@umsai/ums-code 0.0.14-v1 → 0.0.14-v2

package/dist/src/ui/components/shared/TextInput.js

@@ -0,0 +1,127 @@
+import { jsx as _jsx, jsxs as _jsxs } from "react/jsx-runtime";
+/**
+ * @license
+ * Copyright 2025 Qwen
+ * SPDX-License-Identifier: Apache-2.0
+ */
+// no hooks needed beyond keypress handled inside
+import { Box, Text } from 'ink';
+import chalk from 'chalk';
+import stringWidth from 'string-width';
+import { useTextBuffer } from './text-buffer.js';
+import { useKeypress } from '../../hooks/useKeypress.js';
+import { keyMatchers, Command } from '../../keyMatchers.js';
+import { cpSlice, cpLen } from '../../utils/textUtils.js';
+import { theme } from '../../semantic-colors.js';
+import { Colors } from '../../colors.js';
+import { useCallback, useRef, useEffect } from 'react';
+export function TextInput({ value, onChange, onSubmit, placeholder, height = 1, isActive = true, validationErrors = [], inputWidth = 80, }) {
+    const allowMultiline = height > 1;
+    // Stabilize onChange to avoid triggering useTextBuffer's onChange effect every render
+    const onChangeRef = useRef(onChange);
+    useEffect(() => {
+        onChangeRef.current = onChange;
+    }, [onChange]);
+    const stableOnChange = useCallback((text) => {
+        onChangeRef.current?.(text);
+    }, []);
+    const buffer = useTextBuffer({
+        initialText: value || '',
+        viewport: { height, width: inputWidth },
+        isValidPath: () => false,
+        onChange: stableOnChange,
+    });
+    const handleSubmit = () => {
+        if (!onSubmit)
+            return;
+        onSubmit();
+    };
+    useKeypress((key) => {
+        if (!buffer || !isActive)
+            return;
+        // Submit on Enter
+        if (keyMatchers[Command.SUBMIT](key) || key.name === 'return') {
+            if (allowMultiline) {
+                const [row, col] = buffer.cursor;
+                const line = buffer.lines[row];
+                const charBefore = col > 0 ? cpSlice(line, col - 1, col) : '';
+                if (charBefore === '\\') {
+                    buffer.backspace();
+                    buffer.newline();
+                }
+                else {
+                    handleSubmit();
+                }
+            }
+            else {
+                handleSubmit();
+            }
+            return;
+        }
+        // Multiline newline insertion (Shift+Enter etc.)
+        if (allowMultiline && keyMatchers[Command.NEWLINE](key)) {
+            buffer.newline();
+            return;
+        }
+        // Navigation helpers
+        if (keyMatchers[Command.HOME](key)) {
+            buffer.move('home');
+            return;
+        }
+        if (keyMatchers[Command.END](key)) {
+            buffer.move('end');
+            buffer.moveToOffset(cpLen(buffer.text));
+            return;
+        }
+        if (keyMatchers[Command.CLEAR_INPUT](key)) {
+            if (buffer.text.length > 0)
+                buffer.setText('');
+            return;
+        }
+        if (keyMatchers[Command.KILL_LINE_RIGHT](key)) {
+            buffer.killLineRight();
+            return;
+        }
+        if (keyMatchers[Command.KILL_LINE_LEFT](key)) {
+            buffer.killLineLeft();
+            return;
+        }
+        if (keyMatchers[Command.OPEN_EXTERNAL_EDITOR](key)) {
+            buffer.openInExternalEditor();
+            return;
+        }
+        buffer.handleInput(key);
+    }, { isActive });
+    if (!buffer)
+        return null;
