fldfmt(1)(TDH) fldfmt(1)(TDH)
NAMEfldfmt(1) - take a tabular data stream and format, add new field, or
modify a field
SYNOPSISfldfmt [options] [formatstring]
DESCRIPTIONfldfmt is a filter utility that selects fields from a stream of records
and outputs them to standard output. It can be used to build human-
presentable reports, or to process data. Data records should be white-
space delimited, tabular ascii format. One output record is produced
for every input record. Use it in one of these modes:
format mode: Create formatted rows. Fields will be presented as speci‐
fied in a formatstring argument (or format file when -f is used). This
is the default mode. You'll probably want to also use -dp for data
processing purposes.
new field mode: all fields will be output, along with one new field.
See -newfield below
modify field mode: all fields will be output, with one of them changed
to a constant value. See -modfield below
FORMAT MODE
formatstring (or the format file) may contain data field @names or
field @numbers as in the examples below. Inline formatting codes ,
all of which begin with colon (:) may also be used. Everything else in
formatstring is passed through transparently.
formatstring may actually be up to 3 command line arguments, which
fldfmt will concatenate together. This allows long formatstrings to be
split up for readability.
OPTIONS-h
Header mode. Data field names are expected in the first non-
comment non-blank line, separated by white space. The format‐
string can then access fields by @name, instead of just by @num‐
ber.
-t
Input fields expected to be tab-delimited. With -newfield or
-modfield, output fields will also be tab-delimited.
-f formatfile
Output format specification will be taken from formatfile
instead of the command line; same syntax expected. Upper limit
of 3000 characters.
-dp
Data processing mode. Equivalent to -u -c. This is normally
what is desired when doing data processing, ie. when the output
is not for presentation to humans.
-u
Show embedded underscores found in data fields. Normally they
are converted to spaces. Underscore conversion may also be con‐
trolled within formatstring using inline formatting codes :u+
and :u-.
-c
Show nulls literally. Normally nulls are displayed as a zero
length string. Null representation may be controlled from the
project config file.
-newfield fieldpos# contents
Output all fields, and add a new field at position fieldpos#
containing contents. Implies -dp. No formatstring is accepted.
Example: to add a new field containing abc at the beginning of
each output record use fldfmt-newfield 1 "abc"
-modfield fieldpos# contents
Output all fields, and modify the field at position fieldpos# so
that it will contain contents. Implies -dp. No formatstring is
accepted. Example: to change the 3rd field such that it con‐
tains abc on each output record, use fldfmt-modfield 3 "abc"
-l
Turn off evaluation of inline formatting codes
-s
Suppress data-less lines. For lines that contain at least one
variable, and that had all variables evaluate to "" (nothing),
the entire line is suppressed. If format is more than one line,
individual lines are handled separately.
-bh
Output a bogus field name header as the first output line. May
be useful in situations where a header line is expected but not
used.
EXAMPLES
cat mydata | fldfmt "@1 (@2)"
cat mydata | fldfmt "@1 @2 :dec28 @4"
cat scores | fldfmt-h "Name: @firstname @lastname :col40 Score:
@score"
cat scores | fldfmt-newfield 3 "="
cat scores | fldfmt-modfield 6 "d"
The following example makes html table rows containing hyperlinks.
Assumes that CGIPROG is defined to the URL of the CGI executable using
varvalue in the tdh config file.
fldfmt "<td><a href=\"@CGIPROG?rtn=people/details&id=@id\">
@fname @lname</a></td></tr>"
SEE ALSOfldsel(1) , another TDH utility can select and omit fields
19-SEP-2003 TDH scg@jax.org fldfmt(1)(TDH)
