* german: fix case in currency names
Also, add more tests for german language.
* german: default to amount of euros, not cents
If a non-float number is provided, interpret it as a quantity in euros,
not cents.
* german: handle "ein Euro" vs "eins Euro" exception
* german: prefer "einhundert" over "hundert"
"hundert" is usually colloquial (even if widely used).
Exception: "eintausendste" and "hundertste" are special cases, here
"hundertste" and "tausendste" are usually preferred (see Duden).
* Add support for Serbian language
* Fix the billion strings for short scale
Used the long scale name instead of the short one. Simplified the scale and the ones geneder definition.
* Adds Serbian to README
Punt string handling to python Decimal object, this correctly represents both
integers and floats (except with regards to trailing zeros)
Change command line tests to reflect handling of ints
* Refactor tox and travis config files
Separates flake8 and isort into separate tox tasks
Tie flake8 and isort to specific python version so they only run once on Travis
Add python3.7 to tested python versions
* Requires newer Xenial base
* Running Xenial on Travis requires sudo
Move coveralls submission to travis config to avoid triggering during local runs.
* Require all tests to be run
Clean up test generation code form test_fi.py
Perform coverage measurement of tests
Fail tox if coverage of tests directory isn't at 100%
* Further Tox refactoring
Measure project coverage explicitly instead of omitting non-project files from reporting
Add branch measurement to coverage
Only report to coveralls if a .coverage result file exists
Make dependencies for each testenv explict and only include requried deps
* Small readme update to trigger travis
Add overly brief message about installing and running tox
* Fix unescaped flake8 violations
Unescaped \w -> \\w and \e to \\e in regex patterns found with updated flake8
* add support for Portuguese (Portugal) (PT)
* FIX ADD SUPPORT TO PORTUGUESE (PORTUGAL) [PT]
removed venv/ from .gitignore.
correction from "milião" to "milhão" done on merge function.
removed hack for negword, creating a backup of the original one.
* add support PT (Portugal) (PT)
* add support PT (Portugal) (PT)
* add support PT (Portugal) (PT)
* add support PT (Portugal) (PT)
new tests
* increase test coverage - german ordinal_num
* push coverage of german to 100%
* fix line width
* travis ci fix
* remove underscore between numbers due to backwords comp
* removed coverage file
Also update the readme and fix error, on the README.rst file we say tha
code language to use is vn, but to make it work you have to use vi_VN.
Let's refactor this to use the proper iso code `vi` in all places.
* Remove some dead code.
* Overloadable _cents_terse
* One setup should be enough
* Move gen_high_numwords up
* Cards are optional.
* Line length
* DRY set_hight_numwords for EU base.
* Remove unused error message. It has to be an exception anyway.
Negative amounts were not working (when no currency is used), because
`get_digits` method does not expect `-` sign, which crashes conversion.
To avoid that, we split minus sign from number string and prepare its
word to be used with amount words.
closes: #184
* New maintainer 🚀
Since @erozqba no longer work for Savoir-faire Linux, the maintainer has changed.
* Bump version from 0.5.6 -> 0.5.7
* Update CHANGE.rst to publish a new release.
* [FIX] move language specific info to the wiki
The informations about Finnish and Japanese can be found in the [wiki](https://github.com/savoirfairelinux/num2words.wiki.git)
The goal of this commit is to keep the readme short.
* [FIX] add wiki mention in the readme.rst
FI: Add support for Finnish
This is a squash of the following commits :
* The recommendation by Kotus was optional. The code didn't work anyway
because million is 10**6, not 10**9, and the recommendation only applies
to round millions, billions and so on where the multiplier is 10 or
greater.
* FI: currency: use inheritance, add FIM and USD
* FI: add some common currencies
* FI: add inflections
* FI: inflections: small fixes
* FI: fix ordinal 200
* FI: fix comitative
* FI: actually fix ordinal 200
* FI: refactoring
* FI: fix tens
* FI: fix plural for tens and higher
* FI: toinen --> kahdes for numbers > 200
* FI: use collection for options
* FI: add tests
* FI: add minced tests
* FI: trim tests
* FI: finishing touches
* FI: improve cardinal_float, ordinal, add tests
* FI: add documentation to README.rst
* Create lang_CZ.py
* Added czech language to init .
* Update __init__.py
* Update lang_CZ.py
* Added translation to czech language
* Fixed thousands translation
* Added czech language to readme
* Changed pointword in czech translation.
* Create test_cz.py
* Fixed names for very big numbers.
* Fixed currency spelling.
* Created tests for czech language.
* Fixed: Shortened length of some long lines.
* Shortened one more line of code.
* Fixed tests
* Fixed flake-8.
* Changed name of the test class
* Add tests
Starting work on #141, which points out that 2017 should come out
"twenty seventeen"; it currently comes out "two thousand seventeen".
* Handle English years properly.
Resolves#141.
* Whoops -- didn't PEP-8 my tests
* Handle years MORE properly...with suffixes!
- Rendering years between 10-99 was bugged (e.g. "zero fifty").
Fixed that by thinking outside the high/low text box I had made.
- Added suffixes: default none if positive, 'BC' if negative.
User-overrideable.
* add Thai
* change splitby3 to splitbyx
* change lang_th to use function from currency
* make Num2Word_TH inherit from Num2Word_Base
* comment out test failed in 2.7 env
* fix python2.7 error
* add USD EUR for Thai
* pep8 fix
* added Thai
`n % 1 != 0` is not a valid test for float:
In []: 1.1 % 1 == 0
Out[]: False
but:
In []: 1.0 % 1 == 0
Out[]: True
hence it's really necessary to explicitly test for type in this case.
