A newer version of this documentation is available.

View Latest

cbimport csv

      +

      Imports CSV data into Couchbase

      SYNOPSIS

      cbimport csv [--cluster <url>] [--bucket <bucket_name>] [--dataset <path>]
                   [--username <username>] [--password <password>] [--client-cert <path>]
                   [--client-cert-password <password>] [--client-key <path>]
                   [--client-key-password <password>] [--generate-key <key_expr>]
                   [--limit-rows <num>] [--skip-rows <num>] [--field-separator <char>]
                   [--cacert <path>] [--no-ssl-verify] [--threads <num>]
                   [--error-log <path>] [--log-file <path>] [--verbose]
                   [--field-delimiter <char>] [--generator-delimiter <char>]
                   [--ignore-fields <fields>]
                   [--scope-collection-exp <scope_collection_expression>]

      DESCRIPTION

      Imports CSV and other forms of separated value type data into Couchbase. By default data files should contain comma separated values, but if for example you are importing data that is tab separated you can use the --field-separator flag to specify that tabs are used instead of commas.

      The cbimport command also supports custom key-generation for each document in the imported file. Key generation is done with a combination of pre-existing fields in a document and custom generator functions supplied by cbimport. See the KEY GENERATION section below for details about key generators.

      OPTIONS

      Below are a list of required and optional parameters for the cbimport-csv command.

      Required

      -c,--cluster <url>

      The hostname of a node in the cluster to import data into. See the HOST FORMATS section below for details about hostname specification formats.

      -u,--username <username>

      The username for cluster authentication. The user must have the appropriate privileges to take a backup.

      -p,--password <password>

      The password for cluster authentication. The user must have the appropriate privileges to take a backup. If not password is supplied to this option then you will be prompted to enter your password.

      --client-cert <path>

      The path to a client certificate used to authenticate when connecting to a cluster. May be supplied with --client-key as an alternative to the --username and --password flags. See the CERTIFICATE AUTHENTICATION section for more information.

      --client-cert-password <password>

      The password for the certificate provided to the --client-cert flag, when using this flag, the certificate/key pair is expected to be in the PKCS#12 format. See the CERTIFICATE AUTHENTICATION section for more information.

      --client-key <path>

      The path to the client private key whose public key is contained in the certificate provided to the --client-cert flag. May be supplied with --client-cert as an alternative to the --username and --password flags. See the CERTIFICATE AUTHENTICATION section for more information.

      --client-key-password <password>

      The password for the key provided to the --client-key flag, when using this flag, the key is expected to be in the PKCS#8 format. See the CERTIFICATE AUTHENTICATION section for more information.

      -b,--bucket <bucket_name>

      The name of the bucket to import data into.

      -d,--dataset <uri>

      The URI of the dataset to be loaded. cbimport supports loading data from a local file or from a URL. When importing data from a local file the path must be prefixed with file://.

      -g,--generate-key <key_expr>

      Specifies a key expression used for generating a key for each document imported. See the Key Generation section below for more information on specifying key generators. If the resulting key is not unique the values will be overridden, resulting in fewer documents than expected being imported. To ensure that the key is unique add #MONO_INCR# or #UUID# to the key generator expression.

      Optional

      --field-delimiter <char>

      Specifies the character to be used to denote field references in the key generator expression. It defaults to %. See the KEY GENERATION section.

      --generator-delimiter <char>

      Specifies the character to be used to denote generator references in the key generator expression. It defaults to #. See the KEY GENERATION section.

      --field-separator <num>

      Specifies the field separator to use when reading the dataset. By default the separator is a comma. To read tab separated files you can specify a tab in this field. Tabs are specified as \t.

      --limit-rows <num>

      Specifies that the utility should stop loading data after reading a certain amount of rows from the dataset. This option is useful when you have a large dataset and only want to partially load it.

      --skip-rows <num>

      Specifies that the utility should skip some rows before we start importing data. If this flag is used together with the --limit-rows flag then we will import the number of rows specified by --limit-rows after we have skipped the rows specified by --skip-rows.

