Phase 4: Text Extraction

[tlkahn]

Goal

Extract text from PDF pages: full text, positioned characters, styled segments, and search.

Tasks

• Add text FFI to ffi.rs: FPDFText_LoadPage, FPDFText_ClosePage, FPDFText_CountChars, FPDFText_GetUnicode, FPDFText_GetCharBox, FPDFText_GetText, FPDFText_FindStart, FPDFText_FindNext, FPDFText_GetSchResultIndex, FPDFText_GetSchCount, FPDFText_FindClose
• Add safe wrappers
• Implement lazy FPDF_TEXTPAGE creation (cached in PageData)
• Implement Document::page_text() — full text string
• Implement Document::page_chars() — Vec<CharInfo> with positions
• Implement Document::page_text_segments() — runs with shared style
• Implement Document::search_page() — find text with options (case-sensitive, whole word)
• Tests: extract text from test PDFs, verify search results, verify character positions

Exit Criterion

Text extraction matches pdfium-render output on test PDFs. Search works.

Dependencies

• Phase 3: Rendering (MVP) — publish 0.1 #4 (Phase 3 / MVP)
• Independent of Phase 5 and Phase 8