Enum Commands 

pub enum Commands {
    Parse {
        file: String,
        page: Option<u64>,
        verbose: bool,
        no_empty: bool,
        json: bool,
        page_size: Option<u32>,
    },
    Pages {
        file: String,
        page: Option<u64>,
        verbose: bool,
        show_empty: bool,
        list: bool,
        filter_type: Option<String>,
        json: bool,
        page_size: Option<u32>,
    },
    Dump {
        file: String,
        page: Option<u64>,
        offset: Option<u64>,
        length: Option<usize>,
        raw: bool,
        page_size: Option<u32>,
    },
    Corrupt {
        file: String,
        page: Option<u64>,
        bytes: usize,
        header: bool,
        records: bool,
        offset: Option<u64>,
        verify: bool,
        json: bool,
        page_size: Option<u32>,
    },
    Find {
        datadir: String,
        page: u64,
        checksum: Option<u32>,
        space_id: Option<u32>,
        first: bool,
        json: bool,
        page_size: Option<u32>,
    },
    Tsid {
        datadir: String,
        list: bool,
        tablespace_id: Option<u32>,
        json: bool,
        page_size: Option<u32>,
    },
    Sdi {
        file: String,
        pretty: bool,
        page_size: Option<u32>,
    },
    Log {
        file: String,
        blocks: Option<u64>,
        no_empty: bool,
        verbose: bool,
        json: bool,
    },
    Info {
        ibdata: bool,
        lsn_check: bool,
        datadir: Option<String>,
        database: Option<String>,
        table: Option<String>,
        host: Option<String>,
        port: Option<u16>,
        user: Option<String>,
        password: Option<String>,
        defaults_file: Option<String>,
        json: bool,
        page_size: Option<u32>,
    },
    Recover {
        file: String,
        page: Option<u64>,
        verbose: bool,
        json: bool,
        force: bool,
        page_size: Option<u32>,
    },
    Checksum {
        file: String,
        verbose: bool,
        json: bool,
        page_size: Option<u32>,
    },
}

Available subcommands for the inno CLI.

Variants

Parse

Parse .ibd file and display page summary

Reads the 38-byte FIL header of every page in a tablespace, decodes the page type, checksum, LSN, prev/next pointers, and space ID, then prints a per-page breakdown followed by a page-type frequency summary table. Page 0 additionally shows the FSP header (space ID, size, flags). Use --no-empty to skip zero-checksum allocated pages, or -p to inspect a single page in detail. With --verbose, checksum validation and LSN consistency results are included for each page.

Fields

file: String

Path to InnoDB data file (.ibd)

page: Option<u64>

Display a specific page number

verbose: bool

Display additional information

no_empty: bool

Skip empty/allocated pages

json: bool

Output in JSON format

page_size: Option<u32>

Override page size (default: auto-detect)

Pages

Detailed page structure analysis

Goes beyond FIL headers to decode the internal structure of each page type: INDEX pages show the B+Tree index header, FSEG inode pointers, and infimum/supremum system records; UNDO pages show the undo page header and segment state; BLOB/LOB pages show chain pointers and data lengths; and page 0 shows extended FSP header fields including compression and encryption flags. Use -l for a compact one-line-per-page listing, -t INDEX to filter by page type, or -p for a single page deep dive.

Fields

file: String

Path to InnoDB data file (.ibd)

page: Option<u64>

Display a specific page number

verbose: bool

Display additional information

show_empty: bool

Show empty/allocated pages

list: bool

Compact list mode (one line per page)

filter_type: Option<String>

Filter by page type (e.g., INDEX)

json: bool

Output in JSON format

page_size: Option<u32>

Override page size (default: auto-detect)

Dump

Hex dump of raw page bytes

Operates in two modes: page mode (default) reads a full page by number and produces a formatted hex dump with file-relative offsets; offset mode (--offset) reads bytes at an arbitrary file position, useful for inspecting structures that cross page boundaries. Use --length to limit the number of bytes shown, or --raw to emit unformatted binary bytes suitable for piping to other tools.

