This is a failed attempt at parsing the E/CONF.105/13. First I migrated the pdf to text to text with pdftotext -layout. I proceeded to do minimal processing on the PDF with vim. Then I thought I could right a parser on it to get it into a database.
This effort was abandon because the Unicode formatting is very poor, and some of it lacks Unicode entirely; some of it is just text-pictures. This throws off the parser (a problem I may be able to fix and Q/A), but renders the text in the screen shot impossible.
- Sri Lanka, Myanmar and Comoros are examples of screen-shot text.
- Albania / Shqipëri has no text for the first language, and the "short name is actually the language itself"
- All of the names are in ASCII with the exception of the "CÔTE D’IVOIRE"
- There is no way to tell if a column extends down a row, which column it was the first or second.. All we know it that only one column extended down so the other column is empty.
You can find more information here,
* https://opendata.stackexchange.com/questions/13693/does-the-ungegn-release-their-country-names-localized-in-a-format-thats-not-a-p
* https://opendata.stackexchange.com/questions/13692/where-does-iso-3166-get-the-names-and-translations-of-the-countries
* https://dba.stackexchange.com/questions/225996/unicode-storage-of-u202b-rle-and-u202c-pde-in-a-unicode-aware-database
* [UNGEGN](https://unstats.un.org/unsd/geoinfo/UNGEGN/)
* [Working Group on Country Names](https://unstats.un.org/unsd/geoinfo/UNGEGN/wg1.html)