[OSM-ja] reconciling MLIT Shapefile with Japan Post CSV

[田渕 直]

Hi Tom,

The common ID scheme is called
   "全国地方公共団体コード/Zenkoku Chiho-Kokyou Dantai Codes,"
literally meaning "The National Code System for Local Public Bodies."
I don't think there's an official English name for it, but anyway
it's defined as a Japan's industry standard (JIS X 0402).

For the columns of the postal-code csv, let's take the 1st line
as a reference example:
   > 01101,"060 ","0600042","ﾎﾂｶｲﾄﾞｳ","ｻﾂﾎﾟﾛｼﾁﾕｳｵｳｸ","ｵｵﾄﾞｵﾘﾆｼ(1-19ﾁﾖｳ 
ﾒ)","北海道","札幌市中央区","大通西（１〜１９丁目）",1,0,1,0,0,0

   The column 1: "01101" is the district's ID, as you mentioned,
            2-3: "060 ","0600042" are the postal codes for the district,
                 the former is now outdated short form,
                 the latter is canonical,
            7-9: "北海道","札幌市中央区","大通西（１〜１９丁目）" are the
                 prefecture, city and district, street, respectively,
            4-6: "ﾎﾂｶｲﾄﾞｳ","ｻﾂﾎﾟﾛｼﾁﾕｳｵｳｸ","ｵｵﾄﾞｵﾘﾆｼ(1-19ﾁﾖｳﾒ)" are the
                 pronouciation for the pref./city-and-dist./st. resp.

Let me omit the remainder: they are rather involved "flags" and won't be
very important for OSM use.

Lastly, the distinction bw. "oogaki" and "kogaki" wouldn't be serious,
but I believe "kogaki" is a bit human- and machine-friendly;
"kogaki" can easily be canonicalized into "oogaki", but not vice versa.

Hope it helps (and excuse me for a long post.)

Regards,

Noashi Tabuchi (tabuchi_naoshi ＠ trek.co.jp)

(2014/10/12 1:25), Tom Lee wrote:
> Please excuse me for posting to this list in English. I am not a
> Japanese speaker, but have recently been examining the availability of
> Japanese open data for geocoding addresses. This process has been
> difficult due to my unfamiliarity with the language and character set.
> My hope is to share a summary of my work here and receive criticism from
> experts.
>
> My code can be found here:
>
> https://github.com/sbma44/japan_geodata_research
>
> My conclusions and the narrative behind them are present in the
> README.md file, which should display at the bottom of the page when you
> follow that link.
>
> In short: I was surprised to find that Japan Post and MLIT seem to use a
> shared ID scheme, allowing postal codes to be mapped, many-to-one, to
> the boundaries of administrative districts offered in a country-wide
> Shapefile by MLIT.
>
> It might be that this is already common knowledge -- if so, I would be
> glad to know that, too! If not, perhaps this will prove useful to others.
>
> I would be particularly glad to hear suggestions from Japanese speakers
> as to what each column in the postal code CSV should be properly called
> in English and/or in the existing OSM tagging scheme for Japan; and
> whether the difference between the "oogaki" and "kogaki" files offered
> by Japan Post represent a concern.
>
> Please do not hesitate to ask questions or point out problems with my
> analysis.
>
> Tom Lee
>
>
> _______________________________________________
> Talk-ja mailing list
> Talk-ja ＠ openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-ja
>