def corpusMenuInitializer(menu, prevMenu):
menu.text = """
The corpora are used for training new models and testing existing
models. The corpora installable here are from the two BioNLP Shared
Tasks (BioNLP'09 and BioNLP'11) on Event Extraction (organized by
University of Tokyo), and the First Challenge Task: Drug-Drug Interaction
Extraction (DDI'11, organized by Universidad Carlos III de Madrid).
The corpora are downloaded as interaction XML files, generated from the
original Shared Task files. If you need to convert the corpora from
the original files, you can use the convertBioNLP.py and convertDDI.py programs
located at Utils/Convert.
It is also recommended to download the official BioNLP Shared Task evaluator
programs, which will be used by TEES when training or testing on those corpora.
"""
# Mark "skip" as default option, this will be re-marked as install if a corpus is missing
menu.setDefault("s")
# If CORPUS_DIR setting is not set set it now
if not hasattr(Settings, "CORPUS_DIR") or getattr(Settings, "CORPUS_DIR") == None:
Settings.setLocal("CORPUS_DIR", os.path.join(menu.system.defaultInstallDir, "corpora"))
print >>sys.stderr
# Initialize handlers
handlers = []
handlerArgs = []
corpusInstallPath = os.path.join(menu.system.defaultInstallDir, "corpora")
corpusDownloadPath = os.path.join(menu.system.defaultInstallDir, "corpora/download")
# Check which corpora need to be installed
redownload = menu.optDict["1"].toggle
for corpus in ["GE", "EPI", "ID", "BB", "BI", "CO", "REL", "REN"]:
if menu.optDict["2"].toggle or (menu != prevMenu and not checkCorpusInstall(corpus)):
menu.setDefault("i")
menu.optDict["2"].toggle = True
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append(["BIONLP_11_CORPORA", corpusInstallPath, corpusDownloadPath, redownload, True])
break
if menu.optDict["3"].toggle or (menu != prevMenu and not checkCorpusInstall("GE09")):
menu.setDefault("i")
menu.optDict["3"].toggle = True
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append(["BIONLP_09_CORPUS", corpusInstallPath, corpusDownloadPath, redownload, True])
if menu.optDict["4"].toggle or (menu != prevMenu and not checkCorpusInstall("DDI", ("-train.xml", "-devel.xml"))):
menu.setDefault("i")
menu.optDict["4"].toggle = True
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append(["DDI_11_CORPUS", corpusInstallPath, corpusDownloadPath, redownload, True])
# A handler for installing BioNLP'11 evaluators
evaluatorInstallPath = os.path.join(menu.system.defaultInstallDir, "tools/evaluators")
evaluatorDownloadPath = os.path.join(menu.system.defaultInstallDir, "tools/download")
if menu.optDict["5"].toggle or (
menu != prevMenu
and (not hasattr(Settings, "BIONLP_EVALUATOR_DIR") or getattr(Settings, "BIONLP_EVALUATOR_DIR") == None)
):
menu.setDefault("i")
menu.optDict["5"].toggle = True
handlers.append(convertBioNLP.installEvaluators)
handlerArgs.append([evaluatorInstallPath, evaluatorDownloadPath, redownload, True])
# Add the handlers to install option
menu.optDict["i"].handler = handlers
menu.optDict["i"].handlerArgs = handlerArgs
def installPreconverted(url="BIONLP_CORPORA", destPath=None, downloadPath=None, redownload=False, updateLocalSettings=False):
print >> sys.stderr, "---------------", "Downloading preconverted corpora", "---------------"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "corpora")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "corpora/download")
Utils.Download.downloadAndExtract(Settings.URL[url], destPath, downloadPath, redownload=redownload)
Settings.setLocal("CORPUS_DIR", destPath, updateLocalSettings)
def install(destPath=None, redownload=False, updateLocalSettings=True):
if hasattr(Settings, "SE10T8_CORPUS"): # Already installed