+    const linesToRender = buffer.viewportVisualLines;
+    const [cursorVisualRowAbsolute, cursorVisualColAbsolute] = buffer.visualCursor;
+    const scrollVisualRow = buffer.visualScrollRow;
+    return (_jsxs(Box, { flexDirection: "column", gap: 1, children: [_jsxs(Box, { children: [_jsx(Text, { color: theme.text.accent, children: '> ' }), _jsx(Box, { flexGrow: 1, flexDirection: "column", children: buffer.text.length === 0 && placeholder ? (_jsxs(Text, { children: [chalk.inverse(placeholder.slice(0, 1)), _jsx(Text, { color: Colors.Gray, children: placeholder.slice(1) })] })) : (linesToRender.map((lineText, visualIdxInRenderedSet) => {
+                            const cursorVisualRow = cursorVisualRowAbsolute - scrollVisualRow;
+                            let display = cpSlice(lineText, 0, inputWidth);
+                            const currentVisualWidth = stringWidth(display);
+                            if (currentVisualWidth < inputWidth) {
+                                display = display + ' '.repeat(inputWidth - currentVisualWidth);
+                            }
+                            if (visualIdxInRenderedSet === cursorVisualRow) {
+                                const relativeVisualColForHighlight = cursorVisualColAbsolute;
+                                if (relativeVisualColForHighlight >= 0) {
+                                    if (relativeVisualColForHighlight < cpLen(display)) {
+                                        const charToHighlight = cpSlice(display, relativeVisualColForHighlight, relativeVisualColForHighlight + 1) || ' ';
+                                        const highlighted = chalk.inverse(charToHighlight);
+                                        display =
+                                            cpSlice(display, 0, relativeVisualColForHighlight) +
+                                                highlighted +
+                                                cpSlice(display, relativeVisualColForHighlight + 1);
+                                    }
+                                    else if (relativeVisualColForHighlight === cpLen(display) &&
+                                        cpLen(display) === inputWidth) {
+                                        display = display + chalk.inverse(' ');
+                                    }
+                                }
+                            }
+                            return (_jsx(Text, { children: display }, `line-${visualIdxInRenderedSet}`));
+                        })) })] }), validationErrors.length > 0 && (_jsx(Box, { flexDirection: "column", children: validationErrors.map((error, index) => (_jsxs(Text, { color: theme.status.error, children: ["\u26A0 ", error] }, index))) }))] }));
+}
+//# sourceMappingURL=TextInput.js.map

package/dist/src/ui/components/shared/TextInput.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"TextInput.js","sourceRoot":"","sources":["../../../../../src/ui/components/shared/TextInput.tsx"],"names":[],"mappings":";AAAA;;;;GAIG;AAEH,iDAAiD;AACjD,OAAO,EAAE,GAAG,EAAE,IAAI,EAAE,MAAM,KAAK,CAAC;AAChC,OAAO,KAAK,MAAM,OAAO,CAAC;AAC1B,OAAO,WAAW,MAAM,cAAc,CAAC;AACvC,OAAO,EAAE,aAAa,EAAE,MAAM,kBAAkB,CAAC;AACjD,OAAO,EAAE,WAAW,EAAE,MAAM,4BAA4B,CAAC;AACzD,OAAO,EAAE,WAAW,EAAE,OAAO,EAAE,MAAM,sBAAsB,CAAC;AAC5D,OAAO,EAAE,OAAO,EAAE,KAAK,EAAE,MAAM,0BAA0B,CAAC;AAC1D,OAAO,EAAE,KAAK,EAAE,MAAM,0BAA0B,CAAC;AACjD,OAAO,EAAE,MAAM,EAAE,MAAM,iBAAiB,CAAC;AAEzC,OAAO,EAAE,WAAW,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,OAAO,CAAC;AAavD,MAAM,UAAU,SAAS,CAAC,EACxB,KAAK,EACL,QAAQ,EACR,QAAQ,EACR,WAAW,EACX,MAAM,GAAG,CAAC,EACV,QAAQ,GAAG,IAAI,EACf,gBAAgB,GAAG,EAAE,EACrB,UAAU,GAAG,EAAE,GACA;IACf,MAAM,cAAc,GAAG,MAAM,GAAG,CAAC,CAAC;IAElC,sFAAsF;IACtF,MAAM,WAAW,GAAG,MAAM,CAAC,QAAQ,CAAC,CAAC;IACrC,SAAS,CAAC,GAAG,EAAE;QACb,WAAW,CAAC,OAAO,GAAG,QAAQ,CAAC;IACjC,CAAC,EAAE,CAAC,QAAQ,CAAC,CAAC,CAAC;IACf,MAAM,cAAc,GAAG,WAAW,CAAC,CAAC,IAAY,EAAE,EAAE;QAClD,WAAW,CAAC,OAAO,EAAE,CAAC,IAAI,CAAC,CAAC;IAC9B,CAAC,EAAE,EAAE,CAAC,CAAC;IAEP,MAAM,MAAM,GAAG,aAAa,CAAC;QAC3B,WAAW,EAAE,KAAK,IAAI,EAAE;QACxB,QAAQ,EAAE,EAAE,MAAM,EAAE,KAAK,EAAE,UAAU,EAAE;QACvC,WAAW,EAAE,GAAG,EAAE,CAAC,KAAK;QACxB,QAAQ,EAAE,cAAc;KACzB,CAAC,CAAC;IAEH,MAAM,YAAY,GAAG,GAAG,EAAE;QACxB,IAAI,CAAC,QAAQ;YAAE,OAAO;QACtB,QAAQ,EAAE,CAAC;IACb,CAAC,CAAC;IAEF,WAAW,CACT,CAAC,GAAQ,EAAE,EAAE;QACX,IAAI,CAAC,MAAM,IAAI,CAAC,QAAQ;YAAE,OAAO;QAEjC,kBAAkB;QAClB,IAAI,WAAW,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,GAAG,CAAC,IAAI,GAAG,CAAC,IAAI,KAAK,QAAQ,EAAE,CAAC;YAC9D,IAAI,cAAc,EAAE,CAAC;gBACnB,MAAM,CAAC,GAAG,EAAE,GAAG,CAAC,GAAG,MAAM,CAAC,MAAM,CAAC;gBACjC,MAAM,IAAI,GAAG,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;gBAC/B,MAAM,UAAU,GAAG,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,IAAI,EAAE,GAAG,GAAG,CAAC,EAAE,GAAG,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;gBAC9D,IAAI,UAAU,KAAK,IAAI,EAAE,CAAC;oBACxB,MAAM,CAAC,SAAS,EAAE,CAAC;oBACnB,MAAM,CAAC,OAAO,EAAE,CAAC;gBACnB,CAAC;qBAAM,CAAC;oBACN,YAAY,EAAE,CAAC;gBACjB,CAAC;YACH,CAAC;iBAAM,CAAC;gBACN,YAAY,EAAE,CAAC;YACjB,CAAC;YACD,OAAO;QACT,CAAC;QAED,iDAAiD;QACjD,IAAI,cAAc,IAAI,WAAW,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC;YACxD,MAAM,CAAC,OAAO,EAAE,CAAC;YACjB,OAAO;QACT,CAAC;QAED,qBAAqB;QACrB,IAAI,WAAW,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC;YACnC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YACpB,OAAO;QACT,CAAC;QACD,IAAI,WAAW,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC;YAClC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YACnB,MAAM,CAAC,YAAY,CAAC,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC;YACxC,OAAO;QACT,CAAC;QAED,IAAI,WAAW,CAAC,OAAO,CAAC,WAAW,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC;YAC1C,IAAI,MAAM,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC;gBAAE,MAAM,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC;YAC/C,OAAO;QACT,CAAC;QACD,IAAI,WAAW,CAAC,OAAO,CAAC,eAAe,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC;YAC9C,MAAM,CAAC,aAAa,EAAE,CAAC;YACvB,OAAO;QACT,CAAC;QACD,IAAI,WAAW,CAAC,OAAO,CAAC,cAAc,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC;YAC7C,MAAM,CAAC,YAAY,EAAE,CAAC;YACtB,OAAO;QACT,CAAC;QAED,IAAI,WAAW,CAAC,OAAO,CAAC,oBAAoB,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC;YACnD,MAAM,CAAC,oBAAoB,EAAE,CAAC;YAC9B,OAAO;QACT,CAAC;QAED,MAAM,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;IAC1B,CAAC,EACD,EAAE,QAAQ,EAAE,CACb,CAAC;IAEF,IAAI,CAAC,MAAM;QAAE,OAAO,IAAI,CAAC;IAEzB,MAAM,aAAa,GAAG,MAAM,CAAC,mBAAmB,CAAC;IACjD,MAAM,CAAC,uBAAuB,EAAE,uBAAuB,CAAC,GACtD,MAAM,CAAC,YAAY,CAAC;IACtB,MAAM,eAAe,GAAG,MAAM,CAAC,eAAe,CAAC;IAE/C,OAAO,CACL,MAAC,GAAG,IAAC,aAAa,EAAC,QAAQ,EAAC,GAAG,EAAE,CAAC,aAChC,MAAC,GAAG,eACF,KAAC,IAAI,IAAC,KAAK,EAAE,KAAK,CAAC,IAAI,CAAC,MAAM,YAAG,IAAI,GAAQ,EAC7C,KAAC,GAAG,IAAC,QAAQ,EAAE,CAAC,EAAE,aAAa,EAAC,QAAQ,YACrC,MAAM,CAAC,IAAI,CAAC,MAAM,KAAK,CAAC,IAAI,WAAW,CAAC,CAAC,CAAC,CACzC,MAAC,IAAI,eACF,KAAK,CAAC,OAAO,CAAC,WAAW,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EACvC,KAAC,IAAI,IAAC,KAAK,EAAE,MAAM,CAAC,IAAI,YAAG,WAAW,CAAC,KAAK,CAAC,CAAC,