Fields

file: String

Path to InnoDB data file

page: Option<u64>

Page number to dump (default: 0)

offset: Option<u64>

Absolute byte offset to start dumping (bypasses page mode)

length: Option<usize>

Number of bytes to dump (default: page size or 256 for offset mode)

raw: bool

Output raw binary bytes (no formatting)

page_size: Option<u32>

Override page size (default: auto-detect)

Corrupt

Intentionally corrupt pages for testing

Writes random bytes into a tablespace file to simulate data corruption. Targets can be the FIL header (-k), the record data area (-r), or an absolute byte offset (--offset). If no page is specified, one is chosen at random. Use --verify to print before/after checksum comparisons confirming the page is now invalid — useful for verifying that inno checksum correctly detects the damage.

Fields

file: String

Path to data file

page: Option<u64>

Page number to corrupt (random if not specified)

bytes: usize

Number of bytes to corrupt

header: bool

Corrupt the FIL header area

records: bool

Corrupt the record data area

offset: Option<u64>

Absolute byte offset to corrupt (bypasses page calculation)

verify: bool

Show before/after checksum comparison

json: bool

Output in JSON format

page_size: Option<u32>

Override page size (default: auto-detect)

Find

Search for pages across data directory

Recursively discovers all .ibd files under a MySQL data directory, opens each as a tablespace, and reads the FIL header of every page looking for a matching page_number field. Optional --checksum and --space-id filters narrow results when the same page number appears in multiple tablespaces. Use --first to stop after the first match for faster lookups.

Fields

datadir: String

MySQL data directory path

page: u64

Page number to search for

checksum: Option<u32>

Checksum to match

space_id: Option<u32>

Space ID to match

first: bool

Stop at first match

json: bool

Output in JSON format

page_size: Option<u32>

Override page size (default: auto-detect)

Tsid

List/find tablespace IDs

Scans .ibd and .ibu files under a MySQL data directory and reads the space ID from the FSP header (page 0, offset 38) of each file. In list mode (-l) it prints every file and its space ID; in lookup mode (-t <id>) it finds the file that owns a specific tablespace ID. Useful for mapping a space ID seen in error logs or INFORMATION_SCHEMA back to a physical file on disk.

Fields

datadir: String

MySQL data directory path

list: bool

List all tablespace IDs

tablespace_id: Option<u32>

Find table file by tablespace ID

json: bool

Output in JSON format

page_size: Option<u32>

Override page size (default: auto-detect)

Sdi

Extract SDI metadata (MySQL 8.0+)

Locates SDI (Serialized Dictionary Information) pages in a tablespace by scanning for page type 17853, then reassembles multi-page SDI records by following the page chain. The zlib-compressed payload is decompressed and printed as JSON. Each tablespace in MySQL 8.0+ embeds its own table/column/index definitions as SDI records, eliminating the need for the .frm files used in older versions. Use --pretty for indented JSON output.

Fields

file: String

Path to InnoDB data file (.ibd)

pretty: bool

Pretty-print JSON output

page_size: Option<u32>

Override page size (default: auto-detect)

Log

Analyze InnoDB redo log files

Opens an InnoDB redo log file (ib_logfile0/ib_logfile1 for MySQL < 8.0.30, or #ib_redo* files for 8.0.30+) and displays the log file header, both checkpoint records, and per-block details including block number, data length, checkpoint number, and CRC-32C checksum status. With --verbose, MLOG record types within each data block are decoded and summarized. Use --blocks N to limit output to the first N data blocks, or --no-empty to skip blocks that contain no redo data.

