Tika parser with TesseractOCR

[tarunsharma2015]

I am facing problem while extracting content from pdf, the returned content is None in case of pdf images. The same code seems to be working on my local setup whereas failing on aws lambda.

I have followed to install tesseract on AWS lambda as layer still not able to find the error.

Output from tika parser on AWS lambda -

{'metadata': {'Content-Type': 'application/pdf', 'X-Parsed-By': ['org.apache.tika.parser.DefaultParser', 'org.apache.tika.parser.pdf.PDFParser'], 'X-TIKA:embedded_depth': '0', 'X-TIKA:parse_time_millis': '563', 'access_permission:assemble_document': 'true', 'access_permission:can_modify': 'true', 'access_permission:can_print': 'true', 'access_permission:can_print_degraded': 'true', 'access_permission:extract_content': 'true', 'access_permission:extract_for_accessibility': 'true', 'access_permission:fill_in_form': 'true', 'access_permission:modify_annotations': 'true', 'dc:format': 'application/pdf; version=1.4', 'pdf:PDFVersion': '1.4', 'pdf:charsPerPage': '0', 'pdf:encrypted': 'false', 'pdf:hasMarkedContent': 'false', 'pdf:hasXFA': 'false', 'pdf:hasXMP': 'false', 'pdf:unmappedUnicodeCharsPerPage': '0', 'resourceName': "b'test-img-tika-1.pdf'", 'xmpTPg:NPages': '1'}, 'content': None, 'status': 200}

Tesseract Path - All libraries placed under lib folder of zip to be uploaded on AWS lambda.

Libraries - liblept.so.5, libtesseract.so.4,libwebp.so.4
Code -

import boto3
import tika
from tika import parser
import tempfile, os
import requests
from urllib.request import urlretrieve
from tika import detector

SCRIPT_DIR = os.path.dirname(os.path.abspath(file))
LIB_DIR = os.path.join(SCRIPT_DIR, 'lib')
print(f"library path : {LIB_DIR}")
os.system(f'export LD_LIBRARY_PATH={LIB_DIR}')

def lambda_handler(event, context):
tika.initVM()
headers = {"X-Tika-OCRLanguage": "eng"}
parsed = parser.from_file(filepath, headers=headers)
print(parsed)

PS : the same code seems to be working fine if there is no image content in pdf.

[Wiktor]

Hey, have a look at fix I proposed #364, and then set headers in from_file() to headers={'X-Tika-PDFextractInlineImages': 'true'}

[Chris Mattmann]

Yes merged #364 so hopefully this addressed it.