Add tests for custom delimiter.

README.md

@@ -2,7 +2,7 @@
 
 A parser and formatter for delimiter-separated values, most commonly comma-separated values (CSV) or tab-separated values (TSV).
 
-To use in the browser:
+To parse CSV in the browser:
 
 ```html
 <script src="dsv.js"></script>
@@ -13,12 +13,20 @@ console.log(dsv.csv.parse("foo,bar\n1,2")); // [{foo: "1", bar: "2"}]
 </script>
 ```
 
-To use in Node.js, `npm install d3-dsv`, and then:
+To parse CSV in Node.js, `npm install d3-dsv`, and then:
 
 ```js
 var dsv = require("d3-dsv");
 
 console.log(dsv.csv.parse("foo,bar\n1,2")); // [{foo: "1", bar: "2"}]
 ```
 
+To define a new delimiter, use the dsv constructor:
+
+```js
+var psv = dsv.dsv("|");
+
+console.log(psv.parse("foo|bar\n1|2")); // [{foo: "1", bar: "2"}]
+```
+
 For more usage, see [D3’s wiki page](https://github.com/mbostock/d3/wiki/CSV).

index.js

@@ -1,5 +1,10 @@
 import dsv from "./src/dsv";
 
-export var csv = dsv(",");
-export var tsv = dsv("\t");
-export default dsv;
+var csv = dsv(",");
+var tsv = dsv("\t");
+
+export {
+  csv,
+  tsv,
+  dsv
+};

package.json

@@ -1,6 +1,6 @@
 {
   "name": "d3-dsv",
-  "version": "0.1.0",
+  "version": "0.1.1",
   "description": "A parser and formatter for DSV (CSV and TSV) files.",
   "keywords": [
     "d3",

test/data/sample.psv

@@ -0,0 +1,2 @@
+Hello|World
+42|"""fish"""

