Migrate to Python 3.7 (#21)

Migrate to Python 3.7

.circleci/config.yml

@@ -3,7 +3,7 @@ jobs:
   build:
     working_directory: ~/phovea
     docker:
-      - image: caleydo/phovea_circleci_python:v3.0
+      - image: circleci/python:3.7-buster-node-browsers # for node version see Dockerfile on https://hub.docker.com/r/circleci/python
     steps:
       - checkout
       - run:
@@ -48,7 +48,7 @@ jobs:
             npm run dist
       - store_artifacts:
           path: dist
-          prefix: dist
+          destination: dist
 workflows:
   version: 2
 #  build-nightly:

build.py

@@ -19,7 +19,7 @@ def _resolve_plugin(repo, version):
   if os.path.isdir('.git') and repo:
     if repo.endswith('.git'):
       repo = repo[0:-4]
-    return repo + '/commit/' + _git_head('.')
+    return repo + '/commit/' + _git_head('.').decode('utf-8')
   # not a git repo
   return version
 

requirements.txt

@@ -1,3 +1,2 @@
 -e git+https://github.com/phovea/phovea_server.git@develop#egg=phovea_server
-enum==0.4.6
-sklearn==0.0
+sklearn==0.0

requirements_dev.txt

@@ -1,6 +1,6 @@
-flake8==3.0.4
-pep8-naming==0.4.1
-pytest==3.0.3
-pytest-runner==2.9
-Sphinx==1.5.2
-recommonmark==0.4.0
+flake8==3.5.0
+pep8-naming==0.5.0
+pytest==3.5.0
+pytest-runner==4.2
+Sphinx==1.7.2
+recommonmark==0.6.0

setup.py

@@ -3,7 +3,7 @@
 # Copyright (c) The Caleydo Team. All rights reserved.
 # Licensed under the new BSD license, available at http://caleydo.org/license
 ###############################################################################
-from __future__ import with_statement, print_function
+
 from setuptools import setup, find_packages
 from codecs import open
 from os import path
@@ -26,12 +26,12 @@ def read_it(name):
 def packaged(*files):
   r = {}
   global pkg
-  r[pkg['name'].encode('ascii')] = list(files)
+  r[pkg['name']] = list(files)
   return r
 
 
 def requirements(file):
-  return [r.strip().encode('ascii') for r in read_it(file).strip().split('\n') if not r.startswith('-e git+https://')]
+  return [r.strip() for r in read_it(file).strip().split('\n') if not r.startswith('-e git+https://')]
 
 
 def to_version(v):

taco_server/__init__.py

@@ -11,10 +11,9 @@ def phovea(registry):
   :param registry:
   """
   # generator-phovea:begin
-  registry.append('namespace', 'taco', 'taco_server.api',
-                  {
-                      'namespace': '/api/taco'
-                  })
+  registry.append('namespace', 'taco', 'taco_server.api', {
+      'namespace': '/api/taco'
+  })
   # generator-phovea:end
   pass
 

taco_server/api.py

@@ -1,6 +1,6 @@
 from phovea_server import ns
 import timeit
-from src import diff_cache
+from .src import diff_cache
 import logging
 
 
@@ -102,6 +102,7 @@ def create():
   """
   return app
 
+
 if __name__ == '__main__':
   app.debug = True
   app.run(host='0.0.0.0', port=9000)

taco_server/src/diff_cache.py

@@ -3,12 +3,11 @@
 # detail (as detail), middle (as count), overview (as ratios)
 
 
-from __future__ import print_function
-from diff_finder import Table, DiffFinder, Diff, Ratios
+from .diff_finder import Table, DiffFinder, Diff, Ratios
 import phovea_server.dataset as dataset
 import timeit
 import json
-import pandas.json as ujson
+from . import json_encoder
 import os
 import hashlib
 from collections import namedtuple
@@ -37,6 +36,7 @@ def create_cache_dir():
   else:
     _log.info('use existing cache directory: ' + _cache_directory)
 
+
 # run immediately!
 create_cache_dir()
 
@@ -50,7 +50,7 @@ def get_diff_cache(filename):
   file_name = _cache_directory + filename + '.json'
   if os.path.isfile(file_name):
     with open(file_name) as data_file:
-      data = ujson.load(data_file)
+      data = json.load(data_file)
     return data
   # if the file doesn't exist
   return None
@@ -100,16 +100,17 @@ def get_diff_table(id1, id2, direction, ops, jsonit=True):
 
     if isinstance(diffobj, Diff):
       # log the detail
-      json_result = ujson.dumps(diffobj.serialize())
+      serialize = Diff.serialize # noqa E121
+      json_result = (json.dumps(diffobj.__dict__, cls=json_encoder.JsonEncoder))
       set_diff_cache(hash_name, json_result)
     else:
       # todo later find a way to send the error
       # e.g. there's no matching column in this case
-      json_result = ujson.dumps(diffobj)  # which is {} for now!
+      json_result = json.dumps(diffobj, cls=json_encoder.JsonEncoder)  # which is {} for now!
       set_diff_cache(hash_name, json_result)
 
   elif jsonit is False:
-    diffobj = Diff().unserialize(ujson.loads(json_result))
+    diffobj = Diff().unserialize(json.loads(json_result))
 
   if jsonit:
     return json_result
@@ -152,10 +153,10 @@ def get_ratios(id1, id2, direction, ops, bins=1, bins_col=1, jsonit=True):
     # bin == 1 -> timeline bar chart
     # bin == -1 -> 2d ratio plot
     if bins == 1 or bins == -1:
-      json_ratios = ujson.dumps(ratios.serialize())
+      json_ratios = json.dumps(ratios.serialize(), cls=json_encoder.JsonEncoder)
     # bin > 1 -> 2d ratio histogram
     else:
-      json_ratios = ujson.dumps(ratios)
+      json_ratios = json.dumps(ratios, cls=json_encoder.JsonEncoder)
 
     # cache this as overview
     set_diff_cache(hashname, json_ratios)
@@ -198,12 +199,12 @@ def stratify_matrix(m):
     if row_strat is not None:
       rowids = list(m.rowids())
       row_indices = [rowids.index(o) for o in row_strat.rowids()]
-      data = data[row_indices, ...]
+      data = data[row_indices, ...].astype('str')
 
     if col_strat is not None:
       colids = list(m.colids())
       col_indices = [colids.index(o) for o in col_strat.rowids()]
-      data = data[..., col_indices]
+      data = data[..., col_indices].astype('str')
 
     return Table(rows, cols, data)
 
@@ -232,12 +233,12 @@ def create_hashname(id1, id2, bins, bins_col, direction, ops):
   :return:
   """
   name = str(id1) + '_' + str(id2) + '_' + str(bins) + '_' + str(bins_col) + '_' + str(direction) + '_' + str(ops)
-  return hashlib.md5(name).hexdigest()
+  return hashlib.md5(name.encode('utf-8')).hexdigest()
 
 
 def ratio_from_json(jsonobj):
   # idk
-  r = json.loads(jsonobj, object_hook=lambda d: namedtuple('X', d.keys())(*d.values()))
+  r = json.loads(jsonobj, object_hook=lambda d: namedtuple('X', d.keys())(*list(d.values())))
   # todo find a smarter way, really
   cr = 0 if not hasattr(r, "c_ratio") else r.c_ratio
   ar = 0 if not hasattr(r, "a_ratio") else r.a_ratio