def test_pyspark_livy_sql_options():
query = "abc"
sqlquery = SQLQuery(query, samplemethod='take', maxrows=120)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().take(120): print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
sqlquery = SQLQuery(query, samplemethod='take', maxrows=-1)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().collect(): print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
sqlquery = SQLQuery(query, samplemethod='sample', samplefraction=0.25, maxrows=-1)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().sample(False, 0.25).collect(): '
u'print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
sqlquery = SQLQuery(query, samplemethod='sample', samplefraction=0.33, maxrows=3234)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().sample(False, 0.33).take(3234): '
u'print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
def test_pyspark_livy_sql_options():
query = "abc"
sqlquery = SQLQuery(query, samplemethod='take', maxrows=120)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().take(120): print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
sqlquery = SQLQuery(query, samplemethod='take', maxrows=-1)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().collect(): print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
sqlquery = SQLQuery(query,
samplemethod='sample',
samplefraction=0.25,
maxrows=-1)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().sample(False, 0.25).collect(): '
u'print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
sqlquery = SQLQuery(query,
samplemethod='sample',
samplefraction=0.33,
maxrows=3234)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().sample(False, 0.33).take(3234): '
u'print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
def _pyspark_command(self):
command = u'sqlContext.sql(u"""{} """).toJSON()'.format(self.query)
if self.samplemethod == u'sample':
command = u'{}.sample(False, {})'.format(command, self.samplefraction)
if self.maxrows >= 0:
command = u'{}.take({})'.format(command, self.maxrows)
else:
command = u'{}.collect()'.format(command)
command = u'for {} in {}: print({}.encode("{}"))'.format(constants.LONG_RANDOM_VARIABLE_NAME,
command,
constants.LONG_RANDOM_VARIABLE_NAME,
conf.pyspark_sql_encoding())
return Command(command)
def _pyspark_command(self):
command = u'sqlContext.sql(u"""{} """).toJSON()'.format(self.query)
if self.samplemethod == u'sample':
command = u'{}.sample(False, {})'.format(command,
self.samplefraction)
if self.maxrows >= 0:
command = u'{}.take({})'.format(command, self.maxrows)
else:
command = u'{}.collect()'.format(command)
command = u'for {} in {}: print({}.encode("{}"))'.format(
constants.LONG_RANDOM_VARIABLE_NAME, command,
constants.LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())
return Command(command)
def test_unicode_sql():
query = u"SELECT 'è'"
sqlquery = SQLQuery(query, samplemethod='take', maxrows=120)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().take(120): print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
assert_equals(sqlquery._scala_command(),
Command(u'sqlContext.sql("""{}""").toJSON.take(120).foreach(println)'.format(query)))
try:
sqlquery._r_command()
assert False
except NotImplementedError:
pass
def test_unicode_sql():
query = u"SELECT 'è'"
sqlquery = SQLQuery(query, samplemethod='take', maxrows=120)
assert_equals(sqlquery._pyspark_command(),
Command(u'for {} in sqlContext.sql(u"""{} """).toJSON().take(120): print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
assert_equals(
sqlquery._scala_command(),
Command(u'sqlContext.sql("""{}""").toJSON.take(120).foreach(println)'.
format(query)))
try:
sqlquery._r_command()
assert False
except NotImplementedError:
pass
def _pyspark_command(self, sql_context_variable_name, encode_result=True):
command = u'{}.sql(u"""{} """).toJSON()'.format(sql_context_variable_name, self.query)
if self.samplemethod == u'sample':
command = u'{}.sample(False, {})'.format(command, self.samplefraction)
if self.maxrows >= 0:
command = u'{}.take({})'.format(command, self.maxrows)
else:
command = u'{}.collect()'.format(command)
# Unicode support has improved in Python 3 so we don't need to encode.
if encode_result:
print_command = '{}.encode("{}")'.format(constants.LONG_RANDOM_VARIABLE_NAME,
conf.pyspark_sql_encoding())
else:
print_command = constants.LONG_RANDOM_VARIABLE_NAME
command = u'for {} in {}: print({})'.format(constants.LONG_RANDOM_VARIABLE_NAME,
command,
print_command)
return Command(command)
def test_unicode_sql():
query = u"SELECT 'è'"
sqlquery = SQLQuery(query, samplemethod='take', maxrows=120)
assert_equals(sqlquery._pyspark_command("spark"),
Command(u'for {} in spark.sql(u"""{} """).toJSON().take(120): print({}.encode("{}"))'\
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding())))
assert_equals(
sqlquery._scala_command("spark"),
Command(
u'spark.sql("""{}""").toJSON.take(120).foreach(println)'.format(
query)))
assert_equals(
sqlquery._r_command(),
Command(
u'for ({} in (jsonlite:::toJSON(take(sql("{}"),120)))) {{cat({})}}'
.format(LONG_RANDOM_VARIABLE_NAME, query,
LONG_RANDOM_VARIABLE_NAME)))
def test_unicode_sql():
query = u"SELECT 'è'"
sqlquery = SQLQuery(query, samplemethod="take", maxrows=120)
assert_equals(
sqlquery._pyspark_command("spark"),
Command(
u'for {} in spark.sql(u"""{} """).toJSON().take(120): print({}.encode("{}"))'.format(
LONG_RANDOM_VARIABLE_NAME, query, LONG_RANDOM_VARIABLE_NAME, conf.pyspark_sql_encoding()
)
),
)
assert_equals(
sqlquery._scala_command("spark"),
Command(u'spark.sql("""{}""").toJSON.take(120).foreach(println)'.format(query)),
)
assert_equals(
sqlquery._r_command(),
Command(
u'for ({} in (jsonlite:::toJSON(take(sql("{}"),120)))) {{cat({})}}'.format(
LONG_RANDOM_VARIABLE_NAME, query, LONG_RANDOM_VARIABLE_NAME
)
),
)
