how to directly read tika server.jar file from local or server fixed dir, include first time run

[sunweiconfidence]

i debug tika to parse file, found that first time it need to download and install it from remote url, it take long time, if i will deploy the program to app server, no network, how do i deal with it? what method i can solve it by coding it from fixed directory? thanks

[Harshitg10]

Set TIKA_SERVER_JAR before you import tika package os.environ["TIKA_SERVER_JAR"] = << file location >>

[sunweiconfidence]

ok, thanks

[raj5287]

@sunweiconfidence hey, did it resolve the issue? I am doing this :

import os
os.environ["TIKA_SERVER_JAR"] = "/home/user/Downloads/tika-server-1.19.jar"
from tika import parser

raw = parser.from_file('sample.pdf')
print(raw['content'])

but still it is going online to download the file and throwing me an error.

[sunweiconfidence]

@raj5287 yes, i resolve the issue, you need to change the code in tika's source code like this path's file C:\Program Files\Python36\Lib\site-packages\tika\tika.py

[luke4u]

@sunweiconfidence @raj5287 @chrismattmann , I still got error after adding this at top oftika.py os.environ["TIKA_SERVER_JAR"] = r'C:\Users\Desktop\deep_pavlov\tika_server\tika-server.jar' . Not sure what I missed, can you share more insights?

error below: Retrieving C:\Users\Desktop\deep_pavlov\tika_server\tika-server.jar to C:\Users\AppData\Local\Temp\tika-server.jar.

URLError: <urlopen error unknown url type: c>

[chrismattmann]

I think you need to add a file: in front of your string, like this:

os.environ["TIKA_SERVER_JAR"] = r'file:C:\Users\Desktop\deep_pavlov\tika_server\tika-server.jar'

[sunweiconfidence]

@luke4u my complete code change as below: code change path: C:\Program Files\Python36\Lib\site-packages\tika\tika.py code change point as below bold text: log_path = os.getenv('TIKA_LOG_PATH', 'D:/inetpub/wwwroot/bertapi/Tika/') log_file = os.path.join(log_path, 'tika.log')

logFormatter = logging.Formatter("%(asctime)s [%(threadName)-12.12s] [%(levelname)-5.5s] %(message)s") log = logging.getLogger('tika.tika')

File logs

fileHandler = logging.FileHandler(log_file) fileHandler.setFormatter(logFormatter) log.addHandler(fileHandler)

Stdout logs

consoleHandler = logging.StreamHandler() consoleHandler.setFormatter(logFormatter) log.addHandler(consoleHandler)

Log level

log.setLevel(logging.INFO)

Windows = True if platform.system() == "Windows" else False TikaVersion = os.getenv('TIKA_VERSION', '1.19') TikaJarPath = os.getenv('TIKA_PATH', 'D:/inetpub/wwwroot/bertapi/Tika/') TikaFilesPath = "D:/inetpub/wwwroot/bertapi/Tika/" TikaServerLogFilePath = log_path TikaServerJar = os.getenv( 'TIKA_SERVER_JAR', TikaFilesPath + "tika-server"+ ".jar") ServerHost = "localhost"

by the way,D:/inetpub/wwwroot/bertapi/Tika/ this path, you need have tika-server.jar, tika-server.jar.md5, tika-server.log,tika.log file in advance

[luke4u]

Thanks guys. problem solved!

[ruohola]

Thanks for the solutions. I wanted to keep the .jar next to my code. To recap what worked:

File structure:

.
├── lib
│   ├── tika-server.jar
│   └── tika-server.jar.md5
└── main.py

Code in main.py:

import os
from pathlib import Path

ROOT_DIR = Path(__file__).resolve(strict=True).parent

os.environ["TIKA_SERVER_JAR"] = f"file://{ROOT_DIR / 'lib/tika-server.jar'}"

import tika.parser

# Use tika.parser