rea/src/native/NativeMachoInspection.ts

import type { BinaryTarget } from "../domain/binaryTargetTypes.js";
import type { EvidenceLocation } from "../domain/evidence.js";
import type { AnalysisError } from "../domain/analysisErrorBase.js";
import {
  AnalysisOutputError,
  AnalysisUnsupportedTargetError,
} from "../domain/analysisErrorCore.js";
import {
  inspectMachoSchema,
  type NativeCommandInvocation,
} from "../domain/native/nativeInspection.js";
import { jsonValueSchema, type JsonValue } from "../domain/jsonValue.js";
import { machoRuntimeSymbolName } from "../domain/native/machoSymbolName.js";
import { err, ok, type Result } from "../domain/result.js";
import type { NativeCommandCapture } from "./CommandRunner.js";
import {
  MACHO_ARCHITECTURE_POLICIES,
  selectMachoArchitectureName,
} from "./MachoArchitectureSelection.js";
import {
  parseDyldExportAddressRow,
  parseDyldExports,
  parseDyldImports,
} from "./parsers/dyldInfo.js";
import { parseLipoCapture, type LipoArchitecture } from "./parsers/lipo.js";
import { withoutEchoedPathHeader } from "./parsers/echoedPath.js";
import { parseOtoolLoadCommands } from "./parsers/otool.js";
import { reconcileMachoUuid } from "./parsers/machoUuid.js";
import { compareUnicodeCodePoints } from "../domain/unicodeCodePointOrder.js";
import {
  conflictingUuidCommandEvidence,
  isMissingUuidDyldFailure,
  isUnbackedTextDyldFailure,
  nativeMachoCapturedOutput,
  missingUuidLimitation,
  parseDyldFileTypeRefusal,
  unusableNativeCommandOutput,
} from "./NativeMachoCommandDiagnostics.js";

interface NativeMachoObservation {
  readonly result: JsonValue;
  readonly limitations: readonly string[];
  readonly locations: readonly EvidenceLocation[];
}

interface NativeMachoInspectionContext {
  readonly target: BinaryTarget;
  readonly signal?: AbortSignal;
  readonly run: (
    tool: string,
    arguments_: readonly string[],
    signal?: AbortSignal,
    acceptNonZeroIf?: (capture: NativeCommandCapture) => boolean,
  ) => Promise<Result<NativeCommandCapture, AnalysisError>>;
  readonly invocation: (
    capture: NativeCommandCapture,
  ) => NativeCommandInvocation;
}

const DISCOVERY_COMMANDS = [["lipo", ["-detailed_info"]]] as const;

// llvm-nm's default dyld-info pass fakes up symbols the symbol table lacks
// from dyld metadata. It takes seconds on large apps (Keynote arm64: 5.8 s
// versus 0.01 s without it), and under -gU it can only add export-trie names.
const SYMBOL_TABLE_EXPORTS = ["nm", ["-gjU", "--no-dyldinfo"]] as const;
const SYMBOL_TABLE_IMPORTS = ["nm", ["-gju", "--no-dyldinfo"]] as const;
const DYLD_INFO_EXPORTS = ["nm", ["-gjU"]] as const;

const REQUIRED_SLICE_COMMANDS = [
  ["otool", ["-h", "-l"]],
  SYMBOL_TABLE_EXPORTS,
  ["dyld_info", ["-imports"]],
  ["dyld_info", ["-exports"]],
  ["dwarfdump", ["--uuid"]],
] as const;

