diff --git a/src/fdt/fdt-parser-v2.hpp b/src/fdt/fdt-parser-v2.hpp
index ea4d60e..8c03e4b 100644
--- a/src/fdt/fdt-parser-v2.hpp
+++ b/src/fdt/fdt-parser-v2.hpp
@@ -1,10 +1,9 @@
 #pragma once
 
-#include <type_traits>
-#include <variant>
+#include <cstdint>
 #include <string_view>
+#include <variant>
 #include <vector>
-#include <cstdint>
 
 using u32 = std::uint32_t;
 
@@ -37,6 +36,11 @@ struct context {
     std::string_view structs;
     std::string_view strings;
     token_list tokens;
+
+    struct {
+        const char *data{nullptr};
+        int skip{};
+    } state;
 };
 
 template <typename T>
diff --git a/src/fdt/fdt-parser.cpp b/src/fdt/fdt-parser.cpp
index b14fede..9b92689 100644
--- a/src/fdt/fdt-parser.cpp
+++ b/src/fdt/fdt-parser.cpp
@@ -1,12 +1,14 @@
 #include "fdt-parser.hpp"
 #include "fdt/fdt-header.hpp"
+#include "fdt-parser-v2.hpp"
+
+#include <algorithm>
+#include <cstring>
+#include <iostream>
 #include <ostream>
 #include <string_view>
 #include <variant>
 
-#include <cstring>
-#include <concepts>
-#include <algorithm>
 #include <endian-conversions.hpp>
 
 fdt_parser::fdt_parser(const char *data, u64 size, iface_fdt_generator &generator, const QString &default_root_node, const std::vector<fdt_handle_special_property> &handle_special_properties)
@@ -84,10 +86,6 @@ void fdt_parser::parse(const fdt::header header, iface_fdt_generator &generator)
 }
 */
 
-#include "fdt-parser-v2.hpp"
-#include <iostream>
-#include <algorithm>
-
 auto align(const std::size_t size) {
     const auto q = size % sizeof(u32);
     const auto w = size / sizeof(u32);
@@ -99,42 +97,46 @@ namespace fdt::parser {
 using namespace fdt::tokenizer;
 using namespace fdt::tokenizer::types;
 
-auto parse(node_begin &&token, const char *data, int &skip, context &ctx) -> fdt::tokenizer::token {
-    const auto size = std::strlen(data);
-    token.name = std::string_view(data, size);
-    skip += align(size);
+auto parse(node_begin &&token, context &ctx) -> fdt::tokenizer::token {
+    const auto size = std::strlen(ctx.state.data);
+    token.name = std::string_view(ctx.state.data, size);
+    ctx.state.skip += align(size + 1);
     return {std::move(token)};
 }
 
-auto parse(node_end &&token, const char *data, int &skip, context &ctx) -> fdt::tokenizer::token {
+auto parse(node_end &&token, context &ctx) -> fdt::tokenizer::token {
     return {std::move(token)};
 }
 
-auto parse(types::property &&token, const char *data, int &skip, context &ctx) -> fdt::tokenizer::token {
-    const auto header = read_data_32be<fdt::property>(data);
-    skip += align(sizeof(header)) + align(header.len);
-    data += sizeof(header);
+auto parse(types::property &&token, context &ctx) -> fdt::tokenizer::token {
+    const auto header = read_data_32be<fdt::property>(ctx.state.data);
+    ctx.state.skip += align(sizeof(header)) + align(header.len);
+    ctx.state.data += sizeof(header);
 
-    token.name = std::string_view();
-    token.data = std::string_view(data, header.len);
+    const auto property = ctx.strings.data() + header.nameoff;
+
+    token = {
+        .name = std::string_view(property, std::strlen(property)),
+        .data = std::string_view(ctx.state.data, header.len),
+    };
 
     return {std::move(token)};
 }
 
-auto parse(nop &&token, const char *data, int &skip, context &ctx) -> fdt::tokenizer::token {
+auto parse(nop &&token, context &ctx) -> fdt::tokenizer::token {
     return {std::move(token)};
 }
 
