ladybird/Userland/Libraries/LibPDF/Document.cpp
Nico Weber 87112dcbdc LibPDF: Return null for invalid refs, tolerate null objects as outline
https://llvm.org/devmtg/2022-11/slides/TechTalk5-WhatDoesItTakeToRunLLVMBuildbots.pdf
has an xref table that starts like so:

```
xref
0 214
0000000002 65535 f
0000924663 00000 n
0000000003 00000 f
0000000000 00000 f
0000000016 00000 n
0000000160 00000 n
0000000263 00000 n
```

This is a list of objects in the PDF file. The lines ending with 'f'
mean that this object is "free", that is it's not stored in the file.
In this file, objects 0, 2, 3 are free. For free objects, the first
number is the offset of the next free object: Object 0 refers to object
2, 2 to 3, and 3 back to 0 (since it's the last free object).
The lines ending with "n" are actual objects; here the first number is
a byte offset to where that object is stored in the file.

Furthermore, the file contains

```
/Outlines
2
0
R
```

in its root object, meaning that object 2 stores the page outlines.

Since object 2 is set as free, there is no object 2. But the spec
says that an invalid object reference is just the null object.

This patch makes us return null objects for references to free
objects, and it also makes us treat a null object as /Outlines value
the same as not having /Outlines in the first place.

Fixes #23023 -- we can now open that file. (We don't render it super
well, but only for already-known reasons.)

Since I found it a bit confusing: XRefTable has two related methods
here:

1. has_object() returns if an object was explicitly listed in an
   xref table. The first number right after `xref` is the start
   index. So if an xref table were to start with `10`, we'd implicitly
   create 10 trailing objects for which has_object() would return false
2. is_object_in_use() returns true if an object that was in a table
   (i.e. one where has_object() returns true) was listed with 'n' and
   false if it was listed with 'f'.

DocumentParser::parse_object_with_index() should probably return a null
object for the `!has_object()` case as well instead of VERIFY()ing
that has_object() is true. But I haven't seen this in the wild yet,
so keeping as-is for now.
2024-01-31 12:10:19 -05:00