return
print >> sys.stderr, "---------------", "Downloading the SemEval 2010 Task 8 corpus", "---------------"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "resources/SemEval2010_task8_all_data.zip")
Utils.Download.download(Settings.URL["SE10T8_CORPUS"], destPath, addName=False, clear=redownload)
Settings.setLocal("SE10T8_CORPUS", destPath, updateLocalSettings)
def installPreconverted(url="BIONLP_CORPORA", destPath=None, downloadPath=None, redownload=False, updateLocalSettings=False):
print >> sys.stderr, "---------------", "Downloading preconverted corpora", "---------------"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "corpora")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "corpora/download")
Utils.Download.downloadAndExtract(Settings.URL[url], destPath, downloadPath, redownload=redownload)
Settings.setLocal("CORPUS_DIR", destPath, updateLocalSettings)
def installBBData(destPath=None, downloadPath=None, redownload=False, updateLocalSettings=False):
print >>sys.stderr, "---------------", "Downloading TEES data files for BB", "---------------"
print >>sys.stderr, "Bacteria tokens derived from LPSN (http://www.bacterio.cict.fr/)"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "resources")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "resources/download")
Utils.Download.downloadAndExtract(Settings.URL["TEES_RESOURCES"], destPath, downloadPath, redownload=redownload)
Settings.setLocal("TEES_RESOURCES", destPath, updateLocalSettings)
def installEvaluators(destPath=None, downloadPath=None, redownload=False, updateLocalSettings=False):
print >> sys.stderr, "---------------", "Downloading BioNLP Shared Task evaluators", "---------------"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "tools/evaluators")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "tools/download")
Utils.Download.downloadAndExtract(Settings.URL["BIONLP11_EVALUATORS"], destPath, downloadPath, redownload=redownload)
Settings.setLocal("BIONLP_EVALUATOR_DIR", destPath, updateLocalSettings)
Settings.setLocal("BIONLP_EVALUATOR_GOLD_DIR", os.path.join(destPath, "gold"), updateLocalSettings)
def installEvaluators(destPath=None, downloadPath=None, redownload=False, updateLocalSettings=False):
print >> sys.stderr, "---------------", "Downloading BioNLP Shared Task evaluators", "---------------"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "tools/evaluators")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "tools/download")
Utils.Download.downloadAndExtract(Settings.URL["BIONLP11_EVALUATORS"], destPath, downloadPath, redownload=redownload)
Settings.setLocal("BIONLP_EVALUATOR_DIR", destPath, updateLocalSettings)
Settings.setLocal("BIONLP_EVALUATOR_GOLD_DIR", os.path.join(destPath, "gold"), updateLocalSettings)
def finalizeInstall(programs, testCommand={}, programDir=None, settings={}, updateLocalSettings=False):
if checkPrograms(programs, testCommand, programDir):
setVariable = updateLocalSettings
else:
print >> sys.stderr, "All programs may not have been installed correctly"
print >> sys.stderr, "Do not use the following settings if not sure:"
setVariable = False
for key in sorted(settings.keys()):
Settings.setLocal(key, settings[key], setVariable)
def installRENData(destPath=None, downloadPath=None, redownload=False, updateLocalSettings=False):
print >> sys.stderr, "---------------", "Downloading TEES data files for REN", "---------------"
print >> sys.stderr, "These files are derived from UniProt bacsu and SubtiWiki"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "resources")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "resources/download")
Utils.Download.downloadAndExtract(Settings.URL["TEES_RESOURCES"], destPath, downloadPath, redownload=redownload)