CAAC,GAAQ,IAClD,CACR,CAAC,CAAC,CAAC,CACF,aAAa,CAAC,GAAG,CAAC,CAAC,QAAQ,EAAE,sBAAsB,EAAE,EAAE;4BACrD,MAAM,eAAe,GAAG,uBAAuB,GAAG,eAAe,CAAC;4BAClE,IAAI,OAAO,GAAG,OAAO,CAAC,QAAQ,EAAE,CAAC,EAAE,UAAU,CAAC,CAAC;4BAC/C,MAAM,kBAAkB,GAAG,WAAW,CAAC,OAAO,CAAC,CAAC;4BAChD,IAAI,kBAAkB,GAAG,UAAU,EAAE,CAAC;gCACpC,OAAO,GAAG,OAAO,GAAG,GAAG,CAAC,MAAM,CAAC,UAAU,GAAG,kBAAkB,CAAC,CAAC;4BAClE,CAAC;4BAED,IAAI,sBAAsB,KAAK,eAAe,EAAE,CAAC;gCAC/C,MAAM,6BAA6B,GAAG,uBAAuB,CAAC;gCAC9D,IAAI,6BAA6B,IAAI,CAAC,EAAE,CAAC;oCACvC,IAAI,6BAA6B,GAAG,KAAK,CAAC,OAAO,CAAC,EAAE,CAAC;wCACnD,MAAM,eAAe,GACnB,OAAO,CACL,OAAO,EACP,6BAA6B,EAC7B,6BAA6B,GAAG,CAAC,CAClC,IAAI,GAAG,CAAC;wCACX,MAAM,WAAW,GAAG,KAAK,CAAC,OAAO,CAAC,eAAe,CAAC,CAAC;wCACnD,OAAO;4CACL,OAAO,CAAC,OAAO,EAAE,CAAC,EAAE,6BAA6B,CAAC;gDAClD,WAAW;gDACX,OAAO,CAAC,OAAO,EAAE,6BAA6B,GAAG,CAAC,CAAC,CAAC;oCACxD,CAAC;yCAAM,IACL,6BAA6B,KAAK,KAAK,CAAC,OAAO,CAAC;wCAChD,KAAK,CAAC,OAAO,CAAC,KAAK,UAAU,EAC7B,CAAC;wCACD,OAAO,GAAG,OAAO,GAAG,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC;oCACzC,CAAC;gCACH,CAAC;4BACH,CAAC;4BACD,OAAO,CACL,KAAC,IAAI,cAAyC,OAAO,IAA1C,QAAQ,sBAAsB,EAAE,CAAkB,CAC9D,CAAC;wBACJ,CAAC,CAAC,CACH,GACG,IACF,EAEL,gBAAgB,CAAC,MAAM,GAAG,CAAC,IAAI,CAC9B,KAAC,GAAG,IAAC,aAAa,EAAC,QAAQ,YACxB,gBAAgB,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,KAAK,EAAE,EAAE,CAAC,CACtC,MAAC,IAAI,IAAa,KAAK,EAAE,KAAK,CAAC,MAAM,CAAC,KAAK,wBACtC,KAAK,KADC,KAAK,CAET,CACR,CAAC,GACE,CACP,IACG,CACP,CAAC;AACJ,CAAC"}

package/dist/src/ui/components/shared/text-buffer.d.ts

@@ -4,10 +4,23 @@
  * SPDX-License-Identifier: Apache-2.0
  */
 export type Direction = 'left' | 'right' | 'up' | 'down' | 'wordLeft' | 'wordRight' | 'home' | 'end';
-export declare const findNextWordStart: (text: string, currentOffset: number) => number;
-export declare const findPrevWordStart: (text: string, currentOffset: number) => number;
-export declare const findWordEnd: (text: string, currentOffset: number) => number;
-export declare const getOffsetFromPosition: (row: number, col: number, lines: string[]) => number;
+export declare const isWordCharStrict: (char: string) => boolean;
+export declare const isWhitespace: (char: string) => boolean;
+export declare const isCombiningMark: (char: string) => boolean;
+export declare const isWordCharWithCombining: (char: string) => boolean;
+export declare const getCharScript: (char: string) => string;
+export declare const isDifferentScript: (char1: string, char2: string) => boolean;
+export declare const findNextWordStartInLine: (line: string, col: number) => number | null;
+export declare const findPrevWordStartInLine: (line: string, col: number) => number | null;
+export declare const findWordEndInLine: (line: string, col: number) => number | null;
+export declare const findNextWordAcrossLines: (lines: string[], cursorRow: number, cursorCol: number, searchForWordStart: boolean) => {
+    row: number;
+    col: number;
+} | null;
+export declare const findPrevWordAcrossLines: (lines: string[], cursorRow: number, cursorCol: number) => {
+    row: number;
+    col: number;
+} | null;
 export declare const getPositionFromOffsets: (startOffset: number, endOffset: number, lines: string[]) => {
     startRow: number;
     startCol: number;

package/dist/src/ui/components/shared/text-buffer.js

@@ -3,15 +3,14 @@
  * Copyright 2025 Google LLC
  * SPDX-License-Identifier: Apache-2.0
  */
-import stripAnsi from 'strip-ansi';
-import { spawnSync } from 'child_process';
-import fs from 'fs';
-import os from 'os';
-import pathMod from 'path';
