Japan Nationa Postal Code List in Roman
=======================================

Files
-----
* readme.txt: this file
* roman.txt: contains all records, roman characters only.
* roman2.txt: contains all records, roman, hiragana, kanji.


Base Data
---------
The database is based on ken_all.zip which Japan Post Service Co., Ltd. has been publishing for public.

You can download it from this direct link.

http://www.post.japanpost.jp/zipcode/dl/kogaki/zip/ken_all.zip

Japan Post Service Co., Ltd.'s English page is here. http://www.post.japanpost.jp/english/index.html


Roman conversion
----------------
Above data file has Japanese kataana value. We convert it into roman programaticaly.
Roman representation in Japanese has some ways. Kunrei, Hepburn etc.
We use modified Hepburn system of romanization.


Format conversion
-----------------
Original data file has some ugly specs.
For example, one town divided into two or three rows, because of character length.
In our conversion that is merged into one row.
There is four another specs, but its too difficult to explain in english for us.


File format
-----------
* First line is title
* Column seperator: Tab
* Record seperator: CR+LF
* Text encoding: UTF-8

* roman.txt Columns
  1. row number
  2. postal code
  3. prefecture code
  4. city code
  5. town
  6. city
  7. prefecture
  8. area

* roman2.txt Columns
   1. row number
   2. postal code
   3. prefecture code
   4. city code
   5. town
   6. city
   7. prefecture
   8. area
   9. town_kana
  10. city_kana
  11. prefecture_kana
  12. area_kana
  13. town_kanji
  14. city_kanji
  15. prefecture_kanji
  16. area_kanji


* Town column has some special values:
  1. "*" means, when there is not a mention.
  2. "-" means, in case that street numbers come after the city.

* Area division has some alternative ways.

“Japan National Postal Code List in Roman” への17件のフィードバック

1. why “Naka-ku, Nagoya-shi” instead of “Nagoya-shi, Naka-ku”?

   8:11 午後2011年8月4日

2. Hmm…American style. Is it wrong?

   9:41 午後2011年8月4日

3. For example, when sending a letter usually writes the first city in the case Aichi-ken, Nagoya-shi, Naka-ku.
   Province> City> district, in that order.

   3:32 午後2011年8月5日

4. It is also good such rules. I will take it under advisement.
   Thank you.

   3:12 午後2011年8月6日

5. I personally think the western format being used (“Naka-ku, Nagoya-shi”) is more helpfull, specially for those sending letters from outside Japan.

   12:05 午後2012年1月21日

6. Thank you. Currently I also think so.

   12:20 午後2012年1月21日

7. Hi,
   I couldn’t find geographic data in this database 🙁

   There is no latitude/longitude data for the entries??

   9:23 午後2012年2月22日

8. >> There is no latitude/longitude data for the entries??
   Unfotunately, Yes. You can also get geocoding data from Google Maps API.

   9:42 午後2012年2月22日

9. Good point. Thanks.

   10:25 午後2012年2月22日

10. Hello,
    I’ve been checking roman version and I have one question about consistency between some district names in the website (http://yubin.senmon.net/en) and the names in the database.
    There are some names in the web site with double vowels (‘aa’, ‘ee’, ‘ii’, ‘oo’, ‘uu’) that in the data file have lost one vowel. I’ll write just some examples (hiragana – website – data file)
    1) いいしぐん – Iishi-gun – Ishi-gun
    2) ひがしいばらきぐん – Higashiibaraki-gun – Higashibaraki-gun
    3) みなみまつうらぐん – Minamimatsuura-gun – Minamimatsura-gun
    4) しこくちゅうおうし – Shikokuchuuou-shi – Shikokuchuo-shi
    5) もおかし – Mooka-shi – Moka-shi
    So, I would like to know:
    – Why this difference exist?
    – Will be the data file updated with the website version?
    Thank you very much and, by the way, thanks for publishing this data. It’s very useful.

    11:10 午後2012年9月10日

11. Hello,

    – Why this difference exist?
    Japanese romanization has some different rules.
    Web site uses ‘KUNREI’ style. Database version uses ‘Hepburn’ style. This difference originates in the historical reason of our site.
    There are no technical reasons. But some cultural reasons. It’s a long story 🙂

    – Will be the data file updated with the website version?
    I am sorry. Currently we have no plans.
    But I personally think distributing KUNREI version database should be considered.

    Thank you for your help.

    8:01 午後2012年9月15日

12. Davi asked why addresses were not shown from big to small as is customary in Japan.

    In fact, the writing style for international mail was decided by the Bern Convention which Japan joined in 1877. Since then, mail addressed to (or within) Japan in Roman letters has always been done “western style” of small to big.

    Another comment was about using kunrei. Kunrei is fine for Japanese to make “roman” words, but Hepburn is by far preferable for use internationally. There are a few problematic romanizations in kunrei that make it “A Bad Idea” for use by non-Japanese. The troublesome bits in kunrei are mainly zya, zyu, zyo, hu, si, tu and ti.

    6:44 午前2013年8月8日

13. Would it be possible to see the coding for the Hepburn?

    11:02 午前2015年9月3日

14. Hello. Unfortunately we currently have no plan.
    But good news for you. Japan post has been published Hepburn romanized data on their site in Japanese web page!
    http://www.post.japanpost.jp/zipcode/dl/roman.html
    Unfortunately the data is a little outdated and descriptions are in Japanese.
    Try google translate!

    2:57 午後2015年9月3日

15. I mean you create your file following the Postal Code Hepburn procedure right?
    I also did the same but I can’t match it for the Aza.

    5:25 午後2015年9月3日

16. Hello,
    We convert Katakana into roman programaticaly. Hepburn representation in Japanese has some ways. We use modified (for our convinience) Hepburn system of romanization.So our conversion procedure is not *RIGHT*.

    If you need *RIGHT* Hepburn texts, you can use Japan Post’s published version. It may be *RIGHT*.

    On the other hand, We think our modified Hepburn system of romanization is reasonable on most situations.

    Regards.

    9:48 午前2015年9月4日

17. THANK YOU!

    8:31 午後2016年4月18日