def test_numerical_keys_become_strs(self):
# JSON should convert numbers to strings when they are dict keys
self.assertEqual(
(None, {
'3': 4
}),
JSONValueProtocol().read(JSONValueProtocol().write(None, {3: 4})))
def test_bad_keys_and_values(self):
# dictionaries have to have strings as keys
self.assertCantEncode(JSONValueProtocol(), None, {(1, 2): 3})
# only unicodes (or bytes in utf-8) are allowed
self.assertCantEncode(JSONValueProtocol(), None, '\xe9')
# sets don't exist in JSON
self.assertCantEncode(JSONValueProtocol(), None, set())
# Point class has no representation in JSON
self.assertCantEncode(JSONValueProtocol(), None, Point(1, 4))
def reducer_init(self):
self.idfs = {}
for fname in os.listdir(DIRECTORY): # look through file names in the directory
file = open(os.path.join(DIRECTORY, fname)) # open a file
for line in file: # read each line in json file
term_idf = JSONValueProtocol.read(line)[1] # parse the line as a JSON object
self.idfs[term_idf['term']] = term_idf['idf']
def reducer_init(self):
self.idfs = {}
for fname in os.listdir(DIRECTORY): # look through file names in the directory
file = open(os.path.join(DIRECTORY, fname)) # open a file
for line in file: # read each line in json file
term_idf = JSONValueProtocol().read(line)[1] # parse the line as a JSON object
self.idfs[term_idf['term']] = term_idf['idf']
def encode_document(doc_id, text):
"""Encode a document as a JSON dictionary so that MRNgramIDFUtility can read it.
We intend to use `doc_id` as a business/product/entity ID rather than the ID of
an individual review."""
#text = unicode(text) RAWR some amazon reviews won't encode
return JSONValueProtocol.write(
None, {'doc_id': doc_id, 'text': text})
def root_to_json(root_dir, output_file):
walker = EmailWalker(root_dir)
output = open(output_file, "w")
for email in walker:
email['date'] = str(email['date'])
line = JSONValueProtocol.write(None, email) + '\n'
output.write(line)
output.close()
def root_to_json(root_dir, output_file):
walker = EmailWalker(root_dir)
output = open(output_file, "w")
for email in walker:
email['date'] = str(email['date'])
line = JSONValueProtocol.write(None, email) + '\n'
output.write(line)
output.close()
def parse_file(filename):
words = defaultdict(lambda: 0)
with open(filename) as input:
for line in input:
email = JSONValueProtocol.read(line)[1]
for term in get_terms(email['text']):
words[term] += 1
for word, count in words.items():
print word, count
def data(self, minimum=1, **kw):
res = []
mr_job = MRWordFreqJSON()
mr_job.stdin = [JSONValueProtocol().write(None, line) for line in TEXT]
with mr_job.make_runner() as runner:
runner.run()
for line in runner.stream_output():
key, value = mr_job.parse_output_line(line)
if int(value) >= int(minimum):
res.append([key, value])
return dict(data=res)
def encode_document(text, cats=None, id=None):
"""Encode a document as a JSON so that MRTextClassifier can read it.
Args:
text -- the text of the document (as a unicode)
cats -- a dictionary mapping a category name (e.g. 'sports') to True if
the document is in the category, and False if it's not. None indicates
that we have no information about this documents' categories
id -- a unique ID for the document (any kind of JSON-able value should
work). If not specified, we'll auto-generate one.
"""
text = unicode(text, errors="ignore")
cats = dict((unicode(cat), bool(is_in_cat)) for cat, is_in_cat in (cats or {}).iteritems())
return JSONValueProtocol.write(None, {"document": text, "cats": cats, "docid": id, "type": "document"}) + "\n"
def reducer_init(self):
self.idfs = {}
# Iterate through the files in the bucket provided by the user
if self.options.aws_access_key_id and self.options.aws_secret_access_key:
emr = EMRJobRunner(aws_access_key_id=self.options.aws_access_key_id,
aws_secret_access_key=self.options.aws_secret_access_key)
else:
emr = EMRJobRunner()
for key in emr.get_s3_keys("s3://" + self.options.idf_loc):
# Load the whole file first, then read it line-by-line: otherwise,
# chunks may not be even lines
for line in StringIO(key.get_contents_as_string()):
term_idf = JSONValueProtocol.read(line)[1] # parse the line as a JSON object
self.idfs[term_idf['term']] = term_idf['idf']
def reducer_init(self):
self.idfs = {}
# Iterate through the files in the bucket provided by the user
if self.options.aws_access_key_id and self.options.aws_secret_access_key:
emr = EMRJobRunner(
aws_access_key_id=self.options.aws_access_key_id,
aws_secret_access_key=self.options.aws_secret_access_key)
else:
emr = EMRJobRunner()
for key in emr.get_s3_keys("s3://" + self.options.idf_loc):
# Load the whole file first, then read it line-by-line: otherwise,
# chunks may not be even lines
for line in StringIO(key.get_contents_as_string()):
term_idf = JSONValueProtocol.read(line)[
1] # parse the line as a JSON object
self.idfs[term_idf['term']] = term_idf['idf']
def encode_document(text, cats=None, id=None):
"""Encode a document as a JSON so that MRTextClassifier can read it.
