AQUA analysis tools
Using qdbext to extract data
The 'qdbext' script can be used to extract data by block and item
(=column) from a file which is formatted according to the
DBAS file format. Examples of such files are the
'.?rv' (violations) and '.?rc' (counts) files produced by
'aqpc'.
The command syntax is as follows:
qdbext [b=BLOCK_NAME [i=ITEM_NAME] [hdr=none] [id=show|ident] ] FILE
where
- BLOCK_NAME is the name of the data block to be selected.
It can be a regular expression, or: #BLOCK_NUMBER". If not specified,
all data blocks in the file are listed by name and number, and the user
may choose interactively.
- ITEM_NAME is the name of a data item (column) occurring in the
selected block.
- [hdr] is optional flag for printing of a header, consisting of
the full $...
block description (in case of block extraction) or a line beginning with #
(in case of item extraction).
hdr=none: header is not printed
default: header is printed
- [id] is optional flag for printing of ident (see
DBAS file format).
id=show: ident is printed before the data item (item extraction)
id=ident: ident is printed as "$ IDENT ident" line preceding the data
default: ident is not printed
The matches on block and item names are performed case-INsensitive.
Example
To extract the total number of upper bound violations for every structure
from a "violations" file produced by 'aqpc' you might use the command
qdbext b=SUMMARY_per_STRUCTURE i=numuppv myprot.nrv
Remark
If more than one block with the same name occurs in the file, 'qdbext' will
process them all. Intervening "$ IDENT ident" lines (cf.
DBAS file format)
can be used to discriminate between blocks. This facility has been useful
in processing a stream of DBAS data produced by concatening files from a
series of entries.