def tika(conf, attachments):
"""This method updates the attachments results
with the Tika reports.
Args:
attachments (list): all attachments of email
conf (dict): conf of this post processor
Returns:
This method updates the attachments list given
"""
if conf["enabled"]:
from tikapp import TikaApp
tika = TikaApp(file_jar=conf["path_jar"],
memory_allocation=conf["memory_allocation"])
for a in attachments:
if not a.get("is_filtered", False):
if a["Content-Type"] in conf["whitelist_cont_types"]:
payload = a["payload"]
if a["content_transfer_encoding"] != "base64":
payload = payload.encode("base64")
# tika-app only gets payload in base64
a["tika"] = tika.extract_all_content(payload=payload,
convert_to_obj=True)
def main():
args = get_args()
tika = TikaApp(args.jar or os.environ.get("TIKA_APP_JAR", None))
parameters = {
"path": args.file,
"payload": args.payload,
"objectInput": sys.stdin if args.stdin else None
}
try:
if args.detect:
print(tika.detect_content_type(**parameters))
if args.text:
print(tika.extract_only_content(**parameters))
if args.language:
print(tika.detect_language(**parameters))
if args.all:
parameters["pretty_print"] = True
print(tika.extract_all_content(**parameters))
if args.metadata:
parameters["pretty_print"] = True
print(tika.extract_only_metadata(**parameters))
except IOError:
pass
def run(message):
tika_client = TikaApp(file_jar="./tika-app/tika-app-1.21.jar")
tika_result = tika_client.extract_only_content(message["path"])
if (tika_result != None):
processing_dir = "./data/processing/"
identifier = str(uuid.uuid4())
workfile = processing_dir + identifier
with open(workfile, 'wb') as f:
f.write(tika_result.encode('UTF-8'))
new_message = {
"identifier": identifier,
"parent": message["identifier"],
"path": workfile,
"filename": "pdf.txt",
"filetype": "unknown",
"history": [],
"metadata": {},
"original_file": False
}
sendEvent(new_message)
extract_images_pymupdf(message)
sendEvent(message)
def getTextFile(nameOfPDF, jarPath):
'''
getTextFile("nameOfPDF", path) will take the PDF and
output it as a textfile. Path will be taken and used to
specify the .jar file needed for tika.
Note: Needs the tika-app-1.22.jar and tika-app-python files
and folders in the same working directory
'''
# get the Tika Object from current directory
tika_client = TikaApp(file_jar=join(jarPath, "tika-app-1.22.jar"))
# read the pdf
with open(nameOfPDF) as fin:
# get rid of the .pdf to change to .txt for later
if nameOfPDF[-4:] == ".pdf":
foutName = nameOfPDF[:-4]
else:
# if its not .pdf, then just keep the filename
foutName = nameOfPDF
content = tika_client.extract_all_content(objectInput=fin)
# write the pdf to a text file & add .txt to it
with open(foutName + ".txt", "w", encoding='utf-8',
errors='replace') as fout:
fout.write(content)
return foutName + ".txt"
def __init__(self,
index,
es_host="localhost:9200",
file_jar='/Users/laofeng/es_home/tika-app-1.23.jar',
index_content=False,
create_index=True,
force_renew_index=False,
schema_file=None):
self.tika_app = TikaApp(file_jar)
self.es = Elasticsearch(es_host)
self.index = index
self.index_content = index_content
#查询索引是否存在
index_exist = self.es.indices.exists(index)
#if not index_exist and not create_index:
if not index_exist and create_index and schema_file:
with open(schema_file, 'r', encoding='utf-8') as f:
schema = json.load(f)
self.es.indices.create(index, schema)
#如果已经存在，且强制renew，先删除后，再建立
if index_exist and force_renew_index:
self.es.indices.delete(index)
with open(schema_file, 'r', encoding='utf-8') as f:
schema = json.load(f)
self.es.indices.create(index, schema)
def __init__(self,
jar=None,
memory_allocation=None,
valid_content_types=set()):
