SSV stands for separator-separated values. It sounds like a joke, but it can actually be useful.
SSV is a format for saving tabular data (flat-file databases) as delimiter-separated values in a plain text file. In contrast to other formats, e.g. CSV or TSV, delimiter collision is virtually impossible. Instead of using commas (CSV), tabs (TSV) and newlines, SSV relies on the record separator RS and the unit separator US for data structuring, which should not occur at all in textual data.
The format is an example of ASCII delimited text. As there is no official standard, SSV makes decisions that might be considered controversial. In general, simplicity and ease of implementation are preferred to other considerations. For example, SSV does not support escaping of RS and US.
SSV files can be formally described by the following W3C EBNF:
SSV ::= Record (RS Record)*
Record ::= Unit (US Unit)*
Unit ::= UnitChar*
UnitChar ::= Char - (RS | US)
Char ::= /* see http://www.w3.org/TR/xml#NT-Char */
RS ::= #x1E
US ::= #x1F
Table records (rows) are separated by the record separator RS. Each record contains one or more units (fields), which are separated by the unit separator US. A unit may contain zero or more characters, excluding RS and US. As a consequence, an empty file is a valid SSV file and represents a 1-by-1 table containing a single empty field.
import ssv
# simple table
table = [['A1', 'B1', 'C1'],
['A2', 'B2', 'C2']]
# encode table as SSV string
ssv.dumps(table) # 'A1\x1fB1\x1fC1\x1eA2\x1fB2\x1fC2'
# write table to SSV file
ssv.dumpf(table, 'data.ssv')
# load table from SSV file
new_table = ssv.loadf('data.ssv')
assert table == new_table