def ocr_page(
user_id,
document_id,
file_name,
page_num,
lang,
):
logger.debug(f" ocr_page user_id={user_id} doc_id={document_id}"
f" page_num={page_num}")
t1 = time.time()
lang = lang.lower()
doc_path = DocumentPath(
user_id=user_id,
document_id=document_id,
file_name=file_name,
)
mime_type = mime.Mime(default_storage.abspath(doc_path.url()))
logger.debug(f"Mime Type = {mime_type}")
page_type = ''
if mime_type.is_pdf():
ocr_page_pdf(doc_path=doc_path, page_num=page_num, lang=lang)
page_type = 'pdf'
elif mime_type.is_image(): # jpeg, jpeg or png
ocr_page_image(doc_path=doc_path, page_num=page_num, lang=lang)
elif mime_type.is_tiff():
# new filename is a pdf file
logger.debug("TIFF type detected")
new_filename = convert_tiff2pdf(
doc_url=default_storage.abspath(doc_path.url()))
# now .pdf
doc_path.file_name = new_filename
# and continue as usual
ocr_page_pdf(doc_path=doc_path, page_num=page_num, lang=lang)
else:
logger.error(f" user_id={user_id}"
f" doc_id={document_id}"
f" page_num={page_num} error=Unkown file type")
return True
t2 = time.time()
logger.debug(f" user_id={user_id} doc_id={document_id}"
f" page_num={page_num} page_type={page_type}"
f" total_exec_time={t2-t1:.2f}")
return True
def ocr_page(
user_id,
document_id,
file_name,
page_num,
lang,
namespace=None,
):
logger.debug(f" ocr_page user_id={user_id} doc_id={document_id}"
f" page_num={page_num}")
t1 = time.time()
lang = lang.lower()
doc_path = DocumentPath(
user_id=user_id,
document_id=document_id,
file_name=file_name,
)
if not default_storage.exists(doc_path.url()):
# In case of distibuted deployment, document uploaded
# by webapp is not directly available to the worker (which runs on
# separate computer). Thus, if document is not locally available,
# worker will download the document from whatever remote location.
default_storage.download(doc_path_url=doc_path.url(),
namespace=namespace)
mime_type = mime.Mime(default_storage.abspath(doc_path.url()))
logger.debug(f"Mime Type = {mime_type}")
page_type = ''
if mime_type.is_pdf():
ocr_page_pdf(doc_path=doc_path,
page_num=page_num,
lang=lang,
user_id=user_id,
document_id=document_id,
namespace=namespace)
page_type = 'pdf'
elif mime_type.is_image(): # jpeg, jpeg or png
ocr_page_image(doc_path=doc_path,
page_num=page_num,
lang=lang,
user_id=user_id,
document_id=document_id,
namespace=namespace)
elif mime_type.is_tiff():
# new filename is a pdf file
logger.debug("TIFF type detected")
new_filename = convert_tiff2pdf(
doc_url=default_storage.abspath(doc_path.url()))
# now .pdf
orig_file_name = doc_path.file_name
doc_path.file_name = new_filename
# and continue as usual
ocr_page_pdf(
doc_path=doc_path,
page_num=page_num,
lang=lang,
user_id=user_id,
document_id=document_id,
# Pass original file_name i.e. tiff file name as well.
file_name=orig_file_name,
namespace=namespace)
else:
logger.error(f" user_id={user_id}"
f" doc_id={document_id}"
f" page_num={page_num} error=Unkown file type")
return True
t2 = time.time()
logger.debug(f" user_id={user_id} doc_id={document_id}"
f" page_num={page_num} page_type={page_type}"
f" total_exec_time={t2-t1:.2f}")
return True