def ocr_page_image(doc_path, page_num, lang):
"""
image = jpg, jpeg, png
"""
logger.debug("OCR image (jpeg, jpg, png) document")
page_url = PagePath(
document_path=doc_path,
page_num=page_num,
step=Step(1),
# jpeg, jpg, png are 1 page documents
page_count=1)
# resize and eventually convert (png -> jpg)
resize_img(page_url, media_root=settings.MEDIA_ROOT)
extract_txt(page_url, lang=lang, media_root=settings.MEDIA_ROOT)
# First quickly generate preview images
for step in Steps():
page_url.step = step
resize_img(page_url, media_root=settings.MEDIA_ROOT)
# reset page's step
page_url.step = Step(1)
# Now OCR each image
for step in Steps():
if not step.is_thumbnail:
extract_hocr(page_url, lang=lang, media_root=settings.MEDIA_ROOT)
return page_url
def ocr_page_pdf(doc_path, page_num, lang):
"""
doc_path is an mglib.path.DocumentPath instance
"""
logger.debug("OCR PDF document")
page_count = get_pagecount(default_storage.abspath(doc_path.url()))
if page_num <= page_count:
# first quickly generate preview images
page_url = PagePath(document_path=doc_path,
page_num=page_num,
step=Step(1),
page_count=page_count)
for step in Steps():
page_url.step = step
extract_img(page_url, media_root=settings.MEDIA_ROOT)
if page_num <= page_count:
page_url = PagePath(document_path=doc_path,
page_num=page_num,
step=Step(1),
page_count=page_count)
extract_txt(page_url, lang=lang, media_root=settings.MEDIA_ROOT)
for step in Steps():
page_url.step = step
if not step.is_thumbnail:
extract_hocr(page_url,
lang=lang,
media_root=settings.MEDIA_ROOT)
return page_url
def ocr_page_pdf(doc_path, page_num, lang, **kwargs):
"""
doc_path is an mglib.path.DocumentPath instance
On success returns ``mglib.path.PagePath`` instance.
"""
logger.debug("OCR PDF document")
file_name = kwargs.pop('file_name', None)
if not file_name:
file_name = doc_path.file_name
page_count = get_pagecount(default_storage.abspath(doc_path.url()))
if page_num <= page_count:
# first quickly generate preview images
page_path = PagePath(document_path=doc_path,
page_num=page_num,
step=Step(1),
page_count=page_count)
for step in Steps():
page_path.step = step
extract_img(page_path, media_root=settings.MEDIA_ROOT)
notify_pre_page_ocr(page_path,
page_num=page_num,
lang=lang,
file_name=doc_path.file_name,
**kwargs)
if page_num <= page_count:
page_path = PagePath(document_path=doc_path,
page_num=page_num,
step=Step(1),
page_count=page_count)
extract_txt(page_path, lang=lang, media_root=settings.MEDIA_ROOT)
notify_txt_ready(page_path,
page_num=page_num,
lang=lang,
file_name=file_name,
**kwargs)
for step in Steps():
page_path.step = step
if not step.is_thumbnail:
extract_hocr(page_path,
lang=lang,
media_root=settings.MEDIA_ROOT)
notify_hocr_ready(
page_path,
page_num=page_num,
lang=lang,
# step as integer number
step=step.current,
file_name=file_name,
**kwargs)
return page_path
def ocr_page_image(doc_path, page_num, lang, **kwargs):
"""
image = jpg, jpeg, png
On success returns ``mglib.path.PagePath`` instance.
"""
logger.debug("OCR image (jpeg, jpg, png) document")
page_path = PagePath(
document_path=doc_path,
page_num=page_num,
step=Step(1),
# jpeg, jpg, png are 1 page documents
page_count=1)
notify_pre_page_ocr(page_path,
page_num=page_num,
lang=lang,
file_name=doc_path.file_name,
**kwargs)
# resize and eventually convert (png -> jpg)
resize_img(page_path, media_root=settings.MEDIA_ROOT)
extract_txt(page_path, lang=lang, media_root=settings.MEDIA_ROOT)
notify_txt_ready(page_path,
page_num=page_num,
lang=lang,
file_name=doc_path.file_name,
**kwargs)
# First quickly generate preview images
for step in Steps():
page_path.step = step
resize_img(page_path, media_root=settings.MEDIA_ROOT)
# reset page's step
page_path.step = Step(1)
# Now OCR each image
for step in Steps():
if not step.is_thumbnail:
extract_hocr(page_path, lang=lang, media_root=settings.MEDIA_ROOT)
notify_hocr_ready(
page_path,
page_num=page_num,
lang=lang,
# step as integer number
step=step.current,
file_name=doc_path.file_name,
**kwargs)
return page_path
def ocr_page_pdf(doc_path, page_num, lang):
"""
doc_path is an mglib.path.DocumentPath instance
"""
page_count = get_pagecount(default_storage.abspath(doc_path.url()))
if page_num <= page_count:
page_url = PagePath(document_path=doc_path,
page_num=page_num,
step=Step(1),
page_count=page_count)
extract_img(page_url, media_root=settings.MEDIA_ROOT)
extract_txt(page_url, lang=lang, media_root=settings.MEDIA_ROOT)
for step in Steps():
page_url.step = step
extract_img(page_url, media_root=settings.MEDIA_ROOT)
# tesseract unterhalt-1.jpg page-1 -l deu hocr
if not step.is_thumbnail:
extract_hocr(page_url,
lang=lang,
media_root=settings.MEDIA_ROOT)
return page_url