+import { spawnSync } from 'node:child_process';
+import fs from 'node:fs';
+import os from 'node:os';
+import pathMod from 'node:path';
 import { useState, useCallback, useEffect, useMemo, useReducer } from 'react';
 import stringWidth from 'string-width';
 import { unescapePath } from '@umsai/ums-code-core';
-import { toCodePoints, cpLen, cpSlice } from '../../utils/textUtils.js';
+import { toCodePoints, cpLen, cpSlice, stripUnsafeCharacters, } from '../../utils/textUtils.js';
 import { handleVimAction } from './vim-buffer-actions.js';
 // Simple helper for word‑wise ops.
 function isWordChar(ch) {
@@ -20,112 +19,266 @@ function isWordChar(ch) {
     }
     return !/[\s,.;!?]/.test(ch);
 }
-// Vim-specific word boundary functions
-export const findNextWordStart = (text, currentOffset) => {
-    let i = currentOffset;
-    if (i >= text.length)
-        return i;
-    const currentChar = text[i];
+// Helper functions for line-based word navigation
+export const isWordCharStrict = (char) => /[\w\p{L}\p{N}]/u.test(char); // Matches a single character that is any Unicode letter, any Unicode number, or an underscore
+export const isWhitespace = (char) => /\s/.test(char);
+// Check if a character is a combining mark (only diacritics for now)
+export const isCombiningMark = (char) => /\p{M}/u.test(char);
+// Check if a character should be considered part of a word (including combining marks)
+export const isWordCharWithCombining = (char) => isWordCharStrict(char) || isCombiningMark(char);
+// Get the script of a character (simplified for common scripts)
+export const getCharScript = (char) => {
+    if (/[\p{Script=Latin}]/u.test(char))
+        return 'latin'; // All Latin script chars including diacritics
+    if (/[\p{Script=Han}]/u.test(char))
+        return 'han'; // Chinese
+    if (/[\p{Script=Arabic}]/u.test(char))
+        return 'arabic';
+    if (/[\p{Script=Hiragana}]/u.test(char))
+        return 'hiragana';
+    if (/[\p{Script=Katakana}]/u.test(char))
+        return 'katakana';
+    if (/[\p{Script=Cyrillic}]/u.test(char))
+        return 'cyrillic';
+    return 'other';
+};
+// Check if two characters are from different scripts (indicating word boundary)
+export const isDifferentScript = (char1, char2) => {
+    if (!isWordCharStrict(char1) || !isWordCharStrict(char2))
+        return false;
+    return getCharScript(char1) !== getCharScript(char2);
+};
+// Find next word start within a line, starting from col
+export const findNextWordStartInLine = (line, col) => {
+    const chars = toCodePoints(line);
+    let i = col;
+    if (i >= chars.length)
+        return null;
+    const currentChar = chars[i];
     // Skip current word/sequence based on character type
-    if (/\w/.test(currentChar)) {
-        // Skip current word characters
-        while (i < text.length && /\w/.test(text[i])) {
+    if (isWordCharStrict(currentChar)) {
+        while (i < chars.length && isWordCharWithCombining(chars[i])) {
+            // Check for script boundary - if next character is from different script, stop here
+            if (i + 1 < chars.length &&
+                isWordCharStrict(chars[i + 1]) &&
+                isDifferentScript(chars[i], chars[i + 1])) {
+                i++; // Include current character
+                break; // Stop at script boundary
+            }
             i++;
         }
     }
-    else if (!/\s/.test(currentChar)) {
-        // Skip current non-word, non-whitespace characters (like "/", ".", etc.)
