Hanjiang Luo

55 papers A 1B 10C 1Journal 27Unranked 16
YearRankTypeTitle / Venue / Authors
2026 J jnl
Eng. Appl. Artif. Intell.
Huixiang Zhi, Muwei Jian, Hongyu Chen, Wenjing Xu, Changqun Nie, Hanjiang Luo, Xiaoguang Li
2026 J jnl
IEEE Internet Things J.
Yang Zhao, Hanjiang Luo, Hang Tao, Jingjing Wang, Jiehan Zhou, Kaishun Wu
2025 conf
ICHMS
Huixiang Zhi, Muwei Jian, Hongyu Chen, Chunxiao Ren, Wenjing Xu, Changqun Nie, Hanjiang Luo, Xiaoguang Li
2025 conf
ICIT
Hang Tao, Hanjiang Luo, Yang Zhao, Yinyan Wang, Wei Shi, Jingjing Wang
2025 J jnl
IEEE Trans. Netw. Serv. Manag.
Jiangang Liu, Hanjiang Luo, Hang Tao, Jiahong Liu, Jiehan Zhou
2025 conf
ICA3PP (3)
Peijun Dong, Hang Tao, Yang Zhao, Wei Shi, Hanjiang Luo
2025 B conf
ICCCN
Yang Zhao, Hanjiang Luo, Hang Tao, Jinyin Li, Chao Liu, Jiehan Zhou
2025 J jnl
Comput. Mater. Continua
Jianjun Liu, Yang Li, Xiaoting Sun, Xiaohui Wang, Hanjiang Luo
2025 B conf
ICPADS
Jiahong Liu, Hang Tao, Yang Zhao, Chao Liu, Hanjiang Luo
2025 J jnl
Ad Hoc Networks
Hang Tao, Mingyue Shao, Yinyan Wang, Xinxiang Wang, Hanjiang Luo
2024 conf
MSN
Pengyan Dong, Jiahong Liu, Hang Tao, Rukhsana Ruby, Muwei Jian, Hanjiang Luo
2024 J jnl
IEEE Internet Things J.
Hanjiang Luo, Saisai Ma, Hang Tao, Rukhsana Ruby, Jiehan Zhou, Kaishun Wu
2024 J jnl
IEEE Trans. Veh. Technol.
Jinglong Wang, Hanjiang Luo, Rukhsana Ruby, Jiangang Liu, Shengli Wang, Kaishun Wu
2024 conf
ACM TUR-C
Xiaoyu Luo, Daping Liu, Fan Dang, Hanjiang Luo
2024 conf
CCECE
Jiehan Zhou, Yang Zhao, Jiahong Liu, Peijun Dong, Xiaoyu Luo, Hang Tao, Shi Chang, Hanjiang Luo
2023 B conf
ICPADS
Mengzhen Li, Hanjiang Luo, Xiang Li, Rukhsana Ruby, Hang Tao, Kaishun Wu
2023 J jnl
Sensors
Fanfeng Bu, Hanjiang Luo, Saisai Ma, Xiang Li, Rukhsana Ruby, Guangjie Han
2023 conf
ACM TUR-C
Kai Guo, Zhizun Qin, Xiaodong Chen, Haoran Wang, Hanjiang Luo
2023 conf
CBD
Mingyue Shao, Hanjiang Luo, Rukhsana Ruby, Pengyan Dong, Hang Tao
2023 conf
CBD
Kai Guo, Hanjiang Luo, Hang Tao, Rukhsana Ruby, Zhizun Qin, Kui Liu
2023 conf
BIGCOM
Saisai Ma, Mengzhen Li, Tinglong Deng, Jinglong Wang, Rukhsana Ruby, Hanjiang Luo
2023 conf
ACM TUR-C
Mingyue Shao, Xiang Li, Pengyan Dong, Jiangang Liu, Hanjiang Luo
2023 B conf
ICPADS
Jiangang Liu, Hanjiang Luo, Rukhsana Ruby, Haoran Wang, Hang Tao, Kaishun Wu
2023 J jnl
IEEE Trans. Veh. Technol.
Hanjiang Luo, Xu Wang, Fanfeng Bu, Yuting Yang, Rukhsana Ruby, Kaishun Wu
2022 J jnl
Int. J. Distributed Sens. Networks
Hanjiang Luo, Xu Wang, Ziyang Xu, Chao Liu, Jeng-Shyang Pan
2022 conf
IIP
Yuting Yang, Kin-Man Lam, Xin Sun, Junyu Dong, Muwei Jian, Hanjiang Luo
2022 B conf
LCN
Jinglong Wang, Hanjiang Luo, Rukhsana Ruby, Jiangang Liu, Kai Guo, Kaishun Wu
2021 J jnl
IEEE/ACM Trans. Netw.
