Hadi Mardani Kamali

81 papers A* 4A 11B 1C 2Misc 2Journal 42Unranked 19
YearRankTypeTitle / Venue / Authors
2026 J jnl
CoRR
M. Zafir Sadik Khan, Kimia Zamiri Azar, Hadi Mardani Kamali
2026 J jnl
CoRR
Nowfel Mashnoor, Mohammad Akyash, Hadi Mardani Kamali, Kimia Zamiri Azar
2026 J jnl
CoRR
Voktho Das, Kimia Zamiri Azar, Hadi Mardani Kamali
2025 C conf
ICCD
Nowfel Mashnoor, Mohammad Akyash, Hadi Mardani Kamali, Kimia Zamiri Azar
2025 J jnl
CoRR
Nowfel Mashnoor, Mohammad Akyash, Hadi Mardani Kamali, Kimia Zamiri Azar
2025 A conf
ICCAD
Mohammad Akyash, Kimia Zamiri Azar, Hadi Mardani Kamali
2025 J jnl
CoRR
Mohammad Akyash, Kimia Zamiri Azar, Hadi Mardani Kamali
2025 J jnl
IACR Cryptol. ePrint Arch.
Rui Guo, M. Sazadur Rahman, Jingbo Zhou, Hadi Mardani Kamali, Fahim Rahman, Farimah Farahmandi, Mark Tehranipoor
2025 Misc conf
VTS
Nowfel Mashnoor, Mohammad Akyash, Hadi Mardani Kamali, Kimia Zamiri Azar
2025 J jnl
CoRR
Nowfel Mashnoor, Mohammad Akyash, Hadi Mardani Kamali, Kimia Zamiri Azar
2025 conf
ASP-DAC
Md. Saad Ul Haque, Azim Uddin, Jingbo Zhou, Hadi Mardani Kamali, Farimah Farahmandi, Mark Tehranipoor
2025 J jnl
CoRR
Mohammad Akyash, Kimia Zamiri Azar, Hadi Mardani Kamali
2025 C conf
ICCD
M. Zafir Sadik Khan, Nowfel Mashnoor, Mohammad Akyash, Kimia Zamiri Azar, Hadi Mardani Kamali
2025 J jnl
CoRR
M. Zafir Sadik Khan, Nowfel Mashnoor, Mohammad Akyash, Kimia Zamiri Azar, Hadi Mardani Kamali
2025 J jnl
IEEE Access
Voktho Das, Kimia Zamiri Azar, Hadi Mardani Kamali
2025 conf
ASP-DAC
Mohammad Akyash, Hadi Mardani Kamali
2025 J jnl
CoRR
Mohammad Akyash, Kimia Zamiri Azar, Hadi Mardani Kamali
2025 conf
COINS
Nowfel Mashnoor, Mohammad Akyash, Hadi Mardani Kamali, Kimia Zamiri Azar
2025 J jnl
CoRR
Nowfel Mashnoor, Mohammad Akyash, Hadi Mardani Kamali, Kimia Zamiri Azar
2024 J jnl
IEEE Access
Md Sami Ul Islam Sami, Tao Zhang, Amit Mazumder Shuvo, Md. Saad Ul Haque, Paul E. Calzada, Kimia Zamiri Azar, Hadi Mardani Kamali, Fahim Rahman, Farimah Farahmandi, Mark Mohammad Tehranipoor
2024 conf
ACM Great Lakes Symposium on VLSI
Mohammad Akyash, Hadi Mardani Kamali
2024 J jnl
CoRR
Mohammad A. Makhzan, Hadi Mardani Kamali
2024 J jnl
IEEE Trans. Inf. Forensics Secur.
Sree Ranjani Rajendran, Nusrat Farzana Dipu, Shams Tarek, Hadi Mardani Kamali, Farimah Farahmandi, Mark Tehranipoor
2024 A* conf
DAC
Galib Ibne Haidar, Kimia Zamiri Azar, Hadi Mardani Kamali, Mark Tehranipoor, Farimah Farahmandi
2024 J jnl
IEEE Trans. Inf. Forensics Secur.
Kimia Zamiri Azar, Hadi Mardani Kamali, Farimah Farahmandi, Mark Tehranipoor
2024 conf
ISVLSI
Mohammad Akyash, Hadi Mardani Kamali
2024 J jnl
CoRR
Mohammad Akyash, Hadi Mardani Kamali
2024 J jnl
IEEE Trans. Very Large Scale Integr. Syst.
