More Clickteam notes

BinaryObjectScanner/Packer/ClickteamInstallMaker.cs

@@ -4,9 +4,22 @@ namespace BinaryObjectScanner.Packer
     public class ClickteamInstallMaker
     {
         // All data stored in the overlay
+
         // Header is zlib-compressed. Samples seem to indicate that
         // there is a 19-byte "header" before the compressed header
-        // data.
+        // data. 5 bytes before the compressed data appears to be a
+        // 32-bit value that represents the decompressed header size.
+        // It is followed by a single byte
+
+        // Possible layout of pre-compressed header:
+        // 0x00-0x03 - Unknown [77 77 67 54] (1416066935)
+        // 0x04-0x07 - Unknown [29 48 35 14] (339036201)
+        // 0x08-0x09 - Unknown [01 00] (1)
+        // 0x0A-0x0D - Size of compressed data? [54 02 00 00] (596)
+        //      This seems shorter than the expected 604
+        // 0x0E-0x11 - Size of decompressed data [26 08 00 00] (2086)
+        // 0x12      - Unknown [01] (1)
+
         // The decompressed header data has no easily-recognized
         // structures like offsets or sizes.
 
@@ -15,6 +28,20 @@ namespace BinaryObjectScanner.Packer
         // can be found. This results in the entries all being read
         // at the same time and failing to extract as a result.
 
+        // It is possible there is an 8-byte "prefix" before the entries
+        // in the data itself, but no values have been correllated
+        // to either the compressed or extracted sizes yet.
+        // This prefix could also be a data trailer, as multiple values
+        // have been found seemingly-appended to data extracted
+        // from these bz2 entries.
+
+        // Possible layout of pre-compressed entry:
+        // 0x00-0x03 - Unknown [10 9D 01 00] (105744)
+        // 0x04-0x07 - Size of compressed data? [DA 12 02 00] (135898)
+        //      This seems shorter than the expected 137949
+        //      What is in the 2051 bytes remaining?
+        // 0x08      - Unknown [02] (2)
+
         // When manually extracted, each BZ2 block is a correct
         // set of compressed data. Without a notable filename map,
         // the extracted data is difficult to work with.