-        while (i < text.length && !/\w/.test(text[i]) && !/\s/.test(text[i])) {
+    else if (!isWhitespace(currentChar)) {
+        while (i < chars.length &&
+            !isWordCharStrict(chars[i]) &&
+            !isWhitespace(chars[i])) {
             i++;
         }
     }
     // Skip whitespace
-    while (i < text.length && /\s/.test(text[i])) {
+    while (i < chars.length && isWhitespace(chars[i])) {
         i++;
     }
-    // If we reached the end of text and there's no next word,
-    // vim behavior for dw is to delete to the end of the current word
-    if (i >= text.length) {
-        // Go back to find the end of the last word
-        let endOfLastWord = text.length - 1;
-        while (endOfLastWord >= 0 && /\s/.test(text[endOfLastWord])) {
-            endOfLastWord--;
-        }
-        // For dw on last word, return position AFTER the last character to delete entire word
-        return Math.max(currentOffset + 1, endOfLastWord + 1);
-    }
-    return i;
+    return i < chars.length ? i : null;
 };
-export const findPrevWordStart = (text, currentOffset) => {
-    let i = currentOffset;
-    // If at beginning of text, return current position
-    if (i <= 0) {
-        return currentOffset;
-    }
-    // Move back one character to start searching
+// Find previous word start within a line
+export const findPrevWordStartInLine = (line, col) => {
+    const chars = toCodePoints(line);
+    let i = col;
+    if (i <= 0)
+        return null;
     i--;
     // Skip whitespace moving backwards
-    while (i >= 0 && (text[i] === ' ' || text[i] === '\t' || text[i] === '\n')) {
+    while (i >= 0 && isWhitespace(chars[i])) {
         i--;
     }
-    if (i < 0) {
-        return 0; // Reached beginning of text
-    }
-    const charAtI = text[i];
-    if (/\w/.test(charAtI)) {
+    if (i < 0)
+        return null;
+    if (isWordCharStrict(chars[i])) {
         // We're in a word, move to its beginning
-        while (i >= 0 && /\w/.test(text[i])) {
+        while (i >= 0 && isWordCharStrict(chars[i])) {
+            // Check for script boundary - if previous character is from different script, stop here
+            if (i - 1 >= 0 &&
+                isWordCharStrict(chars[i - 1]) &&
+                isDifferentScript(chars[i], chars[i - 1])) {
+                return i; // Return current position at script boundary
+            }
             i--;
         }
-        return i + 1; // Return first character of word
+        return i + 1;
     }
     else {
         // We're in punctuation, move to its beginning
-        while (i >= 0 &&
-            !/\w/.test(text[i]) &&
-            text[i] !== ' ' &&
-            text[i] !== '\t' &&
-            text[i] !== '\n') {
+        while (i >= 0 && !isWordCharStrict(chars[i]) && !isWhitespace(chars[i])) {
             i--;
         }
-        return i + 1; // Return first character of punctuation sequence
+        return i + 1;
     }
 };
-export const findWordEnd = (text, currentOffset) => {
-    let i = currentOffset;
-    // If we're already at the end of a word, advance to next word
-    if (i < text.length &&
-        /\w/.test(text[i]) &&
-        (i + 1 >= text.length || !/\w/.test(text[i + 1]))) {
-        // We're at the end of a word, move forward to find next word
+// Find word end within a line
+export const findWordEndInLine = (line, col) => {
+    const chars = toCodePoints(line);
+    let i = col;
+    // If we're already at the end of a word (including punctuation sequences), advance to next word
+    // This includes both regular word endings and script boundaries
+    const atEndOfWordChar = i < chars.length &&
+        isWordCharWithCombining(chars[i]) &&
+        (i + 1 >= chars.length ||
+            !isWordCharWithCombining(chars[i + 1]) ||
+            (isWordCharStrict(chars[i]) &&
+                i + 1 < chars.length &&
+                isWordCharStrict(chars[i + 1]) &&
+                isDifferentScript(chars[i], chars[i + 1])));
+    const atEndOfPunctuation = i < chars.length &&
+        !isWordCharWithCombining(chars[i]) &&
+        !isWhitespace(chars[i]) &&
