Minor bug fixes (#2128)

- make sure bsp is processed on `l` levels in extraction (caused missing
remaps)
- clean up a few prints in extraction
- handle the <15 byte differences in art group files automatically (no
more errors about file naming)
- fix potential exception thrown by merc2 in a few ways: fixed bad data
in FR3's, check texture index just in case, and handle exceptions a
little bit better (still a crash, but at least you get a print)
- fix mips2 ocean stuff causing ocean far crashes

decompiler/ObjectFile/LinkedObjectFileCreation.cpp

@@ -231,6 +231,7 @@ static void link_v2_or_v4(LinkedObjectFile& f,
                           const std::string& name,
                           DecompilerTypeSystem& dts,
                           GameVersion version) {
+  (void)name;
   const auto* header = (const LinkHeaderV4*)&data.at(0);
   ASSERT(header->version == 4 || header->version == 2);
 
@@ -313,7 +314,8 @@ static void link_v2_or_v4(LinkedObjectFile& f,
           for (uint8_t i = 0; i < count; i++) {
             if (!f.pointer_link_word(0, code_ptr_offset - code_offset, 0,
                                      *((const uint32_t*)(&data.at(code_ptr_offset))))) {
-              lg::error("Skipping link in {} because it is out of range!", name.c_str());
+              // was this just a bug in the linker??
+              // lg::error("Skipping link in {} because it is out of range!", name.c_str());
             }
             f.stats.total_v2_pointers++;
             code_ptr_offset += 4;
@@ -680,13 +682,6 @@ static void link_v3(LinkedObjectFile& f,
         segment_size++;
         adjusted = true;
       }
-
-      if (adjusted) {
-        printf(
-            "Adjusted the size of segment %d in %s, this is fine, but rare (and may indicate a "
-            "bigger problem if it happens often)\n",
-            seg_id, name.c_str());
-      }
     }
 
     auto base_ptr = segment_data_offsets[seg_id];

decompiler/ObjectFile/ObjectFileDB.cpp

@@ -312,6 +312,55 @@ void ObjectFileDB::get_objs_from_dgo(const fs::path& filename, const Config& con
   ASSERT(0 == reader.bytes_left());
 }
 
+/*!
+ * Are two object files the same?
+ * Unfortunately they seemed to have a memory bug in their art-group generator, so there's some
+ * uninitialized padding bytes.
+ */
+bool are_objects_the_same(const std::string& obj_name,
+                          size_t size_a,
+                          const u8* a,
+                          size_t size_b,
+                          const u8* b) {
+  if (size_a != size_b) {
+    return false;
+  }
+
+  // if they are byte-for-byte the same, it's a match
+  if (!memcmp(a, b, size_a)) {
+    return true;
+  }
+
+  // if it's an art group...
+  if (obj_name.size() > 3 && !obj_name.compare(obj_name.length() - 3, 3, "-ag")) {
+    // count up the number of differing bytes, and the location of the first one.
+    size_t first_diff = 0;
+    size_t last_diff = 0;
+    int num_diffs = 0;
+    bool found_first_diff = false;
+    for (size_t i = 0; i < size_a; i++) {
+      if (a[i] != b[i]) {
+        num_diffs++;
+        last_diff = i;
+        if (!found_first_diff) {
+          first_diff = i;
+          found_first_diff = true;
+        }
+      }
+    }
+
+    // find the gap between "code" (really data here) and link table. This has up to 15 bytes of
+    // uninitialized memory.
+    const auto* header = (const LinkHeaderV4*)a;
+    int link_data_offset = header->code_size + sizeof(LinkHeaderV4);
+    int start_off_diff_from_code_end = link_data_offset - (int)first_diff;
+    if (num_diffs < 16 && start_off_diff_from_code_end < 16 && (int)last_diff < link_data_offset) {
+      return true;
+    }
+  }
+  return false;
+}
+
 /*!
  * Add an object file to the ObjectFileDB
  */
@@ -337,10 +386,7 @@ void ObjectFileDB::add_obj_from_dgo(const std::string& obj_name,
   bool duplicated = false;
   // first, check to see if we already got it...
   for (auto& e : obj_files_by_name[obj_name]) {
-    if (e.data.size() == obj_size && e.record.hash == hash) {
-      // just to make sure we don't have a hash collision.
-      ASSERT(!memcmp(obj_data, e.data.data(), obj_size));
-
+    if (are_objects_the_same(obj_name, e.data.size(), e.data.data(), obj_size, obj_data)) {
       // already got it!
       e.reference_count++;
       auto& rec = e.record;