The source of ParPerApp is a database table containing one line per witness at each variant passage. Fields are separated by tab stops. ParPerApp.txt can be imported into a database table with the following structure:

BOOK	int(2)
BC	int(2)
BV	int(2)
BW	int(3)
EW	int(3)
VID	varchar(2)
RDG	varchar(254)
WIT	varchar(10)
EC	int(2)
EV	int(2)
VIDS	varchar(50)


BOOK:	Number of writing - Matthew:1, Mark:2, Luke:3
BC:	BeginChapter - Chapter containing the first word of the variant passage
BV:	BeginVerse - Verse containing the first word of the variant passage
BW:	BeginWord - Number of the first word of the variant passage
EW: 	EndWord - Number of the last word of the variant passage
VID:	VariantIdentifier - Letter identifying the variant or a deficiency (see below)
RDG:	Reading 
WIT:	Witness
EC:	EndChapter - Chapter containing the last word of the variant passage
EV:	EndVerse - Verse containing the last word of the variant passage
VIDS:	VariantIdentifierSuffix - Suffix of VID like 'f' for 'error', 'o' for 'orthographical difference' or, with VID = 'zw', the range of variants possibly supported by a lacunose witness.

ParPerApp.txt can also be imported into a spreadsheet by specifying that fields are separated by tab stops.


VID also contains the letter combinations 'zw', 'zy' and 'zz'.
zw corresponds to the double arrow in the printed volume. It cannot be determined which of the variant readings noted with the double arrow is supported by the witness(es) cited with 'zw'.
zy corresponds to the upward arrow in place of the letter address. The witness(es) listed cannot be cited for a variant, because of an overriding previous note.
zz corresponds to the list of deficient manuscripts given for each variant passage with the minus sign in print. Note that a witness cited there may be additionally cited with 'V' for a variant or under the double arrow, even though the text of the variant lacks one whole word at least.