      --no-ssl-verify

      Skips the SSL verification phase. Specifying this flag will allow a connection using SSL encryption, but will not verify the identity of the server you connect to. You are vulnerable to a man-in-the-middle attack if you use this flag. Either this flag or the --cacert flag must be specified when using an SSL encrypted connection.

      --infer-types

      By default all values in a CSV files are interpreted as strings. If infer types is set then cbimport will look at each value and decide whether it is a string, integer, or boolean value and put the inferred type into the document.

      --omit-empty

      Some values in a CSV row will not contain any data. By default these values are put into the generated JSON document as an empty string. To omit fields that contain empty values specify this flag.

      --cacert <cert_path>

      Specifies a CA certificate that will be used to verify the identity of the server being connecting to. Either this flag or the --no-ssl-verify flag must be specified when using an SSL encrypted connection.

      --scope-collection-exp <scope_collection_expression>

      When importing to a collection aware cluster you may optionally choose to provide a scope/collection expression which will be used to determine which scope/collection to insert the imported document in. This flag closely resembles the behavior/syntax of the --generate-key flag. For example, to use a static scope/collection use --scope-collection-exp scope.collection. To use information from the CSV row, specify the column name between % characters. For example, --scope-collection-exp %scope_column%.%collection_column%. Columns that contain a % character may be escaped using %%. For more information about the accepted format, see the SCOPE/COLLECTION PARSER section.

      -t,--threads <num>

      Specifies the number of concurrent clients to use when importing data. Fewer clients means imports will take longer, but there will be less cluster resources used to complete the import. More clients means faster imports, but at the cost of more cluster resource usage. This parameter defaults to 1 if it is not specified and it is recommended that this parameter is not set to be higher than the number of CPUs on the machine where the import is taking place.

      -e,--errors-log <path>

      Specifies a log file where JSON documents that could not be loaded are written to. A document might not be loaded if a key could not be generated for the document or if the document is not valid JSON. The errors file is written in the "json lines" format (one document per line).

      -l,--log-file <path>

      Specifies a log file for writing debugging information about cbimport execution.

      -v,--verbose

      Specifies that logging should be sent to stdout. If this flag is specified along with the -l/--log-file option then the verbose option is ignored.

      --ignore-fields <fields>

      Specify a comma separated list of field names that will be excluded from the imported document. The field reference syntax is the same as the one used in KEY GENERATORS to refer to fields.

      HOST FORMATS

      When specifying a host/cluster for a command using the -c/--cluster flag, the following formats are accepted:

      • <addr>:<port>

      • http://<addr>:<port>

      • https://<addr>:<port>

      • couchbase://<addr>:<port>

      • couchbases://<addr>:<port>

      • couchbase://<srv>

      • couchbases://<srv>

      • <addr>:<port>,<addr>:<port>

      • <scheme>://<addr>:<port>,<addr>:<port>

      The <port> portion of the host format may be omitted, in which case the default port will be used for the scheme provided. For example, http:// and couchbase:// will both default to 8091 where https:// and couchbases:// will default to 18091. When connecting to a host/cluster using a non-default port, the <port> portion of the host format must be specified.

      Connection Strings (Multiple nodes)

      The -c/--cluster flag accepts multiple nodes in the format of a connection string; this is a comma separated list of <addr>:<port> strings where <scheme> only needs to be specified once. The main advantage of supplying multiple hosts is that in the event of a failure, the next host in the list will be used.

      For example, all of the following are valid connection strings:

      • localhost,[::1]

      • 10.0.0.1,10.0.0.2

      • http://10.0.0.1,10.0.0.2

      • https://10.0.0.1:12345,10.0.0.2

      • couchbase://10.0.0.1,10.0.0.2

      • couchbases://10.0.0.1:12345,10.0.0.2:12345

      SRV Records

      The -c/--cluster flag accepts DNS SRV records in place of a host/cluster address where the SRV record will be resolved into a valid connection string. There are a couple of rules which must be followed when supplying an SRV record which are as follows:

      • The <scheme> portion must be either couchbase:// or couchbases://

      • The <srv> portion should be a hostname with no port

      • The <srv> portion must not be a valid IP address

      For example, all of the following are valid connection string using an SRV record:

      • couchbase://hostname

      • couchbases://hostname

      CERTIFICATE AUTHENTICATION (MTLS AUTHENTICATION)

      This tool supports authenticating against a Couchbase Cluster by using certificate based authentication (mTLS authentication). To use certificate based authentication a certificate/key must be supplied, there a currently multiple ways this may be done.

