From 207474d527a8929a25569b11b1856c0fe2cdab30 Mon Sep 17 00:00:00 2001
From: "Knox S. Long" <long@stsci.edu>
Date: Sun, 3 Sep 2017 15:31:34 -0500
Subject: [PATCH] Added a new python routine regression.py which is intended as
 a program to run a series of srhort python models for testing various aspects
 of the code.

The models that are run are stored (by default) in the directory
examples/regress.  This is intended as way to partially address #292
---
 examples/regress/1d_sn.pf       |   1 +
 examples/regress/cv_standard.pf |   1 +
 py_progs/regression.py          | 302 ++++++++++++++++++++++++++++++++
 3 files changed, 304 insertions(+)
 create mode 120000 examples/regress/1d_sn.pf
 create mode 120000 examples/regress/cv_standard.pf
 create mode 100755 py_progs/regression.py

diff --git a/examples/regress/1d_sn.pf b/examples/regress/1d_sn.pf
new file mode 120000
index 000000000..9bf7145c4
--- /dev/null
+++ b/examples/regress/1d_sn.pf
@@ -0,0 +1 @@
+../core/1d_sn.pf
\ No newline at end of file
diff --git a/examples/regress/cv_standard.pf b/examples/regress/cv_standard.pf
new file mode 120000
index 000000000..a3e3ee7d9
--- /dev/null
+++ b/examples/regress/cv_standard.pf
@@ -0,0 +1 @@
+../core/cv_standard.pf
\ No newline at end of file
diff --git a/py_progs/regression.py b/py_progs/regression.py
new file mode 100755
index 000000000..f57165b60
--- /dev/null
+++ b/py_progs/regression.py
@@ -0,0 +1,302 @@
+#!/usr/bin/env python 
+
+'''
+                    Space Telescope Science Institute
+
+Synopsis:  
+
+Execute a series of short python runs to test whether a version 
+of python is working.
+
+
+Command line usage (if any):
+
+    usage: regression.py [-np 3 -pf_dir test -out_dir foo] version 
+
+    where 
+
+        version         the executable of python
+        -np 3           the number of processors with which to run (default 3)
+        -pf_dir test    the directory containing all of the .pf files which will be run
+                        The defaults is $PYTHON/examples/regress
+        -out_dir foo    The directory (below the current working directory) where the 
+                        tests will run.  The defauld is constructed for the version
+                        and the data
+
+Description:  
+
+    The basic process is as follows
+
+    1.  Create a directory in which to work and intialize it
+    2.  Copy all of the relevant .pf files to this directory
+    3.  Switch to the working directory and run all the models, performing
+        some basic checks 
+
+Primary routines:
+
+    doit:       Internal routine which runs python on all of the pf files of interest.  Use
+                this if working in a python shell
+    steer:      A routine to parse the command lineh
+    check_one:  A routine which oversees checking of the runs
+
+Notes:
+
+    Regression here means to run a series of models. These routines do not compare the
+    models to earlier runs
+                                       
+History:
+
+170903 ksl Coding begun
+
+'''
+
+import sys
+from astropy.io import ascii
+import numpy
+from glob import glob
+import time
+import subprocess
+import os
+import shutil
+
+
+
+
+def sum_errors(root='1d_sn'):
+    '''
+    This sums the errors from various threads
+
+    Note that this effectively duplicates a routine
+    py_error which James wrote, but in Python 3
+
+
+    '''
+    diag_files=glob('diag_%s/%s_*.diag' % (root,root))
+
+    errors=[]
+    for one in diag_files:
+        x=open(one)
+        lines=x.readlines()
+        lines=lines[len(lines)-100:]
+        test=False
+        for one_line in lines:
+            if test and one_line.count(' -- '):
+                errors.append(one_line)
+            if one_line.count('Recurrences --  Description'):
+                test=True
+
+    # At this point we have the error but we need to parse them etc
+
+    i=0
+    error_names=[]
+    error_counts=[]
+    while i<len(errors):
+        errors[i]=errors[i].strip()
+        words=errors[i].split(' -- ')
+        error_names.append(words[1])
+        error_counts.append(int(words[0]))
+        i+=1
+
+    print(errors)
+
+    # Now get the unique names
+
+    unique_errors=set(error_names)
+
+    print(unique_errors)
+
+    # finally add up the errors from the various threads
+
+    records=[]
+    for one in unique_errors:
+        name=one
+        counts=0
+        i=0