-auto parse(end &&token, const char *data, int &skip, context &ctx) -> fdt::tokenizer::token {
+auto parse(end &&token, context &ctx) -> fdt::tokenizer::token {
     return {std::move(token)};
 }
 } // namespace fdt::parser
 
 template <fdt::tokenizer::Tokenizable... Ts>
-auto foreach_token_type(std::variant<Ts...>, const u32 token_id, const char *data, int &skip, fdt::tokenizer::context &ctx) {
+auto foreach_token_type(std::variant<Ts...>, const u32 token_id, fdt::tokenizer::context &ctx) {
     auto conditional_parse = [&](auto &&token) {
         if (fdt::tokenizer::types::id_of(token) == token_id) {
-            ctx.tokens.emplace_back(fdt::parser::parse(std::move(token), data, skip, ctx));
+            ctx.tokens.emplace_back(fdt::parser::parse(std::move(token), ctx));
             return true;
         }
         return false;
@@ -158,7 +160,11 @@ void fdt_parser::parse(const fdt::header header, iface_fdt_generator &generator)
 
     using namespace fdt::tokenizer;
     using namespace fdt::tokenizer::types;
-    context ctx;
+
+    context ctx{
+        .structs = {dt_struct, header.size_dt_struct}, //
+        .strings = {dt_strings, header.size_dt_strings}, //
+    };
 
     const auto begin = reinterpret_cast<const u32 *>(dt_struct);
     const auto end = reinterpret_cast<const u32 *>(dt_struct) + header.size_dt_struct / sizeof(u32);
@@ -167,12 +173,16 @@ void fdt_parser::parse(const fdt::header header, iface_fdt_generator &generator)
 
     for (auto iter = begin; iter != end;) {
         const auto id = static_cast<u32>(convert(*iter));
-        iter++;
-        std::cout << id << std::endl;
+        ctx.state.data = reinterpret_cast<const char *>(++iter);
+        ctx.state.skip = 0;
+
+        if (!foreach_token_type(token{}, id, ctx))
+            break;
+
+        // if (!ctx.tokens.empty() && std::holds_alternative<types::end>(ctx.tokens.back()))
+        //     break;
 
-        auto skip = 0;
-        foreach_token_type(token{}, id, reinterpret_cast<const char *>(iter), skip, ctx);
-        iter += skip;
+        iter += ctx.state.skip;
     }
 
     const auto node_begin_count = std::ranges::count_if(ctx.tokens, [](auto &&v) {
@@ -195,21 +205,29 @@ void fdt_parser::parse(const fdt::header header, iface_fdt_generator &generator)
         return std::holds_alternative<types::end>(v);
     });
 
-    std::cout << "is_aligned       : " << is_aligned << std::endl;
-    std::cout << "node begin  count: " << node_begin_count << std::endl;
-    std::cout << "node end    count: " << node_end_count << std::endl;
-    std::cout << "property    count: " << property_count << std::endl;
-    std::cout << "nop         count: " << nop_count << std::endl;
-    std::cout << "end         count: " << end_count << std::endl;
+    std::cout << "is_aligned : " << is_aligned << std::endl;
+    std::cout << "node begin : " << node_begin_count << std::endl;
+    std::cout << "node end   : " << node_end_count << std::endl;
+    std::cout << "property   : " << property_count << std::endl;
+    std::cout << "nop        : " << nop_count << std::endl;
+    std::cout << "end        : " << end_count << std::endl;
 
     for (auto &&token : ctx.tokens)
         std::visit(overloaded{
                        [&](node_begin &arg) { generator.begin_node(QString::fromUtf8(arg.name.data(), arg.name.size())); },
                        [&](node_end &arg) { generator.end_node(); },
-                       [&](property &arg) { generator.insert_property(fdt_property{
-                                                .name = QString::fromUtf8(arg.name.data(), arg.name.size()),
-                                                .data = QByteArray(arg.data.data(), arg.data.size()),
-                                            }); },
+                       [&](property &arg) {
+                           auto property = fdt_property{
+                               .name = QString::fromUtf8(arg.name.data(), arg.name.size()),
+                               .data = QByteArray(arg.data.data(), arg.data.size()),
+                           };
+
+                           generator.insert_property(property);
+
+                           // for (auto &&handle : m_handle_special_properties)
+                           //     if (handle.name.toStdString() == arg.name)
+                           //         handle.callback(property, generator);
+                       },
                        [&](nop &) {},
                        [&](types::end &) {},
                    },