Fix: Remove hardcoded keys in Analysis

[dhruvan2006]

Removed hard coded keys in the Analysis scraper functions to protect against Yahoo changing the keys in the future.

Changes

End result is more or else the same except:

• Few column names have changed
• Index has names

Earnings Estimate

Before:

After:

Revenue Estimate

Before:

After:

Earnings History

Before:

After:

EPS Trend

Before:

After:

EPS Revisions

Before:

After:

Growth Estimates

Before:

After:

Closes #2188

[R5dan]

Surely it would be quicker and simpler to simply store the keys as lower case strings

[dhruvan2006]

Surely it would be quicker and simpler to simply store the keys as lower case strings

I am aiming to keep backwards compatibility

[ValueRaider]

And what if Yahoo changes another key? Better if most hardcodings go - just unpack the json and use what Yahoo provides as much as possible. Try this:

dd = []
for item in self._earnings_trend:
    ddi = {'period': item['period']}
    for k,v in item['epsRevisions'].items():
        if not isinstance(v, dict) or len(v) == 0:
            continue
        ddi[k] = v['raw']
    dd.append(ddi)
self._eps_revisions = pd.DataFrame(dd).set_index('period')

[R5dan]

And what if Yahoo changes another key?

Yes, especially as they already have added one: 90 days

[ValueRaider]

Do you mind refactoring the other fetch functions?

[dhruvan2006]

Do you mind refactoring the other fetch functions?

Done. Edited PR description to show changes to end result