USAGE : cdx uniq [options…] [file]
| -k | –key Spec | How to compare adjacent lines. Spec is any of the standard Comparisons |
| -c | –count=ColName,Position | Add count of matched lines. See below. |
| -w | –which=Pos,Comp | Which of the matching lines is displayed? See below. |
| -a | –agg=Spec | Merge values from this column, in place. See below. |
| –agg-pre | Merge values from this column, into new column. See below. | |
| –agg-post | Merge values from this column, into new column. See below. | |
| –agg-help | Print help for Aggregators |
If any adjacent lines compare equal, only the first is printed.
Count takes a two part, comma delimited parameter. The first part is th new column name, which defaults to ‘count’. The second part is ‘begin’ or ‘end’, saying that th new column should either be first or last in the output file. The defaul is ‘begin’, so -c , will produce a new first column named ‘count’, and --count Num,End will produce a new final column named ‘Num’.
When several line are equal (according to --key) which line is printed?
Sometimes you don’t want to select one representitive line, but rather you need tom aggregate the values from the input column values into a new column value. There are three ways to specify aggregation
--agg replace a column with its aggregate. It is an error to specify the same column twice in this way. The format is SrcCol,Spec where SrcCol is the column to be aggregated and Spec is the desired Aggregator--agg-pre Create a new column, placed before the columns of the input file. The format is NewCol,SrcCol,Spec where NewCol is the name of the new column, and the rest is as in --agg--agg-post Just like –agg-pre, but the new column is places after the original columns.There can be many different Comparators in opertion at the same time. For example
cdx uniq --key title,lower --which Max,price,float --agg year,max,numeric
The lines of the file are compared by the case-insensitve text of the title field. Among those lines with matching titles, the line
with the highest price is selected, except that the output year column will be the latest year mentioned.