# Init Tika
self._tika_client = TikaApp(file_jar=jar,
memory_allocation=memory_allocation)
self._jar = jar
self._memory_allocation = memory_allocation
self._valid_content_types = valid_content_types
def run(message):
# Text
tika_client = TikaApp(file_jar="./tika-app/tika-app-1.21.jar")
tika_result = tika_client.extract_only_content(message["path"])
if (tika_result != None):
processing_dir = "./data/processing/"
identifier = str(uuid.uuid4())
workfile = processing_dir + identifier
with open(workfile, 'wb') as f:
f.write(tika_result.encode('UTF-8'))
new_message = {
"identifier": identifier,
"parent": message["identifier"],
"path": workfile,
"filename" : "doc.txt",
"filetype": "unknown",
"history": [],
"metadata": {},
"original_file": False
}
sendEvent(new_message)
# Images
with ZipFile(message["path"], 'r') as zipObj:
processng_dir = "./data/processing"
tmp_identifier = os.path.join(processng_dir, str(uuid.uuid4()))
zipObj.extractall(tmp_identifier)
for root, dirs, files in os.walk(tmp_identifier):
for filename in files:
if ".png" in filename or ".jpeg" in filename or ".jpg" in filename:
new_identifier = str(uuid.uuid4())
processing_dest = os.path.join(processng_dir, new_identifier)
new_message = {
"identifier": new_identifier,
"parent": message["identifier"],
"path": processing_dest,
"filename" : filename,
"filetype": "unknown",
"history": [],
"metadata": {},
"original_file": False
}
shutil.move(os.path.join(root, filename), processing_dest)
sendEvent(new_message)
shutil.rmtree(tmp_identifier)
sendEvent(message)
def tika(conf, attachments):
"""This method updates the attachments results
with the Tika reports.
Args:
attachments (list): all attachments of email
conf (dict): conf of this post processor
Returns:
This method updates the attachments list given
"""
if conf["enabled"]:
from tikapp import TikaApp
tika = TikaApp(file_jar=conf["path_jar"],
memory_allocation=conf["memory_allocation"])
wtlist = conf.get("whitelist_content_types", [])
if not wtlist:
log.warning(
"Apache Tika analysis setted, without whitelist content types")
return
for a in attachments:
if not a.get("is_filtered", False):
if a["Content-Type"] in wtlist:
payload = a["payload"]
if a["content_transfer_encoding"] != "base64":
try:
payload = payload.encode("base64")
except UnicodeError:
# content_transfer_encoding': u'x-uuencode'
# it's not binary with strange encoding
continue
# tika-app only gets payload in base64
try:
results = tika.extract_all_content(payload=payload,
convert_to_obj=True)
if results:
a["tika"] = results
except JSONDecodeError:
log.warning(
"JSONDecodeError for {!r} in Tika analysis".format(
a["md5"]))
def main():
args = get_args()
command_line = dict()
if args.jar:
command_line = {"TIKA_APP_JAR": args.jar}
defaults = {"TIKA_APP_JAR": "/opt/tika/tika-app-1.15.jar"}
options = ChainMap(command_line, os.environ, defaults)
tika = TikaApp(options['TIKA_APP_JAR'])
try:
if args.file:
f = args.file
if args.detect:
print(tika.detect_content_type(path=f))
if args.text:
print(tika.extract_only_content(path=f))
if args.language:
print(tika.detect_language(path=f))
if args.all:
print(tika.extract_all_content(path=f, pretty_print=True))
elif args.payload:
p = args.payload
if args.detect:
print(tika.detect_content_type(payload=p))
if args.text:
print(tika.extract_only_content(payload=p))
if args.language:
print(tika.detect_language(payload=p))
if args.all:
print(tika.extract_all_content(payload=p, pretty_print=True))
except IOError:
pass
def __init__(self, tika_jar_path):
self.tika_client = TikaApp(file_jar=tika_jar_path)
def analyse_pdf_archive(pdf_csv, keyword_csv, tika_file_jar, outfile_name):