+        (i + 1 >= chars.length ||
+            isWhitespace(chars[i + 1]) ||
+            isWordCharWithCombining(chars[i + 1]));
+    if (atEndOfWordChar || atEndOfPunctuation) {
+        // We're at the end of a word or punctuation sequence, move forward to find next word
         i++;
-        // Skip whitespace/punctuation to find next word
-        while (i < text.length && !/\w/.test(text[i])) {
+        // Skip whitespace to find next word or punctuation
+        while (i < chars.length && isWhitespace(chars[i])) {
             i++;
         }
     }
-    // If we're not on a word character, find the next word
-    if (i < text.length && !/\w/.test(text[i])) {
-        while (i < text.length && !/\w/.test(text[i])) {
+    // If we're not on a word character, find the next word or punctuation sequence
+    if (i < chars.length && !isWordCharWithCombining(chars[i])) {
+        // Skip whitespace to find next word or punctuation
+        while (i < chars.length && isWhitespace(chars[i])) {
             i++;
         }
     }
-    // Move to end of current word
-    while (i < text.length && /\w/.test(text[i])) {
-        i++;
+    // Move to end of current word (including combining marks, but stop at script boundaries)
+    let foundWord = false;
+    let lastBaseCharPos = -1;
+    if (i < chars.length && isWordCharWithCombining(chars[i])) {
+        // Handle word characters
+        while (i < chars.length && isWordCharWithCombining(chars[i])) {
+            foundWord = true;
+            // Track the position of the last base character (not combining mark)
+            if (isWordCharStrict(chars[i])) {
+                lastBaseCharPos = i;
+            }
+            // Check if next character is from a different script (word boundary)
+            if (i + 1 < chars.length &&
+                isWordCharStrict(chars[i + 1]) &&
+                isDifferentScript(chars[i], chars[i + 1])) {
+                i++; // Include current character
+                if (isWordCharStrict(chars[i - 1])) {
+                    lastBaseCharPos = i - 1;
+                }
+                break; // Stop at script boundary
+            }
+            i++;
+        }
+    }
+    else if (i < chars.length && !isWhitespace(chars[i])) {
+        // Handle punctuation sequences (like ████)
+        while (i < chars.length &&
+            !isWordCharStrict(chars[i]) &&
+            !isWhitespace(chars[i])) {
+            foundWord = true;
+            lastBaseCharPos = i;
+            i++;
+        }
     }
-    // Move back one to be on the last character of the word
-    return Math.max(currentOffset, i - 1);
+    // Only return a position if we actually found a word
+    // Return the position of the last base character, not combining marks
+    if (foundWord && lastBaseCharPos >= col) {
+        return lastBaseCharPos;
+    }
+    return null;
 };
-// Helper functions for vim operations
-export const getOffsetFromPosition = (row, col, lines) => {
-    let offset = 0;
-    for (let i = 0; i < row; i++) {
-        offset += lines[i].length + 1; // +1 for newline
+// Find next word across lines
+export const findNextWordAcrossLines = (lines, cursorRow, cursorCol, searchForWordStart) => {
+    // First try current line
+    const currentLine = lines[cursorRow] || '';
+    const colInCurrentLine = searchForWordStart
+        ? findNextWordStartInLine(currentLine, cursorCol)
+        : findWordEndInLine(currentLine, cursorCol);
+    if (colInCurrentLine !== null) {
+        return { row: cursorRow, col: colInCurrentLine };
     }
-    offset += col;
-    return offset;
+    // Search subsequent lines
+    for (let row = cursorRow + 1; row < lines.length; row++) {
+        const line = lines[row] || '';
+        const chars = toCodePoints(line);
+        // For empty lines, if we haven't found any words yet, return the empty line
+        if (chars.length === 0) {
+            // Check if there are any words in remaining lines
+            let hasWordsInLaterLines = false;
+            for (let laterRow = row + 1; laterRow < lines.length; laterRow++) {
