def __init__(self):
"""
Checks the location of the jar files.
Spawns the server as a process.
"""
print start_corenlp
if VERBOSE:
logger.info(start_corenlp)
self.corenlp = pexpect.spawn(start_corenlp)
# show progress bar while loading the models
widgets = ['Loading Models: ', Fraction()]
pbar = ProgressBar(widgets=widgets, maxval=5,
force_update=True).start()
self.corenlp.expect("done.",
timeout=20) # Load pos tagger model (~5sec)
pbar.update(1)
self.corenlp.expect("done.",
timeout=200) # Load NER-all classifier (~33sec)
pbar.update(2)
self.corenlp.expect("done.",
timeout=600) # Load NER-muc classifier (~60sec)
pbar.update(3)
self.corenlp.expect("done.",
timeout=600) # Load CoNLL classifier (~50sec)
pbar.update(4)
self.corenlp.expect("done.", timeout=200) # Loading PCFG (~3sec)
pbar.update(5)
self.corenlp.expect("Entering interactive shell.")
pbar.finish()
def _spawn_corenlp(self):
if VERBOSE:
print self.start_corenlp
self.corenlp = pexpect.spawn(self.start_corenlp,
timeout=60,
maxread=8192,
searchwindowsize=80)
# show progress bar while loading the models
if VERBOSE:
widgets = ['Loading Models: ', Fraction()]
pbar = ProgressBar(widgets=widgets, maxval=5,
force_update=True).start()
# Model timeouts:
# pos tagger model (~5sec)
# NER-all classifier (~33sec)
# NER-muc classifier (~60sec)
# CoNLL classifier (~50sec)
# PCFG (~3sec)
timeouts = [20, 200, 600, 600, 20]
for i in xrange(5):
self.corenlp.expect("done.", timeout=timeouts[i]) # Load model
pbar.update(i + 1)
self.corenlp.expect("Entering interactive shell.")
pbar.finish()
# interactive shell
self.corenlp.expect("\nNLP> ")
def __init__(self,
corenlp_path=DIRECTORY,
memory="3g",
properties='default.properties'):
"""
Checks the location of the jar files.
Spawns the server as a process.
"""
# spawn the server
start_corenlp = init_corenlp_command(corenlp_path, memory, properties)
if VERBOSE:
print start_corenlp
self.corenlp = pexpect.spawn(start_corenlp)
# show progress bar while loading the models
if VERBOSE:
widgets = ['Loading Models: ', Fraction()]
pbar = ProgressBar(widgets=widgets, maxval=5,
force_update=True).start()
# Model timeouts:
# pos tagger model (~5sec)
# NER-all classifier (~33sec)
# NER-muc classifier (~60sec)
# CoNLL classifier (~50sec)
# PCFG (~3sec)
timeouts = [20, 200, 600, 600, 20]
for i in xrange(5):
self.corenlp.expect("done.", timeout=timeouts[i]) # Load model
pbar.update(i + 1)
self.corenlp.expect("Entering interactive shell.")
pbar.finish()
# interactive shell
self.corenlp.expect("\nNLP> ", timeout=3)
def __init__(self, corenlp_path=None):
"""
Checks the location of the jar files.
Spawns the server as a process.
"""
jars = ["stanford-corenlp-3.4.1.jar",
#"stanford-corenlp-3.4.1-models.jar",
"stanford-chinese-corenlp-2014-02-24-models.jar", # add chinese models
#"joda-time.jar",
#"xom.jar",
#"jollyday.jar"
]
# if CoreNLP libraries are in a different directory,
# change the corenlp_path variable to point to them
if not corenlp_path:
#corenlp_path = "./stanford-corenlp-full-2014-08-27/"
corenlp_path = "/home/kqc/tools/stanford-corenlp-full-2014-08-27/" # my own corenlp dir
java_path = "java"
classname = "edu.stanford.nlp.pipeline.StanfordCoreNLP"
# include the properties file, so you can change defaults
# but any changes in output format will break parse_parser_results()
props = "-props StanfordCoreNLP-chinese.properties" # for chinese
#props = "-props default.properties"
# add and check classpaths
jars = [corenlp_path + jar for jar in jars]
for jar in jars:
if not os.path.exists(jar):
logger.error("Error! Cannot locate %s" % jar)
sys.exit(1)
# spawn the server
#start_corenlp = "%s -Xmx1800m -cp %s %s %s" % (java_path, ':'.join(jars), classname, props)
start_corenlp = "%s -Xmx3g -cp %s %s %s" % (java_path, ':'.join(jars), classname, props) # for chinese
if VERBOSE:
logger.debug(start_corenlp)
self.corenlp = pexpect.spawnu(start_corenlp, encoding='utf8')
# show progress bar while loading the models
widgets = ['Loading Models: ', Fraction()]
pbar = ProgressBar(widgets=widgets, maxval=5, force_update=True).start()
# increase the timeout setting for chinese
self.corenlp.expect(u"done.", timeout=2000) # Load pos tagger model (~5sec)
pbar.update(1)
self.corenlp.expect(u"done.", timeout=2000) # Load NER-all classifier (~33sec)
pbar.update(2)
self.corenlp.expect(u"done.", timeout=6000) # Load NER-muc classifier (~60sec)
pbar.update(3)
self.corenlp.expect(u"done.", timeout=6000) # Load CoNLL classifier (~50sec)
pbar.update(4)
self.corenlp.expect(u"done.", timeout=2000) # Loading PCFG (~3sec)
pbar.update(5)
self.corenlp.expect(u"Entering interactive shell.")