# PDF for each day saved as welt_mmdd. File names listed in csv file 'welt_pdf'. Create list of PDF names.
with open(pdf_csv, 'r', encoding = 'utf-8-sig') as f:
reader = csv.reader(f)
pdf_names = list(reader)
pdf_names = list(itertools.chain(*pdf_names)) # acts as main data frame to contain individual data frames
pdf_list = []
for name in pdf_names:
pdf_list.append(name + '.pdf')
# Create list of keywords from 'keyword_stems.csv'
with open(keyword_csv, 'r', encoding = 'utf-8-sig') as f:
reader = csv.reader(f)
keywords = list(reader)
keywords = list(itertools.chain(*keywords))
tika_client = TikaApp(file_jar=tika_file_jar)
a = 1 #set counter to 1
keyword_counter = []
for pdf in pdf_list:
rawText = tika_client.extract_only_content(pdf)
print("pdf {0} extracted".format(a))
rawList = rawText.split( )
rawList_nopunct = [word.translate(str.maketrans('', '', string.punctuation)) for word in rawList]
counts = Counter(rawList_nopunct)
list_words = counts.most_common()
keyword_hits_list = []
for x in range(0, len(list_words)):
temp = list(list_words[x]) # convert from tuple to list
temp[1] = str(temp[1]) # change number (at index 1) into string
n_temp = [(unicodedata.normalize('NFKD', word).encode('ASCII', 'ignore')).lower().decode() for word in temp] #normalised umlauts in data
#check word (at index 0) against list of keywords, add new column = 1 if match, = 0 otherwise.
hits = 0
for i in range(0, len(keywords)):
if keywords[i] in n_temp[0]:
hits = hits + 1
if hits != 0:
n_temp.append(1)
else:
n_temp.append(0)
keyword_hits = int(n_temp[1])*int(n_temp[2])
keyword_hits_list.append(keyword_hits)
keyword_counts = sum(keyword_hits_list)
keyword_counter.append(keyword_counts)
print("day {0} complete".format(a))
if list_words != []:
a = a + 1
else:
break
df = pd.DataFrame({"id": pdf_names, "keywords": keyword_counter})
df.to_csv(outfile_name, index=False)
def __init__(self, file_process):
# Cliente Tika que utiliza que carga el fichero jar cliente.
self.tika_client = TikaApp(file_jar="tika-app-1.20.jar")
self.file_process = file_process
from tikapp import TikaApp
tika_client = TikaApp(
file_jar="/Users/yma2/Documents/_garage/python/cxm/tika/tika-app-1.20.jar")
analyzeFile = "/Users/yma2/Downloads/Azure_Developer_Guide_eBook_ja-JP.pdf"
print(tika_client.detect_content_type(analyzeFile))
print(tika_client.detect_language(analyzeFile))
print(tika_client.extract_only_content(analyzeFile))
print(tika_client.extract_only_metadata(analyzeFile))
def handle_backup_file(self, context, savepkt):
DebugMessage(context, 100,
"handle_backup_file called with " + str(savepkt) + "\n")
DebugMessage(
context, 100,
"fname: " + savepkt.fname + " Type: " + str(savepkt.type) + "\n")
if (savepkt.type == bFileType['FT_REG']):
DebugMessage(context, 100, "regulaer file, do something now...\n")
# configure your Elasticsearch server here:
es = Elasticsearch([{'host': '192.168.17.2', 'port': 9200}])
# configure your TikaApp jar file here:
try:
tika_client = TikaApp(
file_jar="/usr/local/bin/tika-app-1.20.jar")
except Exception as ex:
JobMessage(
context, bJobMessageType['M_ERROR'],
'Error indexing %s. Tika error: %s' %
(savepkt.fname, str(ex)))
return bRCs['bRC_OK']
# tika_client has several parser options
# Next one is for metadata only:
#result_payload=tika_client.extract_only_metadata(savepkt.fname)
# This one includes file contents as text:
try:
result_payload = tika_client.extract_all_content(savepkt.fname)
except Exception as ex:
JobMessage(
context, bJobMessageType['M_ERROR'],
'Error extracting contents from %s. Tika error: %s' %
(savepkt.fname, str(ex)))
return bRCs['bRC_OK']