const OPTIONAL_VTOOL_COMMAND = ["vtool", ["-show-build"]] as const;
const VTOOL_UNAVAILABLE_LIMITATION =
  "vtool is unavailable; build metadata is normalized from otool only.";
/** Inspect one Mach-O with the required native commands and normalized output. */
export const inspectNativeMacho = async (
  context: NativeMachoInspectionContext,
): Promise<Result<NativeMachoObservation, AnalysisError>> => {
  const captures: NativeCommandCapture[] = [];
  for (const [tool, prefix] of DISCOVERY_COMMANDS) {
    const captured = await context.run(
      tool,
      [...prefix, context.target.path],
      context.signal,
    );
    if (!captured.ok) return err(captured.error);
    captures.push(captured.value);
  }
  const lipo = captures.find(({ tool }) => tool === "lipo");
  if (lipo === undefined) throw new TypeError("Missing lipo capture");
  const parsedLipo = parseLipoCapture(
    lipo,
    "inspect_macho",
    context.target.path,
  );
  if (!parsedLipo.ok) return parsedLipo;
  const { format, architectures } = parsedLipo.value;
  const selection = selectArchitecture(context.target, architectures);
  if (selection.status === "absent")
    return err(
      new AnalysisOutputError(
        "inspect_macho",
        `lipo did not list ${selection.architecture}; refusing to read another slice`,
      ),
    );
  const selectedArchitecture =
    selection.status === "selected" ? selection.architecture : null;
  const runSliceCommand = ([tool, prefix]: readonly [
    string,
    readonly string[],
  ]) =>
    context.run(
      tool,
      [
        ...prefix,
        ...selectedArchitectureArguments(tool, selectedArchitecture),
        context.target.path,
      ],
      context.signal,
      tool === "dyld_info"
        ? (capture) =>
            isMissingUuidDyldFailure(capture) ||
            isUnbackedTextDyldFailure(capture) ||
            parseDyldFileTypeRefusal(capture) !== null
        : undefined,
    );
  // vtool knows no variant subtypes, so no spelling selects an arm64e.x1 slice.
  const vtoolSelectsSlice =
    selectedArchitecture === null ||
    !toolArchitectureName("vtool", selectedArchitecture).includes(".");
  // Slice commands only read the selected artifact, so launch them together.
  // Every launch settles before return; declared order fixes provenance and
  // which failure is reported.
  const [required, optional] = await Promise.all([
    Promise.all(
      REQUIRED_SLICE_COMMANDS.map((command) => runSliceCommand(command)),
    ),
    vtoolSelectsSlice ? runSliceCommand(OPTIONAL_VTOOL_COMMAND) : null,
  ]);
  for (const captured of required) {
    if (!captured.ok) return err(captured.error);
    if (
      (captured.value.exitCode !== 0 || captured.value.signal !== null) &&
      !isMissingUuidDyldFailure(captured.value) &&
      !isUnbackedTextDyldFailure(captured.value) &&
      parseDyldFileTypeRefusal(captured.value) === null
    )
      return err(unusableNativeCommandOutput(captured.value));
    captures.push(captured.value);
  }
  const limitations = architectures.flatMap(
    (architecture) => architecture.limitations,
  );
  if (optional === null)
    limitations.push(
      `vtool cannot select the ${selectedArchitecture} slice; build metadata is normalized from otool only.`,
    );
  else if (optional.ok) captures.push(optional.value);
  else if (optional.error._tag === "AnalysisCapabilityUnavailableError")
    limitations.push(VTOOL_UNAVAILABLE_LIMITATION);
  else return err(optional.error);
  const nmIndex = captures.findIndex(({ tool }) => tool === "nm");
  const dyldExports = captures.filter(({ tool }) => tool === "dyld_info")[1];
  if (nmIndex < 0 || dyldExports === undefined)
    throw new TypeError("Missing nm or dyld_info export capture");
  const otool = captures.find(({ tool }) => tool === "otool");
  if (otool === undefined) throw new TypeError("Missing otool capture");
  let load: ReturnType<typeof parseOtoolLoadCommands>;
  try {
    load = parseOtoolLoadCommands(sliceOutput(otool));
  } catch (cause: unknown) {
    return err(
      new AnalysisOutputError(
        "inspect_macho",
        "Malformed otool load-command output",
        {
          cause,
          capturedOutput: nativeMachoCapturedOutput(otool),
        },
      ),
    );
  }
  const unbackedTextFailure = captures.find(isUnbackedTextDyldFailure);
  if (unbackedTextFailure !== undefined) {
    if (!["10", "DSYM", "MH_DSYM"].includes(load.fileType ?? ""))
