From db0d1f348c3db6c69bb385979742a2e5cea9f649 Mon Sep 17 00:00:00 2001
From: Dave Johnson <zer0dave@outlook.com>
Date: Fri, 16 Oct 2026 17:46:37 +0000
Subject: [PATCH] EBX writer: rewritten arrays keep their own record hash

Each array record carries a hash the original files make specific to
the field or element type; the writer used to give every array the hash
of the first record of the same element type, so a rewrite changed
hundreds of them in bam's level root. Arrays read from a document now
keep their record's hash; only arrays with no source record borrow as
before.
---
 Engine/Resource/ebx_document.cpp | 20 ++++++++++++++++++--
 Engine/Resource/ebx_document.h   |  4 ++++
 Engine/Resource/ebx_writer.cpp   |  8 ++++++--
 3 files changed, 28 insertions(+), 4 deletions(-)

diff --git a/Engine/Resource/ebx_document.cpp b/Engine/Resource/ebx_document.cpp
index efadfaa..a81ef1d 100644
--- a/Engine/Resource/ebx_document.cpp
+++ b/Engine/Resource/ebx_document.cpp
@@ -1,6 +1,7 @@
 #include "ebx_document.h"
 
 #include <algorithm>
+#include <cstring>
 #include <limits>
 #include <stdexcept>
 #include <unordered_map>
@@ -356,7 +357,9 @@ void map_instances(Document& document,
 class ValueReader final {
 public:
     ValueReader(const std::span<const std::byte> bytes, Document& document)
-        : bytes_(bytes), reader_(bytes), document_(document) {}
+        : bytes_(bytes), reader_(bytes), document_(document) {
+        for (const auto& record : document.arrays) arrayHashes_.emplace(record.offset, record.hash);
+    }
 
     void read_instances() {
         for (auto& instance : document_.instances) {
@@ -423,7 +426,19 @@ private:
             }
             const auto position = start + field.dataOffset;
             if (position > document_.dataEnd) throw std::runtime_error("EBX field offset exceeds EBXD");
-            output.fields.push_back({descriptorIndex, field.name, read_field(field, position)});
+            auto value = read_field(field, position);
+            if (field.category() == FieldCategory::array) {
+                const auto* values = std::get_if<Value::Array>(&value.data);
+                if (values && !values->empty()) {
+                    std::int32_t displacement{};
+                    std::memcpy(&displacement, bytes_.data() + position, sizeof(displacement));
+                    const auto target = static_cast<std::int64_t>(position) + displacement -
+                                        static_cast<std::int64_t>(document_.dataStart);
+                    if (const auto found = arrayHashes_.find(target); found != arrayHashes_.end())
+                        value.arrayHash = found->second;
+                }
+            }
+            output.fields.push_back({descriptorIndex, field.name, std::move(value)});
         }
     }
 
@@ -552,6 +567,7 @@ private:
     }
 
     std::span<const std::byte> bytes_;
+    std::unordered_map<std::int64_t, std::uint32_t> arrayHashes_;   // record offset -> its hash
     BinaryReader reader_;
     Document& document_;
     unsigned depth_{};
diff --git a/Engine/Resource/ebx_document.h b/Engine/Resource/ebx_document.h
index ce63e81..947558d 100644
--- a/Engine/Resource/ebx_document.h
+++ b/Engine/Resource/ebx_document.h
@@ -6,6 +6,7 @@
 #include <cstdint>
 #include <span>
 #include <memory>
+#include <optional>
 #include <string>
 #include <variant>
 #include <vector>
@@ -149,6 +150,9 @@ struct Value {
                                  PointerReference, TypeReference, BoxedReference,
                                  std::shared_ptr<struct Object>, Array>;
     Storage data;
+    // An array's own record hash, kept when the document is written again; a new array has none
+    // and borrows one (ebx_writer.cpp, array_hash).
+    std::optional<std::uint32_t> arrayHash;
 };
 
 struct FieldValue {
diff --git a/Engine/Resource/ebx_writer.cpp b/Engine/Resource/ebx_writer.cpp
index 1924d66..23b4ea7 100644
--- a/Engine/Resource/ebx_writer.cpp
+++ b/Engine/Resource/ebx_writer.cpp
@@ -6,6 +6,7 @@
 #include <cmath>
 #include <cstring>
 #include <limits>
+#include <optional>
 #include <ranges>
 #include <stdexcept>
 #include <string_view>
@@ -106,6 +107,7 @@ private:
         std::uint32_t slot{};
         const Value::Array* values{};
         std::size_t descriptor{};
+        std::optional<std::uint32_t> hash;
         bool emitted{};
     };
 
@@ -208,7 +210,7 @@ private:
                 static const Value::Array empty;
                 const auto* array = value ? std::get_if<Value::Array>(&value->data) : nullptr;
                 arraysPending_.push_back({checked_u32(position, "EBX array slot"), array ? array : &empty,
-                                          descriptorIndex});
+                                          descriptorIndex, value ? value->arrayHash : std::nullopt});
             } else if (field.type() == FieldType::structure) {
                 if (field.classRef >= document_.types.size()) throw std::runtime_error("EBX struct type is invalid");
                 const auto nestedStart = align_up(position, std::max<std::size_t>(document_.types[field.classRef].alignment, 1));
@@ -505,8 +507,10 @@ private:
                 write_scalar(field, elementStart, &value);
             }
         }
+        // An array read from a record keeps that record's hash; a new one borrows (array_hash).
+        const auto hash = arraysPending_[index].hash ? *arraysPending_[index].hash : array_hash(field);
         arrays_.push_back({checked_u32(data, "EBX array data"), checked_u32(values->size(), "EBX array count"),
-                           array_hash(field), static_cast<std::uint16_t>(field.encodedType << 1),
+                           hash, static_cast<std::uint16_t>(field.encodedType << 1),
                            field.type() == FieldType::structure
                                ? fixup_slot_for_descriptor(field.classRef)
                                : static_cast<std::uint16_t>(0xFFFF)});