Spaces:
Running
on
T4
Running
on
T4
Update app.py
Browse files
app.py
CHANGED
@@ -4,58 +4,78 @@ import subprocess
|
|
4 |
from PIL import Image
|
5 |
import os
|
6 |
import tempfile
|
|
|
|
|
|
|
|
|
7 |
|
8 |
def save_temp_image(img):
|
9 |
temp_dir = tempfile.mkdtemp()
|
10 |
img_path = os.path.join(temp_dir, "input_image.png")
|
11 |
img.save(img_path)
|
|
|
12 |
return img_path, temp_dir
|
13 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
def ocr_function_cli(img, lang_name):
|
15 |
img_path, temp_dir = save_temp_image(img)
|
16 |
command = f"surya_ocr {img_path} --langs {lang_name} --images --results_dir {temp_dir}"
|
17 |
-
|
18 |
-
subprocess.run(command, shell=True, check=True, encoding='utf-8')
|
19 |
-
except subprocess.CalledProcessError as e:
|
20 |
-
print(f"OCR command failed: {e.output}")
|
21 |
return img, "OCR failed"
|
|
|
22 |
result_img_path = os.path.join(temp_dir, "image_with_text.png")
|
23 |
result_text_path = os.path.join(temp_dir, "results.json")
|
|
|
24 |
if os.path.exists(result_img_path):
|
25 |
result_img = Image.open(result_img_path)
|
26 |
else:
|
27 |
result_img = img
|
|
|
28 |
if os.path.exists(result_text_path):
|
29 |
with open(result_text_path, "r", encoding='utf-8') as file:
|
30 |
result_text = json.load(file)
|
31 |
text_output = "\n".join([str(page) for page in result_text.values()])
|
32 |
else:
|
33 |
text_output = "No text detected"
|
|
|
|
|
34 |
os.remove(img_path)
|
|
|
35 |
return result_img, text_output
|
36 |
|
37 |
def text_line_detection_function_cli(img):
|
38 |
img_path, temp_dir = save_temp_image(img)
|
39 |
command = f"surya_detect {img_path} --images --results_dir {temp_dir}"
|
40 |
-
|
41 |
-
subprocess.run(command, shell=True, check=True, encoding='utf-8')
|
42 |
-
except subprocess.CalledProcessError as e:
|
43 |
-
print(f"Text line detection command failed: {e.output}")
|
44 |
return img, {"error": "Detection failed"}
|
|
|
45 |
result_img_path = os.path.join(temp_dir, "image_with_lines.png")
|
46 |
result_json_path = os.path.join(temp_dir, "results.json")
|
|
|
47 |
if os.path.exists(result_img_path):
|
48 |
result_img = Image.open(result_img_path)
|
49 |
else:
|
50 |
result_img = img
|
|
|
51 |
if os.path.exists(result_json_path):
|
52 |
with open(result_json_path, "r", encoding='utf-8') as file:
|
53 |
result_json = json.load(file)
|
54 |
else:
|
55 |
result_json = {"error": "No detection results found"}
|
|
|
|
|
56 |
os.remove(img_path)
|
|
|
57 |
return result_img, result_json
|
58 |
-
|
59 |
with gr.Blocks() as app:
|
60 |
gr.Markdown("# Surya OCR and Text Line Detection via CLI")
|
61 |
|
|
|
4 |
from PIL import Image
|
5 |
import os
|
6 |
import tempfile
|
7 |
+
import logging
|
8 |
+
|
9 |
+
# Configuração básica de logging
|
10 |
+
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
|
11 |
|
12 |
def save_temp_image(img):
|
13 |
temp_dir = tempfile.mkdtemp()
|
14 |
img_path = os.path.join(temp_dir, "input_image.png")
|
15 |
img.save(img_path)
|
16 |
+
logging.info(f"Imagem salva em {img_path}")
|
17 |
return img_path, temp_dir
|
18 |
|
19 |
+
def run_command(command):
|
20 |
+
try:
|
21 |
+
result = subprocess.check_output(command, shell=True, stderr=subprocess.STDOUT, encoding='utf-8')
|
22 |
+
logging.info("Command Output: " + result)
|
23 |
+
return result
|
24 |
+
except subprocess.CalledProcessError as e:
|
25 |
+
logging.error(f"Command failed with error: {e.output}")
|
26 |
+
return None
|
27 |
+
|
28 |
def ocr_function_cli(img, lang_name):
|
29 |
img_path, temp_dir = save_temp_image(img)
|
30 |
command = f"surya_ocr {img_path} --langs {lang_name} --images --results_dir {temp_dir}"
|
31 |
+
if run_command(command) is None:
|
|
|
|
|
|
|
32 |
return img, "OCR failed"
|
33 |
+
|
34 |
result_img_path = os.path.join(temp_dir, "image_with_text.png")
|
35 |
result_text_path = os.path.join(temp_dir, "results.json")
|
36 |
+
|
37 |
if os.path.exists(result_img_path):
|
38 |
result_img = Image.open(result_img_path)
|
39 |
else:
|
40 |
result_img = img
|
41 |
+
|
42 |
if os.path.exists(result_text_path):
|
43 |
with open(result_text_path, "r", encoding='utf-8') as file:
|
44 |
result_text = json.load(file)
|
45 |
text_output = "\n".join([str(page) for page in result_text.values()])
|
46 |
else:
|
47 |
text_output = "No text detected"
|
48 |
+
|
49 |
+
# Limpeza movida para depois da leitura dos resultados
|
50 |
os.remove(img_path)
|
51 |
+
logging.info(f"Limpeza concluída para {img_path}")
|
52 |
return result_img, text_output
|
53 |
|
54 |
def text_line_detection_function_cli(img):
|
55 |
img_path, temp_dir = save_temp_image(img)
|
56 |
command = f"surya_detect {img_path} --images --results_dir {temp_dir}"
|
57 |
+
if run_command(command) is None:
|
|
|
|
|
|
|
58 |
return img, {"error": "Detection failed"}
|
59 |
+
|
60 |
result_img_path = os.path.join(temp_dir, "image_with_lines.png")
|
61 |
result_json_path = os.path.join(temp_dir, "results.json")
|
62 |
+
|
63 |
if os.path.exists(result_img_path):
|
64 |
result_img = Image.open(result_img_path)
|
65 |
else:
|
66 |
result_img = img
|
67 |
+
|
68 |
if os.path.exists(result_json_path):
|
69 |
with open(result_json_path, "r", encoding='utf-8') as file:
|
70 |
result_json = json.load(file)
|
71 |
else:
|
72 |
result_json = {"error": "No detection results found"}
|
73 |
+
|
74 |
+
# Limpeza movida para depois da leitura dos resultados
|
75 |
os.remove(img_path)
|
76 |
+
logging.info(f"Limpeza concluída para {img_path}")
|
77 |
return result_img, result_json
|
78 |
+
|
79 |
with gr.Blocks() as app:
|
80 |
gr.Markdown("# Surya OCR and Text Line Detection via CLI")
|
81 |
|