      return err(unusableNativeCommandOutput(unbackedTextFailure));
    return err(
      new AnalysisUnsupportedTargetError(
        "inspect_macho",
        context.target.path,
        "The selected Mach-O is a dSYM companion whose unbacked text sections are unsupported by dyld_info inspection",
        {
          capturedOutput: nativeMachoCapturedOutput(unbackedTextFailure),
          remediationAction:
            "Select the matching linked Mach-O executable for native inspection; this operation does not inspect dSYM companions with unbacked text sections.",
        },
      ),
    );
  }
  for (const fileTypeFailure of captures) {
    const refusal = parseDyldFileTypeRefusal(fileTypeFailure);
    if (refusal === null) continue;
    const { fileType, diagnostic } = refusal;
    const typeName = fileType === "4" ? "CORE" : "FVMLIB";
    if (![fileType, typeName, `MH_${typeName}`].includes(load.fileType ?? ""))
      return err(unusableNativeCommandOutput(fileTypeFailure));
    return err(
      new AnalysisUnsupportedTargetError(
        "inspect_macho",
        context.target.path,
        `dyld_info refused the selected Mach-O file type MH_${typeName} (${fileType}): ${diagnostic}`,
        {
          capturedOutput: nativeMachoCapturedOutput(fileTypeFailure),
          remediationAction:
            "Select a linked Mach-O executable or relocatable object for native inspection; dyld_info refused the selected file type.",
        },
      ),
    );
  }
  const uuidlessDyldFailure = captures.find(isMissingUuidDyldFailure);
  if (
    uuidlessDyldFailure !== undefined &&
    load.commands.some(({ kind }) => kind === "LC_UUID")
  )
    return err(conflictingUuidCommandEvidence(uuidlessDyldFailure));
  // A stripped symbol table (strip -T) can omit names the export trie keeps.
  // Only then does llvm-nm's dyld-info pass add names, and its exact names
  // disambiguate dyld_info rows. Refusals above are not export rows.
  if (
    !isMissingUuidDyldFailure(dyldExports) &&
    !namesEveryExport(captures[nmIndex]!.stdout, dyldExports)
  ) {
    const withDyldInfo = await runSliceCommand(DYLD_INFO_EXPORTS);
    if (!withDyldInfo.ok) return err(withDyldInfo.error);
    captures[nmIndex] = withDyldInfo.value;
  }
  const dyldImports = captures.find(
    ({ tool, arguments: arguments_ }) =>
      tool === "dyld_info" && arguments_[0] === "-imports",
  );
  if (dyldImports === undefined)
    throw new TypeError("Missing dyld_info import capture");
  const hasChainedFixups = load.commands.some(
    ({ kind }) => kind === "LC_DYLD_CHAINED_FIXUPS",
  );
  let parsedImports = parseDyldImports(
    sliceOutput(dyldImports),
    hasChainedFixups,
  );
  if (
    !isMissingUuidDyldFailure(dyldImports) &&
    parsedImports.needsNameInventory
  ) {
    const importNames = await runSliceCommand(SYMBOL_TABLE_IMPORTS);
    if (!importNames.ok) return err(importNames.error);
    captures.push(importNames.value);
    parsedImports = parseDyldImports(
      sliceOutput(dyldImports),
      hasChainedFixups,
      new Set(parseNmSymbolNames(importNames.value.stdout)),
    );
  }
  const selectedArchitectureName =
    selectedArchitecture ??
    (architectures.length === 1 ? (architectures[0]?.name ?? null) : null);
  const containerOffset =
    selectedArchitectureName === null
      ? undefined
      : format === "thin"
        ? 0
        : (architectures.find(({ name }) => name === selectedArchitectureName)
            ?.file_offset ?? undefined);
  const normalized = normalizeMacho(
    captures,
    format,
    context.invocation,
    limitations,
    load,
    parsedImports.items,
    captures[nmIndex]!,
    architectures,
    context.target.path,
    {
      name: selectedArchitectureName,
      container_file_offset: containerOffset ?? null,
    },
  );
  const result =
    containerOffset === undefined
      ? inspectMachoSchema.parse({
          ...normalized,
          limitations: [
            ...normalized.limitations,
            selectedArchitectureName === null
              ? "The selected Mach-O slice is unknown; slice-based evidence locations are unavailable."
              : "The selected Mach-O slice's container file offset was not observed; slice-based evidence locations are unavailable.",
          ],
        })
      : normalized;
  return ok({
    result: jsonValueSchema.parse(result),
    limitations: result.limitations,
    locations: fileOffsetLocations(result, containerOffset),
  });
};

