path: root/src/docs/command-line.dox

/*! @page command_line WiredTiger command line utility

WiredTiger includes a command line utility, \c wt.

@section util_global_synopsis SYNOPSIS
<code>wt [-LRVv] [-C config] [-E secretkey ] [-h directory] command [command-specific arguments]</code>

@section util_global_description DESCRIPTION
The \c wt tool is a command-line utility that provides access to
various pieces of the WiredTiger functionality.

@section util_global_options OPTIONS
There are several global options:

@par <code>-C config</code>
Specify configuration strings for the ::wiredtiger_open function.
@par <code>-E secretkey</code>
Specify an encryption secret key for the ::wiredtiger_open function.
@par <code>-h directory</code>
Specify a database home directory.
@par <code>-L</code>
Forcibly turn off logging subsystem for debugging purposes.
@par <code>-R</code>
Run recovery if the underlying database is configured to do so.
@par <code>-V</code>
Display WiredTiger version and exit.
@par <code>-v</code>
Set verbose output.

Unless otherwise described by a \c wt command, the \c wt tool exits zero
on success and non-zero on error.

The \c wt tool supports several commands.  If configured in the underlying
database, some commands will run recovery when opening the database.  If
the user wants to force recovery on any command, use the \c -R option.
In general, commands that modify the database or tables will run recovery
by default and commands that only read data will not run recovery.

<hr>
@section util_backup wt backup
Perform a backup of a database or set of data sources.

The \c backup command performs a backup of the database, copying the
database files to a \c specified directory, which can be subsequently
opened as a WiredTiger database.  See @ref backup for more information,
and @ref file_permissions for specifics on the copied file permissions.

@subsection util_backup_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] backup [-t uri] directory</code>

@subsection util_backup_options Options
The following are command-specific options for the \c backup command:

@par <code>-t uri</code>
By default, the \c backup command does a backup of the entire database;
the \c -t option changes the \c backup command to do a backup of only
the named data sources.

<hr>
@section util_compact wt compact
Compact a table or file.

The \c compact command attempts to rewrite the specified table or file
to consume less disk space.

@subsection util_compact_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] compact uri</code>

@subsection util_compact_options Options
The \c compact command has no command-specific options.

<hr>
@section util_create wt create
Create a table or file.

The \c create command creates the specified \c uri with the specified
configuration.  It is equivalent to a call to WT_SESSION::create with
the specified string arguments.

@subsection util_create_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] create [-c config] uri</code>

@subsection util_create_options Options
The following are command-specific options for the \c create command:

@par <code>-c</code>
Include a configuration string to be passed to WT_SESSION::create.

<hr>
@section util_drop wt drop
Drop a table or file.

The \c drop command drops the specified \c uri.  It is equivalent to a
call to WT_SESSION::drop with the "force" configuration argument.

@subsection util_drop_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] drop uri</code>

@subsection util_drop_options Options
The \c drop command has no command-specific options.

<hr>
@section util_dump wt dump
Export data in a text format.

The \c dump command outputs the specified table in a portable format
which can be re-loaded into a new table using the \c load command.

See @subpage dump_formats for details of the dump file formats.

@subsection util_dump_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] dump [-jrx] [-c checkpoint] [-f output] uri</code>

@subsection util_dump_options Options
The following are command-specific options for the \c dump command:

@par <code>-c</code>
By default, the \c dump command opens the most recent version of the data
source; the \c -c option changes the \c dump command to open the named
checkpoint.

@par <code>-f</code>
By default, the \c dump command output is written to the standard output;
the \c -f option re-directs the output to the specified file.

@par <code>-j</code>
Dump in JSON (<a href="http://www.json.org">JavaScript Object Notation</a>)
format.

@par <code>-r</code>
Dump in reverse order, from largest key to smallest.

@par <code>-x</code>
Dump all characters in a hexadecimal encoding (the default is to leave
printable characters unencoded).

<hr>
@section util_list wt list
List the tables and files in the database.

By default, the \c list command prints out the tables and files stored in
the database.  If a URI is specified as an argument, only information about
that data source is printed.

@subsection util_list_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] list [-cv] [uri]</code>

@subsection util_list_options Options
The following are command-specific options for the \c list command:

@par <code>-c</code>
If the \c -c option is specified, the data source's checkpoints are printed
in a human-readable format.

@par <code>-v</code>
If the \c -v option is specified, the data source's complete schema table
value is printed.

<hr>
@section util_load wt load
Load a table or file from dump output.

The \c load command reads the standard input for data and loads it into
a table or file, creating the table or file if it does not yet exist.
The data should be the format produced by the \c dump command; see
@ref dump_formats for details.

By default, if the table or file already exists, data in the file or
table will be overwritten by the new data (use the \c -n option to
make an attempt to overwrite existing data return an error).

@subsection util_load_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] load [-ajn] [-f input] [-r name] [uri configuration ...]</code>

@subsection util_load_options Options
The following are command-specific options for the \c load command:

@par <code>-a</code>
If the \c -a option is specified, record number keys in the input are
ignored and the data is appended to the data source and assigned new record
number keys.  The \c -a option is only applicable when loading into a
column store.

@par <code>-f</code>
By default, the \c load command reads from the standard input; the \c
-f option reads the input from the specified file.

@par <code>-j</code>
Load input in the JSON (<a href="http://www.json.org">JavaScript Object
Notation</a>) format that was created by the <code>dump -j</code> command.

@par <code>-n</code>
By default, input data will overwrite existing data where the key/value
pair already exists in the data source; the \c -n option causes the \c
load command to fail if there's an attempt to overwrite already existing
data.

