tupelo.csv — tupelo 21.03.25

Utils for reading CSV (comma-separated-value) formatted files.

attrs->entities clj

(attrs->entities cols-map)

Inputs: [cols-map :- tsk/Map] Returns: [tsk/Map]

Converts a map of attribute-vectors into a vector of entity-maps. Not lazy.

entities->attrs clj

(entities->attrs row-maps)

Inputs: [row-maps :- [tsk/Map]] Returns: tsk/Map

Converts a sequence of entity-maps into a map of column-vectors. Not lazy.

parse->attrs clj

(parse->attrs csv-input & {:as opts})

[csv-input & {:as opts} ] Returns a map of attributes constructed from the columns of csv-input. The first line is assumed to be column label strings, which are (safely) converted into keywords. The returned map has one entry for each column label keyword. The corresponding value for each keyword is a vector of string data taken from each subsequent line in the file. See tupelo.csv/parse->entities for options. Not lazy.

parse->entities clj

(parse->entities csv-input & opts)

Inputs: [csv-input & opts] Returns: [tsk/Map]

[csv-input & {:as opts} ] Returns a lazy sequence of maps constructed from csv-input. The first line is assumed to be column label strings, which are (safely) converted into keywords. String data from each subsequent line is paired with the corresponding column keyword to construct a map for that line. Default delimiter is the comma character (i.e. ,) but may be changed using the syntax such as:

   (parse-csv->row-maps <csv-data-src> :delimiter \|)

to select the pipe character (i.e. |) as the delimiter.

<csv-data-source> is either a multi-line-string, or a java.io.Reader.