Rukhsana Ruby, Shuxin Zhong, Basem M. ElHalawany, Hanjiang Luo, Kaishun Wu
2021 B conf
ICPADS
Xu Wang, Hanjiang Luo, Yuting Yang, Rukhsana Ruby, Kaishun Wu
2021 J jnl
Signal Process. Image Commun.
Muwei Jian, Xiangyu Liu, Hanjiang Luo, Xiangwei Lu, Hui Yu, Junyu Dong
2020 B conf
WCNC
Keyong Hu, Xianglin Song, Zhongwei Sun, Hanjiang Luo, Zhongwen Guo
2020 J jnl
IEEE Access
Zhi-Gang Du, Jeng-Shyang Pan, Shu-Chuan Chu, Hanjiang Luo, Pei Hu
2020 conf
ACM TUR-C
Hanjiang Luo, Chao Liu, Yu Luo, Xu Wang, Ziyang Xu, Yongquan Liang
2019 J jnl
IEEE Access
Ziwei Zhang, Wei Shi, Qiuna Niu, Ying Guo, Jingjing Wang, Hanjiang Luo
2019 conf
ACM TUR-C
Hanjiang Luo, Chao Liu, Yongquan Liang
2019 J jnl
Pattern Recognit. Lett.
Muwei Jian, Quan Zhou, Chaoran Cui, Xiushan Nie, Hanjiang Luo, Jianli Zhao, Yilong Yin
2019 J jnl
Sensors
Keyong Hu, Xianglin Song, Zhongwei Sun, Hanjiang Luo, Zhongwen Guo
2019 J jnl
IEEE Access
Hanjiang Luo, Xiumei Xie, Guangjie Han, Rukhsana Ruby, Feng Hong, Yongquan Liang
2019 J jnl
IEEE Trans. Veh. Technol.
Jiaheng Li, Jingjing Wang, Xinjie Wang, Gang Qiao, Hanjiang Luo, T. Aaron Gulliver
2019 J jnl
IEEE Access
Chao Liu, Xi Wang, Hanjiang Luo, Yuan Liu, Zhongwen Guo
2018 J jnl
Sensors
Keyong Hu, Zhongwei Sun, Hanjiang Luo, Wei Zhou, Zhongwen Guo
2018 J jnl
J. Vis. Commun. Image Represent.
Muwei Jian, Runxia Zhao, Xin Sun, Hanjiang Luo, Wenyin Zhang, Huaxiang Zhang, Junyu Dong, Yilong Yin, Kin-Man Lam
2018 J jnl
IEEE Commun. Surv. Tutorials
Hanjiang Luo, Kaishun Wu, Rukhsana Ruby, Yongquan Liang, Zhongwen Guo, Lionel M. Ni
2017 B conf
ICPADS
Hanjiang Luo, Rukhsana Ruby, Xiumei Xie, Yongquan Liang
2017 J jnl
ACM Comput. Surv.
Hanjiang Luo, Kaishun Wu, Rukhsana Ruby, Feng Hong, Zhongwen Guo, Lionel M. Ni
2016 J jnl
IEEE Trans. Veh. Technol.
Hanjiang Luo, Kaishun Wu, Yue-Jiao Gong, Lionel M. Ni
2016 conf
MSN
Hanjiang Luo, Kaishun Wu, Feng Hong
2014 B conf
ICPADS
Hanjiang Luo, Kaishun Wu, Jiang Xiao, Zhongwen Guo
2012 J jnl
IEEE Trans. Parallel Distributed Syst.
Hanjiang Luo, Kaishun Wu, Zhongwen Guo, Lin Gu, Lionel M. Ni
2011 A conf
ICDCS
Hanjiang Luo, Kaishun Wu, Zhongwen Guo, Lin Gu, Zhong Yang, Lionel M. Ni
2010 J jnl
J. Networks
Hanjiang Luo, Zhongwen Guo, Wei Dong, Feng Hong, Yiyang Zhao
2009 J jnl
Sensors
Hanjiang Luo, Zhongwen Guo, Kaishun Wu, Feng Hong, Yuan Feng
2009 conf
PerCom Workshops
Mingxing Jiang, Zhongwen Guo, Feng Hong, Yutao Ma, Hanjiang Luo
2008 C conf
PDCAT
Zhongwen Guo, Hanjiang Luo, Feng Hong, Peng Zhou
2008 B conf
ICPADS
Hanjiang Luo, Yiyang Zhao, Zhongwen Guo, Siyuan Liu, Pengpeng Chen, Lionel M. Ni
docs/new-code-binja-schema.md
← Index docs/new-code-binja-schema.md markdown
#################################
###### CODE RELATED TABLES ######
#################################

