dump does not keep string type for scientific form text

[chalmagr]

When processing a yaml document that has a string value that could be interpreted as a number in scientific form, the dumped value is not quoted, causing problems when composing multiple documents and rendering the outcome as such

Example:

import yaml

document = """
list:
- code: '6E01'
- code: '1234'
- code: ABCD
"""

loaded = yaml.load(document, yaml.BaseLoader)

print(loaded)

print(yaml.dump(loaded, default_flow_style=False))

print(yaml.dump(loaded, default_flow_style=True))

The above code generates the following output

{'list': [{'code': '6E01'}, {'code': '1234'}, {'code': 'ABCD'}]}

list:
- code: 6E01
- code: '1234'
- code: ABCD

{list: [{code: 6E01}, {code: '1234'}, {code: ABCD}]}

Which is incorrect, since now the first code in the list is interpreted as a number (60.0) instead of the string 6E01.

I am expecting this to be

{'list': [{'code': '6E01'}, {'code': '1234'}, {'code': 'ABCD'}]}

list:
- code: '6E01' << quoted
- code: '1234'
- code: ABCD

{list: [{code: '6E01'}, {code: '1234'}, {code: ABCD}]} << quoted '6E01'

[chalmagr]

@nitzmahone will you be able to review the PR that fixes this for any upcoming release? This is quite an annoying defect that causes us to re-process a file that is created by ansible to replace such values

[chalmagr]

Created PR #887 that proposes a fix for this issue