Arguments overview:
‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒

Pattern specification:
Either as first argument, or as --pattern=foo option, or as pattern(s)
in --patterns-from=file option.

 Special markers:
 '/ << bar >> /'          Raku regex indicated by being bounded by / /
 '{ .ends-with("bar") }'  Raku code indicated by being bounded by { }
 '*.starts-with("foo")'   Raku WhateverCode starting with *
 's:foo bar'              Multiple patterns, split on whitespace:
 jp:foo                   JSON path specification (with --json-per-xxx)
 §string   String should match as a word
 ^string   String should match at start of line
 string$   String should match at end of line
 ^string$  String should match with line entirely
 string    String should match anywhere in line
 !string   String should NOT match anywhere in line
 &string   Additional string should ALSO match anywhere in line

 Pattern type overrides:
 --type=auto         Interpret special markers in pattern as above (default)
 --type=regex        Interpret pattern as a regex (without special markers)
 --type=code         Interpret pattern as code (without special markers)
 --type=json-path    Interpret pattern as a JSON path (without jp: prefix)
 --type=words        Pattern must match as a word
 --type=starts-with  Pattern must match at start of line
 --type=ends-with    Pattern must match at end of line
 --type=equal        Pattern must match entirely
 --type=not          Pattern must NOT match
 --type=split        Multiple patterns, split on whitespace
 --type=contains     Pattern must match anywhere (default if no markers)

 Additional patterns:
 --and=bar           Must ALSO match anywhere in line
 --andnot=bar        Must ALSO NOT match anywhere in line
 --or=bar            May ALSO match anywhere in line
 --ornot=bar         May ALSO NOT match anywhere in line

String search pattern modifiers:
 --ignorecase  Ignore distinction between upper, lower and title case letters
 --ignoremark  Only compare base characters, ignore additional marks
 --smartcase   As --ignorecase if pattern does *NOT* contain any uppercase
 --smartmark   As --ignoremark if pattern does *NOT* contain any accents

Item producers:
 --per-line[=producer]  Produce lines for pattern matching
 --per-file[=producer]  Produce whole file for pattern matching
 --encoding[=utf8-c8]   Encoding to be used when reading text files
 --find            Produce absolute paths for pattern matching
 --csv-per-line    Produce rows from Text::CSV object for pattern matching
 --pdf-info        Produce meta-info of PDF file as a single hash
 --pdf-per-file    Produce text of PDF file as a single string
 --pdf-per-line    Produce text of PDF file as lines of text
 --json-per-file   Interpret whole file as a single JSON for matching
 --json-per-elem   Interpret whole file as JSON, match each top level element
 --json-per-line   Interpret each line as JSON for pattern matching
 --blame-per-file  Produce Git::Blame::File object for pattern matcing
 --blame-per-line  Produce Git::Blame::Line objects for pattern matching
 --mbc             Produce one MoarVM::Bytecode object per file for matching
 --mbc-frames      Produce MoarVM::Bytecode::Frame objects for matching
 --mbc-strings     Produce MoarVM::Bytecode string-heap strings for matching
 --unicode         Produce names of codepoints for pattern matching

Code pattern helpers:
 --repository=lib      First look for any modules to load in the "lib" directory
 --module=foo          Load module "foo" before compiling Raku code pattern
 --quietly             Disable warnings when executing pattern (default: False)
 --silently[=out,err]  Capture output when executing pattern (default: False)

Haystack specification:
 path1 path2 ...          All other arguments are paths (default: current dir)
 --paths=specification    Comma separated list of paths instead of ^^
 --files-from=file        Read files to inspect from given file
 --paths-from=file        Read path specifications from given file
 --under-version-control  Select paths that are under version control
 --recurse-symlinked-dir  Recurse into symlinked directories? (default: don't)
 --recurse-unmatched-dir  Recurse into directories not matching (default: don't)
 --eco-code[=name]        All code in local ecosystem cache
 --eco-doc[=name]         All documentation in local ecosystem cache
 --eco-meta[=name]        One or more Raku ecosystem's module's META
 --eco-provides[=name]    All code in modules in local ecosystem cache
 --eco-scripts[=name]     All code in scripts in local ecosystem cache
 --eco-tests[=name]       All code in test files in local ecosystem cache
 --rakudo-all             All files in Rakudo
 --rakudo-c               All C programming language files in Rakudo
 --rakudo-doc             All documentation files
 --rakudo-java            All Java programming language files in Rakudo
 --rakudo-js              All Javascript programming language files in Rakudo
 --rakudo-nqp             All NQP related files in Rakudo
 --rakudo-perl            All Perl programming language files in Rakudo
 --rakudo-raku            All Raku Programming Language files in Rakudo
 --rakudo-shell           All shell scripts / batch files in Rakudo
 --rakudo-test            All testing related files in Rakudo
 --rakudo-yaml            All files containing YAML of some sort in Rakudo