### CREATE FULL REDB DATABASE
def create_decompiled_binja_with_all_tables(client):
    create_phase1_decompiled_binja_tables(client)

#### CREATE FULL TABLES
def create_phase1_decompiled_binja_tables(client):
    create_function_analysis_error_dbtable(client)
    create_binja_decompiled_dbtable(client)
    create_binja_disassembly_dbtable(client)
    create_binja_cfg_dbtable(client)

#################################
### DECOMPILED BINJA INDIVIDUAL TABLES
#################################
def create_function_analysis_error_dbtable(client):
    client.command("""
    CREATE TABLE IF NOT EXISTS function_analysis_errors_binja (
        sha256 FixedString(64),
        function_name Nullable(String) CODEC(ZSTD(3)),
        function_address UInt64,
        error_location LowCardinality(String) CODEC(ZSTD(3)),
        error_message Nullable(String) CODEC(ZSTD(3)),
        error_details Nullable(String) CODEC(ZSTD(3)),
        error_type Nullable(String) CODEC(ZSTD(3)),
        error_hash FixedString(32), -- MD5, To help identify duplicate errors
        status Enum8('new' = 1, 'investigating' = 2, 'fixed' = 3, 'wontfix' = 4) DEFAULT 'new',
        analysis_date DateTime64(3, 'UTC'),
        PRIMARY KEY (sha256, function_address, error_location)
    )
    ENGINE = MergeTree()
    ORDER BY (sha256, function_address, error_location);
    """)

def create_binja_decompiled_dbtable(client):
    # Create decompiled_functions_content table
    client.command("""
    CREATE TABLE IF NOT EXISTS code_binja_decompiled_functions_content (
        decompiled_function_hash FixedString(64),
        decompiled_function String CODEC(ZSTD(3)),
        decompiled_function_lower String MATERIALIZED lower(decompiled_function) CODEC(ZSTD(3)),
        function_type Enum8('USER'=1, 'LIBRARY'=2, 'THUNK'=3, 'EXTERNAL'=4, 'UNKNOWN'=5) DEFAULT 'UNKNOWN',  -- NOT NULL TODO: remove THUNK and EXTERNAL
        flattened_score Nullable(Float64),
        mba_score Nullable(Float64),
        analysis_date DateTime64(3, 'UTC'),

        INDEX idx_function_content_token lower(decompiled_function) TYPE tokenbf_v1(32768, 3, 0) GRANULARITY 1,
        --- INDEX idx_text_decompiled_function decompiled_function TYPE text(tokenizer = 'splitByNonAlpha', preprocessor = lower(decompiled_function)) GRANULARITY 64,
        INDEX idx_func_hash decompiled_function_hash TYPE bloom_filter GRANULARITY 1,
    ) ENGINE = ReplacingMergeTree(analysis_date)
    PRIMARY KEY decompiled_function_hash -- TODO: remove is redundant
    ORDER BY decompiled_function_hash;
    """)