pbar.finish()
def setup(self):
"""
Checks the location of the jar files.
Spawns the server as a process.
"""
jars = [
"stanford-corenlp-3.2.0.jar", "stanford-corenlp-3.2.0-models.jar",
"joda-time.jar", "xom.jar", "jollyday.jar"
]
# if CoreNLP libraries are in a different directory,
# change the corenlp_path variable to point to them
corenlp_path = os.path.relpath(__file__).split(
'/')[0] + "/stanford-corenlp-full-2013-06-20/"
#corenlp_path = "stanford-corenlp-full-2013-06-20/"
java_path = "java"
classname = "edu.stanford.nlp.pipeline.StanfordCoreNLP"
# include the properties file, so you can change defaults
# but any changes in output format will break parse_parser_results()
props = "-props " + os.path.relpath(__file__).split(
'/')[0] + "/default.properties"
# add and check classpaths
jars = [corenlp_path + jar for jar in jars]
for jar in jars:
if not os.path.exists(jar):
print "Error! Cannot locate %s" % jar
sys.exit(1)
#Change from ':' to ';'
# spawn the server
start_corenlp = "%s -Xmx2500m -cp %s %s %s" % (
java_path, ':'.join(jars), classname, props)
if VERBOSE: print start_corenlp
self.corenlp = pexpect.spawn(start_corenlp)
# show progress bar while loading the models
widgets = ['Loading Models: ', Fraction()]
pbar = ProgressBar(widgets=widgets, maxval=4,
force_update=True).start()
self.corenlp.expect("done.",
timeout=20) # Load pos tagger model (~5sec)
pbar.update(1)
self.corenlp.expect("done.",
timeout=200) # Load NER-all classifier (~33sec)
pbar.update(2)
self.corenlp.expect("done.",
timeout=600) # Load NER-muc classifier (~60sec)
pbar.update(3)
self.corenlp.expect("done.",
timeout=600) # Load CoNLL classifier (~50sec)
pbar.update(4)
# self.corenlp.expect("done.", timeout=200) # Loading PCFG (~3sec)
# pbar.update(5)
self.corenlp.expect("Entering interactive shell.")
pbar.finish()
def __init__(self, corenlp_path=None):
"""
Checks the location of the jar files.
Spawns the server as a process.
"""
jars = [
"stanford-corenlp-3.6.0.jar", "stanford-corenlp-3.6.0-models.jar",
"joda-time.jar", "xom.jar", "jollyday.jar", "ejml-0.23.jar",
"slf4j-api.jar", "slf4j-simple.jar"
]
# if CoreNLP libraries are in a different directory,
# change the corenlp_path variable to point to them
if not corenlp_path:
corenlp_path = "./stanford-corenlp-full-2015-12-09/"
java_path = "java"
classname = "edu.stanford.nlp.pipeline.StanfordCoreNLP"
# include the properties file, so you can change defaults
# but any changes in output format will break parse_parser_results()
props = "-props default.properties"
# add and check classpaths
jars = [corenlp_path + jar for jar in jars]
for jar in jars:
if not os.path.exists(jar):
logger.error("Error! Cannot locate %s" % jar)
sys.exit(1)
# spawn the server
start_corenlp = "%s -Xmx3600m -cp %s %s %s" % (
java_path, ':'.join(jars), classname, props)
if VERBOSE:
logger.debug(start_corenlp)
self.corenlp = pexpect.spawn(start_corenlp)
# show progress bar while loading the models
widgets = ['Loading Models: ', Fraction()]
pbar = ProgressBar(widgets=widgets, maxval=5,
force_update=True).start()
self.corenlp.expect("done.",
timeout=20) # Load pos tagger model (~5sec)
pbar.update(1)
self.corenlp.expect("done.",
timeout=200) # Load NER-all classifier (~33sec)
pbar.update(2)
self.corenlp.expect("done.",
timeout=600) # Load NER-muc classifier (~60sec)
pbar.update(3)
self.corenlp.expect("done.",
timeout=600) # Load CoNLL classifier (~50sec)
pbar.update(4)
self.corenlp.expect("done.", timeout=200) # Loading PCFG (~3sec)
pbar.update(5)
self.corenlp.expect("Entering interactive shell.")