const normalizeMacho = (
  captures: readonly NativeCommandCapture[],
  containerFormat: "thin" | "fat",
  toInvocation: (capture: NativeCommandCapture) => NativeCommandInvocation,
  additionalLimitations: readonly string[],
  load: ReturnType<typeof parseOtoolLoadCommands>,
  parsedImports: ReturnType<typeof parseDyldImports>["items"],
  exportNames: NativeCommandCapture,
  architectures: readonly LipoArchitecture[],
  artifactPath: string,
  selectedArchitecture: {
    readonly name: string | null;
    readonly container_file_offset: number | null;
  },
) => {
  const byTool = captureLookup(captures);
  const dyldImportsCapture = byTool("dyld_info", 0);
  const dyldExportsCapture = byTool("dyld_info", 1);
  const missingImportsUuid = isMissingUuidDyldFailure(dyldImportsCapture);
  const missingExportsUuid = isMissingUuidDyldFailure(dyldExportsCapture);
  const imports = missingImportsUuid ? [] : parsedImports.map(withRuntimeName);
  const imageBase =
    load.segments.find(
      (segment) => segment.file_offset === 0 && (segment.file_size ?? 0) > 0,
    )?.vm_address ?? null;
  const nmExports = parseNmExports(exportNames.stdout);
  const dyldExports = missingExportsUuid
    ? []
    : parseDyldExports(
        sliceOutput(dyldExportsCapture),
        imageBase,
        new Set(nmExports.map(({ name }) => name)),
      );
  const exports = uniqueSymbols([...dyldExports, ...nmExports]).map(
    withRuntimeName,
  );
  const dwarfdump = byTool("dwarfdump");
  const uuid = reconcileMachoUuid(
    load.commands,
    dwarfdump.stdout,
    artifactPath,
    selectedArchitecture.name === null
      ? null
      : {
          name: selectedArchitecture.name,
          dwarfdumpName: toolArchitectureName(
            "dwarfdump",
            selectedArchitecture.name,
          ),
        },
  );
  const provenance = captures.map(toInvocation);
  // A thread-state command carries the entrypoint in architecture-specific
  // register state, so an absent decoded offset means unknown, not absent.
  const threadEntrypoint =
    load.entrypoints.length === 0 &&
    load.commands.some(
      (command) =>
        command.kind === "LC_UNIXTHREAD" || command.kind === "LC_THREAD",
    );
  const threadEntrypointLimitation =
    "Thread-state entrypoints (LC_UNIXTHREAD/LC_THREAD) are retained as raw load commands; file offsets are not derived.";
  const limitations = [
    "Imports and exports combine dyld_info and nm; stripped or toolchain-hidden symbols may be absent.",
    ...(captures.some(({ tool }) => tool === "vtool")
      ? [
          "vtool output is retained as provenance but only otool build metadata is normalized.",
        ]
      : []),
    ...additionalLimitations,
    ...(missingImportsUuid ? [missingUuidLimitation("imports")] : []),
    ...(missingExportsUuid ? [missingUuidLimitation("exports")] : []),
    ...uuid.limitations,
    ...(threadEntrypoint ? [threadEntrypointLimitation] : []),
    ...(imageBase === null &&
    /^\s*offset\s+symbol\s*$/mu.test(sliceOutput(byTool("dyld_info", 1)))
      ? [
          "Export virtual addresses are unavailable because no file-backed image base was observed.",
        ]
      : []),
  ];
  return inspectMachoSchema.parse({
    format: containerFormat === "fat" ? "mach-o-universal" : "mach-o",
    selected_architecture: selectedArchitecture,
    endian: load.endian,
    word_size: load.wordSize,
    file_type: load.fileType,
    flags: load.flags,
    uuid: uuid.value,
    entrypoints: covered(
      load.entrypoints,
      !threadEntrypoint,
      threadEntrypoint ? [threadEntrypointLimitation] : [],
    ),
    architectures: covered(architectures, true),
    build_metadata: covered(load.builds, true),
    load_commands: covered(load.commands, true),
    dependencies: covered(
      load.dependencies,
      load.omittedDependencyCommands.length === 0,
      dependencyPayloadLimitations(load.omittedDependencyCommands),
    ),
    imports: covered(imports, false, [
      "dyld_info textual imports may omit chained or toolchain-unsupported metadata.",
      ...(missingImportsUuid ? [missingUuidLimitation("imports")] : []),
      ...(imports.some(({ name_status }) => name_status === "ambiguous")
        ? [
            "Import text contains name, flags, library, or addend boundaries whose split is ambiguous.",
          ]
        : []),
    ]),
    exports: covered(exports, false, [
      ...(missingExportsUuid
        ? [missingUuidLimitation("exports")]
        : [
            "Merged nm/dyld_info results may be incomplete for stripped binaries.",
          ]),
      ...(dyldExports.some(({ sourceSymbol }) => sourceSymbol !== undefined)
        ? [
            "Renamed re-exports are listed under their exported name; the source library's symbol name is not reported.",
          ]
        : []),
    ]),
    segments: covered(
      load.segments.map((segment) => ({
        ...segment,
        sections: covered(segment.sections, true),
      })),
      true,
    ),
    provenance,
    limitations,
  });
};