    # Create decompiled_functions_references table
    client.command("""
    CREATE TABLE IF NOT EXISTS code_binja_decompiled_functions_references (
        sha256 FixedString(64),
        sha1 FixedString(40),
        md5 FixedString(32),
        decompiled_function_hash FixedString(64),
        disassembled_function_hash Nullable(FixedString(64)),
        decompiled_function_name LowCardinality(String),
        decompiled_function_prototype LowCardinality(String),
        decompiled_function_address UInt64,
        functions_caller Array(String),
        functions_call Array(String),

        INDEX idx_sha256 sha256 TYPE bloom_filter GRANULARITY 1,
        INDEX idx_func_hash decompiled_function_hash TYPE bloom_filter GRANULARITY 1,
        INDEX idx_func_name decompiled_function_name TYPE tokenbf_v1(32768, 3, 0) GRANULARITY 1,
        INDEX idx_func_addr decompiled_function_address TYPE minmax GRANULARITY 1,
        analysis_date DateTime64(3, 'UTC')
    ) ENGINE = ReplacingMergeTree(analysis_date)
    ORDER BY (sha256, decompiled_function_hash);
    """)

def create_binja_disassembly_dbtable(client):
    # Create disassembly_functions_content table
    client.command("""
    CREATE TABLE IF NOT EXISTS code_binja_disassembled_functions_content (
        -- Content hashes for each normalization level
        disassembled_function_hash FixedString(64),             -- This is the hash of the disassembled function without addresses
        
        -- The actual function content at different normalization levels
        disassembled_function String CODEC(ZSTD(3)),
        disassembled_function_no_addresses String CODEC(ZSTD(3)),

        -- Function 
        function_type Enum8('USER'=1, 'LIBRARY'=2, 'THUNK'=3, 'EXTERNAL'=4, 'UNKNOWN'=5) DEFAULT 'UNKNOWN',  -- NOT NULL TODO: remove THUNK and EXTERNAL

        -- Analysis fields
        instructions_count UInt32,
        instructions_types Array(LowCardinality(String)), -- Types of instructions used
        control_flow_count UInt32,                      -- Number of control flow instructions
        memory_access_pattern Array(LowCardinality(String)),  -- Patterns of memory access
        register_usage Array(LowCardinality(String)),         -- Types of registers used
        data_references_count UInt32,                   -- Number of data references
        opcode_frequency_vector Array(Float32),-- Not computed yet
        api_calls_vector Array(Float32),-- Not computed yet
        minhash_signature Array(UInt64),-- Not computed yet
        max_block_size Nullable(UInt32),
        num_calls Nullable(UInt32),
        stack_size Nullable(Int32),
        instruction_type_ratios Array(Float32), -- Not computed yet
        instruction_embedding Array(Float32), -- Not computed yet
        analysis_date DateTime64(3, 'UTC'),

        -- Indexes for each normalization level
        INDEX idx_disasm_ngram disassembled_function TYPE ngrambf_v1(3, 32768, 3, 0) GRANULARITY 1,
        INDEX idx_function_type function_type TYPE set(10) GRANULARITY 1,
        INDEX idx_instr_count instructions_count TYPE minmax GRANULARITY 4

    ) ENGINE = ReplacingMergeTree(analysis_date)
    ORDER BY (disassembled_function_hash);
    """)

    client.command("""
    CREATE TABLE code_binja_disassembled_functions_references
    (
        -- Original fields
        sha256 FixedString(64),
        sha1 FixedString(40),
        md5 FixedString(32),
        disassembled_function_hash FixedString(64),
        decompiled_function_hash Nullable(FixedString(64)),

        disassembled_function_name LowCardinality(String),
        disassembled_function_address UInt64,
        analysis_date DateTime64(3, 'UTC'),
        
        -- New hash fields moved from content table
        ssdeep_disassembly Nullable(String),
        tlsh_disassembly Nullable(FixedString(72)),
        ssdeep_llil Nullable(String),
        tlsh_llil Nullable(FixedString(72)),
        