@par <code>-r</code>
By default, the \c load command uses the table or file name taken from the
input; the \c -r option renames the data source.

Additionally, \c uri and \c configuration pairs may be specified to the
\c load command.  These configuration pairs can be used to modify the
configuration values from the dump header passed to the WT_SESSION::create
call.

The \c uri part of the configuration pair should match only one of the
objects being loaded, but may be a prefix of the object being matched.
For example, the following two sets of configuration pairs are
equivalent in the case of loading a single table named \c xxx.

@code
table block_allocation=first
table:xxx block_allocation=first
@endcode

It's an error, however, to specify a matching prefix that matches more
than a single object being loaded.

Multiple \c configuration arguments may be specified. For example, the
following two sets of configuration pairs are equivalent:

@code
table:xxx block_allocation=first,prefix_compress=false
table:xxx block_allocation=first table:xxx prefix_compress=false
@endcode

<hr>
@section util_loadtext wt loadtext
Load text into a table or file.

The \c loadtext command reads the standard input for text and loads it
into a table or file.  The input data should be printable characters,
with newline delimiters for each key or value.

The \c loadtext command does not create the file if it does not yet
exist.

In the case of inserting values into a column-store table or file, each
value is appended to the table or file; in the case of inserting values
into a row-store table or file, lines are handled in pairs, where the
first line is the key and the second line is the value.  If the
row-store table or file already exists, data in the table or file will
be overwritten by the new data.

@subsection util_loadtext_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] loadtext [-f input]</code>

@subsection util_loadtext_options Options
The following are command-specific options for the \c loadtext command:

@par <code>-f</code>
By default, the \c loadtext command reads from the standard input; the
\c -f option reads the input from the specified file.

<hr>
@section util_printlog wt printlog
Display the database log.

The \c printlog command outputs the database log.

@subsection util_printlog_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] printlog [-x] [-f output]</code>

@subsection util_printlog_options Options
The following are command-specific options for the \c printlog command:

@par <code>-f</code>
By default, the \c printlog command output is written to the standard
output; the \c -f option re-directs the output to the specified file.

@par <code>-x</code>
Keys and value items in the log are printed in hex format in addition
to the default string format.

<hr>
@section util_read wt read
Read records from a table or file.

The \c read command prints out the records associated with the specified
keys from the specified data source.  The data source must be configured
with string or record number keys and string values.

The \c read command exits non-zero if a specified record is not found.

@subsection util_read_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] read uri key ...</code>

@subsection util_read_options Options
The \c read command has no command-specific options.

<hr>
@section util_rename wt rename
Rename a table or file.

The \c rename command renames the specified table or file.

@subsection util_rename_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] rename uri name</code>

@subsection util_rename_options Options
The \c rename command has no command-specific options.

<hr>
@section util_salvage wt salvage
Recover data from a corrupted file.

The \c salvage command salvages the specified data source, discarding any
data that cannot be recovered.  Underlying files are re-written in
place, overwriting the original file contents.

@subsection util_salvage_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] salvage [-F force] uri</code>

@subsection util_salvage_options Options
The following are command-specific options for the \c salvage command:

@par <code>-F</code>
By default, salvage will refuse to salvage files that fail basic tests
(for example, files that don't appear to be in a WiredTiger format).
The \c -F option forces the salvage of the file, regardless.

<hr>
@section util_stat wt stat
Display database or data source statistics.

The \c stat command outputs run-time statistics for the WiredTiger
engine, or, if specified, for the URI on the command-line.

@subsection util_stat_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] stat [-f] [uri]</code>

@subsection util_stat_options Options
The following are command-specific options for the \c stat command:

@par <code>-f</code>
Include only "fast" statistics in the output (equivalent to passing
<code>statistics=(fast)</code>) to WT_SESSION::open_cursor.

<hr>
@section util_upgrade wt upgrade
Upgrade a table or file.

The \c upgrade command upgrades the specified table or file, exiting
success if the data source is up-to-date, and failure if the data source
cannot be upgraded.

@subsection util_upgrade_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] upgrade uri</code>

@subsection util_upgrade_options Options
The \c upgrade command has no command-specific options.

<hr>
@section util_verify wt verify
Check the structural integrity of a table or file.

The \c verify command verifies the specified table or file, exiting
success if the data source is correct, and failure if the data source is
corrupted.

@subsection util_verify_synopsis Synopsis
<code>wt [-RVv] [-C config] [-E secretkey ] [-h directory] verify uri</code>

@subsection util_verify_options Options
The \c verify command has no command-specific options.

<hr>
@section util_write wt write
Write records to a table or file.

The \c write command stores records into the specified data source.
The data source must be configured with string or record number keys and
string values.

If the \c write command is called with the \c -a option, each
command-line argument is a single value to be appended to the specified
column-store data source.  If the \c write command is not called with
the \c -a option, the command-line arguments are key/value pairs.

Attempting to overwrite an already existing record will fail.

@subsection util_write_synopsis Synopsis
<code>
wt [-RVv] [-C config] [-E secretkey ] [-h directory] write -a uri value ...
<br>
wt [-RVv] [-C config] [-E secretkey ] [-h directory] write [-o] uri key value ...
</code>

@subsection util_write_options Options
The following are command-specific options for the \c write command:

@par <code>-a</code>
Append each value as a new record in the data source.

@par <code>-o</code>
By default, attempting to overwrite an already existing record will
fail.  The \c -o option changes \c write to overwrite previously
existing records.

*/