A fast, helpful, and open-source document parser
nit: reduce comments
Logan Markewich3c8ea79ab2a5
diff --git a/crates/liteparse/src/ocr_merge.rs b/crates/liteparse/src/ocr_merge.rsindex 20cf967..eec4339 100644--- a/crates/liteparse/src/ocr_merge.rs+++ b/crates/liteparse/src/ocr_merge.rs@@ -43,20 +43,13 @@ pub(crate) struct RenderedPage {pub pixels: Vec<u8>,pub width: u32,pub height: u32,- /// DPI this page was actually rendered at. Differs from the configured- /// DPI when the long-edge cap kicked in (see `MAX_OCR_RENDER_LONG_EDGE_PX`);- /// OCR pixel coordinates must be mapped back to PDF points with THIS value.+ /// DPI this page was actually rendered at.pub dpi: f32,}/// Why a page was flagged as needing more than the cheap text-only path./// Multiple reasons can apply to one page (e.g. a sparse page whose little/// text is also garbled). Empty exactly when `needs_ocr` is false.-///-/// This is the discriminator a caller routes on: a scan goes to OCR, dense-/// vector text to a vector-aware pass, and so on. New variants will be added-/// as the routing function learns to recommend heavier pipelines (tables,-/// charts, LLM passes), so callers should treat unknown variants leniently.#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize)]#[serde(rename_all = "kebab-case")]pub enum ComplexityReason {@@ -418,10 +411,9 @@ fn count_columns(region: &Region, total_items: usize) -> usize {////// The pdfium `Document` holds raw pointers that are not `Send`, so callers must/// drop it before awaiting the OCR engine.-/// Cap on the rendered long edge (in pixels) for OCR page rasters.///-/// Large-format pages (architectural / engineering sheets, e.g. 24x36in) at the-/// configured DPI decode to multi-hundred-MB uncompressed rasters;+/// Cap on the rendered long edge (in pixels) for OCR page rasters+/// to avoid excessive memory usage.pub(crate) const MAX_OCR_RENDER_LONG_EDGE_PX: f32 = 4096.0;pub(crate) fn render_pages_for_ocr(