        -- OPTIMIZED INDEXES for all search patterns
        INDEX idx_func_addr disassembled_function_address TYPE minmax GRANULARITY 1,
        INDEX idx_sha256 sha256 TYPE bloom_filter GRANULARITY 1,
        INDEX idx_tlsh_disasm tlsh_disassembly TYPE bloom_filter GRANULARITY 1,
        INDEX idx_ssdeep_disasm ssdeep_disassembly TYPE bloom_filter GRANULARITY 1,
        INDEX idx_tlsh_llil tlsh_llil TYPE bloom_filter GRANULARITY 1,
        INDEX idx_ssdeep_norm ssdeep_llil TYPE bloom_filter GRANULARITY 1,
        INDEX idx_func_name disassembled_function_name TYPE tokenbf_v1(32768, 3, 0) GRANULARITY 2
    )
    ENGINE = ReplacingMergeTree(analysis_date)
    -- OPTIMIZED ORDER BY: Primary query pattern first, then hash fields for locality
    ORDER BY (sha256, disassembled_function_hash)
    SETTINGS index_granularity = 8192
    """)

def create_binja_llil_dbtable(client):
    # Create llil_functions_content table
    client.command("""
    CREATE TABLE IF NOT EXISTS code_binja_llil_functions_content (
        -- Content hashes for each normalization level
        llil_function_hash FixedString(64),             -- sha256_llil This is the hash of the disassembled function without addresses

        -- Function 
        function_type Enum8('USER'=1, 'LIBRARY'=2, 'THUNK'=3, 'EXTERNAL'=4, 'UNKNOWN'=5) DEFAULT 'UNKNOWN',  -- NOT NULL TODO: remove THUNK and EXTERNAL

        -- Analysis fields
        instructions_types_llil Array(LowCardinality(String)), -- Types of instructions used
        control_flow_count_llil UInt32,                      -- Number of control flow instructions
        memory_access_pattern_llil Array(LowCardinality(String)),  -- Patterns of memory access
        register_usage_llil Map(LowCardinality(String), Tuple(UInt32, UInt32)),         -- Types of registers used
        total_reg_reads UInt32,
        total_reg_written UInt32,
        data_references_count UInt32,                   -- Number of data references
        max_block_size Nullable(UInt32),
        num_calls Nullable(UInt32),
        stack_size Nullable(Int32),
        body_llil_vector Array(Tuple(UInt32, Array(UInt16))),
        analysis_date DateTime64(3, 'UTC'),

        -- Indexes for each normalization level
        INDEX function_type_idx function_type TYPE set(10) GRANULARITY 1,
        INDEX idx_reg_usage mapKeys(register_usage_llil) TYPE bloom_filter(0.01) GRANULARITY 1

    ) ENGINE = ReplacingMergeTree(analysis_date)
    ORDER BY llil_function_hash;
    """)

    client.command("""
    CREATE TABLE code_binja_llil_functions_references
    (
        -- Original fields
        sha256 FixedString(64),
        sha1 FixedString(40),
        md5 FixedString(32),
        llil_function_hash Nullable(FixedString(64)),
        disassembled_function_hash FixedString(64),
        
        -- New hash fields moved from content table
        ssdeep_disassembly Nullable(String),
        tlsh_disassembly Nullable(FixedString(72)),
        ssdeep_llil Nullable(String),
        tlsh_llil Nullable(FixedString(72)),

        analysis_date DateTime64(3, 'UTC'),
        
        -- OPTIMIZED INDEXES for all search patterns
        INDEX idx_sha256 sha256 TYPE bloom_filter GRANULARITY 1,
        INDEX idx_tlsh_disasm tlsh_disassembly TYPE bloom_filter GRANULARITY 1,
        INDEX idx_ssdeep_disasm ssdeep_disassembly TYPE bloom_filter GRANULARITY 1,
        INDEX idx_tlsh_llil tlsh_llil TYPE bloom_filter GRANULARITY 1,
        INDEX idx_ssdeep_llil ssdeep_llil TYPE bloom_filter GRANULARITY 1,
    )
    ENGINE = ReplacingMergeTree(analysis_date)
    -- OPTIMIZED ORDER BY: Primary query pattern first, then hash fields for locality
    ORDER BY (sha256, function_address)
    SETTINGS index_granularity = 8192
    """)

