DEPR: correct locations to access public json/parser objects in depr message

[jorisvandenbossche]

Another one remaining is pd.parser.na_values

In [2]: pd.parser.na_values
/home/joris/miniconda3/envs/dev/bin/ipython:1: FutureWarning: pandas.parser.na_values is deprecated. 
Please use pandas.io.libparsers.na_values instead.
  #!/home/joris/miniconda3/envs/dev/bin/python
Out[2]: 
{numpy.bool_: 255,
 dtype('bool'): 255,
 numpy.uint64: 18446744073709551615,
 numpy.int64: -9223372036854775808,
 dtype('int16'): -32768,
 dtype('uint64'): 18446744073709551615,
 numpy.int8: -128,
 dtype('int32'): -2147483648,
 numpy.uint8: 255,
 dtype('int8'): -128,
 numpy.object_: nan,
 numpy.float64: nan,
 numpy.uint32: 4294967295,
 dtype('O'): nan,
 numpy.int32: -2147483648,
 dtype('uint32'): 4294967295,
 dtype('uint16'): 65535,
 dtype('uint8'): 255,
 dtype('int64'): -9223372036854775808,
 dtype('float64'): nan,
 numpy.uint16: 65535,
 numpy.int16: -32768}

I don't think we should refer to libparsers, but for now it is not imported in pandas.io.parsers as alternative. I could add it there, but the question is maybe whether this is actually useful for users? Maybe we should say that it will be removed without giving alternative?

[codecov]

Codecov Report

Merging #15909 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #15909      +/-   ##
==========================================
+ Coverage   90.95%   90.96%   +<.01%     
==========================================
  Files         145      145              
  Lines       49535    49535              
==========================================
+ Hits        45057    45058       +1     
+ Misses       4478     4477       -1

Flag	Coverage Δ
#multiple	88.72% <100%> (ø)	⬆️
#single	40.62% <100%> (ø)	⬆️

Impacted Files	Coverage Δ
pandas/__init__.py	92.68% <100%> (ø)	⬆️
pandas/core/common.py	91.03% <0%> (+0.34%)	⬆️

---

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 e4e87ec...49d032b. Read the comment docs.

[codecov]

Codecov Report

❗ No coverage uploaded for pull request base (master@88bed54). Click here to learn what that means.
The diff coverage is 100%.

@@            Coverage Diff            @@
##             master   #15909   +/-   ##
=========================================
  Coverage          ?   90.96%           
=========================================
  Files             ?      145           
  Lines             ?    49475           
  Branches          ?        0           
=========================================
  Hits              ?    45007           
  Misses            ?     4468           
  Partials          ?        0

Flag	Coverage Δ
#multiple	88.72% <100%> (?)
#single	40.68% <100%> (?)

Impacted Files	Coverage Δ
pandas/util/depr_module.py	80.48% <ø> (ø)
pandas/tslib.py	100% <ø> (ø)
pandas/__init__.py	92.68% <100%> (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 88bed54...01f8c38. Read the comment docs.

[jreback]

hmm, I don't remember why I added compat for this. let me have a look.

[jreback]

yeah I think we should just remove na_values entirely. any objections? (as I said I don't remember why I had added this).

[jorisvandenbossche]

I don't know why you would need na_values, so OK for me (@gfyoung ?)

[jorisvandenbossche]

Added removal of na_values and correct redirect for CParserError

[jreback]

you might be able to remove the deprmodto argument entirely (which I added to handle moving modules).

[gfyoung]

@jorisvandenbossche :

I don't know why you would need na_values, so OK for me

Makes sense to me. Hopefully it doesn't break any tests.

[jorisvandenbossche]

@gfyoung to be clear, it is not removed internally, only the public exposure is deprecated. So it is about whether external users would want to use it

[gfyoung]

@jorisvandenbossche :

to be clear, it is not removed internally

Yes, absolutely! I meant that hopefully no tests were accessing this variable externally.

[jreback]

ok, used the script to rebase you locally and push up. seemed to work.

[jreback]

thanks!

[jorisvandenbossche]

ok, used the script to rebase you locally and push up. seemed to work.

Great!