Synced repo using 'sync_with_huggingface' Github Action

.dockerignore

@@ -0,0 +1,9 @@
+html/*
+.DS_Store
+
+.venv
+*.swp
+.mypy_cache
+.pytest_cache
+.ipynb_checkpoints
+__pycache__

Dockerfile

@@ -0,0 +1,66 @@
+FROM ubuntu:22.04 AS base
+
+ARG PYTHON_VERSION=3.10
+
+ENV DEBIAN_FRONTEND=noninteractive
+ENV WORKDIR /app/
+
+WORKDIR /opt
+
+# install dev tools
+RUN apt-get update && apt-get install -y \
+  vim neovim nano \
+  git git-lfs \
+  zip unzip \
+  curl wget make build-essential xz-utils file tree \
+  sudo \
+  dnsutils \
+  tzdata language-pack-ja \
+  && apt-get clean \
+  && rm -rf /var/lib/apt/lists/*
+
+# for Japanese settings
+# ENV TZ Asia/Tokyo
+# ENV LANG ja_JP.utf8
+
+# for US settings
+ENV LANG en_US.UTF-8
+ENV LANGUAGE en_US
+
+# install Python
+RUN apt-get update && apt-get -yV upgrade && DEBIAN_FRONTEND=noninteractive apt-get -yV install \
+  build-essential libssl-dev libffi-dev \
+  python${PYTHON_VERSION} python${PYTHON_VERSION}-distutils python${PYTHON_VERSION}-dev \
+  && ln -s /usr/bin/python${PYTHON_VERSION} /usr/local/bin/python3 \
+  && ln -s /usr/bin/python${PYTHON_VERSION} /usr/local/bin/python \
+  && apt-get clean \
+  && rm -rf /var/lib/apt/lists/*
+
+## install pip
+RUN curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py \
+  && python3 get-pip.py \
+  && pip3 --no-cache-dir install --upgrade pip
+
+## install Poetry
+RUN curl -sSL https://install.python-poetry.org | python3 -
+ENV PATH $PATH:/root/.local/bin
+RUN poetry config virtualenvs.create true \
+  && poetry config virtualenvs.in-project false
+
+WORKDIR ${WORKDIR}
+
+# install python packages
+COPY poetry.lock pyproject.toml ./
+COPY src ./src
+RUN poetry install --no-dev
+
+FROM base AS dev
+WORKDIR ${WORKDIR}
+
+# install python packages
+COPY poetry.lock pyproject.toml ./
+COPY src ./src
+RUN poetry install
+
+# Hugging Face Hub Settings
+CMD ["poetry", "run", "streamlit", "run", "src/app.py", "--server.port", "7860"]

compose.yaml

@@ -0,0 +1,29 @@
+version: "3.2"
+services:
+  unlock_pdf_password:
+    tty: true
+    stdin_open: true
+    user: root
+    working_dir: /app
+    build:
+      context: .
+      dockerfile: docker/cpu/Dockerfile
+      target: dev
+      # secrets:
+      #   - github_token
+      args:
+        progress: plain
+    volumes:
+      - type: bind
+        source: ./
+        target: /app
+    ports:
+      - "8501:8501"
+    command:
+      poetry run streamlit run src/app.py
+    environment:
+        PYTHONPATH: "/app/src"
+        PYTHONUNBUFFERED: 1
+# secrets:
+#   github_token:
+#     file: ${HOME}/.git-credentials

docker/cpu/Dockerfile

@@ -0,0 +1,63 @@
+FROM ubuntu:22.04 AS base
+
+ARG PYTHON_VERSION=3.10
+
+ENV DEBIAN_FRONTEND=noninteractive
+ENV WORKDIR /app/
+
+WORKDIR /opt
+
+# install dev tools
+RUN apt-get update && apt-get install -y \
+  vim neovim nano \
+  git git-lfs \
+  zip unzip \
+  curl wget make build-essential xz-utils file tree \
+  sudo \
+  dnsutils \
+  tzdata language-pack-ja \
+  && apt-get clean \
+  && rm -rf /var/lib/apt/lists/*
+
+# for Japanese settings
+# ENV TZ Asia/Tokyo
+# ENV LANG ja_JP.utf8
+
+# for US settings
+ENV LANG en_US.UTF-8
+ENV LANGUAGE en_US
+
+# install Python
+RUN apt-get update && apt-get -yV upgrade && DEBIAN_FRONTEND=noninteractive apt-get -yV install \
+  build-essential libssl-dev libffi-dev \
+  python${PYTHON_VERSION} python${PYTHON_VERSION}-distutils python${PYTHON_VERSION}-dev \
+  && ln -s /usr/bin/python${PYTHON_VERSION} /usr/local/bin/python3 \
+  && ln -s /usr/bin/python${PYTHON_VERSION} /usr/local/bin/python \
+  && apt-get clean \
+  && rm -rf /var/lib/apt/lists/*
+
+## install pip
+RUN curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py \
+  && python3 get-pip.py \
+  && pip3 --no-cache-dir install --upgrade pip
+
+## install Poetry
+RUN curl -sSL https://install.python-poetry.org | python3 -
+ENV PATH $PATH:/root/.local/bin
+RUN poetry config virtualenvs.create true \
+  && poetry config virtualenvs.in-project false
+
+WORKDIR ${WORKDIR}
+
+# install python packages
+COPY poetry.lock pyproject.toml ./
+COPY src ./src
+RUN poetry install --no-dev
+
+FROM base AS dev
+WORKDIR ${WORKDIR}
+
+# install python packages
+COPY poetry.lock pyproject.toml ./
+COPY src ./src
+RUN poetry install