Filesystem filters:
 --accessed=condition    Check on epoch when file was last accessed
 --accept=code           Given an IO of a file, return if file acceptable
 --auto-decompress       File may be gzipped (.gz) or bzip2ed (.bz2)
 --blocks=condition      Number of filesystem blocks used by file
 --created=condition     Check on epoch when file was created
 --deny=code             Given an IO of a file, return if **NOT** acceptable
 --device-number         Number of device on which file is located
 --dir=expression        Directory basename filter, default: not ^.
 --exec=program          Run program, include if successful
 --extensions=spec       By group (e.g. #raku) and/or comma separated
 --file=expression       File basename filter, default: all
 --filesize=condition    Number of bytes of data used by file
 --group=selector        File is owned by group names / expression
 --gid=condition         File is owned by given gid
 --hard-links=condition  Number of hard-links to file on file system
 --has-setgid            Has SETGID bit set in attributes
 --has-setuid            Has SETUID bit set in attributes
 --inode=condition       Inode number of file on file system
 --is-empty              File does (not) contain any data
 --is-executable         File can (not) be executed by owner
 --is-group-executable   File can (not) be executed by group members of owner
 --is-group-readable     File can (not) be read by group members of owner
 --is-group-writable     File can (not) be written to by group members of owner
 --is-moarvm             File is (not) a MoarVM bytecode file
 --is-owned-by-group     File is (not) owned by group of user
 --is-owned-by-user      File is (not) owned by user
 --is-pdf                File is (not) a PDF file
 --is-readable           File can (not) be read by owner
 --is-sticky             Has STICKY bit set in attributes
 --is-symbolic-link      File is (not) a symbolic link
 --is-text               File contains (does not contain) text
 --is-world-executable   File can (not) be executed by anybody
 --is-world-readable     File can (not) be read by anybody
 --is-world-writable     File can (not) be written to by anybody
 --is-writable           File can (not) be written to by owner
 --meta-modified=cond    Check on epoch when meta-info of file was last changed
 --mode=condition        The full mode value of the file
 --modified=condition    Check on epoch when file was last changed
 --shell=action          Run shell command, include if successful
 --user=selector         File is owned by user names / expression
 --uid=condition         File is owned by given uid

Result modifiers:
 --also-first=N            List first N lines if there is a match
 --always-first=N          List first N lines always
 --context=N               List N lines around any line with a match
 --before-context=N        List N lines before any line with a match
 --after-context=N         List N lines after any line with a match
 --paragraph-context       List all lines in the same paragraph around a match
 --passthru-context        List all lines if there is at least one match
 --passthru                List all lines always
 --invert-match            Select lines that do **not** match
 --max-matches-per-file=N  Maximum number of matches per file, default: all
 --count-only              Only return count of matches
 --stats-only              Only return statistics of the search
 --files-with-matches      Only return filenames with matches
 --files-without-matches   Only return filenames without matches
 --matches-only            Only produce the actual matches
 --unique                  Only return unique lines
 --frequencies             Only return lines and their frequencies
 --classify=classifier     Classify lines by a key (or not)
 --categorize=categorizer  Classify lines by zero or more keys

Listing modifiers:
 --human                     Force defaults as if a human is watching
 --absolute                  Show paths as absolute paths
 --break[=string]            String to show between different files
 --group-matches             Show filename only once on a separate line
 --show-blame                Show git blame information if possible
 --show-filename             Show the filename in which match occurred
 --file-separator-null       Separate filenames with null bytes
 --show-item-number          Show the item number in which match occurred
 --summary-if-larger-than=N  Summarize matching line if longer than N chars
 --trim                      Remove whitespace, true if no context
 --highlight                 Highlight any matches
 --highlight-before=xxx      String to put before match
 --highlight-after=yyy       String to put after match
 --proximate[=N]             Grouping of matched lines
 --only-first[=N]            Only N results, default N = 1, absence = 1000
 --output-file=filename      Store output to indicated file (default: STDOUT)
 --output-dir=directory      Store output ino indicated directory / group
 --pager                     Pager to use (default: RAK_PAGER // none)
 --stats                     Also show statistics

Content modifiers:
 --modify-files        Change content of files using the Callable pattern
 --backup[=extension]  Make backup of file with extension (default: .bak)
 --rename-files        Rename files using the Callable pattern
 --dryrun              Don't actually make any changes, just run the process
 --no-modifications    Do not allow any modifications to be made

Resource usage:
 --batch[=N]          Max number of files to process in a thread (default: 64)
 --degree[=N | code]  Max workers to use for processing (default: cores - 1)

Special options:
 --edit[=editor]        Go edit the result in an editor, (default EDITOR or vim)
 --vimgrep              Produce output in :vimgrep format
 --checkout=string      Checkout branch if pattern matches single branch name
 --sourcery             Treat pattern as code, use candidates as result
 --backtrace            Interpret text as a Raku backtrace, use as result
 --execute-raku[=code]  Run raku code for exception, use as result

Option management:
 --save=name            Translate --name to all other options specified,
                        remove if --save was the only option specified
 --description=text     Add description to custom option
 --ack                  Activate "ack" compatibility options
 --list-custom-options  List all previously saved options

General options:
 --help          Show this
 --help=subject  Show additional help about subject, out of:
   argument   | code       | content  | debugging | examples | faq
   filesystem | general    | haystack | item      | listing  | option
   pattern    | philosophy | resource | result    | special  | string
 --help foo      Search for "foo" in all help subjects
 --help=faq foo  Search for "foo" in subject "faq"
 --progress      Show search progress on STDERR
 --version       Show version information
 --verbose       Be more verbose, if applicable

Debugging options:
 --dont-catch             Let code exceptions be thrown with backtrace
 --list-known-extensions  Show all known extensions and their group namep
 --list-expanded-options  List all options given after having been expanded
 --rak                    Show arguments to "rak" plumbing before running

Argument format:
 foo        Positional argument
 --foo      Option "foo" is True
 --/foo     Option "foo" is False
 --no-foo   Option "foo" is False
 --foo=bar  Option "foo" is "bar"
 -fo        Single letter options "f" and "o" True
 -/ba       Single letter options "b" and "a" False
 -j2        Single letter option "j" with value 2
