hirano-satoshi / price-level-calculator Goto Github PK

Proofreading is in progress with a contributor. Some corrections have been reflected and deployed.

Bouvet Island which is not used.

Although exchange rates are fetched, they are not used during conversion from Luncho to local currencies. Luncho calculation reflects only PPP values.

To reflect the latest exchange rate of a currency in a country, we need the exchange rate at the moment when International Comparison Program measured the price levels in the country.

In STATISTICAL APPENDIX of World Economic Outlook 2021, they wrote:

• Real effective exchange rates for the advanced economies are assumed to remain constant at their average levels measured during January 18, 2021–February 15, 2021. For 2021 and 2022 these assumptions imply average US dollar–special drawing right (SDR) conversion rates of 1.445 and 1.458,

So, we have to calculate the average exchange rate of the currency during the period above mentioned. And we can reflect the latest exchange rate against the average exchange rate.

We need a database. The current implementation runs on Google App Engine Standard (GAE) but it should run on any Docker container. If I use free Datastore on GAE, it won't run on other environment.

Currently Python and JS use different method names.

Found on the log.

We could use aurelia-meta but it seems lacking i18n.

Done

From USD to JPY for example.

Currently USD based exchange rates are stored in FixerExchangeRate as it is a dist, but it is not actually.

I wrote a local file version and it is commented out for new. And some test code is also commented out.

def test_server_api_error():
    #XXX if os.path.exists(conf.Last_Fixer_Exchange_File):
    #    os.remove(conf.Last_Fixer_Exchange_File)

• app
• aurelia
• fetch

"validation errors for LunchoData" occurs and requests fail with 500 error.

That occurs when requests.get() for a Forex data provider does not return response.ok.
The requests.get() in ppp_data.update() is called from main.init() which is called from post_worker_init() in configure_gunicorn.py.

Such failure may occur temporarily or permanently with the shutdown of the provider.

We have to save the last forex data to a file and load it if the get() fails.

• write a test
• lock global data
• save the last forex data
• load it
• log failure of the fetching of the forex data.
• test

• Website https://data.imf.org/?sk=4FFB52B2-3653-409A-B471-D47B46D904B5

• document https://data.imf.org/?sk=4FFB52B2-3653-409A-B471-D47B46D904B5&sId=1485878846789

• map https://www.imf.org/external/datamapper/PCPIEPCH@WEO/CAQ

• data https://data.imf.org/regular.aspx?key=61015892

• Harmonized Indexes https://data.imf.org/regular.aspx?key=61057839

  • Only EU and UK

The default routing policy of Aurelia uses fragment URL with hash (#) such as https://luncho-index.org/#/about.html. Such URL scheme is not indexed by Google.

Aurelia supports pushState to remove # from URLs.

Catch-all is needed on FastAPI to use the pushState.

It would be nice if the single page has a map of the country with many lunch images with Luncho prices.

• A location based image service API for lunch images
  • test
  • code
• Leaflet with OpenStreet map on the single page
• Login with Land ID
• The entry page