Commit 9a842919 authored by Vít Starý Novotný's avatar Vít Starý Novotný
Browse files

Fix imports

parent b5188a20

ahisto_ocr/cli.py

+1 −1
Original line number Original line Diff line number Diff line
@@ -85,10 +85,10 @@ def run_ocr(input_images: List[Path], output_dir: Path, google_vision_ai: bool, 
        run_tesseract(client, postprocessing_volume, tesseract_volume)

        run_tesseract(client, postprocessing_volume, tesseract_volume)





        if google_vision_ai:

        if google_vision_ai:

            LOGGER.info('Skipping Google Vision AI')

            run_google_vision_ai(client, postprocessing_volume, google_vision_ai_volume)

            run_google_vision_ai(client, postprocessing_volume, google_vision_ai_volume)

            combine_results(client, tesseract_volume, google_vision_ai_volume, output_volume)

            combine_results(client, tesseract_volume, google_vision_ai_volume, output_volume)

        else:

        else:

            LOGGER.info('Skipping Google Vision AI')

            output_volume = tesseract_volume

            output_volume = tesseract_volume





        copy_output_from(client, output_volume, input_images, output_dir)
 
        copy_output_from(client, output_volume, input_images, output_dir)

ahisto_ocr/ocr.py

+1 −2
Original line number Original line Diff line number Diff line 
from logging import getLogger

from logging import getLogger

from pathlib import Path





from .config import CONFIG as _CONFIG

from .config import CONFIG as _CONFIG

from .util import (

from .util import (
@@ -103,7 +102,7 @@ def run_tesseract_second_pass(client, preprocessing_volume, first_pass_volume, 
    run_docker_container(client, 'ahisto/ocr-fileformat', command=command, volumes=volumes)

    run_docker_container(client, 'ahisto/ocr-fileformat', command=command, volumes=volumes)









def run_google_vision_ai(client, postprocessing_volume, google_vision_ai_volume) -> None:

def run_google_vision_ai(client, preprocessing_volume, google_vision_ai_volume) -> None:

    LOGGER.info('Running Google Vision AI')

    LOGGER.info('Running Google Vision AI')





    volumes = {

    volumes = {

ahisto_ocr/util.py

+3 −1
Original line number Original line Diff line number Diff line
@@ -9,6 +9,7 @@ from contextlib import closing 
from io import BytesIO

from io import BytesIO

import os

import os

import sys

import sys

from pathlib import Path





from docker.types import DeviceRequest

from docker.types import DeviceRequest

from docker.errors import ContainerError

from docker.errors import ContainerError
@@ -77,5 +78,6 @@ def extract_text_file_from_container(container, filename: str) -> str: 
        for chunk in bits:

        for chunk in bits:

            tf.write(chunk)

            tf.write(chunk)

        tf.seek(0)

        tf.seek(0)

        text = tf.read().decode('utf-8')

        with tarfile.open(fileobj=tf, mode='r') as tar:

            text = tar.extractfile(Path(filename).name).read().decode('utf-8')

    return text
 
    return text

ahisto_ocr/volume.py

+1 −1
Original line number Original line Diff line number Diff line
@@ -4,7 +4,7 @@ from logging import getLogger 
from tempfile import TemporaryFile

from tempfile import TemporaryFile

import tarfile

import tarfile





from .util import create_temporary_docker_container, add_text_file_to_tarfile

from .util import create_temporary_docker_container, add_text_file_to_tarfile, extract_text_file_from_container









LOGGER = getLogger(__name__)

LOGGER = getLogger(__name__)