Hand written ISO parser for Timedelta construction

[WillAyd]

• closes PERF: Hand Written Parser for Timedelta ISO 8601 Construction #19103
• tests added / passed
• passes git diff upstream/master -u -- "*.py" | flake8 --diff
• whatsnew entry

Running 16 total benchmarks (2 commits * 1 environments * 8 benchmarks) [ 0.00%] · For pandas commit hash 76475dd8: [ 0.00%] ·· Building for conda-py3.6-Cython-matplotlib-numexpr-numpy-openpyxl-pytables-pytest-scipy-sqlalchemy-xlrd-xlsxwriter-xlwt............................................................... [ 0.00%] ·· Benchmarking conda-py3.6-Cython-matplotlib-numexpr-numpy-openpyxl-pytables-pytest-scipy-sqlalchemy-xlrd-xlsxwriter-xlwt [ 6.25%] ··· Running timedelta.TimedeltaConstructor.time_from_components 15.4±0.4μs [ 12.50%] ··· Running timedelta.TimedeltaConstructor.time_from_datetime_timedelta 8.06±0.3μs [ 18.75%] ··· Running timedelta.TimedeltaConstructor.time_from_int 6.24±0.2μs [ 25.00%] ··· Running timedelta.TimedeltaConstructor.time_from_iso_format 12.1±0.4μs [ 31.25%] ··· Running timedelta.TimedeltaConstructor.time_from_missing 2.07±0.03μs [ 37.50%] ··· Running timedelta.TimedeltaConstructor.time_from_np_timedelta 5.06±0.09μs [ 43.75%] ··· Running timedelta.TimedeltaConstructor.time_from_string 6.52±0.1μs [ 50.00%] ··· Running timedelta.TimedeltaConstructor.time_from_unit 6.66±0.1μs [ 50.00%] · For pandas commit hash 8acdf801: [ 50.00%] ·· Building for conda-py3.6-Cython-matplotlib-numexpr-numpy-openpyxl-pytables-pytest-scipy-sqlalchemy-xlrd-xlsxwriter-xlwt... [ 50.00%] ·· Benchmarking conda-py3.6-Cython-matplotlib-numexpr-numpy-openpyxl-pytables-pytest-scipy-sqlalchemy-xlrd-xlsxwriter-xlwt [ 56.25%] ··· Running timedelta.TimedeltaConstructor.time_from_components 15.5±0.3μs [ 62.50%] ··· Running timedelta.TimedeltaConstructor.time_from_datetime_timedelta 7.90±0.2μs [ 68.75%] ··· Running timedelta.TimedeltaConstructor.time_from_int 6.11±0.2μs [ 75.00%] ··· Running timedelta.TimedeltaConstructor.time_from_iso_format 25.0±0.7μs [ 81.25%] ··· Running timedelta.TimedeltaConstructor.time_from_missing 2.27±0.06μs [ 87.50%] ··· Running timedelta.TimedeltaConstructor.time_from_np_timedelta 5.04±0.1μs [ 93.75%] ··· Running timedelta.TimedeltaConstructor.time_from_string 7.04±0.2μs [100.00%] ··· Running timedelta.TimedeltaConstructor.time_from_unit 6.51±0.1μs before after ratio [8acdf801] [76475dd8] - 25.0±0.7μs 12.1±0.4μs 0.48 timedelta.TimedeltaConstructor.time_from_iso_format SOME BENCHMARKS HAVE CHANGED SIGNIFICANTLY. 

[codecov]

Codecov Report

Merging #19191 into master will not change coverage.
The diff coverage is n/a.

@@ Coverage Diff @@ ## master #19191 +/- ## ======================================= Coverage 91.52% 91.52% ======================================= Files 147 147 Lines 48775 48775 ======================================= Hits 44639 44639 Misses 4136 4136

Flag	Coverage Δ
#multiple	89.89% <ø> (ø)	⬆️
#single	41.62% <ø> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 8acdf80...bfba8db. Read the comment docs.

[jreback]

lgtm. ping on green (even if the 3.5 conda build fails that's ok, its a travis issue).

[WillAyd]

@jreback all good save the 3.5 build you mentioned

[jreback]

thanks!