Phase 3 (Organize + robust save) + Phase 4 (Edit content); update docs

Engine: - organize.rs: build_document(plan) core — insert/delete/rotate/crop/extract/merge/split - watermark.rs: watermark + header/footer via PdfPages::watermark, CID Unicode fonts - qpdf.rs: QPDF CLI wrapper — repair corrupt files, re-encrypt after edits, ensure_openable - edit.rs: MOAT — list_objects + apply_edits (SetText VN, Transform, Delete, Add/ReplaceImage) - Tests: organize/watermark/qpdf_safety/edit roundtrip — 43/43 engine tests green UI (Tauri): "Tổ chức trang" mode (grid + drag-reorder + multi-select), "Sửa nội dung" mode (WYSIWYG object overlay, edit-in-place), global undo/redo, robust file open. Corpus: add encrypted.pdf + corrupt-truncated.pdf fixtures. Docs: mark Phase 1-4 done in roadmap; document QPDF setup + engine module layout in 05-dev-setup; add 10-13 phase summaries & user-test checklists; README status.
2026-07-14 22:07:38 +07:00
parent 274012f854
commit e442a9f678
26 changed files with 3924 additions and 40 deletions
@@ -0,0 +1,173 @@
+//! Round-trip test cho thao tác tổ chức trang (Phase 3): xoá/xoay/trích/
+//! merge/split/crop/chèn trang trắng — đường GHI file rủi ro nhất nên test
+//! kỹ bằng cách thao tác rồi đọc lại, so khớp số trang/thứ tự/nội dung.
+
+use std::path::PathBuf;
+
+use ff_engine::{PagePlanEntry, Rect};
+
+fn workspace_root() -> PathBuf {
+    PathBuf::from(env!("CARGO_MANIFEST_DIR"))
+        .join("..")
+        .join("..")
+        .canonicalize()
+        .expect("workspace root")
+}
+
+fn pdfium() -> pdfium_render::prelude::Pdfium {
+    if std::env::var("FOFREEXIT_PDFIUM_PATH").is_err() {
+        std::env::set_var("FOFREEXIT_PDFIUM_PATH", workspace_root());
+    }
+    ff_engine::bind_pdfium().expect("nạp PDFium")
+}
+
+fn tmp(name: &str) -> PathBuf {
+    let p = std::env::temp_dir().join(name);
+    let _ = std::fs::remove_file(&p);
+    p
+}
+
+#[test]
+fn delete_pages_keeps_remaining_content_and_order() {
+    let pdf = pdfium();
+    let input = workspace_root().join("corpus").join("sample-multipage.pdf");
+    let output = tmp("ff_org_delete.pdf");
+
+    // sample-multipage.pdf: 3 trang, outline Chapter 1/2/3 -> trang 0/1/2.
+    ff_engine::delete_pages(&pdf, &input, &[1], &output, None).expect("delete_pages");
+
+    assert_eq!(ff_engine::page_count(&pdf, &output, None).expect("count"), 2);
+    let p0 = ff_engine::extract_text(&pdf, &output, 0, None).expect("text p0");
+    let p1 = ff_engine::extract_text(&pdf, &output, 1, None).expect("text p1");
+    assert!(p0.contains("1") || !p0.is_empty(), "trang 0 gốc phải còn nguyên: {p0:?}");
+    assert!(!p1.is_empty(), "trang 2 gốc (giờ là trang 1) phải còn nguyên: {p1:?}");
+    // Trang giữa (index 1 gốc) không còn xuất hiện ở bất kỳ đâu trong output.
+    assert_ne!(p0, p1);
+}
+
+#[test]
+fn delete_all_pages_is_rejected() {
+    let pdf = pdfium();
+    let input = workspace_root().join("corpus").join("sample-multipage.pdf");
+    let output = tmp("ff_org_delete_all.pdf");
+    let err = ff_engine::delete_pages(&pdf, &input, &[0, 1, 2], &output, None);
+    assert!(err.is_err(), "không được cho xoá hết toàn bộ trang");
+}
+
+#[test]
+fn rotate_pages_persists_in_saved_file() {
+    let pdf = pdfium();
+    let input = workspace_root().join("corpus").join("sample-multipage.pdf");
+    let output = tmp("ff_org_rotate.pdf");
+
+    ff_engine::rotate_pages(&pdf, &input, &[0], 90, &output, None).expect("rotate_pages");
+
+    let doc = pdf.load_pdf_from_file(&output, None).expect("mở lại");
+    let page0 = doc.pages().get(0).expect("trang 0");
+    let page1 = doc.pages().get(1).expect("trang 1");
+    assert_eq!(
+        page0.rotation().expect("rotation p0"),
+        pdfium_render::prelude::PdfPageRenderRotation::Degrees90,
+        "trang 0 phải xoay 90 độ"
+    );
+    assert_eq!(
+        page1.rotation().expect("rotation p1"),
+        pdfium_render::prelude::PdfPageRenderRotation::None,
+        "trang 1 không bị xoay"
+    );
+}
+
+#[test]
+fn extract_pages_creates_new_file_without_touching_input() {
+    let pdf = pdfium();
+    let input = workspace_root().join("corpus").join("sample-multipage.pdf");
+    let output = tmp("ff_org_extract.pdf");
+
+    ff_engine::extract_pages(&pdf, &input, &[2, 0], &output, None).expect("extract_pages");
+
+    assert_eq!(ff_engine::page_count(&pdf, &output, None).expect("count"), 2);
+    assert_eq!(ff_engine::page_count(&pdf, &input, None).expect("count input"), 3, "input không bị sửa");
+
+    // Thứ tự đúng theo `pages` truyền vào: [2, 0] -> output trang 0 = input trang 2.
