sos-import-csv 

Date:: May 14 2019

NAME 

sos-import-csv - Import data from a comma-separated text file into a SOS Container

SYNOPSIS 

sos-import-csv –path SOS-PATH –csv CSV-FILE–schema SCHEMA-FILE –map MAP-FILE [ –sep SEP-STR –status ]

The sos-import-csv commannd parses CSV text data and imports this data into a SOS container. There are three files read by the sos-import-csv command: the CSV text file containing the data to import, a schema-file that defines the type of object created for each line of the CSV text file, and a map-file that specifies how object attribute values are obtained from columns in the CSV file.

OPTIONS 

–path SOS-PATH: Specifies the path to the SOS container. The container must already exist.
–schema SCHEMA: The name of the SOS schema used to create objects. The schema must exist in the container.
–csv CSV-FILE: The path to the CSV text file containing the data to import.
–map MAP-FILE: The path to a JSON formatted text file specifying how columns in the CSV file map to attributes in the object schema.
–sep SEP-STR: A string specifying the characters that will be interpretted as column separators. By default this is a comma (“,”).
–status: If the status option is present, import progress status will be provided as the CSV file is processed.

CSV file format 

The CSV text file must contain newline terminated lines of text. Each line may contain multiple columns separated by the separator character. By default, the character separating each column is a comma, however, this can be overriden with the –sep command line option.

Map File Format 

The map file contains a single list object where each element in the list is an action-specification. An action-specification is an object that tells the sos-import-cmd what to do with each line of the input CSV file.

Each action-specification contains two attributes, a target, and a source. The target specifies the attribute in the object to be assigned a value. The value of the target attribute can be a string, in which it is the name of the object attribute from the schema, or an integer in which case is the attribute id.

The source-specification defines the value that will be assigned to the target attribute and is specified as a JSON object. The JSON object contains only one of four possible attribute names: “value”, “column”, “list”, and “range”.

If the attribute named “value” is present, the value of the “value” attribute is assigned to the target attribute. This is useful for assigning values to object attributes that are not contained in the CSV file.

If the attribute named “column” is present, the value of the “column” attribute is an ordinal column number in the CSV file. In this case, the value contained in that column of the text file is cast to the target attribute type and assigned to the target attribute. If a conversion from the column text string to the target attribute type is not possible, an error is given and processing continues.

If the attribute named “list” is present, the target attribute must be an array, and the value of the “list” attribute is a JSON list of columns that will be assigned to each element in the target array beginning with 0. Text to target attribute type conversion is performed for each column specifed in the list.

If the attribute named “range” is present, the target attribute must be an array and the value of the “range” attribute is a list of two column numbers. All values in the range from the first column to the last column are assigned to the elements of the target list attribute. The range is inclusive.Text to target attribute type conversion is performed for each column in the range.

[
    { "target" : "component_id", "source" : { "value" : 10000 } },
    { "target" : 0, "source" : { "column" : 0 } },
    { "target" : 2, "source" : { "list" : [ 1, 3, 5 } },
    { "target" : 3, "source" : { "range" : [ 1, 5 } }
]