      PEM ENCODED CERTIFICATE/KEY

      An unencrypted PEM encoded certificate/key may be supplied by using: - --client-cert <path> - --client-key <path>

      The file passed to --client-cert must contain the client certificate, and an optional chain required to authenticate the client certificate.

      The file passed to --client-key must contain at most one private key, the key can be in one of the following formats: - PKCS#1 - PKCS#8 - EC

      Currently, only the following key types are supported: - RSA - ECDSA - ED25519

      PEM ENCODED CERTIFICATE/PEM OR DER ENCRYPTED PKCS#8 KEY

      An encrypted PKCS#8 formatted key may be provided using: - --client-cert <path> - --client-key <path> - --client-key-password <password>

      The file passed to --client-cert must contain the client certificate, and an optional chain required to authenticate the client certificate.

      Currently, only the following key types are supported: - RSA - ECDSA - ED25519

      ENCRYPTED PKCS#12 CERTIFICATE/KEY

      An encrypted PKCS#12 certificate/key may be provided using: - --client-cert <path> - --client-cert-password <password>

      The file passed to --client-cert must contain the client certificate and exactly one private key. It may also contain the chain required to authenticate the client certificate.

      Currently, only the following key types are supported: - RSA - ECDSA - ED25519

      KEY GENERATORS

      Key generators are used in order to generate a key for each document loaded. Keys can be generated by using a combination of characters, the values of a given row in a document, and custom generators. Row substitutions are done by wrapping the column name in "%" and custom generators are wrapped in "#". Below is an example of a key generation expression.

      Given the CSV dataset:

      fname,age
      alice,40
      barry,36

      Key Generator Expression:

      --generate-key key::%fname%::#MONO_INCR#

      The following keys would be generated:

      key::alice::1
      key::barry::2

      In the example above we generate a key using the value in each row of the fname column and a custom generator. To specify that we want to substitute the value in the fname column we put the name of the column fname between two percent signs. This is an example of field substitution and it allows the ability to build keys out of data that is already in the dataset.

      This example also contains a generator function MONO_INCR which will increment by 1 each time the key generator is called. Since this is the first time this key generator was executed it returns 1. If we executed the key generator again it would return 2 and so on. The starting value of the MONO_INCR generator is 1 by default, but it can be changed by specifying a number in brackets after the MONO_INCR generator name. To start generating monotonically incrementing values starting at 100 for example, the generator MONO_INCR[100] would be specified. The cbimport command current contains a monotonic increment generator (MONO_INCR) and a UUID generator (UUID).

      Any text that isn’t wrapped in "%" or "#" is static text and will be in the result of all generated keys. If a key needs to contain a "%" or "#" in static text then they need to be escaped by providing a double "%" or "#" (ex. "%%" or "##"). The delimiter characters can be changed to avoid having to escape them by using the --field-delimiter and --generator-delimiter flags.

      If a key cannot be generated because the field specified in the key generator is not present in the document then the key will be skipped. To see a list of document that were not imported due to failed key generation users can specify the --errors-log <path> parameter to dump a list of all documents that could not be imported to a file.

      SCOPE/COLLECTION PARSER

      Scope/collection parsers are used in order to determine which scope/collection to insert documents into. There are currently two supported parsers text/field.

      Given the CSV dataset:

       product, stock, type, subtype
      apple, 100, produce, fruit

      Scope/collection expression:

      --scope-collection-exp %type%.%subtype%

      The row would be inserted into the 'fruit' collection inside the 'produce' scope.

      Given the CSV dataset:

      fname,age
      alice,40
      barry,36

      Scope/collection expression:

      --scope-collection-exp uk.manchester

      In this case, no fields in the row will be used to determine the scope or collection; all the rows would be inserted into the 'manchester' collection inside the 'uk' scope.

