Skip to content

FileFormat.md

Latest commit

 

History

History
42 lines (37 loc) · 1.67 KB

FileFormat.md

File metadata and controls

42 lines (37 loc) · 1.67 KB

#AdVersion File Format Adversion decomposes ADCIRC meshes across multiple partitions. The following file describes the formats used to create the file format

##Creating Hash Values One critical part of the process for creating unique identifiers for use with the versioning software is creating cryptographic hashes. The format specifiers for this process are critical to ensure that two programs can arrive at the same hash. The hash functions used within this code are embedded within the QADCModules library, but this code dictates how it was implemented there.

###Nodes Nodes are hashed by passing the X and Y (not Z) to the chosen hash algorithm. The x and y values are written to a string using the following format:

%+18.12e

The hash is constructed by successively passing data to the hash algorithm. The code used is below

hash.addData(x_string.toUtf8(),hashSeed1.length());
hash.addData(y_string.toUtf8(),hashSeed2.length());

###Elements Elements are hashed by passing the node hashes for each node into the hash algorithm in its raw form. The code used is below:

hash.addData(node1->positionHash);
hash.addData(node2->positionHash);
hash.addData(node3->positionHash);

###Boundaries Boundaries are created so that any change along the boundary ensures that a new hash is created. The boundary hash is constructed using the following steps:

  1. Hash the boundary code integer
  2. For each node on the boundary, the data is hashed in the order that most closely mathes the standard Adcirc format. Add the data to the hash in this way:
  3. Node 1 hash
  4. Node 2 hash [if applicable]
  5. Weir parameters in the standard Adcirc ordering (Formatted using %+018.12e)