Args:
text -- the text of the document (as a unicode)
cats -- a dictionary mapping a category name (e.g. 'sports') to True if
the document is in the category, and False if it's not. None indicates
that we have no information about this documents' categories
id -- a unique ID for the document (any kind of JSON-able value should
work). If not specified, we'll auto-generate one.
"""
text = unicode(text)
cats = dict((unicode(cat), bool(is_in_cat))
for cat, is_in_cat
in (cats or {}).iteritems())
return JSONValueProtocol.write(
None, {'text': text, 'cats': cats, 'id': id}) + '\n'
def test_numerical_keys_become_strs(self):
# JSON should convert numbers to strings when they are dict keys
self.assertEqual(
(None, {'3': 4}),
JSONValueProtocol.read(JSONValueProtocol.write(None, {3: 4})))
def test_tuples_become_lists(self):
# JSON should convert tuples into lists
self.assertEqual(
(None, [3, 4]),
JSONValueProtocol.read(JSONValueProtocol.write(None, (3, 4))))
def test_uses_json_format(self):
VALUE = {'foo': {'bar': 3}, 'baz': None, 'quz': ['a', 1]}
ENCODED = '{"foo": {"bar": 3}, "baz": null, "quz": ["a", 1]}'
self.assertEqual((None, VALUE), JSONValueProtocol.read(ENCODED))
self.assertEqual(ENCODED, JSONValueProtocol.write(None, VALUE))
#s3_input_path = "s3://joeloren//iceval_out//input//datasets//"
tmp_dir_out = "s3://joeloren/interim_out/"
tmp_dir_in = "s3://joeloren/interim_in/"
tmp_dir_in_relative = "interim_in/"
tmp_dir_out_relative = "interim_out/"
from mrjob.protocol import JSONValueProtocol, JSONProtocol
jvp = JSONValueProtocol()
jp = JSONProtocol()
from boto.s3.connection import S3Connection
import sys
c = S3Connection('AKIAI4OZ3HY56BTOHA3A',
'6isbkZjBM8kt3PIk53EXVIf76VOPxOH8rNleGc6B')
bucket = c.get_bucket("joeloren")
datasets_bucket = c.get_bucket('joel_datasets')
def test_tuples_become_lists(self):
# JSON should convert tuples into lists
self.assertEqual(
(None, [3, 4]),
JSONValueProtocol.read(JSONValueProtocol.write(None, (3, 4))))
def test_uses_json_format(self):
VALUE = {'foo': {'bar': 3}, 'baz': None, 'quz': ['a', 1]}
ENCODED = '{"foo": {"bar": 3}, "baz": null, "quz": ["a", 1]}'
self.assertEqual((None, VALUE), JSONValueProtocol.read(ENCODED))
self.assertEqual(ENCODED, JSONValueProtocol.write(None, VALUE))
def test_uses_json_format(self):
VALUE = {'foo': 'bar'}
ENCODED = b'{"foo": "bar"}'
self.assertEqual((None, VALUE), JSONValueProtocol().read(ENCODED))
self.assertEqual(ENCODED, JSONValueProtocol().write(None, VALUE))
def test_bad_data(self):
self.assertCantDecode(JSONValueProtocol(), '{@#$@#!^&*$%^')
def test_round_trip_with_trailing_tab(self):
for _, v in JSON_KEYS_AND_VALUES:
self.assertRoundTripWithTrailingTabOK(JSONValueProtocol(), None, v)
def test_round_trip(self):
for _, v in JSON_KEYS_AND_VALUES:
self.assertRoundTripOK(JSONValueProtocol(), None, v)
import re
import sys
from collections import defaultdict
from mrjob.protocol import JSONValueProtocol
from term_tools import get_terms
input = open(sys.argv[1])
words = defaultdict(lambda: 0)
for line in input:
email = JSONValueProtocol.read(line)[1]
for term in get_terms(email['text']):
words[term] += 1
for word, count in words.items():
print word, count
