From 691b9316b8ec53741ca5876c784f23058d93a721 Mon Sep 17 00:00:00 2001 From: c-bommu Date: Tue, 5 Dec 2023 17:13:33 -0500 Subject: [PATCH] Attempted to fix file path issue so Docker can find uploaded file --- ACMAS/app/ACMAS_Web/ocr_files/ocr.py | 29 ++++++++-------------------- 1 file changed, 8 insertions(+), 21 deletions(-) diff --git a/ACMAS/app/ACMAS_Web/ocr_files/ocr.py b/ACMAS/app/ACMAS_Web/ocr_files/ocr.py index 6a78d80..659a082 100644 --- a/ACMAS/app/ACMAS_Web/ocr_files/ocr.py +++ b/ACMAS/app/ACMAS_Web/ocr_files/ocr.py @@ -1,25 +1,11 @@ -''' import os from pdf2image import convert_from_path import pytesseract -def ocr_driver(pdf_name): - if ending_type(pdf_name) == 'pdf': - images = 'Hi' - else: - images = pdf_name - ocr_output = images - return pdf_name -''' - -import os -from pdf2image import convert_from_path -import pytesseract - -#absolute_path = os.path.dirname(__file__) -#relative_path = "../ocr_files/" -#full_path = os.path.join(absolute_path, relative_path) +absolute_path = os.path.dirname(__file__) +relative_path = "../ocr_files" +full_path = os.path.join(absolute_path, relative_path) # Gets the ending types of files @@ -37,16 +23,16 @@ def png_conversion(pdf_name): print("CONVERTING PNG") - if not os.path.isdir('mediafiles/ocr_images'): - os.system("mkdir mediafiles/ocr_images") + if not os.path.isdir('acmas_media_files/ocr_images'): + os.system("mkdir acmas_media_files/ocr_images") image_names = [] images_from_path = convert_from_path(pdf_name, 600) for i in range(len(images_from_path)): # Save pages as images in the pdf - images_from_path[i].save('mediafiles/ocr_images/' + 'page' + str(i) + '.jpg', 'JPEG') - image_names.append('mediafiles/ocr_images/' + 'page' + str(i) + '.jpg') + images_from_path[i].save('acmas_media_files/ocr_images/' + 'page' + str(i) + '.jpg', 'JPEG') + image_names.append('acmas_media_files/ocr_images/' + 'page' + str(i) + '.jpg') return image_names @@ -60,6 +46,7 @@ def run_ocr(image_name): + def ocr_driver(pdf_name): if ending_type(pdf_name) == 'pdf': images = png_conversion(pdf_name)