+                const laterLine = lines[laterRow] || '';
+                const laterChars = toCodePoints(laterLine);
+                let firstNonWhitespace = 0;
+                while (firstNonWhitespace < laterChars.length &&
+                    isWhitespace(laterChars[firstNonWhitespace])) {
+                    firstNonWhitespace++;
+                }
+                if (firstNonWhitespace < laterChars.length) {
+                    hasWordsInLaterLines = true;
+                    break;
+                }
+            }
+            // If no words in later lines, return the empty line
+            if (!hasWordsInLaterLines) {
+                return { row, col: 0 };
+            }
+            continue;
+        }
+        // Find first non-whitespace
+        let firstNonWhitespace = 0;
+        while (firstNonWhitespace < chars.length &&
+            isWhitespace(chars[firstNonWhitespace])) {
+            firstNonWhitespace++;
+        }
+        if (firstNonWhitespace < chars.length) {
+            if (searchForWordStart) {
+                return { row, col: firstNonWhitespace };
+            }
+            else {
+                // For word end, find the end of the first word
+                const endCol = findWordEndInLine(line, firstNonWhitespace);
+                if (endCol !== null) {
+                    return { row, col: endCol };
+                }
+            }
+        }
+    }
+    return null;
 };
+// Find previous word across lines
+export const findPrevWordAcrossLines = (lines, cursorRow, cursorCol) => {
+    // First try current line
+    const currentLine = lines[cursorRow] || '';
+    const colInCurrentLine = findPrevWordStartInLine(currentLine, cursorCol);
+    if (colInCurrentLine !== null) {
+        return { row: cursorRow, col: colInCurrentLine };
+    }
+    // Search previous lines
+    for (let row = cursorRow - 1; row >= 0; row--) {
+        const line = lines[row] || '';
+        const chars = toCodePoints(line);
+        if (chars.length === 0)
+            continue;
+        // Find last word start
+        let lastWordStart = chars.length;
+        while (lastWordStart > 0 && isWhitespace(chars[lastWordStart - 1])) {
+            lastWordStart--;
+        }
+        if (lastWordStart > 0) {
+            // Find start of this word
+            const wordStart = findPrevWordStartInLine(line, lastWordStart);
+            if (wordStart !== null) {
+                return { row, col: wordStart };
+            }
+        }
+    }
+    return null;
+};
+// Helper functions for vim line operations
 export const getPositionFromOffsets = (startOffset, endOffset, lines) => {
     let offset = 0;
     let startRow = 0;
@@ -218,27 +371,6 @@ export const replaceRangeInternal = (state, startRow, startCol, endRow, endCol,
         preferredCol: null,
     };
 };
-/**
- * Strip characters that can break terminal rendering.
- *
- * Strip ANSI escape codes and control characters except for line breaks.
- * Control characters such as delete break terminal UI rendering.
- */
-function stripUnsafeCharacters(str) {
-    const stripped = stripAnsi(str);
-    return toCodePoints(stripped)
-        .filter((char) => {
-        if (char.length > 1)
-            return false;
-        const code = char.codePointAt(0);
-        if (code === undefined) {
-            return false;
-        }
-        const isUnsafe = code === 127 || (code <= 31 && code !== 13 && code !== 10);
-        return !isUnsafe;
-    })
-        .join('');
-}
 function clamp(v, min, max) {
     return v < min ? min : v > max ? max : v;
 }
@@ -1219,6 +1351,12 @@ export function useTextBuffer({ initialText = '', initialCursorOffset = 0, viewp
     }, [text, stdin, setRawMode]);
     const handleInput = useCallback((key) => {
         const { sequence: input } = key;
+        if (key.paste) {
+            // Do not do any other processing on pastes so ensure we handle them
+            // before all other cases.
+            insert(input, { paste: key.paste });
+            return;
+        }
         if (key.name === 'return' ||
             input === '\r' ||
             input === '\n' ||