[<<Previous Entry] [^^Up^^] [Next Entry>>] [Menu] [About The Guide]
Structure of the .fdx file
--------------------------

        The fdx file is a B+ tree containing 3 major types of nodes; root,
branch and leaf. The root node contains chksum bytes so that the Force runtime
system can verify the index and a pointer to the 1st branch. Branches can point
to either another branch or to a leaf. Leaf nodes point to nothing an contain
the actual key data.

        The nodes, as of this writing, are 4096 in size.

        Each node contains a header. The structure of the header is as follows:

Structure of Node Header
-------------------------

        byte            meaning
        -----------------------
         0-1            word: node type.
                                1 = branch
                                2 = leaf
                                3 = root

         2-3            word: number of Key Records in node
         4-7            dword: used at runtime
         8-11           dword: used at runtime. If the root node,
                                points to the 1st branch
                

        Following the Node Header are a sequence of N Key Records. Each
Key Record has the following structure:

        byte            meaning
        -----------------------
         0-3            a) leaf node:
                                contains the record number for the key
                        b) branch node:
                                contains the file offset of the leaf
                                containing the key
                        c) root node:
                                (does not contain key records)

         4-N            Key value
         N+1-N+4        contains record number of key

See Also: HEADER_EQUATES NODE_EQUATES COMPILER_RECORD
This page created by ng2html v1.05, the Norton guide to HTML conversion utility. Written by Dave Pearson