def create_binja_cfg_dbtable(client):
    client.command("""
    CREATE TABLE IF NOT EXISTS code_binja_cfg_blocks_content (
        block_instructions_hash FixedString(64), -- PK
        block_size UInt32,
        instructions_count UInt16,
        branch_type Enum8('DIRECT'=1, 'CONDITIONAL'=2, 'CALL'=3, 'RETURN'=4, 'FALLTHROUGH'=5, 'INDIRECT'=6, 'UNKNOWN'=7),
        block_type Enum8('CODE'=1, 'DATA'=2, 'THUNK'=3),
        flags Array(String),
        analysis_date DateTime64(3, 'UTC'),

        -- Indexes for filtering/analysis
        INDEX idx_branch_type branch_type TYPE set(10) GRANULARITY 4,
        INDEX idx_block_type block_type TYPE set(5) GRANULARITY 4,
        INDEX idx_instr_count instructions_count TYPE minmax GRANULARITY 4,
        INDEX idx_block_size block_size TYPE minmax GRANULARITY 4
    )
    ENGINE = ReplacingMergeTree(analysis_date)
    ORDER BY (block_instructions_hash);
    """)
  
    client.command("""
    CREATE TABLE IF NOT EXISTS code_binja_cfg_blocks_references (
        sha256 FixedString(64),
        sha1 FixedString(40),
        md5 FixedString(32),
        block_instructions_hash FixedString(64),
        disassembled_function_hash FixedString(64),
        
        -- Address information (for CFG traversal and debugging)
        function_address UInt64,
        block_start_address UInt64,
        block_end_address UInt64,
        
        -- CFG structure (addresses refer to block_start_address)
        predecessor_blocks Array(UInt64),
        successor_blocks Array(UInt64),
        depth UInt16,
        position UInt32,
        
        -- Dominator analysis
        dominators Array(UInt16),
        post_dominators Array(UInt16),
        
        analysis_date DateTime64(3, 'UTC'),
        
        -- OPTIMIZED INDEXES based on query patterns
        INDEX idx_sha256 sha256 TYPE bloom_filter GRANULARITY 1,
        INDEX idx_block_hash block_instructions_hash TYPE bloom_filter GRANULARITY 1,
        INDEX idx_func_hash disassembled_function_hash TYPE bloom_filter GRANULARITY 1,
        INDEX idx_func_addr function_address TYPE minmax GRANULARITY 1,
        INDEX idx_block_start block_start_address TYPE minmax GRANULARITY 1,
        INDEX idx_depth depth TYPE minmax GRANULARITY 4,
        INDEX idx_position position TYPE minmax GRANULARITY 4
    )
    ENGINE = ReplacingMergeTree(analysis_date)
    ORDER BY (sha256, function_address, block_start_address)
    SETTINGS index_granularity = 8192;
    """)

def create_binja_strings_dbtable(client):
    # 1. Raw table - INSERT target
    client.command("""
    CREATE TABLE IF NOT EXISTS code_binja_strings_raw (
        sha256 FixedString(64),
        string String CODEC(ZSTD(3)),              -- Decoded UTF-8 string (for searching)
        string_raw String CODEC(ZSTD(3)),          -- Raw bytes from binary (ground truth)
        string_encoding LowCardinality(String),    -- Detected/guessed encoding by binja
        string_offset UInt64,
        string_length UInt32,                      -- Length of decoded string
        string_raw_length UInt32,                  -- Length of raw bytes
        string_entropy Float32
    )
    ENGINE = Null;
    """)
   
    # 3a. Target table for reverse lookup (with indexes)
    client.command("""
    CREATE TABLE IF NOT EXISTS code_binja_strings_by_binary (
        sha256 FixedString(64),
        string_offset UInt64,
        string String CODEC(ZSTD(3)),
        string_raw String CODEC(ZSTD(3)),
        string_encoding LowCardinality(String),
        string_length UInt32,
        string_raw_length UInt32,
        string_entropy Float32,
        