/** Slice output without the echoed operand path, which can contain any text. */
const sliceOutput = (capture: NativeCommandCapture): string =>
  withoutEchoedPathHeader(capture.stdout, capture.arguments.at(-1));

const captureLookup =
  (captures: readonly NativeCommandCapture[]) =>
  (tool: string, occurrence = 0): NativeCommandCapture => {
    const capture = captures.filter((item) => item.tool === tool)[occurrence];
    if (capture === undefined) throw new TypeError(`Missing ${tool} capture`);
    return capture;
  };

/**
 * Whether every dyld_info export row names a symbol-table export. A row this
 * parser cannot read counts as uncovered, so an unfamiliar format falls back.
 */
const namesEveryExport = (
  nmOutput: string,
  dyldExports: NativeCommandCapture,
): boolean => {
  const names = new Set(parseNmSymbolNames(nmOutput));
  return sliceOutput(dyldExports)
    .split("\n")
    .every((rawLine, index) => {
      // Keep a trailing carriage return: it belongs to the exported name.
      const line = rawLine.trimStart();
      if (
        line.length === 0 ||
        (index === 0 && line.endsWith(":")) ||
        /^-?exports:$/u.test(line) ||
        /^offset\s+symbol$/u.test(line)
      )
        return true;
      const rows = parseDyldExports(line, null, names);
      // The parser withholds a row only when nm names both of its readings.
      if (rows.length === 0)
        return names.has(parseDyldExportAddressRow(line)?.text ?? "");
      return rows.length === 1 && names.has(rows[0]!.name);
    });
};

// nm -j ends each name with LF; a carriage return before it is part of the name.
const parseNmSymbolNames = (output: string): string[] =>
  output.split("\n").filter((name) => name.length > 0);

const parseNmExports = (output: string) =>
  parseNmSymbolNames(output).map((name) => ({
    name,
    address: null,
    weak: null,
    reexport: null,
    source: "nm",
  }));

const dependencyPayloadLimitations = (
  commands: readonly string[],
): string[] => {
  if (commands.length === 0) return [];
  const counts = new Map<string, number>();
  for (const command of commands)
    counts.set(command, (counts.get(command) ?? 0) + 1);
  const described = [...counts].map(
    ([command, count]) => `${count} ${command}`,
  );
  return [
    `Dependency inventory omits ${described.join(", ")} because otool -l does not print that command's linkedit install name. trace_dylib_resolution reads the payload.`,
  ];
};