655 lines
25 KiB
C++

/*
* Copyright (c) 2021-2022, Matthew Olsson <mattco@serenityos.org>
*
* SPDX-License-Identifier: BSD-2-Clause
*/
#include <LibPDF/CommonNames.h>
#include <LibPDF/Document.h>
#include <LibPDF/Parser.h>
#include <LibTextCodec/Decoder.h>
namespace PDF {
ByteString OutlineItem::to_byte_string(int indent) const
{
auto indent_str = ByteString::repeated(" "sv, indent + 1);
StringBuilder child_builder;
child_builder.append('[');
for (auto& child : children)
child_builder.appendff("{}\n", child->to_byte_string(indent + 1));
child_builder.appendff("{}]", indent_str);
StringBuilder builder;
builder.append("OutlineItem {{\n"sv);
builder.appendff("{}title={}\n", indent_str, title);
builder.appendff("{}count={}\n", indent_str, count);
builder.appendff("{}dest={}\n", indent_str, dest);
builder.appendff("{}color={}\n", indent_str, color);
builder.appendff("{}italic={}\n", indent_str, italic);
builder.appendff("{}bold={}\n", indent_str, bold);
builder.appendff("{}children={}\n", indent_str, child_builder.to_byte_string());
builder.appendff("{}}}", ByteString::repeated(" "sv, indent));
return builder.to_byte_string();
}
PDFErrorOr<Optional<ByteString>> InfoDict::title() const
{
return get_text(CommonNames::Title);
}
PDFErrorOr<Optional<ByteString>> InfoDict::author() const
{
return get_text(CommonNames::Author);
}
PDFErrorOr<Optional<ByteString>> InfoDict::subject() const
{
return get_text(CommonNames::Subject);
}
PDFErrorOr<Optional<ByteString>> InfoDict::keywords() const
{
return get_text(CommonNames::Keywords);
}
PDFErrorOr<Optional<ByteString>> InfoDict::creator() const
{
return get_text(CommonNames::Creator);
}
PDFErrorOr<Optional<ByteString>> InfoDict::producer() const
{
return get_text(CommonNames::Producer);
}
PDFErrorOr<Optional<ByteString>> InfoDict::creation_date() const
{
return get(CommonNames::CreationDate);
}
PDFErrorOr<Optional<ByteString>> InfoDict::modification_date() const
{
return get(CommonNames::ModDate);
}
PDFErrorOr<Optional<ByteString>> InfoDict::get_text(DeprecatedFlyString const& name) const
{
return TRY(get(name)).map(Document::text_string_to_utf8);
}
ByteString Document::text_string_to_utf8(ByteString const& text_string)
{
if (text_string.bytes().starts_with(Array<u8, 2> { 0xfe, 0xff })) {
// The string is encoded in UTF16-BE
return TextCodec::decoder_for("utf-16be"sv)->to_utf8(text_string).release_value_but_fixme_should_propagate_errors().to_byte_string();
}
if (text_string.bytes().starts_with(Array<u8, 3> { 239, 187, 191 })) {
// The string is encoded in UTF-8.
return text_string.substring(3);
}
return TextCodec::decoder_for("PDFDocEncoding"sv)->to_utf8(text_string).release_value_but_fixme_should_propagate_errors().to_byte_string();
}
PDFErrorOr<NonnullRefPtr<Document>> Document::create(ReadonlyBytes bytes)
{
size_t offset_to_start = TRY(DocumentParser::scan_for_header_start(bytes));
if (offset_to_start != 0) {
dbgln("warning: PDF header not at start of file, skipping {} bytes", offset_to_start);
bytes = bytes.slice(offset_to_start);
}
auto parser = adopt_ref(*new DocumentParser({}, bytes));
auto document = adopt_ref(*new Document(parser));
document->m_version = TRY(parser->initialize());
document->m_trailer = parser->trailer();
document->m_catalog = TRY(parser->trailer()->get_dict(document, CommonNames::Root));
if (document->m_trailer->contains(CommonNames::Encrypt)) {
auto encryption_dict = TRY(document->m_trailer->get_dict(document, CommonNames::Encrypt));
document->m_security_handler = TRY(SecurityHandler::create(document, encryption_dict));
// Automatically attempt to unencrypt the document with the empty string. The
// result is not important; it is the caller's responsibility to ensure the
// document is unencrypted before calling initialize().
document->m_security_handler->try_provide_user_password(""sv);
}
return document;
}
Document::Document(NonnullRefPtr<DocumentParser> const& parser)
: m_parser(parser)
{
m_parser->set_document(this);
}
PDFErrorOr<void> Document::initialize()
{
if (m_security_handler)
VERIFY(m_security_handler->has_user_password());
TRY(build_page_tree());
TRY(build_outline());
