LibXML: Add a fairly basic XML parser

Currently this can parse XML and resolve external resources/references,
and read a DTD (but not apply or verify its rules).
That's good enough for _most_ XHTML documents as the HTML 5 spec
enforces its own rules about document well-formedness, and does not make
use of XML DTDs (aside from a list of predefined entities).

An accompanying `xml` utility is provided that can read and dump XML
documents, and can also run the XML conformance test suite.

Userland/Libraries/LibXML/DOM/Document.h

@@ -0,0 +1,53 @@
+/*
+ * Copyright (c) 2022, Ali Mohammad Pur <mpfard@serenityos.org>
+ *
+ * SPDX-License-Identifier: BSD-2-Clause
+ */
+
+#pragma once
+
+#include <AK/HashMap.h>
+#include <AK/NonnullOwnPtr.h>
+#include <LibXML/DOM/DocumentTypeDeclaration.h>
+#include <LibXML/DOM/Node.h>
+#include <LibXML/Forward.h>
+
+namespace XML {
+
+enum class Version {
+    Version10,
+    Version11,
+};
+
+struct Doctype {
+    String type;
+    Vector<MarkupDeclaration> markup_declarations;
+    Optional<ExternalID> external_id;
+};
+
+class Document {
+public:
+    explicit Document(NonnullOwnPtr<Node> root, Optional<Doctype> doctype, HashMap<Name, String> processing_instructions, Version version)
+        : m_root(move(root))
+        , m_processing_instructions(move(processing_instructions))
+        , m_version(version)
+        , m_explicit_doctype(move(doctype))
+    {
+    }
+
+    Node& root() { return *m_root; }
+    Node const& root() const { return *m_root; }
+
+    HashMap<Name, String> const& processing_instructions() const { return m_processing_instructions; }
+
+    Version version() const { return m_version; }
+
+    Optional<Doctype> const& doctype() const { return m_explicit_doctype; }
+
+private:
+    NonnullOwnPtr<Node> m_root;
+    HashMap<Name, String> m_processing_instructions;
+    Version m_version;
+    Optional<Doctype> m_explicit_doctype;
+};
+}

Userland/Libraries/LibXML/DOM/DocumentTypeDeclaration.h

@@ -0,0 +1,138 @@
+/*
+ * Copyright (c) 2022, Ali Mohammad Pur <mpfard@serenityos.org>
+ *
+ * SPDX-License-Identifier: BSD-2-Clause
+ */
+
+#pragma once
+
+#include <AK/HashTable.h>
+#include <AK/String.h>
+#include <AK/Variant.h>
+#include <AK/Vector.h>
+#include <LibXML/FundamentalTypes.h>
+
+namespace XML {
+
+struct ElementDeclaration {
+    struct Empty {
+    };
+    struct Any {
+    };
+    struct Mixed {
+        HashTable<Name> types;
+        bool many;
+    };
+    struct Children {
+        struct Entry;
+        enum class Qualifier {
+            ExactlyOnce,
+            Optional,
+            Any,
+            OneOrMore,
+        };
+
+        struct Choice {
+            Vector<Entry> entries;
+            Qualifier qualifier;
+        };
+        struct Sequence {
+            Vector<Entry> entries;
+            Qualifier qualifier;
+        };
+
+        struct Entry {
+            Variant<Name, Choice, Sequence> sub_entries;
+            Qualifier qualifier;
+        };
+
+        Variant<Choice, Sequence> contents;
+        Qualifier qualifier;
+    };
+    using ContentSpec = Variant<Empty, Any, Mixed, Children>;
+
+    Name type;
+    ContentSpec content_spec;
+};
+
+struct AttributeListDeclaration {
+    enum class StringType {
+        CData,
+    };
+    enum class TokenizedType {
+        ID,
+        IDRef,
+        IDRefs,
+        Entity,
+        Entities,
+        NMToken,
+        NMTokens,
+    };
+    struct NotationType {
+        HashTable<Name> names;
+    };
+    struct Enumeration {
+        // FIXME: NMToken
+        HashTable<String> tokens;
+    };
+    using Type = Variant<StringType, TokenizedType, NotationType, Enumeration>;
+
+    struct Required {
+    };
+    struct Implied {
+    };
+    struct Fixed {
+        String value;
+    };
+    struct DefaultValue {
+        String value;
+    };
+
+    using Default = Variant<Required, Implied, Fixed, DefaultValue>;
+
+    struct Definition {
+        Name name;
+        Type type;
+        Default default_;
+    };
+    Name type;
+    Vector<Definition> attributes;
+};
+
+struct PublicID {
+    String public_literal;
+};
+
+struct SystemID {
+    String system_literal;
+};
+
+struct ExternalID {
+    Optional<PublicID> public_id;
+    SystemID system_id;
+};
+
+struct EntityDefinition {
+    ExternalID id;
+    Optional<Name> notation;
+};
+
+struct GEDeclaration {
+    Name name;
+    Variant<String, EntityDefinition> definition;
+};
+
+struct PEDeclaration {
+    Name name;
+    Variant<String, ExternalID> definition;
+};
+
+using EntityDeclaration = Variant<GEDeclaration, PEDeclaration>;
+
+struct NotationDeclaration {
+    Name name;
+    Variant<ExternalID, PublicID> notation;
+};
+
+using MarkupDeclaration = Variant<ElementDeclaration, AttributeListDeclaration, EntityDeclaration, NotationDeclaration>;
+}