pbar.finish()
def __init__(self,
corenlp_path="stanford-corenlp-full-2013-04-04/",
memory="3g"):
"""
Checks the location of the jar files.
Spawns the server as a process.
"""
# spawn the server
start_corenlp = init_corenlp_command(corenlp_path, memory)
if VERBOSE: print start_corenlp
self.corenlp = pexpect.spawn(start_corenlp)
# show progress bar while loading the models
if VERBOSE:
widgets = ['Loading Models: ', Fraction()]
pbar = ProgressBar(widgets=widgets, maxval=5,
force_update=True).start()
self.corenlp.expect("done.",
timeout=20) # Load pos tagger model (~5sec)
if VERBOSE: pbar.update(1)
self.corenlp.expect("done.",
timeout=200) # Load NER-all classifier (~33sec)
if VERBOSE: pbar.update(2)
self.corenlp.expect("done.",
timeout=600) # Load NER-muc classifier (~60sec)
if VERBOSE: pbar.update(3)
self.corenlp.expect("done.",
timeout=600) # Load CoNLL classifier (~50sec)
if VERBOSE: pbar.update(4)
self.corenlp.expect("done.", timeout=200) # Loading PCFG (~3sec)
if VERBOSE: pbar.update(5)
self.corenlp.expect("Entering interactive shell.")
if VERBOSE: pbar.finish()
# interactive shell
self.corenlp.expect("\nNLP> ", timeout=3)
def __init__(self, corenlp_path=None):
"""
Checks the location of the jar files.
Spawns the server as a process.
"""
jars = [
"stanford-corenlp-2017-04-14-build.jar",
"stanford-corenlp-models-current.jar",
"stanford-chinese-corenlp-models-current.jar",
#"stanford-english-corenlp-models-current.jar",
#"stanford-english-kbp-corenlp-models-current.jar",
"joda-time.jar",
"xom.jar",
"jollyday.jar"
]
# if CoreNLP libraries are in a different directory,
# change the corenlp_path variable to point to them
if not corenlp_path:
corenlp_path = "./CoreNLP/"
java_path = "java"
classname = "edu.stanford.nlp.pipeline.StanfordCoreNLP"
# include the properties file, so you can change defaults
# but any changes in output format will break parse_parser_results()
#props = "-props default.properties"
props = "-props StanfordCoreNLP-chinese.properties"
# add and check classpaths
jars = [corenlp_path + jar for jar in jars]
for jar in jars:
if not os.path.exists(jar):
logger.error("Error! Cannot locate %s" % jar)
sys.exit(1)
# spawn the server
start_corenlp = "%s -Xmx5g -cp %s %s %s" % (java_path, ':'.join(jars),
classname, props)
#start_corenlp = 'java -Xmx100g -cp "*" edu.stanford.nlp.pipeline.StanfordCoreNLPServer -serverProperties StanfordCoreNL-chinese.properties -port 14444 -timeout 100000'
if VERBOSE:
logger.debug(start_corenlp)
self.corenlp = pexpect.spawnu(start_corenlp)
# show progress bar while loading the models
widgets = ['Loading Models: ', Fraction()]
pbar = ProgressBar(widgets=widgets, maxval=4,
force_update=True).start()
self.corenlp.expect([u"done.", pexpect.EOF],
timeout=200) # Load pos tagger model (~5sec)
pbar.update(1)
self.corenlp.expect([u"done.", pexpect.EOF],
timeout=2000) # Load NER-all classifier (~33sec)
pbar.update(2)
self.corenlp.expect([u"done.", pexpect.EOF],
timeout=6000) # Load NER-muc classifier (~60sec)
pbar.update(3)
self.corenlp.expect([u"done.", pexpect.EOF],
timeout=6000) # Load CoNLL classifier (~50sec)
pbar.update(4)
#self.corenlp.expect([u"done.", pexpect.EOF], timeout=2000) # Loading PCFG (~3sec)
#pbar.update(5)
self.corenlp.expect([u"Entering interactive shell.", pexpect.EOF])
pbar.finish()