[LNT] Python 3 support: adapt to renaming of raw_input to input

[thopre]

Summary

Replace calls to raw_input by calls to input() which is not evaluated as
an expression in Python3. Import input from builtins to maintain
current behavior on Python2. This was produced by running
futurize's stage2 lib2to3.fixes.fix_raw_input.

[hubert.reinterpretcast]

I don't see how this maintains compatibility with Python 2.

[hubert.reinterpretcast]

I believe adding this to the files should work:

try:
    from builtins import input as raw_input
except ImportError:
    pass

[thopre]

Import input from builtins for input to behave like raw_input in Python 2

[thopre]

In D67818#1677598, @hubert.reinterpretcast wrote:

I believe adding this to the files should work:

try:
    from builtins import input as raw_input
except ImportError:
    pass

The import from builtins without renaming ought to be sufficient. It basically means input on Python2 behaves like raw_input. It's weird that 2to3's fix_raw_input did this transformation without adding the import. I'm glad you were more careful evaluating the change than me.

[hubert.reinterpretcast]

lnt/external/stats/stats.py

227

This is dependent on the "future" package. I guess the update to the package dependency list needs to move earlier in patch set to this patch?

[thopre]

Add dependency on future

[hubert.reinterpretcast]

LGTM.

lnt/external/stats/stats.py

 ## 11/12/98 ... added functionality to amean, aharmonicmean, ageometricmean
 ##              added 'asum' function (added functionality to N.add.reduce)
 ##              fixed both moment and amoment (two errors)
 ##              changed name of skewness and askewness to skew and askew
 ##              fixed (a)histogram (which sometimes counted points <lowerlimit)
 
 from __future__ import absolute_import
 from __future__ import print_function
+from builtins import input
 from . import pstat        # required 3rd party module
 import math, string, copy  # required python modules
 from types import *
 
 __version__ = 0.6
 
 ############# DISPATCH CODE ##############
 

 appropriated statistic for paired group data.
 
 Usage:   lpaired(x,y)
 Returns: appropriate statistic name, value, and probability
 """
     samples = ''
     while samples not in ['i','r','I','R','c','C']:
         print('\nIndependent or related samples, or correlation (i,r,c): ', end=' ')
-        samples = raw_input()
+        samples = input()
 
     if samples in ['i','I','r','R']:
         print('\nComparing variances ...', end=' ')
 # USE O'BRIEN'S TEST FOR HOMOGENEITY OF VARIANCE, Maxwell & delaney, p.112
         r = obrientransform(x,y)
         f,p = F_oneway(pstat.colex(r,0),pstat.colex(r,1))
         if p<0.05:
             vartype='unequal, p='+str(round(p,4))
# USE O'BRIEN'S TEST FOR HOMOGENEITY OF VARIANCE, Maxwell & delaney, p.112
                 print('\nRelated samples t-test:  ', round(t, 4), round(p, 4))
             else:
                 t,p = ranksums(x,y)
                 print('\nWilcoxon T-test (NONparametric):  ', round(t, 4), round(p, 4))
     else:  # CORRELATION ANALYSIS
         corrtype = ''
         while corrtype not in ['c','C','r','R','d','D']:
             print('\nIs the data Continuous, Ranked, or Dichotomous (c,r,d): ', end=' ')
-            corrtype = raw_input()
+            corrtype = input()
         if corrtype in ['c','C']:
             m,b,r,p,see = linregress(x,y)
             print('\nLinear regression for continuous variables ...')
             lol = [['Slope','Intercept','r','Prob','SEestimate'],[round(m,4),round(b,4),round(r,4),round(p,4),round(see,4)]]
             pstat.printcc(lol)
         elif corrtype in ['r','R']:
             r,p = spearmanr(x,y)
             print('\nCorrelation for ranked variables ...')

 appropriated statistic for paired group data.
 
 Usage:   apaired(x,y)     x,y = the two arrays of values to be compared
 Returns: appropriate statistic name, value, and probability
 """
     samples = ''
     while samples not in ['i','r','I','R','c','C']:
         print('\nIndependent or related samples, or correlation (i,r,c): ', end=' ')
-        samples = raw_input()
+        samples = input()
 
     if samples in ['i','I','r','R']:
         print('\nComparing variances ...', end=' ')
 # USE O'BRIEN'S TEST FOR HOMOGENEITY OF VARIANCE, Maxwell & delaney, p.112
         r = obrientransform(x,y)
         f,p = F_oneway(pstat.colex(r,0),pstat.colex(r,1))
         if p<0.05:
             vartype='unequal, p='+str(round(p,4))
# USE O'BRIEN'S TEST FOR HOMOGENEITY OF VARIANCE, Maxwell & delaney, p.112
                 print('\nRelated samples t-test:  ', round(t, 4), round(p, 4))
             else:
                 t,p = ranksums(x,y)
                 print('\nWilcoxon T-test (NONparametric):  ', round(t, 4), round(p, 4))
     else:  # CORRELATION ANALYSIS
         corrtype = ''
         while corrtype not in ['c','C','r','R','d','D']:
             print('\nIs the data Continuous, Ranked, or Dichotomous (c,r,d): ', end=' ')
-            corrtype = raw_input()
+            corrtype = input()
         if corrtype in ['c','C']:
             m,b,r,p,see = linregress(x,y)
             print('\nLinear regression for continuous variables ...')
             lol = [['Slope','Intercept','r','Prob','SEestimate'],[round(m,4),round(b,4),round(r,4),round(p,4),round(see,4)]]
             pstat.printcc(lol)
         elif corrtype in ['r','R']:
             r,p = spearmanr(x,y)
             print('\nCorrelation for ranked variables ...')

requirements.client.txt

 python-gnupg==0.3.7
 pytz==2016.10
 WTForms==2.0.2
 Flask-WTF==0.12
 typing
 click==6.7
 pyyaml==3.12
 requests
+future
 -e svn+https://llvm.org/svn/llvm-project/llvm/trunk/utils/lit/#egg=lit