Publication date potentially corrected from the consolidation service seems to be lost in a limbo

[lfoppiano]

Related to #759.
I tried to solve the problem by enabling the consolidation, however, the correct publication date was not retrieved correctly.

The BiblioItem object from the consolidation service had:

• publicationDate = null
• normalisedPublicationDate with the correct value

(I wonder if this is the correct behaviour)

however, in the BiblioItem, after the correct(...) method is invoked, the normalisedPublication date is overridden with the first raw publication date:

// normalization of dates
                if (resHeader != null) {
                    if (resHeader.getPublicationDate() != null) {
                        List<Date> dates = parsers.getDateParser().processing(resHeader.getPublicationDate());
                        // most basic heuristic, we take the first date - to be
                        // revised...
                        if (dates != null) {
                            if (dates.size() > 0) {
                                resHeader.setNormalizedPublicationDate(dates.get(0));
                            }
                        }
                    }

This will inevitably lose the good normalised date from the consolidation service. I'm not sure what is the correct behaviour, I supply this issue mostly for information as I assume there many different use cases 😉

[kermitt2]

Good catch Luca !!
It's a bug yes, we should not normalize the publication date if (resHeader.getNormalizedPublicationDate() != null) already.

(I wonder if this is the correct behaviour)

yes it's the correct behavior, crossref record has only a normalized publication date (no "raw" publication date).