def test_version():
e = ocr.Engine("")
version = e.get_version()
version_regex = re.compile("[0-9][.][0-9]{1,2}[.][0-9]{1,2}")
print(version)
assert version_regex.match(version)
def set_tess_path(cls, path: str = None) -> None:
if path is None:
path = locate_tessdata()
assert path is not None
assert os.path.exists(path)
cls.engine = ocr.Engine(path)
assert cls.engine is not None
class GenerateOCRFileTask(speedwagon.tasks.Subtask):
engine = ocr.Engine(locate_tessdata())
def __init__(self,
source_image,
out_text_file,
lang="eng",
tesseract_path=None) -> None:
super().__init__()
self._source = source_image
self._output_text_file = out_text_file
self._lang = lang
GenerateOCRFileTask.set_tess_path(tesseract_path or locate_tessdata())
assert self.engine is not None
@classmethod
def set_tess_path(cls, path=locate_tessdata()):
assert path is not None
assert os.path.exists(path)
cls.engine = ocr.Engine(path)
assert cls.engine is not None
def work(self) -> bool:
# Get the ocr text reader for the proper language
reader = self.engine.get_reader(self._lang)
self.log("Reading {}".format(os.path.normcase(self._source)))
f = io.StringIO()
with contextlib.redirect_stderr(f):
# Capture the warning messages
resulting_text = reader.read(self._source)
stderr_messages = f.getvalue()
if stderr_messages:
# Log any error messages
self.log(stderr_messages.strip())
# Generate a text file from the text data extracted from the image
self.log("Writing to {}".format(self._output_text_file))
with open(self._output_text_file, "w", encoding="utf8") as wf:
wf.write(resulting_text)
result = {"text": resulting_text, "source": self._source}
self.set_results(result)
return True
def test_no_osd_file(tmpdir_factory):
e = ocr.Engine("")
version = e.get_version()
english_data_url = "{}/{}/{}".format(TESSDATA_SOURCE_URL_BASE, "4.0.0",
"eng.traineddata")
tessdata_path = tmpdir_factory.mktemp("no_osd_tessdata", numbered=False)
if not os.path.exists(tessdata_path):
os.makedirs(tessdata_path)
download_data(english_data_url, destination=tessdata_path)
with pytest.raises(FileNotFoundError):
reader = ocr.Reader(language_code="eng",
tesseract_data_path=tessdata_path)
shutil.rmtree(tessdata_path)
def set_tess_path(cls, path=locate_tessdata()):
assert path is not None
assert os.path.exists(path)
cls.engine = ocr.Engine(path)
assert cls.engine is not None
class GenerateOCRFileTask(speedwagon.tasks.Subtask):
engine = ocr.Engine(locate_tessdata())
def __init__(self,
source_image: str,
out_text_file: str,
lang: str = "eng",
tesseract_path: str = None) -> None:
super().__init__()
self._source = source_image
self._output_text_file = out_text_file
self._lang = lang
self._tesseract_path = tesseract_path
GenerateOCRFileTask.set_tess_path(tesseract_path or locate_tessdata())
assert self.engine is not None
@classmethod
def set_tess_path(cls, path: str = None) -> None:
if path is None:
path = locate_tessdata()
assert path is not None
assert os.path.exists(path)
cls.engine = ocr.Engine(path)
assert cls.engine is not None
def work(self) -> bool:
resulting_text = self.read_image(self._source, self._lang)
# Generate a text file from the text data extracted from the image
self.log("Writing to {}".format(self._output_text_file))
with open(self._output_text_file, "w", encoding="utf8") as write_file:
write_file.write(resulting_text)
result = {
"text": resulting_text,
"source": self._source
}
self.set_results(result)
return True
def read_image(self, file: str, lang: str) -> str:
if self.engine.data_set_path is None:
self.engine.data_set_path = self._tesseract_path
# Get the ocr text reader for the proper language
reader = self.engine.get_reader(lang)
self.log("Reading {}".format(os.path.normcase(file)))
file_handle = io.StringIO()
with contextlib.redirect_stderr(file_handle):
# Capture the warning messages
try:
resulting_text = reader.read(file)
except ocr.tesseractwrap.TesseractGlueException as error:
raise SpeedwagonException(f"Unable to read {file}") from error
stderr_messages = file_handle.getvalue()
if stderr_messages:
# Log any error messages
self.log(stderr_messages.strip())
return resulting_text
def ocr_engine(tess_path):
tess_engine = ocr.Engine(tess_path)
return tess_engine