OutputExplanations.txt

For Hecate to work, in each project  folder 
-------------------------------------------------------------
Input: schemata
  ... go through Hecate ...
Output: results

Input: 
a list of .sql files, each with the DDL statements of the version. 
The names of the files must reflect their time order (ideally, the name of the file is the Unix time of its commit to github)

Output:
You can find several files inside results

/* A. Detailed reporting on changes */
- transitions.xml: the xml-encoded version of the following 
- transitions.csv: for each transition, for each table affected, for each attribute affected, what was the change

Explanations
----------------------
Insertion: New table => attr born with new table
Insertion: Update table => attr injected to existing  table
Deletion: DeleteTable => attr dies with dying table
Deletion: Update table => attr ejected from existing  table
Update: TypeChange => data type of attr changes
Update: KeyChange => attr involved in primary key that changes


/* B. Summaries of changes. These two  tell us cumulative stats for (a) tables and (b) transitions */
- tablesDetailedStats.tsv: for each table, cumulative stats
- SchemaHeartbeat.tsv: for each transition, cumulative stats
// you can also find their old versions, metrics.csv and tables_stats.csv

/* C. super detailed stats. Per type of change, per table, per transition */
_perVersion*.csv: One file per type of change. Per table (row), per transition (column) how many attributed where changed.
The only possibly unclear: ConcatAllMetrics.csv
For example, if for a certain table, for a certain transition, the cell is
     7[0|0|0|0] 
it means that this table, at this transition had 7 attributes@schema [0 attr's inserted | 0 attr's deleted | 0 attr's with attrTypeChange | 0 attr's with keyChange]
		---
		writeText(metricsOverVersion.getSize(i) + "[" + c.toString() + "]" + ";");
		---
		return this.insertions + "|" + this.deletions + "|" +
		       this.attributeTypeChange + "|" + this.keyChange