Struct TextPositionIndex 

pub struct TextPositionIndex { /* private fields */ }

Maps character offsets in concatenated document text to PDF coordinates.

Built from per-page ExtractedText (with preserve_layout: true), this index allows mapping a character range (like those in DocumentChunk) back to the physical locations on PDF pages.

Implementations

impl TextPositionIndex

pub fn build(pages: &[ExtractedText]) -> Self

Build an index from per-page extracted text.

Pages are concatenated with "\n\n" separators (matching the chunker). For each page’s TextFragment, we find its position in the page text and compute the global character offset.

pub fn fragments_for_range( &self, start: usize, end: usize, ) -> Vec<&IndexedFragment>

Find all fragments whose character range overlaps with [start, end).

pub fn page_offset(&self, page: usize) -> Option<usize>

Get the character offset where a given page starts.

pub fn len(&self) -> usize

Total number of indexed fragments.

pub fn is_empty(&self) -> bool

Whether the index is empty.

pub fn entries(&self) -> &[IndexedFragment]

All indexed entries (for inspection/testing).

Trait Implementations

impl Debug for TextPositionIndex

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.