This is information for ippToPsps for ingesting the PV2 version of the 3pi database.  Substantial changes have been made.


== Batch types ==

There are now 8 types of batches, currently based off of SchemaNew16.0.tar  - there are only minor corrections of typos / datatypes / addition of gcobjID 
 
|| 2 digit || type  || schema ||
||IN || init batch ||  [wiki:ippToPspsPV2_IN]  ||
|| P2 ||single detections|| [wiki:ippToPspsPV2_P2]  ||
||ST ||stack|| [wiki:ippToPspsPV2_ST]   ||
||OB ||object||  [wiki:ippToPspsPV2_OB]  ||
||FW || forced warp measurement|| [wiki:ippToPspsPV2_FW]  ||
||FO || forced object|| [wiki:ippToPspsPV2_FO]  ||
||DF ||diff detection||  [wiki:ippToPspsPV2_DF]  ||
||DO || diff object|| [wiki:ippToPspsPV2_DO]   ||


== Schema Manipulation ==

There are multiple versions / types of the schema

 * (schemaNewXX.0.tar )  -  These are not complete ( the filters are not expanded), there is extra text, units/datatypes may not be consistent.  This is what Heather works from when editing the VO Tables. The column names and orders are definitely correct, use caution with definitions/datatypes/etc.

 * VO Tables  - What ipptopsps uses - Heather has edited them to contain all the new types of batches, edited the descriptions to make them more accurate, this is the definitive list of everything.

 * wiki /csv file - generated from a parsing script using VO Tables as input.  These are what conrad/thomas work from

=== Conversion of datatypes ===

The parsing script takes the datatypes in the votables (which is what ipptopsps uses in jython) and converts them to microsoft. This is the conversion table:

|| jython datatype   || jython array size  ||     Microsofted  ||
|| float       ||  1 || REAL 4 ||
|| double   || 1 || FLOAT 8 ||
||int ||1|| INT 4||
||short ||1|| SMALLINT 2||
||unsignedByte || 1 || TINYINT 1||
||long || 1 || BIGINT 8||
||char || num || VARCHAR(num) ||