feat: add splitLines method

lib/index.js

@@ -4,6 +4,7 @@ const stringWidth = require('./stringWidth'),
     charWidths = require('./charWidths'),
     stripAnsi = require('./stripAnsi'),
     codePointWidth = require('./codePoint'),
+    splitLines = require('./splitLines'),
     { sliceChars, sliceColumns } = require('./sliceString'),
     wordWrap = require('./wordWrap');
 
@@ -14,6 +15,7 @@ module.exports = {
     charWidths,
     stripAnsi,
     codePointWidth,
+    splitLines,
     sliceChars,
     sliceColumns,
     wordWrap,

lib/splitLines.js

@@ -0,0 +1,55 @@
+const parseAnsi = require('./parseAnsi'),
+    { parseEscape, closeEscapes } = require('./utils');
+
+/**
+ * Split a string with ANSI escape codes into an array of lines. Supports both `CRLF` and `LF` newlines.
+ * @param {string} string - string to split
+ * @returns {string[]} - array of lines in the input string
+ */
+module.exports = function splitLines(string) {
+    // ansi escapes stack, items in the form [seq, isLink, close, [i, j]]
+    const ansiStack = [],
+        // result array to hold processed lines
+        result = [],
+        // split input string into lines
+        lines = String(string).split(/\r?\n/g);
+    // iterate through each line in the input string
+    for (let i = 0, n = lines.length; i < n; i += 1) {
+        // the processed line
+        let line = '',
+            // intraline index
+            j = -1;
+        // match all ansi escape codes in the input line
+        for (const [chunk, isEscape] of parseAnsi(lines[i])) {
+            // check if chunk is an escape sequence
+            if (isEscape) {
+                // process this escape sequence
+                const closed = parseEscape(ansiStack, chunk, [i, j]);
+                if (closed && j >= 0) {
+                    // append escape if it closes an active item in the stack
+                    const [xi, xj] = closed;
+                    if (xi < i || (xi === i && xj < j)) line += chunk;
+                } else if (closed === null) {
+                    // escape is not a SGR/hyperlink escape
+                    line += chunk;
+                }
+                continue;
+            }
+            // append any new escape sequences from the ansi stack
+            line += (j < 0 ? ansiStack : ansiStack.filter(([,,, [xi, xj]]) => (xi === i && xj === j)))
+                .map(([s]) => s)
+                .join('');
+            // append this chunk
+            line += chunk;
+            // increment the intraline index
+            j += 1;
+        }
+        // close open escape sequences if line is not empty
+        if (j >= 0) {
+            line += closeEscapes(ansiStack.filter(([,,, [xi, xj]]) => (xi < i || (xi === i && xj < j))));
+        }
+        // add proccessed line to the result array
+        result.push(line);
+    }
+    return result;
+};

test/splitLines.test.js

@@ -0,0 +1,63 @@
+const splitLines = require('../lib/splitLines');
+
+describe('splitLines', () => {
+    test('supports both LF and CRLF newline types', () => {
+        expect(splitLines('AA\nBB\r\nCC')).toEqual(['AA', 'BB', 'CC']);
+    });
+
+    test('handles non-string inputs', () => {
+        expect(splitLines(10)).toEqual(['10']);
+    });
+
+    test('splits style escapes that span multiple lines', () => {
+        expect(splitLines('\u001b[41mAAA\u001b[33mBBB\nCCC\u001b[39mDDD\u001b[49m')).toEqual([
+            '\u001b[41mAAA\u001b[33mBBB\u001b[39m\u001b[49m',
+            '\u001b[41m\u001b[33mCCC\u001b[39mDDD\u001b[49m',
+        ]);
+    });
+
+    test('splits style escapes that span a single line', () => {
+        expect(splitLines('\u001b[41mAAAA\u001b[49m\u001b[33m\nBBBB\u001b[39m')).toEqual([
+            '\u001b[41mAAAA\u001b[49m',
+            '\u001b[33mBBBB\u001b[39m',
+        ]);
+    });
+
+    test('splits style escape sequences that overlap across a line break', () => {
+        expect(splitLines('\u001b[41mAAAA\u001b[33m\n\u001b[49mBBBB\u001b[39m')).toEqual([
+            '\u001b[41mAAAA\u001b[49m',
+            '\u001b[33mBBBB\u001b[39m',
+        ]);
+    });
+
+    test('splits style escape sequences that span empty lines', () => {
+        expect(splitLines('\u001b[41m\nAAA\n\nBBB\u001b[49m')).toEqual([
+            '',
+            '\u001b[41mAAA\u001b[49m',
+            '',
+            '\u001b[41mBBB\u001b[49m',
+        ]);
+    });
+
+    test('scrubs empty escape sequences', () => {
+        expect(splitLines('AA\u001b[41m\u001b[49mA\nBB')).toEqual(['AAA', 'BB']);
+    });
+
+    test('scrubs escape sequences that span only line breaks', () => {
+        expect(splitLines('AAAA\u001b[41m\n\u001b[49mBBBB')).toEqual(['AAAA', 'BBBB']);
+    });
+
+    test('supports ansi hyperlink escapes', () => {
+        expect(splitLines('\u001b]8;;link\u0007AA\nB\u001b]8;;\u0007b')).toEqual([
+            '\u001b]8;;link\u0007AA\u001b]8;;\u0007',
+            '\u001b]8;;link\u0007B\u001b]8;;\u0007b',
+        ]);
+    });
+
+    test('handles non-SGR/non-hyperlink ansi escape sequences', () => {
+        expect(splitLines('AA\u001B]0;window_title\u0007\nBB')).toEqual([
+            'AA\u001B]0;window_title\u0007',
+            'BB',
+        ]);
+    });
+});