From fde990ead8bde644a27eeb190b82397753cb3f78 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Julian=20Offenh=C3=A4user?= <offenhaeuser@protonmail.com>
Date: Fri, 24 Mar 2023 22:07:03 +0100
Subject: [PATCH] LibPDF: Allow optional inheritable page attributes

Previously, get_inheritable_object would always try to find the object
and throw an error if it couldn't. The spec tells us that some page
attributes, like CropBox, are optional but also inheritable. Others,
like the media box and resources, are technically required by the spec,
but omitted by some documents.

In both cases, we are now able to search for inheritable objects and
find a suitable replacement if there wasn't one.
---
 Userland/Libraries/LibPDF/Document.cpp | 50 ++++++++++++++++++--------
 Userland/Libraries/LibPDF/Document.h   |  2 +-
 2 files changed, 37 insertions(+), 15 deletions(-)

diff --git a/Userland/Libraries/LibPDF/Document.cpp b/Userland/Libraries/LibPDF/Document.cpp
index 99d70127ca..de1a373a95 100644
--- a/Userland/Libraries/LibPDF/Document.cpp
+++ b/Userland/Libraries/LibPDF/Document.cpp
@@ -109,26 +109,46 @@ PDFErrorOr<Page> Document::get_page(u32 index)
     auto page_object = TRY(get_or_load_value(page_object_index));
     auto raw_page_object = TRY(resolve_to<DictObject>(page_object));
 
-    auto resources = TRY(get_inheritable_object(CommonNames::Resources, raw_page_object))->cast<DictObject>();
+    RefPtr<DictObject> resources;
+    auto maybe_resources_object = TRY(get_inheritable_object(CommonNames::Resources, raw_page_object));
+    if (maybe_resources_object.has_value())
+        resources = maybe_resources_object.value()->cast<DictObject>();
+    else
+        resources = adopt_ref(*new DictObject({}));
+
     auto contents = TRY(raw_page_object->get_object(this, CommonNames::Contents));
 
-    auto media_box_array = TRY(get_inheritable_object(CommonNames::MediaBox, raw_page_object))->cast<ArrayObject>();
-    auto media_box = Rectangle {
-        media_box_array->at(0).to_float(),
-        media_box_array->at(1).to_float(),
-        media_box_array->at(2).to_float(),
-        media_box_array->at(3).to_float(),
-    };
+    Rectangle media_box;
+    auto maybe_media_box_object = TRY(get_inheritable_object(CommonNames::MediaBox, raw_page_object));
+    if (maybe_media_box_object.has_value()) {
+        auto media_box_array = maybe_media_box_object.value()->cast<ArrayObject>();
+        media_box = Rectangle {
+            media_box_array->at(0).to_float(),
+            media_box_array->at(1).to_float(),
+            media_box_array->at(2).to_float(),
+            media_box_array->at(3).to_float(),
+        };
+    } else {
+        // As most other libraries seem to do, we default to the standard
+        // US letter size of 8.5" x 11" (612 x 792 Postscript units).
+        media_box = Rectangle {
+            0, 0,
+            612, 792
+        };
+    }
 
-    auto crop_box = media_box;
-    if (raw_page_object->contains(CommonNames::CropBox)) {
-        auto crop_box_array = TRY(raw_page_object->get_array(this, CommonNames::CropBox));
+    Rectangle crop_box;
+    auto maybe_crop_box_object = TRY(get_inheritable_object(CommonNames::CropBox, raw_page_object));
+    if (maybe_crop_box_object.has_value()) {
+        auto crop_box_array = maybe_crop_box_object.value()->cast<ArrayObject>();
         crop_box = Rectangle {
             crop_box_array->at(0).to_float(),
             crop_box_array->at(1).to_float(),
             crop_box_array->at(2).to_float(),
             crop_box_array->at(3).to_float(),
         };
+    } else {
+        crop_box = media_box;
     }
 
     float user_unit = 1.0f;
@@ -141,7 +161,7 @@ PDFErrorOr<Page> Document::get_page(u32 index)
         VERIFY(rotate % 90 == 0);
     }
 
-    Page page { move(resources), move(contents), media_box, crop_box, user_unit, rotate };
+    Page page { resources.release_nonnull(), move(contents), media_box, crop_box, user_unit, rotate };
     m_pages.set(index, page);
     return page;
 }
@@ -306,13 +326,15 @@ PDFErrorOr<Destination> Document::create_destination_from_parameters(NonnullRefP
     return Destination { type, page_number_by_index_ref.get(page_ref.as_ref_index()), parameters };
 }
 
-PDFErrorOr<NonnullRefPtr<Object>> Document::get_inheritable_object(DeprecatedFlyString const& name, NonnullRefPtr<DictObject> object)
+PDFErrorOr<Optional<NonnullRefPtr<Object>>> Document::get_inheritable_object(DeprecatedFlyString const& name, NonnullRefPtr<DictObject> object)
 {
     if (!object->contains(name)) {
+        if (!object->contains(CommonNames::Parent))
+            return { OptionalNone() };
         auto parent = TRY(object->get_dict(this, CommonNames::Parent));
         return get_inheritable_object(name, parent);
     }
-    return object->get_object(this, name);
+    return TRY(object->get_object(this, name));
 }
 
 PDFErrorOr<Destination> Document::create_destination_from_dictionary_entry(NonnullRefPtr<Object> const& entry, HashMap<u32, u32> const& page_number_by_index_ref)
diff --git a/Userland/Libraries/LibPDF/Document.h b/Userland/Libraries/LibPDF/Document.h
index 7805a20017..8880c653a2 100644
--- a/Userland/Libraries/LibPDF/Document.h
+++ b/Userland/Libraries/LibPDF/Document.h
@@ -143,7 +143,7 @@ private:
     PDFErrorOr<Destination> create_destination_from_parameters(NonnullRefPtr<ArrayObject>, HashMap<u32, u32> const&);
     PDFErrorOr<Destination> create_destination_from_dictionary_entry(NonnullRefPtr<Object> const& entry, HashMap<u32, u32> const& page_number_by_index_ref);
 
-    PDFErrorOr<NonnullRefPtr<Object>> get_inheritable_object(DeprecatedFlyString const& name, NonnullRefPtr<DictObject>);
+    PDFErrorOr<Optional<NonnullRefPtr<Object>>> get_inheritable_object(DeprecatedFlyString const& name, NonnullRefPtr<DictObject>);
 
     PDFErrorOr<NonnullRefPtr<Object>> find_in_name_tree(NonnullRefPtr<DictObject> root, DeprecatedFlyString name);
     PDFErrorOr<NonnullRefPtr<Object>> find_in_name_tree_nodes(NonnullRefPtr<ArrayObject> siblings, DeprecatedFlyString name);