return {};
}
PDFErrorOr<Value> Document::get_or_load_value(u32 index)
{
auto value = get_value(index);
if (!value.has<Empty>()) // FIXME: Use Optional instead?
return value;
auto object = TRY(m_parser->parse_object_with_index(index));
m_values.set(index, object);
return object;
}
u32 Document::get_first_page_index() const
{
// FIXME: A PDF can have a different default first page, which
// should be fetched and returned here
return 0;
}
u32 Document::get_page_count() const
{
return m_page_object_indices.size();
}
static ErrorOr<void> collect_referenced_indices(Value const& value, Vector<int>& referenced_indices)
{
TRY(value.visit(
[&](Empty const&) -> ErrorOr<void> { return {}; },
[&](nullptr_t const&) -> ErrorOr<void> { return {}; },
[&](bool const&) -> ErrorOr<void> { return {}; },
[&](int const&) -> ErrorOr<void> { return {}; },
[&](float const&) -> ErrorOr<void> { return {}; },
[&](Reference const& ref) -> ErrorOr<void> {
TRY(referenced_indices.try_append(ref.as_ref_index()));
return {};
},
[&](NonnullRefPtr<Object> const& object) -> ErrorOr<void> {
if (object->is<ArrayObject>()) {
for (auto& element : object->cast<ArrayObject>()->elements())
TRY(collect_referenced_indices(element, referenced_indices));
} else if (object->is<DictObject>()) {
for (auto& [key, value] : object->cast<DictObject>()->map()) {
if (key != CommonNames::Parent)
TRY(collect_referenced_indices(value, referenced_indices));
}
} else if (object->is<StreamObject>()) {
for (auto& [key, value] : object->cast<StreamObject>()->dict()->map()) {
if (key != CommonNames::Parent)
TRY(collect_referenced_indices(value, referenced_indices));
}
}
return {};
}));
return {};
}
static PDFErrorOr<void> dump_tree(Document& document, size_t index, HashTable<int>& seen)
{
if (seen.contains(index))
return {};
seen.set(index);
auto const& value = TRY(document.get_or_load_value(index));
outln("{} 0 obj", index);
outln("{}", value.to_byte_string(0));
outln("endobj");
Vector<int> referenced_indices;
TRY(collect_referenced_indices(value, referenced_indices));
for (auto index : referenced_indices)
TRY(dump_tree(document, index, seen));
return {};
}
PDFErrorOr<void> Document::dump_page(u32 index)
{
VERIFY(index < m_page_object_indices.size());
auto page_object_index = m_page_object_indices[index];
HashTable<int> seen;
TRY(dump_tree(*this, page_object_index, seen));
return {};
}
PDFErrorOr<Page> Document::get_page(u32 index)
{
VERIFY(index < m_page_object_indices.size());
auto cached_page = m_pages.get(index);
if (cached_page.has_value())
return cached_page.value();
auto page_object_index = m_page_object_indices[index];
auto page_object = TRY(get_or_load_value(page_object_index));
auto raw_page_object = TRY(resolve_to<DictObject>(page_object));
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 = make_object<DictObject>(HashMap<DeprecatedFlyString, Value> {});
RefPtr<Object> contents;
if (raw_page_object->contains(CommonNames::Contents))
contents = TRY(raw_page_object->get_object(this, CommonNames::Contents));
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
};
}
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;
if (raw_page_object->contains(CommonNames::UserUnit))
user_unit = raw_page_object->get_value(CommonNames::UserUnit).to_float();
int rotate = 0;
auto maybe_rotate = TRY(get_inheritable_value(CommonNames::Rotate, raw_page_object));
if (maybe_rotate.has_value()) {
rotate = TRY(resolve_to<int>(maybe_rotate.value()));
VERIFY(rotate % 90 == 0);
}
Page page { resources.release_nonnull(), move(contents), media_box, crop_box, user_unit, rotate };
m_pages.set(index, page);
return page;
}
PDFErrorOr<Value> Document::resolve(Value const& value)
{
if (value.has<Reference>()) {
// FIXME: Surely indirect PDF objects can't contain another indirect PDF object,
// right? Unsure from the spec, but if they can, these return values would have
// to be wrapped with another resolve() call.
return get_or_load_value(value.as_ref_index());
}
if (!value.has<NonnullRefPtr<Object>>())
return value;
auto& obj = value.get<NonnullRefPtr<Object>>();
if (obj->is<IndirectValue>())
return static_ptr_cast<IndirectValue>(obj)->value();