# result_payload is a list of json-strings. Nested structes like
# tar-files or emails with attachments or inline documents are
# returned as distinct json string.
# The first string [0] contains information for the main file
# TODO: care about nested structures, for now we only takte the first/main file
try:
data = json.loads(result_payload)[0]
except Exception as ex:
JobMessage(
context, bJobMessageType['M_ERROR'],
'Error reading json fields delivered by Tika examining file %s. Json error: %s'
% (savepkt.fname, str(ex)))
return bRCs['bRC_OK']
# Tika eventually adds "Unkonwn Tags (id)", with id as increasing number, which
# could lead to exceed the keyword limit in elasticsearch indices, we
# remove those tags
for data_keyword in data.keys():
if data_keyword.startswith("Unknown tag ("):
del data[data_keyword]
# Tika adds some emptylines at the beginning of content, we strip it here
if 'X-TIKA:content' in data:
data['X-TIKA:content'] = data['X-TIKA:content'].strip()
data['bareos_jobId'] = self.jobId
data['bareos_fdname'] = self.fdname
data['bareos_joblevel'] = unichr(self.level)
data['bareos_directory'] = os.path.dirname(savepkt.fname)
try:
esRes = es.index(index="bareos-test",
doc_type='_doc',
body=data)
except Exception as ex:
JobMessage(
context, bJobMessageType['M_ERROR'],
'Error indexing %s. Elastic error: %s' %
(savepkt.fname, str(ex)))
return bRCs['bRC_OK']
def extract_text(filename):
jar_path = os.path.abspath(os.path.join("lib", "tika-app-1.28.jar"))
tika_client = TikaApp(file_jar=jar_path)
parsed = tika_client.extract_only_content(filename)
return parsed
def tika_extract_only_content(memory=None):
tika_client = TikaApp(file_jar=TIKA_APP_JAR, memory_allocation=memory)
output = tika_client.extract_only_content(path=test_zip)
return output
def tika_detect_language():
tika_client = TikaApp(file_jar=TIKA_APP_JAR)
output = tika_client.detect_language(path=test_zip)
return output
def __init__(self, **kwargs):
super(TikaProcessing, self).__init__(**kwargs)
# Init Tika
self._tika_client = TikaApp(file_jar=self.jar,
memory_allocation=self.memory_allocation)
# -*- coding: utf-8 -*-
# @Time : 2020/12/11 10:25
# @Author : 王西亚
# @File : c_doc.py
from tikapp import TikaApp
tika_client = TikaApp(file_jar="/usr/local/Cellar/tika/1.24.1_1/libexec/tika-app-1.24.1.jar")
metadata = tika_client.extract_only_metadata("/Users/wangxiya/Downloads/000101020062805119-00.pdf")
print(type(metadata))
print(metadata)
# from tika import parser
# parsed = parser.from_file('/path/to/file')
# print(parsed["metadata"])
# print(parsed["content"])
# parsed = parser.from_file('/Users/wangxiya/Downloads/000101020062805119-00.pdf', 'http://localhost:9998/tika')
# metadata = parsed["metadata"]
# print(type(metadata))
# print(metadata)
def convert_Tika(self,fname):
tika_client = TikaApp(file_jar=os.getcwd()+'/tika-app-1.20.jar')
return tika_client.extract_only_content(fname)
from tikapp import TikaApp
from image_mod import convert_image_to_string
import os
tika_client = TikaApp(
os.path.join(os.path.dirname(os.path.realpath(__file__)), 'src',
'tika-app-1.22.jar'))
def receive_text_from_file(path: str, ext=None):
text = tika_client.extract_only_content(path)
if text == "" and ext == 'pdf':
return convert_image_to_string(path, ext=ext)
return text
def tika_content_type():
tika_client = TikaApp(file_jar=TIKA_APP_JAR)
output = tika_client.detect_content_type(path=test_zip)
return output