        INDEX idx_sha256 sha256 TYPE bloom_filter GRANULARITY 1,
        INDEX idx_string_ngram lower(string) TYPE ngrambf_v1(3, 32768, 3, 0) GRANULARITY 1,
        --- INDEX idx_string string TYPE tokenbf_v1(32768, 3, 0) GRANULARITY 1,
        --- INDEX idx_string_raw string_raw TYPE tokenbf_v1(32768, 3, 0) GRANULARITY 1,
        INDEX idx_entropy string_entropy TYPE minmax GRANULARITY 4
    )
    ENGINE = ReplacingMergeTree()
    ORDER BY (sha256, string_offset);
    """)
    
    # 3b. Materialized view -> target table
    client.command("""
    CREATE MATERIALIZED VIEW IF NOT EXISTS code_binja_strings_by_binary_mv
    TO code_binja_strings_by_binary
    AS SELECT
        sha256,
        string_offset,
        string,
        string_raw,
        string_encoding,
        string_length,
        string_raw_length,
        string_entropy
    FROM code_binja_strings_raw;
    """)

    client.command("""
    -- 1. Create the new public-only popularity table
    CREATE TABLE mv_string_popularity_public_table (
        string_xxhash64 UInt64,
        sample_count AggregateFunction(uniq, FixedString(64))
    ) ENGINE = AggregatingMergeTree()
    ORDER BY string_xxhash64;
    """)

    client.command("""
    -- 2. Create the MV for new data going forward
    CREATE MATERIALIZED VIEW mv_string_popularity_public
    TO mv_string_popularity_public_table AS
    SELECT
        xxHash64(s.string) AS string_xxhash64,
        uniqState(s.sha256) AS sample_count
    FROM code_binja_strings_raw s
    INNER JOIN catalog_visibility_lookup_v2 v 
        ON s.sha256 = v.sha256 AND v.visibility_scope = 'public'
    GROUP BY string_xxhash64;
    """)

    # client.command("""
    # INSERT INTO mv_string_popularity_public_table
    # SELECT
    #     xxHash64(s.string) AS string_xxhash64,
    #     uniqState(s.sha256) AS sample_count
    # FROM code_binja_strings_by_binary s
    # INNER JOIN catalog_visibility_lookup_v2 v 
    #     ON s.sha256 = v.sha256 AND v.visibility_scope = 'public'
    # GROUP BY string_xxhash64;
    # """)

def create_binja_function_similarity_metrics_dbtable(client):
    client.command("""
    CREATE TABLE IF NOT EXISTS code_binja_function_similarity_metrics (
        disassembled_function_hash FixedString(64),
        
        -- Complexity metrics
        cyclomatic_complexity Nullable(UInt16),
        
        -- Fuzzy hashes (duplicated from reference tables for performance)
        ssdeep_disassembly Nullable(String),
        tlsh_disassembly Nullable(FixedString(72)),
        ssdeep_llil Nullable(String),
        tlsh_llil Nullable(FixedString(72)),
        
        -- Similarity hashes
        minhash Array(UInt8),
        
        analysis_date DateTime64(3, 'UTC'),
        
        -- Indexes
        INDEX idx_complexity cyclomatic_complexity TYPE minmax GRANULARITY 4,
        INDEX idx_ssdeep_disasm ssdeep_disassembly TYPE bloom_filter GRANULARITY 1,
        INDEX idx_tlsh_disasm tlsh_disassembly TYPE bloom_filter GRANULARITY 1,
        INDEX idx_ssdeep_llil ssdeep_llil TYPE bloom_filter GRANULARITY 1,
        INDEX idx_tlsh_llil tlsh_llil TYPE bloom_filter GRANULARITY 1
    )
    ENGINE = ReplacingMergeTree(analysis_date)
    ORDER BY disassembled_function_hash;
    """)

def create_golang_metadata_dbtable(client):
    client.command("""
    CREATE TABLE IF NOT EXISTS redb_golang_metadata (
        sha256 FixedString(64),
        goresym JSON,
        analysis_date DateTime64(3, 'UTC')
    ) ENGINE = ReplacingMergeTree(analysis_date)
    ORDER BY sha256
    """)