return value;
}
PDFErrorOr<Optional<InfoDict>> Document::info_dict()
{
if (!trailer()->contains(CommonNames::Info))
return OptionalNone {};
return InfoDict(this, TRY(trailer()->get_dict(this, CommonNames::Info)));
}
PDFErrorOr<Vector<DeprecatedFlyString>> Document::read_filters(NonnullRefPtr<DictObject> dict)
{
Vector<DeprecatedFlyString> filters;
// We may either get a single filter or an array of cascading filters
auto filter_object = TRY(dict->get_object(this, CommonNames::Filter));
if (filter_object->is<ArrayObject>()) {
auto filter_array = filter_object->cast<ArrayObject>();
for (size_t i = 0; i < filter_array->size(); ++i)
filters.append(TRY(filter_array->get_name_at(this, i))->name());
} else {
filters.append(filter_object->cast<NameObject>()->name());
}
return filters;
}
PDFErrorOr<void> Document::build_page_tree()
{
auto page_tree = TRY(m_catalog->get_dict(this, CommonNames::Pages));
return add_page_tree_node_to_page_tree(page_tree);
}
PDFErrorOr<void> Document::add_page_tree_node_to_page_tree(NonnullRefPtr<DictObject> const& page_tree)
{
auto kids_array = TRY(page_tree->get_array(this, CommonNames::Kids));
auto page_count = page_tree->get(CommonNames::Count).value().get<int>();
if (static_cast<size_t>(page_count) != kids_array->elements().size()) {
// This page tree contains child page trees, so we recursively add
// these pages to the overall page tree
for (auto& value : *kids_array) {
auto reference_index = value.as_ref_index();
auto maybe_page_tree_node = TRY(m_parser->conditionally_parse_page_tree_node(reference_index));
if (maybe_page_tree_node) {
TRY(add_page_tree_node_to_page_tree(maybe_page_tree_node.release_nonnull()));
} else {
m_page_object_indices.append(reference_index);
}
}
} else {
// We know all of the kids are leaf nodes
for (auto& value : *kids_array)
m_page_object_indices.append(value.as_ref_index());
}
return {};
}
PDFErrorOr<NonnullRefPtr<Object>> Document::find_in_name_tree(NonnullRefPtr<DictObject> tree, DeprecatedFlyString name)
{
if (tree->contains(CommonNames::Kids)) {
return find_in_name_tree_nodes(tree->get_array(CommonNames::Kids), name);
}
if (!tree->contains(CommonNames::Names))
return Error { Error::Type::MalformedPDF, "name tree has neither Kids nor Names" };
auto key_value_names_array = TRY(tree->get_array(this, CommonNames::Names));
return find_in_key_value_array(key_value_names_array, name);
}
PDFErrorOr<NonnullRefPtr<Object>> Document::find_in_name_tree_nodes(NonnullRefPtr<ArrayObject> siblings, DeprecatedFlyString name)
{
for (size_t i = 0; i < siblings->size(); i++) {
auto sibling = TRY(resolve_to<DictObject>(siblings->at(i)));
auto limits = sibling->get_array(CommonNames::Limits);
if (limits->size() != 2)
return Error { Error::Type::MalformedPDF, "Expected 2-element Limits array" };
auto start = limits->get_string_at(0);
auto end = limits->get_string_at(1);
if (start->string() <= name && end->string() >= name) {
return find_in_name_tree(sibling, name);
}
}
return Error { Error::Type::MalformedPDF, ByteString::formatted("Didn't find node in name tree containing name {}", name) };
}
PDFErrorOr<NonnullRefPtr<Object>> Document::find_in_key_value_array(NonnullRefPtr<ArrayObject> key_value_array, DeprecatedFlyString name)
{
if (key_value_array->size() % 2 == 1)
return Error { Error::Type::MalformedPDF, "key/value array has dangling key" };
for (size_t i = 0; i < key_value_array->size() / 2; i++) {
auto key = key_value_array->get_string_at(2 * i);
if (key->string() == name) {
return key_value_array->get_object_at(this, 2 * i + 1);
}
}
return Error { Error::Type::MalformedPDF, ByteString::formatted("Didn't find expected name {} in key/value array", name) };
}
PDFErrorOr<void> Document::build_outline()
{
if (!m_catalog->contains(CommonNames::Outlines))
return {};
auto outlines = TRY(resolve(m_catalog->get_value(CommonNames::Outlines)));
if (outlines.has<nullptr_t>())
return {};
auto outline_dict = cast_to<DictObject>(outlines);
if (!outline_dict->contains(CommonNames::First))
return {};
if (!outline_dict->contains(CommonNames::Last))
return {};
HashMap<u32, u32> page_number_by_index_ref;
