Performance fixes (#37, #47, #51)

[LemonBoy]

Cherry pick #52 too

[kenorb]

CI fail with:

Traceback (most recent call last):
  File "/home/travis/build/FX31337/FX-BT-Scripts/convert_csv_to_mt.py", line 426, in <module>
    HST574(CSV(args.inputFile), outputPath, timeframe, symbol)
  File "/home/travis/build/FX31337/FX-BT-Scripts/convert_csv_to_mt.py", line 51, in __init__
    self._map_obj = mmap.mmap(self.path.fileno(), 0, prot=mmap.PROT_READ)
ValueError: cannot mmap an empty file

[kenorb]

@LemonBoy There was missing xargs variable. Fixed in b9b469d3eace3784efd06f4a7e07fa887b30d4df.

[kenorb]

I think the tests are fixed now in 119289959.

[kenorb]

You can try now to rebase with master.

[LemonBoy]

ehh, this is your branch, feel free to rebase it yourself

[kenorb]

Master is now passing fine, ok. I did rebase, but CI fails.

[kenorb]

Current CI results:

+python3 convert_csv_to_mt.py --help
Traceback (most recent call last):
  File "convert_csv_to_mt.py", line 68, in <module>
    class CSV(Input):
NameError: name 'Input' is not defined
make: *** [test-help] Error 1

You can checkout that branch in your fork and send another PR against master when it's fixed.

[LemonBoy]

sigh, the two speed improving PR are actually incompatible, hence why I recommended to run travis for each PR in it's own branch

[kenorb]

As part of your PR, you can try to revert the incompatible code, so once CI pass the tests, I can re-test on local again.

[kenorb]

@LemonBoy Is this branch still useful, or can be lost?

[LemonBoy]

You can safely delete this