Settings.setLocal("TEES_RESOURCES", destPath, updateLocalSettings)
def installDrugBank(destPath=None, downloadPath=None, redownload=False, updateLocalSettings=False):
print >> sys.stderr, "---------------", "Downloading Drug Bank XML", "---------------"
print >> sys.stderr, "See http://www.drugbank.ca/downloads for conditions of use"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "resources")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "resources/download")
filenames = Utils.Download.downloadAndExtract(Settings.URL["DRUG_BANK_XML"], destPath, downloadPath, redownload=redownload)
assert len(filenames) == 1
Settings.setLocal("DRUG_BANK_XML", os.path.join(destPath, filenames[0]), updateLocalSettings)
def installDrugBank(destPath=None, downloadPath=None, redownload=False, updateLocalSettings=False):
print >> sys.stderr, "---------------", "Downloading Drug Bank XML", "---------------"
print >> sys.stderr, "See http://www.drugbank.ca/downloads for conditions of use"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "resources")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "resources/download")
filenames = Utils.Download.downloadAndExtract(Settings.URL["DRUG_BANK_XML"], destPath, downloadPath, redownload=redownload)
assert len(filenames) == 1
Settings.setLocal("DRUG_BANK_XML", os.path.join(destPath, filenames[0]), updateLocalSettings)
def finalizeInstall(programs,
testCommand={},
programDir=None,
settings={},
updateLocalSettings=False):
if checkPrograms(programs, testCommand, programDir):
setVariable = updateLocalSettings
else:
print >> sys.stderr, "All programs may not have been installed correctly"
print >> sys.stderr, "Do not use the following settings if not sure:"
setVariable = False
for key in sorted(settings.keys()):
Settings.setLocal(key, settings[key], setVariable)
def installRENData(destPath=None,
downloadPath=None,
redownload=False,
updateLocalSettings=False):
print >> sys.stderr, "---------------", "Downloading TEES data files for REN", "---------------"
print >> sys.stderr, "These files are derived from UniProt bacsu and SubtiWiki"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "resources")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "resources/download")
Utils.Download.downloadAndExtract(Settings.URL["TEES_RESOURCES"],
destPath,
downloadPath,
redownload=redownload)
Settings.setLocal("TEES_RESOURCES", destPath, updateLocalSettings)
def installBBData(destPath=None,
downloadPath=None,
redownload=False,
updateLocalSettings=False):
print >> sys.stderr, "---------------", "Downloading TEES data files for BB", "---------------"
print >> sys.stderr, "Bacteria tokens derived from LPSN (http://www.bacterio.cict.fr/)"
if destPath == None:
destPath = os.path.join(Settings.DATAPATH, "resources")
if downloadPath == None:
downloadPath = os.path.join(Settings.DATAPATH, "resources/download")
Utils.Download.downloadAndExtract(Settings.URL["TEES_RESOURCES"],
destPath,
downloadPath,
redownload=redownload)
Settings.setLocal("TEES_RESOURCES", destPath, updateLocalSettings)
def finalizeInstall(programs, testCommand={}, programDir=None, settings={}, updateLocalSettings=False):
installOK = checkPrograms(programs, testCommand, programDir)
if installOK:
setVariable = updateLocalSettings
else:
print >> sys.stderr, "All programs may not have been installed correctly"
print >> sys.stderr, "Do not use the following settings if not sure:"
setVariable = False
for key in sorted(settings.keys()):
if settings[key] != None:
#raise Exception("Local setting " + str(key) + " is undefined")
Settings.setLocal(key, settings[key], setVariable)
else:
print >> sys.stderr, "Warning, local setting " + str(key) + " is undefined"
