Conversion of INFORM/NINFOM into Information Type

[Christian-Shom]

This question is mainly for conversion tools providers. In order to optimize S-101 encoding and avoid time-consuming manual editing, could it be possible that a converter:

• identifies all the S-57 features (possibly with the same acronym) in a product or database that have exactly the same (not empty) INFORM/NINFOM attributes;
• creates one S-101 Information type based on S-57 INFORM/NINFOM;
• creates the association between the Information Type and all related Geo features;
• remove INFORM/NINFOM from the S-101 features.

[tdepuyt]

The short answer is yes, it can logically be done during conversion.

• Why did you say "possibly with the same acronym"? Does it matter if they are the same or not? The simplest approach would be to evaluate the INFORM/NINFOM attribute only and not consider the feature. I can't think of any exceptions to that but if there are, those would need to be identified and clearly documented.
• Formatted INFORM text for feature conversion would have to be excluded from the search criteria as well.
• If not supported, there's no requirement that you MUST create a single Information Type, just that it is recommended (should). You could do the manual editing over time or not at all as well but I understand the desire for optimizing your information attributes.

[TDYCARHugh]

In the CARIS conversion tools a new Information Type is only created if it is different than what was already created.

[kusala9]

I think this should be the default behaviour (to share common INFORM values across different features). I seem to remember when we made the case for including information as an inline attribute in addition to allowing associations with Nautical information it was because information types were created to facilitate sharing. So, conversion should really support that by creating shared information types where possible - maybe this is an "optimisation" of the conversion process but it supports the original justification for allowing both information attributes and information types. S-65 Annex B could recommend this approach (while noting that it isn't invalid as an encoding to do it in other ways). I suspect S-101 validation may pick up duplicates which aren't shared as a validation warning....