Userland/Libraries/LibXML/DOM/Node.cpp

@@ -0,0 +1,54 @@
+/*
+ * Copyright (c) 2022, Ali Mohammad Pur <mpfard@serenityos.org>
+ *
+ * SPDX-License-Identifier: BSD-2-Clause
+ */
+
+#include <AK/HashMap.h>
+#include <LibXML/DOM/Node.h>
+
+namespace XML {
+
+bool Node::operator==(Node const& other) const
+{
+    return content.visit(
+        [&](Text const& text) -> bool {
+            auto other_text = other.content.get_pointer<Text>();
+            if (!other_text)
+                return false;
+            return text.builder.string_view() == other_text->builder.string_view();
+        },
+        [&](Comment const& comment) -> bool {
+            auto other_comment = other.content.get_pointer<Comment>();
+            if (!other_comment)
+                return false;
+            return comment.text == other_comment->text;
+        },
+        [&](Element const& element) -> bool {
+            auto other_element = other.content.get_pointer<Element>();
+            if (!other_element)
+                return false;
+            if (element.name != other_element->name)
+                return false;
+            if (element.attributes.size() != other_element->attributes.size())
+                return false;
+
+            for (auto& entry : element.attributes) {
+                auto it = other_element->attributes.find(entry.key);
+                if (it == other_element->attributes.end())
+                    return false;
+                if (it->value != entry.value)
+                    return false;
+            }
+
+            if (element.children.size() != other_element->children.size())
+                return false;
+            for (size_t i = 0; i < element.children.size(); ++i) {
+                if (element.children[i] != other_element->children[i])
+                    return false;
+            }
+            return true;
+        });
+}
+
+}

Userland/Libraries/LibXML/DOM/Node.h

@@ -0,0 +1,40 @@
+/*
+ * Copyright (c) 2022, Ali Mohammad Pur <mpfard@serenityos.org>
+ *
+ * SPDX-License-Identifier: BSD-2-Clause
+ */
+
+#pragma once
+
+#include <AK/NonnullOwnPtrVector.h>
+#include <AK/String.h>
+#include <AK/Variant.h>
+#include <AK/Vector.h>
+#include <LibXML/FundamentalTypes.h>
+
+namespace XML {
+
+struct Attribute {
+    Name name;
+    String value;
+};
+
+struct Node {
+    struct Text {
+        StringBuilder builder;
+    };
+    struct Comment {
+        String text;
+    };
+    struct Element {
+        Name name;
+        HashMap<Name, String> attributes;
+        NonnullOwnPtrVector<Node> children;
+    };
+
+    bool operator==(Node const&) const;
+
+    Variant<Text, Comment, Element> content;
+    Node* parent { nullptr };
+};
+}