mirror of
https://github.com/zebrajr/node.git
synced 2026-01-15 12:15:26 +00:00
0da2084d7f
The option is now set to true by default. Most terminals do not have full emoji support and visualize emojis with zero width joiners as individual emojis. Also verify that at least one argument is always passed through to the function and remove support for passing through code points. Only accept strings from now on to simplify the API. PR-URL: https://github.com/nodejs/node/pull/31112 Reviewed-By: Michaël Zasso <targos@protonmail.com> Reviewed-By: James M Snell <jasnell@gmail.com>
90 lines
4.0 KiB
JavaScript
90 lines
4.0 KiB
JavaScript
// Flags: --expose_internals
|
|
'use strict';
|
|
const common = require('../common');
|
|
|
|
if (!common.hasIntl)
|
|
common.skip('missing Intl');
|
|
|
|
const assert = require('assert');
|
|
const readline = require('internal/readline/utils');
|
|
|
|
// Test column width
|
|
|
|
// Ll (Lowercase Letter): LATIN SMALL LETTER A
|
|
assert.strictEqual(readline.getStringWidth('a'), 1);
|
|
assert.strictEqual(readline.getStringWidth(String.fromCharCode(0x0061)), 1);
|
|
// Lo (Other Letter)
|
|
assert.strictEqual(readline.getStringWidth('丁'), 2);
|
|
assert.strictEqual(readline.getStringWidth(String.fromCharCode(0x4E01)), 2);
|
|
// Surrogate pairs
|
|
assert.strictEqual(readline.getStringWidth('\ud83d\udc78\ud83c\udfff'), 4);
|
|
assert.strictEqual(readline.getStringWidth('👅'), 2);
|
|
// Cs (Surrogate): High Surrogate
|
|
assert.strictEqual(readline.getStringWidth('\ud83d'), 1);
|
|
// Cs (Surrogate): Low Surrogate
|
|
assert.strictEqual(readline.getStringWidth('\udc78'), 1);
|
|
// Cc (Control): NULL
|
|
assert.strictEqual(readline.getStringWidth('\u0000'), 0);
|
|
// Cc (Control): BELL
|
|
assert.strictEqual(readline.getStringWidth(String.fromCharCode(0x0007)), 0);
|
|
// Cc (Control): LINE FEED
|
|
assert.strictEqual(readline.getStringWidth('\n'), 0);
|
|
// Cf (Format): SOFT HYPHEN
|
|
assert.strictEqual(readline.getStringWidth(String.fromCharCode(0x00AD)), 1);
|
|
// Cf (Format): LEFT-TO-RIGHT MARK
|
|
// Cf (Format): RIGHT-TO-LEFT MARK
|
|
assert.strictEqual(readline.getStringWidth('\u200Ef\u200F'), 1);
|
|
// Cn (Unassigned): Not a character
|
|
assert.strictEqual(readline.getStringWidth(String.fromCharCode(0x10FFEF)), 1);
|
|
// Cn (Unassigned): Not a character (but in a CJK range)
|
|
assert.strictEqual(readline.getStringWidth(String.fromCharCode(0x3FFEF)), 1);
|
|
// Mn (Nonspacing Mark): COMBINING ACUTE ACCENT
|
|
assert.strictEqual(readline.getStringWidth(String.fromCharCode(0x0301)), 0);
|
|
// Mc (Spacing Mark): BALINESE ADEG ADEG
|
|
// Chosen as its Canonical_Combining_Class is not 0, but is not a 0-width
|
|
// character.
|
|
assert.strictEqual(readline.getStringWidth(String.fromCharCode(0x1B44)), 1);
|
|
// Me (Enclosing Mark): COMBINING ENCLOSING CIRCLE
|
|
assert.strictEqual(readline.getStringWidth(String.fromCharCode(0x20DD)), 0);
|
|
|
|
// The following is an emoji sequence with ZWJ (zero-width-joiner). In some
|
|
// implementations, it is represented as a single glyph, in other
|
|
// implementations as a sequence of individual glyphs. By default, each
|
|
// component will be counted individually, since not a lot of systems support
|
|
// these fully.
|
|
// See https://www.unicode.org/reports/tr51/tr51-16.html#Emoji_ZWJ_Sequences
|
|
assert.strictEqual(readline.getStringWidth('👩👩👧👧'), 8);
|
|
// TODO(BridgeAR): This should have a width of two and six. The heart contains
|
|
// the \uFE0F variation selector that indicates that it should be displayed as
|
|
// emoji instead of as text. Emojis are all full width characters when not being
|
|
// rendered as text.
|
|
// https://en.wikipedia.org/wiki/Variation_Selectors_(Unicode_block)
|
|
assert.strictEqual(readline.getStringWidth('❤️'), 1);
|
|
assert.strictEqual(readline.getStringWidth('👩❤️👩'), 5);
|
|
// The length of one is correct. It is an emoji treated as text.
|
|
assert.strictEqual(readline.getStringWidth('❤'), 1);
|
|
|
|
// By default, unicode characters whose width is considered ambiguous will
|
|
// be considered half-width. For these characters, getStringWidth will return
|
|
// 1. In some contexts, however, it is more appropriate to consider them full
|
|
// width. By default, the algorithm will assume half width.
|
|
assert.strictEqual(readline.getStringWidth('\u01d4'), 1);
|
|
|
|
// Control chars and combining chars are zero
|
|
assert.strictEqual(readline.getStringWidth('\u200E\n\u220A\u20D2'), 1);
|
|
|
|
// Test that the fast path for ASCII characters yields results consistent
|
|
// with the 'slow' path.
|
|
for (let i = 0; i < 256; i++) {
|
|
const char = String.fromCharCode(i);
|
|
assert.strictEqual(
|
|
readline.getStringWidth(char + '🎉'),
|
|
readline.getStringWidth(char) + 2);
|
|
|
|
if (i < 32 || (i >= 127 && i < 160)) { // Control character
|
|
assert.strictEqual(readline.getStringWidth(char), 0);
|
|
} else { // Regular ASCII character
|
|
assert.strictEqual(readline.getStringWidth(char), 1);
|
|
}
|
|
}
|