+    let out_p0 = ff_engine::extract_text(&pdf, &output, 0, None).expect("text");
+    let in_p2 = ff_engine::extract_text(&pdf, &input, 2, None).expect("text");
+    assert_eq!(out_p0, in_p2, "trích đúng thứ tự đã chọn");
+}
+
+#[test]
+fn merge_files_concatenates_in_order() {
+    let pdf = pdfium();
+    let a = workspace_root().join("corpus").join("hello.pdf");
+    let b = workspace_root().join("corpus").join("sample-multipage.pdf");
+    let output = tmp("ff_org_merge.pdf");
+
+    ff_engine::merge_files(&pdf, &[a.clone(), b.clone()], &output).expect("merge_files");
+
+    let count_a = ff_engine::page_count(&pdf, &a, None).expect("count a");
+    let count_b = ff_engine::page_count(&pdf, &b, None).expect("count b");
+    assert_eq!(
+        ff_engine::page_count(&pdf, &output, None).expect("count out"),
+        count_a + count_b
+    );
+
+    let merged_first = ff_engine::extract_text(&pdf, &output, 0, None).expect("text");
+    let a_first = ff_engine::extract_text(&pdf, &a, 0, None).expect("text");
+    assert_eq!(merged_first, a_first, "trang đầu output phải khớp file a");
+}
+
+#[test]
+fn split_by_page_count_produces_correct_chunks() {
+    let pdf = pdfium();
+    let input = workspace_root().join("corpus").join("sample-multipage.pdf"); // 3 trang
+    let out_dir = std::env::temp_dir();
+
+    let parts = ff_engine::split_by_page_count(&pdf, &input, 2, &out_dir, "ff_org_split", None)
+        .expect("split_by_page_count");
+
+    assert_eq!(parts.len(), 2, "3 trang / 2 mỗi file = 2 phần (2+1)");
+    assert_eq!(ff_engine::page_count(&pdf, &parts[0], None).expect("count"), 2);
+    assert_eq!(ff_engine::page_count(&pdf, &parts[1], None).expect("count"), 1);
+
+    for p in &parts {
+        let _ = std::fs::remove_file(p);
+    }
+}
+
+#[test]
+fn crop_sets_cropbox_on_saved_file() {
+    let pdf = pdfium();
+    let input = workspace_root().join("corpus").join("sample-multipage.pdf");
+    let output = tmp("ff_org_crop.pdf");
+
+    let mut plan = ff_engine::identity_plan(&pdf, &input, None).expect("identity_plan");
+    plan[0].crop = Some(Rect { left: 50.0, bottom: 60.0, right: 500.0, top: 700.0 });
+    ff_engine::build_document(&pdf, &input, &plan, &output, None).expect("build_document crop");
+
+    let doc = pdf.load_pdf_from_file(&output, None).expect("mở lại");
+    let page0 = doc.pages().get(0).expect("trang 0");
+    let crop = page0.boundaries().crop().expect("crop box");
+    assert!((crop.bounds.left().value - 50.0).abs() < 0.5);
+    assert!((crop.bounds.bottom().value - 60.0).abs() < 0.5);
+    assert!((crop.bounds.right().value - 500.0).abs() < 0.5);
+    assert!((crop.bounds.top().value - 700.0).abs() < 0.5);
+}
+
+#[test]
+fn insert_blank_page_has_correct_size_and_no_text() {
+    let pdf = pdfium();
+    let input = workspace_root().join("corpus").join("sample-multipage.pdf");
+    let output = tmp("ff_org_blank.pdf");
+
+    let mut plan = ff_engine::identity_plan(&pdf, &input, None).expect("identity_plan");
+    plan.insert(1, PagePlanEntry::blank(300.0, 400.0));
+    ff_engine::build_document(&pdf, &input, &plan, &output, None).expect("build_document blank");
+
+    assert_eq!(ff_engine::page_count(&pdf, &output, None).expect("count"), 4);
+    let doc = pdf.load_pdf_from_file(&output, None).expect("mở lại");
+    let blank = doc.pages().get(1).expect("trang trắng");
+    assert!((blank.width().value - 300.0).abs() < 0.5);
+    assert!((blank.height().value - 400.0).abs() < 0.5);
+    drop(doc);
+    let text = ff_engine::extract_text(&pdf, &output, 1, None).expect("text trang trắng");
+    assert!(text.trim().is_empty(), "trang trắng không được có text: {text:?}");
+}