for (u32 page_number = 0; page_number < m_page_object_indices.size(); ++page_number) {
page_number_by_index_ref.set(m_page_object_indices[page_number], page_number);
}
auto first_ref = outline_dict->get_value(CommonNames::First);
auto children = TRY(build_outline_item_chain(first_ref, page_number_by_index_ref));
m_outline = adopt_ref(*new OutlineDict());
m_outline->children = move(children);
if (outline_dict->contains(CommonNames::Count))
m_outline->count = outline_dict->get_value(CommonNames::Count).get<int>();
return {};
}
PDFErrorOr<Destination> Document::create_destination_from_parameters(NonnullRefPtr<ArrayObject> array, HashMap<u32, u32> const& page_number_by_index_ref)
{
auto page_ref = array->at(0);
if (page_ref.has<nullptr_t>())
return Destination { Destination::Type::XYZ, {}, {} };
auto type_name = TRY(array->get_name_at(this, 1))->name();
Vector<Optional<float>> parameters;
TRY(parameters.try_ensure_capacity(array->size() - 2));
for (size_t i = 2; i < array->size(); i++) {
auto& param = array->at(i);
if (param.has<nullptr_t>())
parameters.unchecked_append({});
else
parameters.append(param.to_float());
}
Destination::Type type;
if (type_name == CommonNames::XYZ) {
type = Destination::Type::XYZ;
} else if (type_name == CommonNames::Fit) {
type = Destination::Type::Fit;
} else if (type_name == CommonNames::FitH) {
type = Destination::Type::FitH;
} else if (type_name == CommonNames::FitV) {
type = Destination::Type::FitV;
} else if (type_name == CommonNames::FitR) {
type = Destination::Type::FitR;
} else if (type_name == CommonNames::FitB) {
type = Destination::Type::FitB;
} else if (type_name == CommonNames::FitBH) {
type = Destination::Type::FitBH;
} else if (type_name == CommonNames::FitBV) {
type = Destination::Type::FitBV;
} else {
VERIFY_NOT_REACHED();
}
// The spec requires page_ref to be an indirect reference to a page object,
// but in practice it's sometimes a page index.
Optional<u32> page_number;
if (page_ref.has<int>())
page_number = page_ref.get<int>();
else
page_number = page_number_by_index_ref.get(page_ref.as_ref_index());
return Destination { type, page_number, parameters };
}
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 TRY(object->get_object(this, name));
}
PDFErrorOr<Optional<Value>> Document::get_inheritable_value(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_value(name, parent);
}
return object->get(name);
}
PDFErrorOr<Destination> Document::create_destination_from_dictionary_entry(NonnullRefPtr<Object> const& entry, HashMap<u32, u32> const& page_number_by_index_ref)
{
if (entry->is<ArrayObject>()) {
auto entry_array = entry->cast<ArrayObject>();
return create_destination_from_parameters(entry_array, page_number_by_index_ref);
}
auto entry_dictionary = entry->cast<DictObject>();
auto d_array = MUST(entry_dictionary->get_array(this, CommonNames::D));
return create_destination_from_parameters(d_array, page_number_by_index_ref);
}
PDFErrorOr<Destination> Document::create_destination_from_object(NonnullRefPtr<Object> const& dest_obj, HashMap<u32, u32> const& page_number_by_index_ref)
{
// PDF 1.7 spec, "8.2.1 Destinations"
if (dest_obj->is<ArrayObject>()) {
auto dest_arr = dest_obj->cast<ArrayObject>();
return TRY(create_destination_from_parameters(dest_arr, page_number_by_index_ref));
}
if (dest_obj->is<NameObject>() || dest_obj->is<StringObject>()) {
DeprecatedFlyString dest_name;
if (dest_obj->is<NameObject>())
dest_name = dest_obj->cast<NameObject>()->name();
else
dest_name = dest_obj->cast<StringObject>()->string();
if (auto dests_value = m_catalog->get(CommonNames::Dests); dests_value.has_value()) {
auto dests = TRY(resolve_to<DictObject>(dests_value.value()));
auto entry = MUST(dests->get_object(this, dest_name));
return TRY(create_destination_from_dictionary_entry(entry, page_number_by_index_ref));
}
if (auto names_value = m_catalog->get(CommonNames::Names); names_value.has_value()) {
auto names = TRY(resolve_to<DictObject>(names_value.release_value()));
if (!names->contains(CommonNames::Dests))
return Error { Error::Type::MalformedPDF, "Missing Dests key in document catalogue's Names dictionary" };
