1 (** csv.mli - comma separated values parser
3 * $Id: csv.mli,v 1.10 2006-11-24 15:49:24 rich Exp $
6 type t = string list list
7 (** Representation of CSV files. *)
9 exception Bad_CSV_file of string
10 (** Badly formed CSV files throw this exception. *)
13 (** Work out the number of lines in a CSV file. *)
15 val columns : t -> int
16 (** Work out the (maximum) number of columns in a CSV file. Note that each
17 * line may be a different length, so this finds the one with the most
21 val load_in : ?separator:char -> in_channel -> t
23 * @param chan Input file stream
26 val load : ?separator:char -> string -> t
28 * @param filename CSV filename.
29 * If [filename] is ["-"] then load from [stdin].
32 val load_rows : ?separator:char -> (string list -> unit) -> in_channel -> unit
33 (** For very large CSV files which cannot be processed in memory at once,
34 * this function is appropriate. It parses the input one row at a time and
35 * calls your function once for each row.
37 * Note that if you CSV file contains cells which have embedded
38 * line feeds, then it is non-trivial to parse these lines and
39 * pass them correctly to [load_rows].
41 * @param f Callout function.
42 * @param chan Input file stream.
45 val trim : ?top:bool -> ?left:bool -> ?right:bool -> ?bottom:bool -> t -> t
46 (** This takes a CSV file and trims empty cells.
48 * All four of the option arguments ([~top], [~left], [~right], [~bottom])
51 * The exact behaviour is:
53 * [~right]: If true, remove any empty cells at the right hand end of
54 * any row. The number of columns in the resulting CSV structure will
55 * not necessarily be the same for each row.
57 * [~top]: If true, remove any empty rows (no cells, or containing just empty
58 * cells) from the top of the CSV structure.
60 * [~bottom]: If true, remove any empty rows from the bottom of the
63 * [~left]: If true, remove any empty columns from the left of the
64 * CSV structure. Note that [~left] and [~right] are quite different:
65 * [~left] considers the whole CSV structure, whereas [~right] considers
66 * each row in isolation.
70 (** Make the CSV data "square" (actually rectangular). This pads out
71 * each row with empty cells so that all rows are the same length as
72 * the longest row. After this operation, every row will have length
76 val is_square : t -> bool
77 (** Return true iff the CSV is "square" (actually rectangular). This
78 * means that each row has the same number of cells.
81 val set_columns : int -> t -> t
82 (** [set_columns cols csv] makes the CSV data square by forcing the width
83 * to the given number of [cols]. Any short rows are padded with blank
84 * cells. Any long rows are truncated.
87 val set_rows : int -> t -> t
88 (** [set_rows rows csv] makes the CSV data have exactly [rows] rows
89 * by adding empty rows or truncating rows as necessary.
91 * Note that [set_rows] does not make the CSV square. If you want it
92 * to be square, call either {!Csv.square} or {!Csv.set_columns} after.
95 val set_size : int -> int -> t -> t
96 (** [set_size rows cols csv] makes the CSV data square by forcing the
97 * size to [rows * cols], adding blank cells or truncating as necessary.
98 * It is the same as calling [set_columns cols (set_rows rows csv)]
101 val sub : int -> int -> int -> int -> t -> t
102 (** [sub r c rows cols csv] returns a subset of [csv]. The subset is
103 * defined as having top left corner at row [r], column [c] (counting
104 * from [0]) and being [rows] deep and [cols] wide.
106 * The returned CSV will be square.
109 val compare : t -> t -> int
110 (** Compare two CSV files for equality, ignoring blank cells at the end
111 * of a row, and empty rows appended to one or the other. This is
112 * "semantic" equality - roughly speaking, the two CSV files would
113 * look the same if opened in a spreadsheet program.
116 val concat : t list -> t
117 (** Concatenate CSV files so that they appear side by side, arranged
118 * left to right across the page. Each CSV file (except the final
119 * one) is first squared.
121 * (To concatenate CSV files so that they appear from top to bottom,
122 * just use {!List.concat}).
125 val to_array : t -> string array array
126 val of_array : string array array -> t
127 (** Convenience functions to convert to and from a matrix representation.
128 * [to_array] will produce a ragged matrix (not all rows will have the
129 * same length) unless you call {!Csv.square} first.
132 val associate : string list -> t -> (string * string) list list
133 (** [associate header data] takes a block of data and converts each
134 * row in turn into an assoc list which maps column header to data cell.
136 * Typically a spreadsheet will have the format:
138 * header1 header2 header3
139 * data11 data12 data13
140 * data21 data22 data23
144 * This function arranges the data into a more usable form which is
145 * robust against changes in column ordering. The output of the
148 * [ ["header1", "data11"; "header2", "data12"; "header3", "data13"];
149 * ["header1", "data21"; "header2", "data22"; "header3", "data23"];
153 * Each row is turned into an assoc list (see [List.assoc]).
155 * If a row is too short, it is padded with empty cells ([""]). If
156 * a row is too long, it is truncated.
158 * You would typically call this function as:
161 * let header, data = match csv with h :: d -> h, d | [] -> assert false;;
162 * let data = Csv.associate header data;;
165 * The header strings are shared, so the actual space in memory consumed
166 * by the spreadsheet is not much larger.
169 val print : ?separator:char -> t -> unit
170 (** Print string list list - same as [save_out stdout] *)
172 val save_out : ?separator:char -> out_channel -> t -> unit
173 (** Save string list list to a channel. *)
175 val save : ?separator:char -> string -> t -> unit
176 (** Save string list list to a file. *)
178 val print_readable : t -> unit
179 (** Print the CSV data to [stdout] in a human-readable format. Not much
180 * is guaranteed about how the CSV is printed, except that it will be
181 * easier to follow than a "raw" output done with {!Csv.print}. This is
182 * a one-way operation. There is no easy way to parse the output of
183 * this command back into CSV data.
185 val save_out_readable : out_channel -> t -> unit
186 (** As for {!Csv.print_readable}, allowing the output to be sent to a channel.