if not installOK:
raise Exception("Error installing programs: " + ", ".join(programs))
def finalizeInstall(programs,
testCommand={},
programDir=None,
settings={},
updateLocalSettings=False):
installOK = checkPrograms(programs, testCommand, programDir)
if installOK:
setVariable = updateLocalSettings
else:
print >> sys.stderr, "All programs may not have been installed correctly"
print >> sys.stderr, "Do not use the following settings if not sure:"
setVariable = False
for key in sorted(settings.keys()):
if settings[key] != None:
#raise Exception("Local setting " + str(key) + " is undefined")
Settings.setLocal(key, settings[key], setVariable)
else:
print >> sys.stderr, "Warning, local setting " + str(
key) + " is undefined"
if not installOK:
raise Exception("Error installing programs: " + ", ".join(programs))
def setKey(key, value):
if value == None:
print >> sys.stderr, "No value defined for key '" + str(key) + "'"
return
if key not in Settings.KEY_TYPE:
print >> sys.stderr, "Unknown key '" + str(key) + "'"
return
if Settings.KEY_TYPE[key].get("type") == "file":
fullPath = os.path.abspath(value)
if not os.path.exists(os.path.abspath(value)):
print >> sys.stderr, "No file at '" + fullPath + "'"
return
if not os.path.isfile(fullPath):
print >> sys.stderr, "'" + fullPath + "' is not a file"
return
expectedMD5 = Settings.KEY_TYPE[key].get("md5")
if Settings.KEY_TYPE[key].get("md5") != None:
print >> sys.stderr, "Determining MD5 for '" + fullPath + "'"
md5 = FileUtils.getFileMd5(fullPath)
if md5 != expectedMD5:
print >> sys.stderr, "MD5 '" + md5 + "' does not match expected value '" + expectedMD5 + "'"
return
#print >> sys.stderr, "Defining value '" + fullPath + "' for key '" + key + "'"
Settings.setLocal(key, value)
def setKey(key, value):
if value == None:
print >> sys.stderr, "No value defined for key '" + str(key) + "'"
return
if key not in Settings.KEY_TYPE:
print >> sys.stderr, "Unknown key '" + str(key) + "'"
return
if Settings.KEY_TYPE[key].get("type") == "file":
fullPath = os.path.abspath(value)
if not os.path.exists(os.path.abspath(value)):
print >> sys.stderr, "No file at '" + fullPath + "'"
return
if not os.path.isfile(fullPath):
print >> sys.stderr, "'" + fullPath + "' is not a file"
return
expectedMD5 = Settings.KEY_TYPE[key].get("md5")
if Settings.KEY_TYPE[key].get("md5") != None:
print >> sys.stderr, "Determining MD5 for '" + fullPath + "'"
md5 = FileUtils.getFileMd5(fullPath)
if md5 != expectedMD5:
print >> sys.stderr, "MD5 '" + md5 + "' does not match expected value '" + expectedMD5 + "'"
return
#print >> sys.stderr, "Defining value '" + fullPath + "' for key '" + key + "'"
Settings.setLocal(key, value)
def corpusMenuInitializer(menu, prevMenu):
menu.text = """
The corpora are used for training new models and testing existing
models. The corpora installable here are from the three BioNLP Shared
Tasks (2009, 2011 and 2013) on Event Extraction (organized by
University of Tokyo), and the two Drug-Drug Interaction Extraction
tasks (DDI'11 and 13, organized by Universidad Carlos III de Madrid).
The 2009 and 2011 corpora are downloaded as interaction XML files,
generated from the original Shared Task files. If you need to convert
the corpora from the original files, you can use the convertBioNLP.py,
convertDDI.py and convertDDI13.py programs located at Utils/Convert.
The 2013 corpora will be converted to interaction XML from the official
corpus files, downloaded automatically from the task websites. Installing
the BioNLP'13 corpora will take about 10 minutes.
It is also recommended to download the official BioNLP Shared Task evaluator
programs, which will be used by TEES when training or testing on those corpora.