Tao Zhang, Md Latifur Rahman, Hadi Mardani Kamali, Kimia Zamiri Azar, Farimah Farahmandi
2023 A* conf
DAC
Zahin Ibnat, M. Sazadur Rahman, Mridha Md Mashahedur Rahman, Hadi Mardani Kamali, Mark Tehranipoor, Farimah Farahmandi
2023 J jnl
IEEE Des. Test
Md Sami Ul Islam Sami, Hadi Mardani Kamali, Farimah Farahmandi, Fahim Rahman, Mark Tehranipoor
2023 A conf
DATE
Rui Guo, M. Sazadur Rahman, Hadi Mardani Kamali, Fahim Rahman, Farimah Farahmandi, Mark Tehranipoor
2023 J jnl
IEEE Trans. Comput. Aided Des. Integr. Circuits Syst.
Md Rafid Muttaki, Roshanak Mohammadivojdan, Hadi Mardani Kamali, Mark Tehranipoor, Farimah Farahmandi
2023 A conf
DATE
Sree Ranjani Rajendran, Shams Tarek, Benjamin M. Hicks, Hadi Mardani Kamali, Farimah Farahmandi, Mark Tehranipoor
2023 conf
DFT
Zahin Ibnat, Hadi Mardani Kamali, Farimah Farahmandi
2023 conf
ACM Great Lakes Symposium on VLSI
Mohammad Sazadur Rahman, Kimia Zamiri Azar, Farimah Farahmandi, Hadi Mardani Kamali
2023 A conf
DATE
Md Rafid Muttaki, Shyvagata Saha, Hadi Mardani Kamali, Fahim Rahman, Mark Tehranipoor, Farimah Farahmandi
2023 J jnl
IEEE Access
M. Sazadur Rahman, Rui Guo, Hadi Mardani Kamali, Fahim Rahman, Farimah Farahmandi, Mark Tehranipoor
2023 conf
ASP-DAC
Shang Shi, Nitin Pundir, Hadi Mardani Kamali, Mark Tehranipoor, Farimah Farahmandi
2023 conf
MWSCAS
Md Rafid Muttaki, Zahin Ibnat, Shang Shi, Hadi Mardani Kamali, Farimah Farahmandi
2023 A conf
DATE
Hadi Mardani Kamali, Kimia Zamiri Azar, Farimah Farahmandi, Mark Tehranipoor
2022 J jnl
IACR Cryptol. ePrint Arch.
Hadi Mardani Kamali, Kimia Zamiri Azar, Farimah Farahmandi, Mark Tehranipoor
2022 A* conf
DAC
M. Sazadur Rahman, Rui Guo, Hadi Mardani Kamali, Fahim Rahman, Farimah Farahmandi, Mohamed Abdel-Moneum, Mark Tehranipoor
2022 J jnl
IACR Cryptol. ePrint Arch.
Hadi Mardani Kamali
2022 conf
HOST
Kimia Zamiri Azar, Hadi Mardani Kamali, Farimah Farahmandi, Mark Tehranipoor
2021 conf
ISQED
Hadi Mardani Kamali, Kimia Zamiri Azar, Houman Homayoun, Avesta Sasan
2021 J jnl
IEEE Trans. Very Large Scale Integr. Syst.
Kimia Zamiri Azar, Hadi Mardani Kamali, Shervin Roshanisefat, Houman Homayoun, Christos P. Sotiriou, Avesta Sasan
2021 J jnl
Frontiers Big Data
Zhiqian Chen, Lei Zhang, Gaurav Kolhe, Hadi Mardani Kamali, Setareh Rafatirad, Sai Manoj Pudukotai Dinakarrao, Houman Homayoun, Chang-Tien Lu, Liang Zhao
2021 J jnl
IEEE Access
Kimia Zamiri Azar, Hadi Mardani Kamali, Houman Homayoun, Avesta Sasan
2021 conf
ACM Great Lakes Symposium on VLSI
Shervin Roshanisefat, Hadi Mardani Kamali, Houman Homayoun, Avesta Sasan
2020 Misc conf
VTS
Shervin Roshanisefat, Hadi Mardani Kamali, Kimia Zamiri Azar, Sai Manoj Pudukotai Dinakarrao, Naghmeh Karimi, Houman Homayoun, Avesta Sasan
2020 J jnl
CoRR
Shervin Roshanisefat, Hadi Mardani Kamali, Kimia Zamiri Azar, Sai Manoj Pudukotai Dinakarrao, Naghmeh Karimi, Houman Homayoun, Avesta Sasan
2020 J jnl
CoRR
Hadi Mardani Kamali, Kimia Zamiri Azar, Shervin Roshanisefat, Ashkan Vakil, Avesta Sasan
2020 A conf
ICCAD
Hadi Mardani Kamali, Kimia Zamiri Azar, Houman Homayoun, Avesta Sasan
2020 J jnl
CoRR
Hadi Mardani Kamali, Kimia Zamiri Azar, Houman Homayoun, Avesta Sasan
2020 A conf
ICCAD
Kimia Zamiri Azar, Hadi Mardani Kamali, Houman Homayoun, Avesta Sasan
2020 J jnl
CoRR
Kimia Zamiri Azar, Hadi Mardani Kamali, Houman Homayoun, Avesta Sasan
2020 conf
ACM Great Lakes Symposium on VLSI
Hadi Mardani Kamali, Kimia Zamiri Azar, Houman Homayoun, Avesta Sasan
2020 J jnl
CoRR
Hadi Mardani Kamali, Kimia Zamiri Azar, Houman Homayoun, Avesta Sasan
2020 J jnl
IEEE Trans. Very Large Scale Integr. Syst.
