Spaces:
Running
Running
Update Dockerfile
Browse files- Dockerfile +11 -18
Dockerfile
CHANGED
@@ -1,43 +1,36 @@
|
|
1 |
-
#
|
2 |
FROM python:3.10-slim
|
3 |
|
4 |
-
#
|
5 |
RUN apt-get update && apt-get install -y \
|
6 |
tesseract-ocr \
|
7 |
libglib2.0-0 \
|
8 |
libgl1 \
|
9 |
-
libpoppler-cpp-dev \
|
10 |
-
poppler-utils \
|
11 |
libsm6 \
|
12 |
libxext6 \
|
13 |
libxrender-dev \
|
|
|
14 |
ffmpeg \
|
15 |
build-essential \
|
16 |
git \
|
17 |
-
|
18 |
-
|
19 |
-
# Ensure tesseract is in PATH and test it
|
20 |
-
RUN which tesseract && tesseract --version
|
21 |
|
22 |
-
#
|
23 |
-
ENV
|
24 |
-
|
25 |
-
HF_HUB_DISABLE_SYMLINKS_WARNING=1 \
|
26 |
-
TESSERACT_PATH=/usr/bin/tesseract \
|
27 |
-
PATH="/usr/bin:$PATH"
|
28 |
|
29 |
-
#
|
30 |
WORKDIR /home/user/app
|
31 |
|
32 |
-
#
|
33 |
COPY requirements.txt .
|
34 |
RUN pip install --upgrade pip && pip install -r requirements.txt
|
35 |
|
36 |
# Copy app files
|
37 |
COPY . .
|
38 |
|
39 |
-
# Download NLTK data
|
40 |
RUN python -m nltk.downloader punkt
|
41 |
|
42 |
-
#
|
43 |
CMD ["python", "app.py"]
|
|
|
1 |
+
# Use a base image with Python and system utilities
|
2 |
FROM python:3.10-slim
|
3 |
|
4 |
+
# Install system packages including Tesseract
|
5 |
RUN apt-get update && apt-get install -y \
|
6 |
tesseract-ocr \
|
7 |
libglib2.0-0 \
|
8 |
libgl1 \
|
|
|
|
|
9 |
libsm6 \
|
10 |
libxext6 \
|
11 |
libxrender-dev \
|
12 |
+
poppler-utils \
|
13 |
ffmpeg \
|
14 |
build-essential \
|
15 |
git \
|
16 |
+
&& rm -rf /var/lib/apt/lists/*
|
|
|
|
|
|
|
17 |
|
18 |
+
# 🔧 Make sure Tesseract is in PATH
|
19 |
+
ENV TESSERACT_PATH=/usr/bin/tesseract
|
20 |
+
ENV PATH="${TESSERACT_PATH}:${PATH}"
|
|
|
|
|
|
|
21 |
|
22 |
+
# Work directory
|
23 |
WORKDIR /home/user/app
|
24 |
|
25 |
+
# Python packages
|
26 |
COPY requirements.txt .
|
27 |
RUN pip install --upgrade pip && pip install -r requirements.txt
|
28 |
|
29 |
# Copy app files
|
30 |
COPY . .
|
31 |
|
32 |
+
# Download NLTK tokenizer data
|
33 |
RUN python -m nltk.downloader punkt
|
34 |
|
35 |
+
# Run the app
|
36 |
CMD ["python", "app.py"]
|