FME User Conference 2017

Helping a UK-based French 2017 election campaign using nationbuilder.com, the Liste Electorale Consulaire required four address columns to be normalised for upload. A previous FME WT presentation, to scrape web data and rank music playlists, inspired a similar usage against non-spatial data. Using three transformers, we got the first matches of address strings, assigned next address to empty address strings, and used country name as constant last string to built normalised string sets backward from it. A fourth transformer allowed to compare lists for QC and ongoing data uploads.