Shervin Roshanisefat, Hadi Mardani Kamali, Houman Homayoun, Avesta Sasan
2020 J jnl
CoRR
Shervin Roshanisefat, Hadi Mardani Kamali, Houman Homayoun, Avesta Sasan
2020 conf
ISVLSI
Hadi Mardani Kamali, Kimia Zamiri Azar, Houman Homayoun, Avesta Sasan
2020 J jnl
CoRR
Hadi Mardani Kamali, Kimia Zamiri Azar, Houman Homayoun, Avesta Sasan
2019 A conf
RAID
Kimia Zamiri Azar, Farnoud Farahmand, Hadi Mardani Kamali, Shervin Roshanisefat, Houman Homayoun, William Diehl, Kris Gaj, Avesta Sasan
2019 J jnl
CoRR
Kimia Zamiri Azar, Farnoud Farahmand, Hadi Mardani Kamali, Shervin Roshanisefat, Houman Homayoun, William Diehl, Kris Gaj, Avesta Sasan
2019 A* conf
DAC
Hadi Mardani Kamali, Kimia Zamiri Azar, Houman Homayoun, Avesta Sasan
2019 A conf
ISLPED
Hossein Farrokhbakht, Hadi Mardani Kamali, Natalie D. Enright Jerger
2019 J jnl
IACR Trans. Cryptogr. Hardw. Embed. Syst.
Kimia Zamiri Azar, Hadi Mardani Kamali, Houman Homayoun, Avesta Sasan
2019 A conf
ICCAD
Gaurav Kolhe, Hadi Mardani Kamali, Miklesh Naicker, Tyler David Sheaves, Hamid Mahmoodi, Sai Manoj P. D., Houman Homayoun, Setareh Rafatirad, Avesta Sasan
2019 conf
ACM Great Lakes Symposium on VLSI
Kimia Zamiri Azar, Hadi Mardani Kamali, Houman Homayoun, Avesta Sasan
2019 J jnl
CoRR
Kimia Zamiri Azar, Hadi Mardani Kamali, Houman Homayoun, Avesta Sasan
2018 J jnl
IEEE Trans. Computers
Hadi Mardani Kamali, Kimia Zamiri Azar, Shaahin Hessabi
2018 conf
ISVLSI
Hadi Mardani Kamali, Kimia Zamiri Azar, Kris Gaj, Houman Homayoun, Avesta Sasan
2018 J jnl
CoRR
Hadi Mardani Kamali, Kimia Zamiri Azar, Kris Gaj, Houman Homayoun, Avesta Sasan
2018 conf
ACM Great Lakes Symposium on VLSI
Hadi Mardani Kamali, Avesta Sasan
2018 A conf
ISLPED
Hossein Farrokhbakht, Hadi Mardani Kamali, Natalie D. Enright Jerger, Shaahin Hessabi
2018 conf
ACM Great Lakes Symposium on VLSI
Shervin Roshanisefat, Hadi Mardani Kamali, Avesta Sasan
2018 J jnl
CoRR
Shervin Roshanisefat, Hadi Mardani Kamali, Avesta Sasan
2018 J jnl
CoRR
Hadi Mardani Kamali
2017 conf
NOCS
Hossein Farrokhbakht, Hadi Mardani Kamali, Shaahin Hessabi
2016 J jnl
J. Circuits Syst. Comput.
Hadi Mardani Kamali, Shaahin Hessabi
2016 B conf
FPL
Hadi Mardani Kamali, Shaahin Hessabi
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
    """)