      There is nothing stopping the mixture of text/field expressions the following are all valid expressions.

      uk.%city%
      uk.%city%-5
      %country%.%city%::%town%

      EXAMPLES

      In the examples below we will show examples for importing data from the files below.

      ./data/people.csv
      
        fname,age
        alice,40
        barry,36
      
      ./data/people.tsv
      
        fname  age
        alice  40
        barry  36

      To import data from /data/people.csv using a key containing the fname column and utilizing 4 threads the following command can be run.

      $ cbimport csv -c couchbase://127.0.0.1 -u Administrator -p password \
       -b default -d file:///data/people.csv -g key::%fname% -t 4

      To import data from /data/people.tsv using a key containing the fname column and the UUID generator the following command would be run.

      $ cbimport csv -c couchbase://127.0.0.1 -u Administrator -p password \
       -b default -d file:///data/people.tsv --field-separator $_\t_ \
       -g key::%fname%::#UUID# -t 4

      To import data from /data/list.csv using a key containing the "name" column and then a unique id separated by a # we could use the --generator-delimiter flag to avoid escaping the # sign. An example would be:

      $ cbimport csv -c couchbase://127.0.0.1 -u Administrator -p password \
       -b default -d file:///data/list.csv --generator-delimiter '£' \
      -g key::%name%#£UUID£ -t 4

      If the dataset in not available on the local machine where the command is run, but is available via an HTTP URL we can still import the data using cbimport. If we assume that the data is located at http://data.org/people.csv then we can import the data with the command below.

      $ cbimport csv -c couchbase://127.0.0.1 -u Administrator -p password \
       -b default -d http://data.org/people.csv -g key::%fname%::#UUID# -t 4

      If the CSV dataset contains information which would allow importing into scopes/collections then an command like the one below could be used.

      product, stock, type, subtype
      apple, 100, produce, fruit
      $ cbimport csv -c couchbase://127.0.0.1 -u Administrator -p password
       -b default -d file://data/list.csv -g %product%
       --scope-collection-exp %type%.%subtype%

      This command would place the row into the fruit collection inside the produce scope.

      DISCUSSION

      The cbimport-csv command is used to quickly import data from various files containing CSV, TSV or other separated format data. While importing CSV the cbimport command only utilizes a single reader. As a result importing large dataset may benefit from being partitioned into multiple files and running a separate cbimport process on each file.

      ENVIRONMENT AND CONFIGURATION VARIABLES

      CB_CLUSTER

      Specifies the hostname of the Couchbase cluster to connect to. If the hostname is supplied as a command line argument then this value is overridden.

      CB_USERNAME

      Specifies the username for authentication to a Couchbase cluster. If the username is supplied as a command line argument then this value is overridden.

      CB_PASSWORD

      Specifies the password for authentication to a Couchbase cluster. If the password is supplied as a command line argument then this value is overridden.

      CB_CLIENT_CERT

      The path to a client certificate used to authenticate when connecting to a cluster. May be supplied with CB_CLIENT_KEY as an alternative to the CB_USERNAME and CB_PASSWORD variables. See the CERTIFICATE AUTHENTICATION section for more information.

      CB_CLIENT_CERT_PASSWORD

      The password for the certificate provided to the CB_CLIENT_CERT variable, when using this variable, the certificate/key pair is expected to be in the PKCS#12 format. See the CERTIFICATE AUTHENTICATION section for more information.

      CB_CLIENT_KEY

      The path to the client private key whose public key is contained in the certificate provided to the CB_CLIENT_CERT variable. May be supplied with CB_CLIENT_CERT as an alternative to the CB_USERNAME and CB_PASSWORD variables. See the CERTIFICATE AUTHENTICATION section for more information.

      CB_CLIENT_KEY_PASSWORD

      The password for the key provided to the CB_CLIENT_KEY variable, when using this variable, the key is expected to be in the PKCS#8 format. See the CERTIFICATE AUTHENTICATION section for more information.

      SEE ALSO

      CBIMPORT

      Part of the cbimport suite