+        while i<len(error_names):
+            if name==error_names[i]:
+                counts+=error_counts[i]
+            i+=1
+        records.append([name,counts])
+    return(records)
+
+
+def check_one(xfile,root):
+    '''
+    Perform some checks on what happened in a run
+    '''
+
+    g=open(root+'.sig')
+    lines=g.readlines()
+    last_line=lines[len(lines)-1]
+    words=last_line.split()
+    if last_line.count('COMPLETE'):
+        xfile.write('The model ran to completion in %s s\n' % (words[5]))
+    else:
+        xfile.write('The models stopped before completing\n')
+
+    errors=sum_errors(root)
+
+    xfile.write('\n The reported errors in all threads was as follows: \n')
+    for one in errors:
+        xfile.write('%10d -- %s\n' % (one[1],one[0]))
+
+    return
+    
+    
+
+
+def doit(version='py',pf_dir='',out_dir='',np=3,outputfile='Summary.txt'):
+    '''
+    Test a specific version of python against a series of models
+
+    Description:
+
+    Notes:
+
+    History:
+
+        170903 ksl Bagan work
+
+    '''
+
+    date=time.strftime("%y%m%d", time.gmtime())
+
+    if out_dir=='':
+        out_dir='%s_%s' % (version,date)
+
+    if os.path.exists(out_dir)==False:
+        os.mkdir(out_dir)
+
+    print(date)
+
+    # Get the PYTHON environment variable
+
+    PYTHON=os.environ['PYTHON']
+    print(PYTHON)
+
+    if pf_dir=='':
+        pf_files=glob(PYTHON+'/examples/regress/*pf')
+    else:
+        pf_files=glob(pf_dir+'/*pf')
+
+    print(pf_files)
+
+
+    commands=[]
+    root_names=[]
+    for one in pf_files:
+        shutil.copy(one,out_dir)
+        words=one.split('/')
+        pf=(words[len(words)-1])
+        root_name=pf.replace('.pf','')
+        if np<=1:
+            command='%s %s' % (version,pf)
+        else:
+            command='mpirun -np %d %s %s >%s.stdout.txt' % (np,version,pf,root_name)
+        commands.append(command)
+
+        root_names.append(root_name)
+
+    print(commands)
+
+    # Switch to the work directory
+    cwd=os.getcwd()
+    os.chdir(out_dir)
+
+    proc=subprocess.Popen('Setup_Py_Dir',shell=True,stdout=subprocess.PIPE,stderr=subprocess.PIPE)
+
+    f=open(outputfile,'w')
+
+    i=0
+    while i<len(commands):
+        one=commands[i]
+        string='Running %s' % one
+        print(string)
+        f.write('%s\n'% string)
+
+        proc=subprocess.Popen(one,shell=True,stdout=subprocess.PIPE,stderr=subprocess.PIPE)
+        stdout,stderr=proc.communicate()
+        if len(stderr):
+            string='Stderrs were reported!!'
+            f.write('%s\n'% string)
+            print(string)
+            print(stderr)
+            stderr=str(stderr)
+            stderr.replace('\\n','\n')
+            g=open(root_names[i]+'.stderr.txt','w')
+            g.write(stderr)
+            g.close()
+        else:
+            string='No stderrs were reported'
+            print(string)
+            f.write('%s\n' % string)
+            check_one(f,root_names[i])
+
+        string='Finished %s' % one
+        print(string)
+        f.write('%s\n\n'% string)
+        i+=1
+
+
+    f.close()
+    # Return to the place where the code was made from
+    os.chdir(cwd)
+
+
+
+
+
+    return
+
+
+def steer(argv):
+    '''
+    This is just a steering routine so that switches can be processed 
+    from the command line
+    '''
+    pf_dir=''
+    out_dir=''
+    np=3
+
+    i=1
+    words=[]
+    while i<len(argv):
+        if argv[i]=='-h':
+            print(__doc__)
+            return
+        elif argv[i]=='-np':
+            i=i+1
+            np=int(argv[i])
+        elif argv[i]=='-pf_dir':
+            i=i+1
+            pf_dir=(argv[i])
+        elif argv[i]=='-out_dir':
+            i=i+1
+            out_dir=(argv[i])
+        elif argv[i][0]=='-':
+            print('Error: Unknown switch ---  %s' % argv[i])
+            return
+        else:
+            words.append(argv[i])
+        i+=1
+
+    if(len(words)==0):
+        print('Error: Consumed of command line without a python executable')
+        return
+
+    for one in words:
+        doit(version=one,pf_dir=pf_dir,out_dir=out_dir,np=np,outputfile='Summary.txt')
+
+
+
+
+
+
+
+# Next lines permit one to run the routine from the command line
+if __name__ == "__main__":
+    import sys
+    if len(sys.argv)>1:
+        steer(sys.argv)
+    else:
+        print(__doc__)