---

This file is part of Logtalk https://logtalk.org/ SPDX-FileCopyrightText: 2021-2022 Paulo Moura <pmoura@logtalk.org> SPDX-FileCopyrightText: 2021 Jacinto Dávila Quintero <jdavila@optimusprime.ai> SPDX-License-Identifier: Apache-2.0

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

csv

The csv library provides predicates for reading and writing CSV files and streams:

https://www.rfc-editor.org/rfc/rfc4180.txt

The main object, csv/3, is a parametric object allowing passing options for the handling of the header of the file, the fields separator, and the handling of double-quoted fields. The cvs object extends the csv/3 parametric object using default option values.

The library also include predicates to guess the separator and guess the number of columns in a given CSV file.

Files and streams can be read into a list of rows (with each row being represented by a list of fields) or asserted using a user-defined dynamic predicate. Reading can be done by first loading the whole file (using the read_file/2-3 predicates) into memory or line by line (using the read_file_by_line/2-3 predicates). Reading line by line is usually the best option for parsing large CSV files.

Data can be saved to a CSV file or stream by providing the object and predicate for accessing the data plus the name of the destination file or the stream handle or alias.

API documentation

Open the [../../docs/library_index.html#csv](../../docs/library_index.html#csv) link in a web browser.

Loading

To load all entities in this library, load the loader.lgt file:

| ?- logtalk_load(csv(loader)).

Testing

To test this library predicates, load the tester.lgt file:

| ?- logtalk_load(csv(tester)).

Usage

The csv(Header, Separator, IgnoreQuotes) parametric object allows passing the following options:

1. Header: possible values are missing, skip, and keep.
2. Separator: possible values are comma, tab, semicolon, and colon.
3. IgnoreQuotes: possible values are true to ignore double quotes surrounding field data and false to preserve the double quotes.

   The csv object uses the default values keep, comma, and false.

   When writing CSV files or streams, set the quoted fields option to false to write all non-numeric fields double-quoted (i.e. escaped).

   The library objects can also be used to guess the separator used in a CSV file if necessary. For example:

   | ?- csv::guess_separator('test_files/crlf_ending.csv', Separator). Is this the proper reading of a line of this file (y/n)? [aaa,bb,ccc] |> y.

   Separator = comma ?

   This information can then be used to read the CSV file returning a list of rows:

   | ?- csv(keep, comma, true)::read_file('test_files/crlf_ending.csv', Rows).

   Rows = [[aaa,bbb,ccc],[zzz,yyy,xxx]] ?

   Alternatively, The CSV data can be saved using a public and dynamic object predicate (that must be previously declared). For example:

   | ?- assertz(p(_,_,_)), retractall(p(_,_,_)). yes

   | ?- csv(keep, comma, true)::read_file('test_files/crlf_ending.csv', user, p/3). yes

   | ?- p(A,B,C).

   A = aaa B = bbb C = ccc ? ;

   A = zzz B = yyy C = xxx

   Given a predicate representing a table, the predicate data can be written to a file or stream. For example:

   | ?- csv(keep, comma, true)::write_file('output.csv', user, p/3). yes

   leaving the content just as the original file thanks to the use of true for the IgnoreQuotes option:

   aaa,bbb,ccc
   zzz,yyy,xxx

   Otherwise:

   | ?- csv(keep, comma, false)::write_file('output.csv', user, p/3). yes

   results in the following file content:

   "aaa","bbb","ccc"
   "zzz","yyy","xxx"

   The guess_arity/2 method, to identify the arity, i. e. the number of fields or columns per record in a given CSV file, for example:

   | ?- csv(keep, comma, false)::guess_arity('test_files/crlf_ending.csv', Arity). Is this the proper reading of a line of this file (y/n)? [aaa,bbb,ccc] |> y.

   Arity = 3