Fields

file: String

Path to redo log file (ib_logfile0, ib_logfile1, or #ib_redo*)

blocks: Option<u64>

Limit to first N data blocks

no_empty: bool

Skip empty blocks

verbose: bool

Display additional information

json: bool

Output in JSON format

Info

Show InnoDB file and system information

Operates in three modes. --ibdata reads the ibdata1 page 0 FIL header and redo log checkpoint LSNs. --lsn-check compares the ibdata1 header LSN with the latest redo log checkpoint LSN to detect whether the system tablespace and redo log are in sync (useful for diagnosing crash-recovery state). -D/-t queries a live MySQL instance via INFORMATION_SCHEMA.INNODB_TABLES and INNODB_INDEXES for tablespace IDs, table IDs, index root pages, and key InnoDB status metrics (requires the mysql feature).

Fields

ibdata: bool

Inspect ibdata1 page 0 header

lsn_check: bool

Compare ibdata1 and redo log LSNs

datadir: Option<String>

MySQL data directory path

database: Option<String>

Database name (for table/index info)

table: Option<String>

Table name (for table/index info)

host: Option<String>

MySQL host

port: Option<u16>

MySQL port

user: Option<String>

MySQL user

password: Option<String>

MySQL password

defaults_file: Option<String>

Path to MySQL defaults file (.my.cnf)

json: bool

Output in JSON format

page_size: Option<u32>

Override page size (default: auto-detect)

Recover

Recover data from corrupt/damaged tablespace files

Scans a tablespace file and classifies each page as intact, corrupt, empty, or unreadable. For INDEX pages, counts recoverable user records by walking the compact record chain. Produces a recovery assessment showing how many pages and records can be salvaged.

Use --force to also extract records from pages with bad checksums but valid-looking headers — useful when data is partially damaged but the record chain is still intact. Use --page-size to override page size detection when page 0 is corrupt.

With --verbose, per-page details are shown including page type, status, LSN, and record count. With --json, a structured report is emitted including optional per-record detail when combined with --verbose.

Fields

file: String

Path to InnoDB data file (.ibd)

page: Option<u64>

Analyze a single page instead of full scan

verbose: bool

Show per-page details

json: bool

Output in JSON format

force: bool

Extract records from corrupt pages with valid headers

page_size: Option<u32>

Override page size (critical when page 0 is corrupt)

Checksum

Validate page checksums

Reads every page in a tablespace and validates its stored checksum against both CRC-32C (MySQL 5.7.7+) and legacy InnoDB algorithms. Also checks that the header LSN low-32 bits match the FIL trailer. All-zero pages are counted as empty and skipped. With --verbose, per-page results are printed including the detected algorithm and stored vs. calculated values. Exits with code 1 if any page has an invalid checksum, making it suitable for use in scripts and CI.

Fields

file: String

Path to InnoDB data file (.ibd)

verbose: bool

Show per-page checksum details

json: bool

Output in JSON format

page_size: Option<u32>

Override page size (default: auto-detect)

Trait Implementations

impl FromArgMatches for Commands

fn from_arg_matches(__clap_arg_matches: &ArgMatches) -> Result<Self, Error>

Instantiate Self from ArgMatches, parsing the arguments as needed.

fn from_arg_matches_mut( __clap_arg_matches: &mut ArgMatches, ) -> Result<Self, Error>

Instantiate Self from ArgMatches, parsing the arguments as needed.

fn update_from_arg_matches( &mut self, __clap_arg_matches: &ArgMatches, ) -> Result<(), Error>

Assign values from ArgMatches to self.

fn update_from_arg_matches_mut<'b>( &mut self, __clap_arg_matches: &mut ArgMatches, ) -> Result<(), Error>

Assign values from ArgMatches to self.

impl Subcommand for Commands

fn augment_subcommands<'b>(__clap_app: Command) -> Command

Append to Command so it can instantiate Self via FromArgMatches::from_arg_matches_mut

fn augment_subcommands_for_update<'b>(__clap_app: Command) -> Command

Append to Command so it can instantiate self via FromArgMatches::update_from_arg_matches_mut

fn has_subcommand(__clap_name: &str) -> bool

Test whether Self can parse a specific subcommand