test/dsv-test.js

@@ -0,0 +1,171 @@
+var tape = require("tape"),
+    dsv = require("../"),
+    fs = require("fs");
+
+var psv = dsv.dsv("|");
+
+tape("dsv(\"|\").parse(string) returns the expected objects", function(test) {
+  test.deepEqual(psv.parse("a|b|c\n1|2|3\n"), [{a: "1", b: "2", c: "3"}]);
+  test.deepEqual(psv.parse(fs.readFileSync("test/data/sample.psv", "utf-8")), [{Hello: "42", World: "\"fish\""}]);
+  test.end();
+});
+
+tape("dsv(\"|\").parse(string) does not strip whitespace", function(test) {
+  test.deepEqual(psv.parse("a|b|c\n 1| 2|3\n"), [{a: " 1", b: " 2", c: "3"}]);
+  test.end();
+});
+
+tape("dsv(\"|\").parse(string) parses quoted values", function(test) {
+  test.deepEqual(psv.parse("a|b|c\n\"1\"|2|3"), [{a: "1", b: "2", c: "3"}]);
+  test.deepEqual(psv.parse("a|b|c\n\"1\"|2|3\n"), [{a: "1", b: "2", c: "3"}]);
+  test.end();
+});
+
+tape("dsv(\"|\").parse(string) parses quoted values with quotes", function(test) {
+  test.deepEqual(psv.parse("a\n\"\"\"hello\"\"\""), [{a: "\"hello\""}]);
+  test.end();
+});
+
+tape("dsv(\"|\").parse(string) parses quoted values with newlines", function(test) {
+  test.deepEqual(psv.parse("a\n\"new\nline\""), [{a: "new\nline"}]);
+  test.deepEqual(psv.parse("a\n\"new\rline\""), [{a: "new\rline"}]);
+  test.deepEqual(psv.parse("a\n\"new\r\nline\""), [{a: "new\r\nline"}]);
+  test.end();
+});
+
+tape("dsv(\"|\").parse(string) observes Unix, Mac and DOS newlines", function(test) {
+  test.deepEqual(psv.parse("a|b|c\n1|2|3\n4|5|\"6\"\n7|8|9"), [{a: "1", b: "2", c: "3"}, {a: "4", b: "5", c: "6"}, {a: "7", b: "8", c: "9"}]);
+  test.deepEqual(psv.parse("a|b|c\r1|2|3\r4|5|\"6\"\r7|8|9"), [{a: "1", b: "2", c: "3"}, {a: "4", b: "5", c: "6"}, {a: "7", b: "8", c: "9"}]);
+  test.deepEqual(psv.parse("a|b|c\r\n1|2|3\r\n4|5|\"6\"\r\n7|8|9"), [{a: "1", b: "2", c: "3"}, {a: "4", b: "5", c: "6"}, {a: "7", b: "8", c: "9"}]);
+  test.end();
+});
+
+tape("dsv(\"|\").parse(string, convert) returns the expected converted objects", function(test) {
+  function convert(row) { row.Hello = -row.Hello; return row; }
+  test.deepEqual(psv.parse(fs.readFileSync("test/data/sample.psv", "utf-8"), convert), [{Hello: -42, World: "\"fish\""}]);
+  test.deepEqual(psv.parse("a|b|c\n1|2|3\n", function(row) { return row; }), [{a: "1", b: "2", c: "3"}]);
+  test.end();
+});
+
+tape("dsv(\"|\").parse(string, convert) skips rows if convert returns null or undefined", function(test) {
+  function convert(row, i) { return [row, null, undefined, false][i]; }
+  test.deepEqual(psv.parse("field\n42\n\n\n\n", convert), [{field: "42"}, false]);
+  test.deepEqual(psv.parse("a|b|c\n1|2|3\n2|3|4", function(row) { return row.a & 1 ? null : row; }), [{a: "2", b: "3", c: "4"}]);
+  test.deepEqual(psv.parse("a|b|c\n1|2|3\n2|3|4", function(row) { return row.a & 1 ? undefined : row; }), [{a: "2", b: "3", c: "4"}]);
+  test.end();
+});
+
+tape("dsv(\"|\").parse(string, convert) invokes convert for every row in order", function(test) {
+  var rows = [];
+  psv.parse("a\n1\n2\n3\n4", function(d, i) { rows.push({d: d, i: i}); });
+  test.deepEqual(rows, [{d: {a: "1"}, i: 0}, {d: {a: "2"}, i: 1}, {d: {a: "3"}, i: 2}, {d: {a: "4"}, i: 3}]);
+  test.end();
+});
+
+tape("dsv(\"|\").parseRows(string) returns the expected array of array of string", function(test) {
+  test.deepEqual(psv.parseRows("a|b|c\n"), [["a", "b", "c"]]);
+  test.end();
+});
+
+tape("dsv(\"|\").parseRows(string) parses quoted values", function(test) {
+  test.deepEqual(psv.parseRows("\"1\"|2|3\n"), [["1", "2", "3"]]);
+  test.deepEqual(psv.parseRows("\"hello\""), [["hello"]]);
+  test.end();
+});
+
+tape("dsv(\"|\").parseRows(string) parses quoted values with quotes", function(test) {
+  test.deepEqual(psv.parseRows("\"\"\"hello\"\"\""), [["\"hello\""]]);
+  test.end();
+});
+
+tape("dsv(\"|\").parseRows(string) parses quoted values with newlines", function(test) {
+  test.deepEqual(psv.parseRows("\"new\nline\""), [["new\nline"]]);
+  test.deepEqual(psv.parseRows("\"new\rline\""), [["new\rline"]]);
+  test.deepEqual(psv.parseRows("\"new\r\nline\""), [["new\r\nline"]]);
