split_piped_string - ObjectVision/GeoDMS GitHub Wiki

String functions split_piped_string

The split_piped_string function splits strings by the pipe character (|) into separate elements.

split_piped_string(strings: E->String) -> unit F with:
    - attribute<String> (F): the split parts
    - attribute<E> org_rel (F): relation back to original strings

Splits each input string at pipe (|) characters, creating a new element for each part. The result is a new domain containing all the split parts, with a relation back to the original string.

This is commonly used for:

• Parsing delimited data files
• Extracting multiple values from a single field
• Processing pipe-separated lists

Returns a container with:

• A new unit F (the result domain, one element per split part)
• The split string values (F->String)
• org_rel: relation from F back to E (which original string each part came from)

Time complexity: O(n × L) where n is the number of strings and L is the average string length.

Memory usage scales with the total number of split parts produced.

• Empty strings produce no output elements
• Consecutive pipes (||) produce empty string elements
• Leading/trailing pipes produce empty string elements

unit<uint32> Records: nrofrows = 3;
attribute<String> categories (Records) := union_data(Records,
    'red|green|blue',
    'apple|orange',
    'single'
);

container split := split_piped_string(categories);
// split unit has 6 elements: 'red', 'green', 'blue', 'apple', 'orange', 'single'
// split/org_rel = {0, 0, 0, 1, 1, 2}

// Count categories per record
attribute<uint32> category_count (Records) := pcount(split/org_rel);
// category_count = {3, 2, 1}

• ReadLines - splits by line endings
• split_multi_linestring - splits geometry sequences
• String functions
• asList - inverse operation (combines strings with separator)

7.100