const covered = <Value>(
  items: readonly Value[],
  exhaustive: boolean,
  limitations: readonly string[] = [],
) => ({
  items: [...items],
  total: exhaustive ? items.length : null,
  exhaustive,
  limitations: [...limitations],
});

const withRuntimeName = <Value extends { readonly name: string | null }>(
  symbol: Value,
): Value & { readonly runtime_name: string | null } => ({
  ...symbol,
  runtime_name: machoRuntimeSymbolName(symbol.name),
});

const uniqueSymbols = <Value extends { readonly name: string }>(
  items: readonly Value[],
): Value[] => {
  const unique = new Map<string, Value>();
  for (const item of items)
    if (!unique.has(item.name)) unique.set(item.name, item);
  return [...unique.values()].sort((left, right) =>
    compareUnicodeCodePoints(left.name, right.name),
  );
};

const fileOffsetLocations = (
  macho: ReturnType<typeof inspectMachoSchema.parse>,
  containerOffset: number | undefined,
): EvidenceLocation[] => {
  const locations = architectureLocations(macho.architectures.items);
  if (containerOffset === undefined) return locations;
  for (const entrypoint of macho.entrypoints.items)
    locations.push({
      kind: "file-offset",
      offset: containerOffset + entrypoint.file_offset,
    });
  for (const segment of macho.segments.items) {
    if (
      segment.file_offset === null ||
      segment.file_size === null ||
      segment.file_size === 0
    )
      continue;
    locations.push({
      kind: "file-offset-range",
      start: containerOffset + segment.file_offset,
      end: containerOffset + segment.file_offset + segment.file_size,
    });
  }
  return locations;
};

/** Project architecture slices into evidence file-offset locations. */
export const architectureLocations = (
  architectures: readonly {
    readonly file_offset: number | null;
    readonly size: number | null;
  }[],
): EvidenceLocation[] => {
  const locations: EvidenceLocation[] = [];
  for (const { file_offset: offset, size } of architectures) {
    if (offset === null) continue;
    locations.push(
      size === null
        ? { kind: "file-offset", offset }
        : { kind: "file-offset-range", start: offset, end: offset + size },
    );
  }
  return locations;
};

/**
 * lipo appends the pointer-auth ABI version to arm64e (`arm64e.v1`). Only
 * dyld_info accepts that spelling; the other Apple tools select the slice by
 * its CPU-family name. Other variants are distinct subtypes (`arm64e.x1`)
 * that the tools address only by their exact name.
 */
const toolArchitectureName = (tool: string, architecture: string): string =>
  tool === "dyld_info" ? architecture : architecture.replace(/\.v\d+$/u, "");

const selectedArchitectureArguments = (
  tool: string,
  architecture: string | null,
): string[] => {
  if (architecture === null) return [];
  const toolArchitecture = toolArchitectureName(tool, architecture);
  return tool === "dwarfdump"
    ? [`--arch=${toolArchitecture}`]
    : ["-arch", toolArchitecture];
};

type ArchitectureSelection =
  | { readonly status: "unfiltered" }
  | { readonly status: "selected"; readonly architecture: string }
  | { readonly status: "absent"; readonly architecture: string };

const selectArchitecture = (
  target: BinaryTarget,
  available: readonly LipoArchitecture[],
): ArchitectureSelection => {
  // BinaryTarget lists CPU families and lipo lists slices. A universal file
  // can hold several slices of one family (arm64e and arm64e.x1), and slice
  // tools without -arch print every slice.
  if (
    target.kind !== "executable" ||
    (target.availableArchitectures.length < 2 && available.length < 2)
  )
    return { status: "unfiltered" };
  const normalized =
    target.architecture === "x86" ? "i386" : target.architecture;
  const policy = MACHO_ARCHITECTURE_POLICIES[target.architecture];
  const selected =
    policy === undefined
      ? available.find(({ name }) => name === normalized)?.name
      : selectMachoArchitectureName(policy, available);
  if (selected !== undefined)
    return { status: "selected", architecture: selected };
  return { status: "absent", architecture: normalized };
};