auto dest_obj = TRY(find_in_name_tree(TRY(names->get_dict(this, CommonNames::Dests)), dest_name));
return TRY(create_destination_from_dictionary_entry(dest_obj, page_number_by_index_ref));
}
}
return Error { Error::Type::MalformedPDF, "Malformed outline destination" };
}
PDFErrorOr<NonnullRefPtr<OutlineItem>> Document::build_outline_item(NonnullRefPtr<DictObject> const& outline_item_dict, HashMap<u32, u32> const& page_number_by_index_ref)
{
auto outline_item = adopt_ref(*new OutlineItem {});
if (outline_item_dict->contains(CommonNames::First)) {
VERIFY(outline_item_dict->contains(CommonNames::Last));
auto first_ref = outline_item_dict->get_value(CommonNames::First);
auto children = TRY(build_outline_item_chain(first_ref, page_number_by_index_ref));
for (auto& child : children) {
child->parent = outline_item;
}
outline_item->children = move(children);
}
outline_item->title = text_string_to_utf8(TRY(outline_item_dict->get_string(this, CommonNames::Title))->string());
if (outline_item_dict->contains(CommonNames::Count))
outline_item->count = outline_item_dict->get_value(CommonNames::Count).get<int>();
if (outline_item_dict->contains(CommonNames::Dest)) {
auto dest_obj = TRY(outline_item_dict->get_object(this, CommonNames::Dest));
outline_item->dest = TRY(create_destination_from_object(dest_obj, page_number_by_index_ref));
} else if (outline_item_dict->contains(CommonNames::A)) {
// PDF 1.7 spec, "8.5 Actions"
auto action_dict = TRY(outline_item_dict->get_dict(this, CommonNames::A));
if (action_dict->contains(CommonNames::S)) {
// PDF 1.7 spec, "TABLE 8.48 Action types"
auto action_type = TRY(action_dict->get_name(this, CommonNames::S))->name();
if (action_type == "GoTo") {
// PDF 1.7 spec, "Go-To Actions"
if (action_dict->contains(CommonNames::D)) {
auto dest_obj = TRY(action_dict->get_object(this, CommonNames::D));
outline_item->dest = TRY(create_destination_from_object(dest_obj, page_number_by_index_ref));
}
} else {
dbgln("Unhandled action type {}", action_type);
}
}
}
if (outline_item_dict->contains(CommonNames::C)) {
auto color_array = TRY(outline_item_dict->get_array(this, CommonNames::C));
auto r = static_cast<int>(255.0f * color_array->at(0).to_float());
auto g = static_cast<int>(255.0f * color_array->at(1).to_float());
auto b = static_cast<int>(255.0f * color_array->at(2).to_float());
outline_item->color = Color(r, g, b);
}
if (outline_item_dict->contains(CommonNames::F)) {
auto bitfield = outline_item_dict->get_value(CommonNames::F).get<int>();
outline_item->italic = bitfield & 0x1;
outline_item->bold = bitfield & 0x2;
}
return outline_item;
}
PDFErrorOr<Vector<NonnullRefPtr<OutlineItem>>> Document::build_outline_item_chain(Value const& first_ref, HashMap<u32, u32> const& page_number_by_index_ref)
{
// We used to receive a last_ref parameter, which was what the parent of this chain
// thought was this chain's last child. There are documents out there in the wild
// where this cross-references don't match though, and it seems like simply following
// the /First and /Next links is the way to go to construct the whole Outline
// (we already ignore the /Parent attribute too, which can also be out of sync).
VERIFY(first_ref.has<Reference>());
Vector<NonnullRefPtr<OutlineItem>> children;
auto first_value = TRY(get_or_load_value(first_ref.as_ref_index())).get<NonnullRefPtr<Object>>();
auto first_dict = first_value->cast<DictObject>();
auto first = TRY(build_outline_item(first_dict, page_number_by_index_ref));
children.append(first);
auto current_child_dict = first_dict;
u32 current_child_index = first_ref.as_ref_index();
while (current_child_dict->contains(CommonNames::Next)) {
auto next_child_dict_ref = current_child_dict->get_value(CommonNames::Next);
current_child_index = next_child_dict_ref.as_ref_index();
auto next_child_value = TRY(get_or_load_value(current_child_index)).get<NonnullRefPtr<Object>>();
auto next_child_dict = next_child_value->cast<DictObject>();
auto next_child = TRY(build_outline_item(next_child_dict, page_number_by_index_ref));
children.append(next_child);
current_child_dict = move(next_child_dict);
}
return children;
}
}