dpsapi/e2e/harness/gostrings.ts

// SPDX-License-Identifier: GPL-2.0-or-later
// Copyright (C) 2026 Bryan Joshua Pedini

/**
 * Go string quoting/unquoting compatible with strconv.Quote / strconv.Unquote,
 * as used by `token list` (DESCRIPTION column), slog's text handler and the
 * configuration error messages.
 */

/** Characters Go's strconv.IsPrint accepts: letters, marks, numbers, punctuation, symbols, and ASCII space. */
const PRINTABLE = /^[\p{L}\p{M}\p{N}\p{P}\p{S} ]$/u;

/**
 * Go strconv.Quote of a byte string (Buffer) or a JS string (encoded as UTF-8).
 * Note: relies on Node's Unicode tables; avoid code points newer than Go's tables in tests.
 */
export function goQuote(input: string | Buffer): string {
  const bytes = typeof input === 'string' ? Buffer.from(input, 'utf8') : input;
  let out = '"';
  let i = 0;
  while (i < bytes.length) {
    const b = bytes[i];
    if (b < 0x80) {
      i++;
      const c = String.fromCharCode(b);
      switch (c) {
        case '"':
          out += '\\"';
          continue;
        case '\\':
          out += '\\\\';
          continue;
        case '\x07':
          out += '\\a';
          continue;
        case '\b':
          out += '\\b';
          continue;
        case '\f':
          out += '\\f';
          continue;
        case '\n':
          out += '\\n';
          continue;
        case '\r':
          out += '\\r';
          continue;
        case '\t':
          out += '\\t';
          continue;
        case '\v':
          out += '\\v';
          continue;
      }
      if (b < 0x20 || b === 0x7f) out += '\\x' + hex2(b);
      else out += c;
      continue;
    }
    // Multi-byte: decode one UTF-8 sequence strictly
    const cp = decodeUtf8At(bytes, i);
    if (!cp) {
      out += '\\x' + hex2(b);
      i++;
      continue;
    }
    const ch = String.fromCodePoint(cp.codePoint);
    if (PRINTABLE.test(ch)) out += ch;
    else if (cp.codePoint < 0x10000) out += '\\u' + cp.codePoint.toString(16).padStart(4, '0');
    else out += '\\U' + cp.codePoint.toString(16).padStart(8, '0');
    i += cp.length;
  }
  return out + '"';
}

function hex2(b: number): string {
  return b.toString(16).padStart(2, '0');
}

/** Strict UTF-8 decode of the sequence at i (rejects overlongs, surrogates, > U+10FFFF). */
function decodeUtf8At(buf: Buffer, i: number): { codePoint: number; length: number } | undefined {
  const b0 = buf[i];
  let len: number;
  let cp: number;
  let min: number;
  if (b0 >= 0xc2 && b0 <= 0xdf) {
    len = 2;
    cp = b0 & 0x1f;
    min = 0x80;
  } else if (b0 >= 0xe0 && b0 <= 0xef) {
    len = 3;
    cp = b0 & 0x0f;
    min = 0x800;
  } else if (b0 >= 0xf0 && b0 <= 0xf4) {
    len = 4;
    cp = b0 & 0x07;
    min = 0x10000;
  } else return undefined;
  if (i + len > buf.length) return undefined;
  for (let k = 1; k < len; k++) {
    const b = buf[i + k];
    if ((b & 0xc0) !== 0x80) return undefined;
    cp = (cp << 6) | (b & 0x3f);
  }
  if (cp < min || cp > 0x10ffff || (cp >= 0xd800 && cp <= 0xdfff)) return undefined;
  return { codePoint: cp, length: len };
}

/**
 * Go strconv.Unquote for double-quoted strings. Returns the raw bytes (descriptions may
 * contain invalid UTF-8). Throws on malformed input.
 */
export function goUnquoteBytes(quoted: string): Buffer {
  if (quoted.length < 2 || quoted[0] !== '"' || quoted[quoted.length - 1] !== '"') {
    throw new Error(`not a Go double-quoted string: ${JSON.stringify(quoted)}`);
  }
  const s = quoted.slice(1, -1);
  const out: number[] = [];
  const pushStr = (str: string) => out.push(...Buffer.from(str, 'utf8'));
  for (let i = 0; i < s.length; ) {
    const c = s[i];
    if (c === '"') throw new Error(`unescaped quote in Go string: ${JSON.stringify(quoted)}`);
    if (c !== '\\') {
      const cp = s.codePointAt(i)!;
      const ch = String.fromCodePoint(cp);
      pushStr(ch);
      i += ch.length;
      continue;
    }
    const e = s[i + 1];
    const simple: Record<string, number> = { a: 7, b: 8, f: 12, n: 10, r: 13, t: 9, v: 11, '\\': 92, '"': 34, "'": 39 };
    if (e in simple) {
      out.push(simple[e]);
      i += 2;
    } else if (e === 'x') {
      out.push(parseHex(s.slice(i + 2, i + 4), quoted));
      i += 4;
    } else if (e === 'u') {
      pushStr(String.fromCodePoint(parseHex(s.slice(i + 2, i + 6), quoted)));
      i += 6;
    } else if (e === 'U') {
      pushStr(String.fromCodePoint(parseHex(s.slice(i + 2, i + 10), quoted)));
      i += 10;
    } else if (e >= '0' && e <= '7') {
      const oct = s.slice(i + 1, i + 4);
      if (!/^[0-7]{3}$/.test(oct)) throw new Error(`bad octal escape in ${JSON.stringify(quoted)}`);
      out.push(parseInt(oct, 8));
      i += 4;
    } else {
      throw new Error(`bad escape \\${e} in Go string ${JSON.stringify(quoted)}`);
    }
  }
  return Buffer.from(out);
}

/** goUnquoteBytes decoded as UTF-8 (invalid sequences become U+FFFD). */
export function goUnquote(quoted: string): string {
  return goUnquoteBytes(quoted).toString('utf8');
}

function parseHex(h: string, ctx: string): number {
  if (!/^[0-9a-fA-F]+$/.test(h)) throw new Error(`bad hex escape in ${JSON.stringify(ctx)}`);
  return parseInt(h, 16);
}

/** POSIX sh single-quoting: the result is always one shell word with the exact bytes of s. */
export function shellQuote(s: string): string {
  return `'${s.split("'").join(`'\\''`)}'`;
}