+  test.end();
+});
+
+tape("dsv(\"|\").parseRows(string) parses Unix, Mac and DOS newlines", function(test) {
+  test.deepEqual(psv.parseRows("a|b|c\n1|2|3\n4|5|\"6\"\n7|8|9"), [["a", "b", "c"], ["1", "2", "3"], ["4", "5", "6"], ["7", "8", "9"]]);
+  test.deepEqual(psv.parseRows("a|b|c\r1|2|3\r4|5|\"6\"\r7|8|9"), [["a", "b", "c"], ["1", "2", "3"], ["4", "5", "6"], ["7", "8", "9"]]);
+  test.deepEqual(psv.parseRows("a|b|c\r\n1|2|3\r\n4|5|\"6\"\r\n7|8|9"), [["a", "b", "c"], ["1", "2", "3"], ["4", "5", "6"], ["7", "8", "9"]]);
+  test.end();
+});
+
+tape("dsv(\"|\").parseRows(string, convert) returns the expected converted array of array of string", function(test) {
+  function convert(row, i) { if (i) row[0] = -row[0]; return row; }
+  test.deepEqual(psv.parseRows(fs.readFileSync("test/data/sample.psv", "utf-8"), convert), [["Hello", "World"], [-42, "\"fish\""]]);
+  test.deepEqual(psv.parseRows("a|b|c\n1|2|3\n", function(row) { return row; }), [["a", "b", "c"], ["1", "2", "3"]]);
+  test.end();
+});
+
+tape("dsv(\"|\").parseRows(string, convert) skips rows if convert returns null or undefined", function(test) {
+  function convert(row, i) { return [row, null, undefined, false][i]; }
+  test.deepEqual(psv.parseRows("field\n42\n\n\n", convert), [["field"], false]);
+  test.deepEqual(psv.parseRows("a|b|c\n1|2|3\n2|3|4", function(row, i) { return i & 1 ? null : row; }), [["a", "b", "c"], ["2", "3", "4"]]);
+  test.deepEqual(psv.parseRows("a|b|c\n1|2|3\n2|3|4", function(row, i) { return i & 1 ? undefined : row; }), [["a", "b", "c"], ["2", "3", "4"]]);
+  test.end();
+});
+
+tape("dsv(\"|\").parseRows(string, convert) invokes convert for every row in order", function(test) {
+  var rows = [];
+  psv.parseRows("a\n1\n2\n3\n4", function(d, i) { rows.push({d: d, i: i}); });
+  test.deepEqual(rows, [{d: ["a"], i: 0}, {d: ["1"], i: 1}, {d: ["2"], i: 2}, {d: ["3"], i: 3}, {d: ["4"], i: 4}]);
+  test.end();
+});
+
+tape("dsv(\"|\").format(array) takes an array of objects as input", function(test) {
+  test.deepEqual(psv.format([{a: 1, b: 2, c: 3}]), "a|b|c\n1|2|3");
+  test.end();
+});
+
+tape("dsv(\"|\").format(array) escapes field names and values containing delimiters", function(test) {
+  test.deepEqual(psv.format([{"foo|bar": true}]), "\"foo|bar\"\ntrue");
+  test.deepEqual(psv.format([{field: "foo|bar"}]), "field\n\"foo|bar\"");
+  test.end();
+});
+
+tape("dsv(\"|\").format(array) computes the union of all fields", function(test) {
+  test.deepEqual(psv.format([{a: 1}, {a: 1, b: 2}, {a: 1, b: 2, c: 3}, {b: 1, c: 2}, {c: 1}]), "a|b|c\n1||\n1|2|\n1|2|3\n|1|2\n||1");
+  test.end();
+});
+
+tape("dsv(\"|\").format(array) orders fields by first-seen", function(test) {
+  test.deepEqual(psv.format([{a: 1, b: 2}, {c: 3, b: 4}, {c: 5, a: 1, b: 2}]), "a|b|c\n1|2|\n|4|3\n1|2|5");
+  test.end();
+});
+
+tape("dsv(\"|\").formatRows(array) takes an array of array of string as input", function(test) {
+  test.deepEqual(psv.formatRows([["a", "b", "c"], ["1", "2", "3"]]), "a|b|c\n1|2|3");
+  test.end();
+});
+
+tape("dsv(\"|\").formatRows(array) separates lines using Unix newline", function(test) {
+  test.deepEqual(psv.formatRows([[], []]), "\n");
+  test.end();
+});
+
+tape("dsv(\"|\").formatRows(array) does not strip whitespace", function(test) {
+  test.deepEqual(psv.formatRows([["a ", " b", "c"], ["1", "2", "3 "]]), "a | b|c\n1|2|3 ");
+  test.end();
+});
+
+tape("dsv(\"|\").formatRows(array) does not quote simple values", function(test) {
+  test.deepEqual(psv.formatRows([["a"], [1]]), "a\n1");
+  test.end();
+});
+
+tape("dsv(\"|\").formatRows(array) escapes double quotes", function(test) {
+  test.deepEqual(psv.formatRows([["\"fish\""]]), "\"\"\"fish\"\"\"");
+  test.end();
+});
+
+tape("dsv(\"|\").formatRows(array) escapes Unix newlines", function(test) {
+  test.deepEqual(psv.formatRows([["new\nline"]]), "\"new\nline\"");
+  test.end();
+});
+
+tape("dsv(\"|\").formatRows(array) escapes values containing delimiters", function(test) {
+  test.deepEqual(psv.formatRows([["oxford|tab"]]), "\"oxford|tab\"");
+  test.end();
+});