Skip to content

OPTIONS

The OPTIONS statement configures optional features of the text intelligence engine that are used to best match the characteristics of a given project. The statement can be placed in any source files, however, it is recommended to place it in the config.cr file.

The syntax is:

OPTIONS
{
    option[,
    option,
    ...]
}

where option can be one of the followings:

  • FIXED_SCORE
  • STATIC_SCORE
  • CHILD_TO_FATHER

  • FIXED_SEQUENCE_MODE

    The first three options affect the categorization scoring mechanism, while the last one affects the count of distances between the elements in a sentence at a token level, even though there are atom-based attributes after the positional sequences that should guarantee an atom-based count.

  • (Advanced use) EXPAND_SYNCON_CACHE=false: disables the cache used to store the expansion of ancestor syncons. The evaluation of every ANCESTOR attribute requires the creation of an in-memory list of descendant syncons. By default, this list is created and cached at compile time, then used at run-time. When this option is set, the list is created at run-time and only when really needed.

  • (Advanced use) EXPAND_SYNCON_LIMIT=value: the value in this option is a positive integer number between 10000 (default) and 10000000. It influences the creation of the list of descendant syncons when ANCESTOR attributes are evaluated (see the option above). If the number of descendants exceeds this limit, the cache will not be created at compile time.
  • (Advanced use) PRECALC_PATTERN: if your Studio project makes an extensive use of the PATTERN attribute and the document analysis is slow, use this option to make it faster.
  • (Advanced use) DISABLE_SEGMENT_MERGE: in case of overlapping instances of a single segment, a single bigger segment occurrence is created that merges all its instances. This option disables the segment merge creating as many segment occurrences as its instances. More details in the dedicated page.