"""
# Mark "skip" as default option, this will be re-marked as install if a corpus is missing
menu.setDefault("s")
# If CORPUS_DIR setting is not set set it now
if not hasattr(Settings, "CORPUS_DIR") or getattr(Settings,
"CORPUS_DIR") == None:
Settings.setLocal(
"CORPUS_DIR", os.path.join(menu.system.defaultInstallDir,
"corpora"))
print >> sys.stderr
# Initialize handlers
handlers = []
handlerArgs = []
corpusInstallPath = os.path.join(menu.system.defaultInstallDir, "corpora")
corpusDownloadPath = os.path.join(menu.system.defaultInstallDir,
"corpora/download")
# Check which corpora need to be installed
redownload = menu.optDict["1"].toggle
# 2009-2011 corpora
for corpus in [
"GE11", "EPI11", "ID11", "BB11", "BI11", "CO11", "REL11", "REN11"
]:
if menu.optDict["2"].toggle or (menu != prevMenu
and not checkCorpusInstall(corpus)):
menu.setDefault("i")
menu.optDict["2"].toggle = True
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append([
"BIONLP_11_CORPORA", corpusInstallPath, corpusDownloadPath,
redownload, True
])
break
if menu.optDict["3"].toggle or (menu != prevMenu
and not checkCorpusInstall("GE09")):
menu.setDefault("i")
menu.optDict["3"].toggle = True
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append([
"BIONLP_09_CORPUS", corpusInstallPath, corpusDownloadPath,
redownload, True
])
if menu.optDict["4"].toggle or (
menu != prevMenu
and not checkCorpusInstall("DDI11", ("-train.xml", "-devel.xml"))):
menu.setDefault("i")
menu.optDict["4"].toggle = True
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append([
"DDI_11_CORPUS", corpusInstallPath, corpusDownloadPath, redownload,
True
])
# 2013 corpora
bioNLP13Corpora = [
"GE13", "CG13", "PC13", "GRO13", "GRN13", "BB13T2", "BB13T3"
]
for corpus in bioNLP13Corpora:
if menu.optDict["5"].toggle or (menu != prevMenu
and not checkCorpusInstall(corpus)):
menu.setDefault("i")
menu.optDict["5"].toggle = True
handlers.append(convertBioNLP.convert)
handlerArgs.append([
bioNLP13Corpora, corpusInstallPath, corpusDownloadPath,
redownload, False
])
break
if menu.optDict["6"].toggle or (
menu != prevMenu
and not checkCorpusInstall("DDI13", ("-train.xml", ))):
menu.setDefault("i")
menu.optDict["6"].toggle = False #True
handlers.append(convertDDI13.convertDDI13)
handlerArgs.append([corpusInstallPath, corpusDownloadPath, redownload])
# A handler for installing BioNLP'11 evaluators
evaluatorInstallPath = os.path.join(menu.system.defaultInstallDir,
"tools/evaluators")
evaluatorDownloadPath = os.path.join(menu.system.defaultInstallDir,
"tools/download")
if menu.optDict["7"].toggle or (
menu != prevMenu and
(not hasattr(Settings, "BIONLP_EVALUATOR_DIR")
or getattr(Settings, "BIONLP_EVALUATOR_DIR") == None)):
menu.setDefault("i")
menu.optDict["7"].toggle = True
handlers.append(convertBioNLP.installEvaluators)
handlerArgs.append(
[evaluatorInstallPath, evaluatorDownloadPath, redownload, True])
# Add the handlers to install option
menu.optDict["i"].handler = handlers
menu.optDict["i"].handlerArgs = handlerArgs
def corpusMenuInitializer(menu, prevMenu):
menu.text = """
The corpora are used for training new models and testing existing
models. The corpora installable here are from the three BioNLP Shared
Tasks (2009, 2011 and 2013) on Event Extraction (organized by
University of Tokyo), and the two Drug-Drug Interaction Extraction
tasks (DDI'11 and 13, organized by Universidad Carlos III de Madrid).
The 2009 and 2011 corpora are downloaded as interaction XML files,
generated from the original Shared Task files. If you need to convert
the corpora from the original files, you can use the convertBioNLP.py,
convertDDI.py and convertDDI13.py programs located at Utils/Convert.
The 2013 corpora will be converted to interaction XML from the official
corpus files, downloaded automatically from the task websites. Installing
the BioNLP'13 corpora will take about 10 minutes.
