Jack Park

14 papers C 2Unranked 9
YearRankTypeTitle / Venue / Authors
2019 C conf
ICVS
Justin van Heek, Gideon Woo, Jack Park, Herbert H. Tsang
2019 conf
SSCI
Justin van Heek, Gideon Woo, Jack Park, Herbert H. Tsang
2018 conf
SSCI
Justin van Heek, Jack Park, Xavier Yu, Herbert H. Tsang
2010 C conf
COMMA
Simon Buckingham Shum, Maarten Sierhuis, Jack Park, Matthew Brown
2007 conf
TMRA
Jack Park
2006 ed.
TMRA
Lutz Maicher, Jack Park
2006 ed.
Semantic Desktop Workshop
Stefan Decker, Jack Park, Dennis Quan, Leo Sauermann
2006 ed.
SemDesk
Stefan Decker, Jack Park, Leo Sauermann, Sören Auer, Siegfried Handschuh
2006 conf
SemDesk
Jack Park
2006 conf
TMRA
Jack Park
2005 conf
Semantic Desktop Workshop
Adam Cheyer, Jack Park, Richard Giuli
2005 conf
TMRA
Jack Park, Adam Cheyer
2005 conf
TMRA
Jack Park
2002 conf
Extreme Markup Languages®
Jack Park
redb/extractors/macho_extractors/macho_similarity_hashes.py
← Index redb/extractors/macho_extractors/macho_similarity_hashes.py python
import inspect
from datetime import datetime, timezone
from typing import Any

from redb.extractors.enum import Tag
from redb.extractors.macho_extractor import MachOExtractor


class MachOSimilarityHashExtractor(MachOExtractor):
    """Extract Mach-O similarity hashes using machofile API.

    Similarity hashes are MD5 fingerprints of sorted, deduplicated binary components:
    - dylib_hash: MD5 of dynamic library names
    - import_hash: MD5 of imported function names
    - export_hash: MD5 of exported symbol names
    - entitlement_hash: MD5 of entitlement names and array values
    - symhash: MD5 of external undefined symbols

    For FAT binaries:
    - Inserts one row per architecture slice with per-slice hashes
    - Inserts one row for the FAT container with combined hashes

    For single-arch binaries:
    - Inserts one row with that architecture's hashes

    Note: parent_sha256 and architecture relationships are tracked in redb_basic_properties,
    not duplicated here. Use JOIN with redb_basic_properties when needed.
    """

    def __init__(
        self,
        filepath,
        log,
        exporters=None,
        index_prefix=None,
        elastic_index=None,
        known_benign=False,
        known_malicious=False,
        macho=None,
    ):
        super().__init__(
            filepath,
            log,
            exporters,
            index_prefix,
            elastic_index,
            known_benign,
            known_malicious,
            macho,
        )
        self.elastic_index = self.index_prefix + "-macho_hashes"
        self.log.debug(inspect.currentframe().f_code.co_name)

    def tag(self):
        return Tag.MACHO_HASHES.value

    def _extract_similarity_hashes(self, arch_name=None):
        """Extract similarity hashes for a specific architecture."""
        self.log.debug(inspect.currentframe().f_code.co_name)

        if not self.macho:
            return None

        try:
            similarity_hashes = self.macho.get_similarity_hashes(arch=arch_name)
            return similarity_hashes if similarity_hashes else None
        except Exception as e:
            self.log.error(f"Error extracting similarity hashes for arch {arch_name}: {e}")
            return None

    def extract(self):
        self.log.debug(inspect.currentframe().f_code.co_name)
        try:
            if not self.macho:
                return None

            architectures = self.macho.get_architectures()
            if not architectures:
                return None

            if len(architectures) > 1:
                # FAT binary - return combined hashes + per-arch hashes
                results = []

                # First add combined hashes for the FAT container
                all_hashes = self.macho.get_similarity_hashes()
                combined_hashes = all_hashes.get('combined', {}) if all_hashes else {}
                if combined_hashes:
                    combined_hashes['arch_identifier'] = 'fat'
                    results.append(combined_hashes)

                # Then add per-arch hashes
                for arch_name in architectures:
                    hashes = self._extract_similarity_hashes(arch_name)
                    if hashes:
                        hashes['arch_identifier'] = arch_name
                        results.append(hashes)
                return results
            else:
                # Single architecture - return single result
                return self._extract_similarity_hashes(architectures[0])
        except Exception as e:
            self.log.error(f"Error extracting similarity hashes: {e}")
            return None

    def prepare_export_data(self, exporter_type: str) -> Any:
        if exporter_type == "ElasticsearchExporter":
            return self.extract()
        elif exporter_type == "ClickHouseExporter":
            if not self.macho:
                return None

            try:
                architectures = self.macho.get_architectures()
                is_fat = len(architectures) > 1
            except Exception as e:
                self.log.error(f"Could not get architectures: {e}")
                return None

            data = []
            current_time = datetime.now(timezone.utc)

            # For FAT binaries, first insert a row for the container with combined hashes
            if is_fat:
                all_hashes = self.macho.get_similarity_hashes()  # Without arch returns all including 'combined'
                combined_hashes = all_hashes.get('combined', {}) if all_hashes else {}
                if combined_hashes:
                    data.append([
                        self.sha256,                                    # sha256 (FAT container)
                        combined_hashes.get('dylib_hash'),              # dylib_hash
                        combined_hashes.get('import_hash'),             # import_hash
                        combined_hashes.get('export_hash'),             # export_hash
                        combined_hashes.get('entitlement_hash'),        # entitlement_hash
                        combined_hashes.get('symhash'),                 # symhash
                        current_time,                                   # analysis_date
                    ])

            # Insert rows for each architecture slice
            for arch_name in architectures:
                # Get architecture-specific sha256
                try:
                    arch_general_info = self.macho.get_general_info(arch=arch_name)
                    arch_sha256 = arch_general_info.get('SHA256', self.sha256)
                except Exception as e:
                    self.log.warning(f"Could not get arch-specific sha256 for {arch_name}: {e}")
                    arch_sha256 = self.sha256

                # Get similarity hashes for this architecture
                similarity_hashes = self._extract_similarity_hashes(arch_name)
                if not similarity_hashes:
                    continue

                data.append([
                    arch_sha256,                                    # sha256 (arch-specific)
                    similarity_hashes.get('dylib_hash'),            # dylib_hash
                    similarity_hashes.get('import_hash'),           # import_hash
                    similarity_hashes.get('export_hash'),           # export_hash
                    similarity_hashes.get('entitlement_hash'),      # entitlement_hash
                    similarity_hashes.get('symhash'),               # symhash
                    current_time,                                   # analysis_date
                ])

            if not data:
                return None

            column_names = [
                'sha256',
                'macho_dylib_hash', 'macho_import_hash', 'macho_export_hash',
                'macho_entitlement_hash', 'macho_symhash',
                'analysis_date'
            ]

            column_type_names = [
                'FixedString(64)',
                'Nullable(FixedString(32))', 'Nullable(FixedString(32))', 'Nullable(FixedString(32))',
                'Nullable(FixedString(32))', 'Nullable(FixedString(32))',
                'DateTime64(3, \'UTC\')'
            ]

            return (data, column_names, column_type_names)

        return None

    def get_clickhouse_table(self) -> str:
        return "redb_hashes"