"))]
#[allow(unused_imports)]
pub(crate) use scoring::{NativeTextStats, OcrFallbackDecision, evaluate_native_text_for_ocr};

// ~keep Consumed only by `extractors::pdf::layout_runner`'s own tests, which additionally
// require `feature = "layout-detection"`; a `test, ocr, pdf` build without it compiles this
// re-export with no reader. `#[allow(unused_imports)]` documents that on purpose.
#[cfg(all(test, feature = "ocr", feature = "pdf"))]
#[allow(unused_imports)]
pub(crate) use rendering::render_selected_pages_for_ocr;

#[cfg(any(feature = "ocr", feature = "ocr-pipeline"))]
pub(crate) use document::{
    boundaries_after_replacements, merge_ocr_pages_into_internal_document,
    merge_structured_ocr_pages_into_internal_document,
};

#[cfg(all(any(feature = "ocr", feature = "ocr-pipeline"), feature = "pdf"))]
pub(crate) use pipeline::extract_mixed_ocr_native;
#[cfg(any(feature = "ocr", feature = "ocr-pipeline"))]
pub(crate) use pipeline::{extract_with_ocr, run_ocr_pipeline};

#[cfg(all(test, any(feature = "ocr", feature = "ocr-pipeline")))]
mod recognition_noise_tests;
#[cfg(all(test, feature = "ocr"))]
mod tests;