It is also recommended to download the official BioNLP Shared Task evaluator
programs, which will be used by TEES when training or testing on those corpora.
"""
# Mark "skip" as default option, this will be re-marked as install if a corpus is missing
menu.setDefault("s")
# If CORPUS_DIR setting is not set set it now
if not hasattr(Settings, "CORPUS_DIR") or getattr(Settings, "CORPUS_DIR") == None:
Settings.setLocal("CORPUS_DIR", os.path.join(menu.system.defaultInstallDir, "corpora"))
print >> sys.stderr
# Initialize handlers
handlers = []
handlerArgs = []
corpusInstallPath = os.path.join(menu.system.defaultInstallDir, "corpora")
corpusDownloadPath = os.path.join(menu.system.defaultInstallDir, "corpora/download")
# Check which corpora need to be installed
redownload = menu.optDict["1"].toggle
# 2009-2011 corpora
for corpus in ["GE11", "EPI11", "ID11", "BB11", "BI11", "CO11", "REL11", "REN11"]:
if menu.optDict["2"].toggle or (menu != prevMenu and not checkCorpusInstall(corpus)):
menu.setDefault("i")
menu.optDict["2"].toggle = True
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append(["BIONLP_11_CORPORA", corpusInstallPath, corpusDownloadPath, redownload, True])
break
if menu.optDict["3"].toggle or (menu != prevMenu and not checkCorpusInstall("GE09")):
menu.setDefault("i")
menu.optDict["3"].toggle = True
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append(["BIONLP_09_CORPUS", corpusInstallPath, corpusDownloadPath, redownload, True])
if menu.optDict["4"].toggle or (menu != prevMenu and not checkCorpusInstall("DDI11", ("-train.xml", "-devel.xml"))):
menu.setDefault("i")
menu.optDict["4"].toggle = True
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append(["DDI_11_CORPUS", corpusInstallPath, corpusDownloadPath, redownload, True])
# 2013 corpora
bioNLP13Corpora = ["GE13", "CG13", "PC13", "GRO13", "GRN13", "BB13T2", "BB13T3"]
for corpus in bioNLP13Corpora:
if menu.optDict["5"].toggle or (menu != prevMenu and not checkCorpusInstall(corpus)):
menu.setDefault("i")
menu.optDict["5"].toggle = True
#handlers.append(convertBioNLP.convert)
#handlerArgs.append([bioNLP13Corpora, corpusInstallPath, corpusDownloadPath, redownload, False])
handlers.append(convertBioNLP.installPreconverted)
handlerArgs.append(["BIONLP_13_CORPORA", corpusInstallPath, corpusDownloadPath, redownload, True])
break
if menu.optDict["6"].toggle or (menu != prevMenu and not checkCorpusInstall("DDI13", ("-train.xml",))):
menu.setDefault("i")
menu.optDict["6"].toggle = False #True
handlers.append(convertDDI13.convertDDI13)
handlerArgs.append([corpusInstallPath, corpusDownloadPath, ["DDI13_TRAIN", "DDI13_TEST_TASK_9.1", "DDI13_TEST_TASK_9.2"], redownload])
# A handler for installing BioNLP'11 evaluators
evaluatorInstallPath = os.path.join(menu.system.defaultInstallDir, "tools/evaluators")
evaluatorDownloadPath = os.path.join(menu.system.defaultInstallDir, "tools/download")
if menu.optDict["7"].toggle or (menu != prevMenu and (not hasattr(Settings, "BIONLP_EVALUATOR_DIR") or getattr(Settings, "BIONLP_EVALUATOR_DIR") == None)):
menu.setDefault("i")
menu.optDict["7"].toggle = True
handlers.append(convertBioNLP.installEvaluators)
handlerArgs.append([evaluatorInstallPath, evaluatorDownloadPath, redownload, True])
# Add the handlers to install option
menu.optDict["i"].handler = handlers
menu.optDict["i"].handlerArgs = handlerArgs
