def test_extract_words_from(self):
hocr_file = os.path.join(BASE_DIR, "data", "page-1.hocr")
hocr = Hocr(hocr_file_path=hocr_file)
try:
json.dumps({
'hocr': hocr.good_json_words(),
'hocr_meta': hocr.get_meta()
})
except TypeError:
self.assertTrue(False, "Unserializable result")
def test_extract_img_size(self):
hocr_file = os.path.join(BASE_DIR, "data", "page-1.hocr")
hocr = Hocr(hocr_file_path=hocr_file)
self.assertEqual(hocr.width, 1240)
self.assertEqual(hocr.height, 1754)
def hocr(request, id, step=None, page="1"):
logger.debug(f"hocr for doc_id={id}, step={step}, page={page}")
try:
doc = Document.objects.get(id=id)
except Document.DoesNotExist:
raise Http404("Document does not exists")
doc_ep = doc.doc_ep
if request.user.has_perm(Access.PERM_READ, doc):
if not doc_ep.exists():
download(doc_ep)
page_count = get_pagecount(doc_ep.url())
if page > page_count or page < 0:
raise Http404("Page does not exists")
page_ep = doc.page_eps[page]
logger.debug(f"Extract words from {page_ep.hocr_url()}")
if not page_ep.hocr_exists():
# check if HOCR data exists on S3
if settings.S3 and page_ep.hocr_exists(ep=Endpoint.S3):
# ok, it should be able to download it.
download_hocr(page_ep)
else:
# normal scenario, HOCR is not yet ready
raise Http404("HOCR data not yet ready.")
# At this point local HOCR data should be available.
hocr = Hocr(
hocr_file_path=page_ep.hocr_url()
)
return HttpResponse(
json.dumps({
'hocr': hocr.good_json_words(),
'hocr_meta': hocr.get_meta()
}),
content_type="application/json",
)
return HttpResponseForbidden()
def hocr(request, id, step=None, page="1"):
logger.debug(f"hocr for doc_id={id}, step={step}, page={page}")
try:
doc = Document.objects.get(id=id)
except Document.DoesNotExist:
raise Http404("Document does not exists")
doc_path = doc.path
if request.user.has_perm(Access.PERM_READ, doc):
# document absolute path
doc_abs_path = default_storage.abspath(doc_path.url())
if not os.path.exists(
doc_abs_path
):
raise Http404("HOCR data not yet ready.")
page_count = get_pagecount(doc_abs_path)
if page > page_count or page < 0:
raise Http404("Page does not exists")
page_path = doc.page_paths[page]
hocr_abs_path = default_storage.abspath(page_path.hocr_url())
logger.debug(f"Extract words from {hocr_abs_path}")
if not os.path.exists(hocr_abs_path):
raise Http404("HOCR data not yet ready.")
# At this point local HOCR data should be available.
hocr = Hocr(
hocr_file_path=hocr_abs_path
)
return HttpResponse(
json.dumps({
'hocr': hocr.good_json_words(),
'hocr_meta': hocr.get_meta()
}),
content_type="application/json",
)
return HttpResponseForbidden()
def test_empty_file_hocr(self):
"""
If empty or invalid file is provided then
json_good_words() and get_meta()
will return empty list.
"""
file = tempfile.NamedTemporaryFile(mode="r+t")
hocr = Hocr(hocr_file_path=file.name)
self.assertEqual(hocr.good_json_words(), [])
meta = hocr.get_meta()
self.assertEqual(
meta, {
'count_all': 0,
'bad_words': [],
'count_good': 0,
'count_bad': 0,
'count_non_empty': 0,
'count_low_wconf': 0,
'width': 0,
'height': 0,
'min_wconf': 30
})
file.close()