notebook/helloworld.ipynb

@@ -0,0 +1,51 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "18edcb70-64a5-4d17-94c0-a86ecc435be4",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "hello world\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(\"hello world\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b90e3e1f-5aa1-43a0-b9e2-1d1bcaf9ff94",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

pyproject.toml

@@ -0,0 +1,58 @@
+[tool.poetry]
+name = "unlock_pdf_password"
+version = "0.1.0"
+description = ""
+authors = ["iamtatsuki05 <tatsukio0522@gmail.com>"]
+packages = [
+    { include = "unlock_pdf_password", from = "src/" },
+]
+
+[tool.poetry.dependencies]
+python = "^3.10"
+python-dotenv = "^1.0.0"
+setuptools = "^69.0.3"
+fire = "^0.5.0"
+pydantic = "^2.5.3"
+beautifulsoup4 = "^4.12.2"
+selenium = "^4.16.0"
+fastapi = "^0.108.0"
+uvicorn = "^0.25.0"
+matplotlib = "^3.5.1"
+pandas = "^1.4.2"
+seaborn = "^0.11.2"
+japanize-matplotlib = "^1.1.3"
+numpy = "^1.22.3"
+jupyterlab = "^3.3.4"
+tqdm = "^4.64.0"
+scikit-learn = "^1.1.1"
+openpyxl = "^3.1.2"
+pikepdf = "^8.15.1"
+streamlit = "^1.34.0"
+
+[tool.poetry.group.dev.dependencies]
+pytest = "^7.0.0"
+ipykernel = ">=6.13.0"
+autopep8 = ">=1.6.0"
+autoflake = ">=1.4"
+flake8 = ">=4.0.1"
+flake8-isort = ">=4.1.1"
+flake8-quotes = ">=3.3.1"
+flake8-print = ">=4.0.0"
+isort = ">=5.10.1"
+black = ">=22.10.0"
+mypy = ">=0.971"
+tox = ">=3.25.1"
+pre-commit = ">=3.3.3"
+nbstripout = "0.6.1"
+
+[tool.isort]
+line_length = 88
+multi_line_output = 3
+include_trailing_comma = true
+
+[tool.black]
+skip-string-normalization = true
+
+[build-system]
+requires = ["poetry-core>=1.0.0"]
+build-backend = "poetry.core.masonry.api"

scripts/unlock_pdf_password.py

@@ -0,0 +1,46 @@
+import logging
+from typing import Union, Optional
+from pathlib import Path
+
+import fire
+
+from unlock_pdf_password.unlock_pdf_password import unlock_pdf_password, unlock_pdf
+
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+def unlock_pdf_password_wrapper(
+    input_pdf_path: Union[str, Path],
+    output_pdf_path: Union[str, Path],
+    password_length: int = None,
+    password_start_length: int = None,
+    password_in_number: Optional[bool] = True,
+    password_in_lower: Optional[bool] = True,
+    password_in_upper: Optional[bool] = True,
+    password_in_sign: Optional[bool] = True,
+) -> None:
+    input_pdf_path = Path(input_pdf_path)
+    output_pdf_path = Path(output_pdf_path)
+    output_pdf_path.parent.mkdir(parents=True, exist_ok=True)
+
+    password = unlock_pdf_password(
+        input_pdf_path=input_pdf_path,
+        password_length=password_length,
+        password_start_length=password_start_length,
+        password_in_number=password_in_number,
+        password_in_lower=password_in_lower,
+        password_in_upper=password_in_upper,
+        password_in_sign=password_in_sign,
+    )
+
+    if password:
+        unlock_pdf(
+            input_pdf_path=input_pdf_path,
+            output_pdf_path=output_pdf_path,
+            password=password,
+        )
+        logger.info(f'PDF unlocked successfully! Password: {password}')
+    logger.error('Password not found')
+
+if __name__ == '__main__':
+    fire.Fire(unlock_pdf_password)

src/app.py

@@ -0,0 +1,70 @@
+import tempfile
+from pathlib import Path
+
+import streamlit as st
+
+from unlock_pdf_password.unlock_pdf_password import unlock_pdf, unlock_pdf_password
+
+
+def app():
+    st.title("PDF Unlocker")
+
+    uploaded_file = st.file_uploader("Upload your PDF", type="pdf")
+    password_length = st.number_input(
+        "Enter maximum password length:", min_value=1, value=4
+    )
+    password_start_length = st.number_input(
+        "Enter starting password length:", min_value=0, value=0
+    )
+    password_in_number = st.checkbox("Include numbers", True)
+    password_in_lower = st.checkbox("Include lowercase letters", True)
+    password_in_upper = st.checkbox("Include uppercase letters", True)
+    password_in_sign = st.checkbox("Include special characters", True)
+
+    if st.button("Unlock PDF"):
+        if uploaded_file is None:
+            st.error("Please upload a PDF file.")
+        else:
+            with tempfile.NamedTemporaryFile(
+                delete=False, suffix=".pdf"
+            ) as temp_input_file:
+                temp_input_file.write(uploaded_file.getvalue())
+                input_pdf_path = temp_input_file.name
+
+        password = unlock_pdf_password(
+            input_pdf_path=input_pdf_path,
+            password_length=password_length,
+            password_start_length=password_start_length,
+            password_in_number=password_in_number,
+            password_in_lower=password_in_lower,
+            password_in_upper=password_in_upper,
+            password_in_sign=password_in_sign,
+        )
+        if password:
+            result = unlock_pdf(
+                input_pdf_path=input_pdf_path,
+                password=password,
+                output_pdf_path=None,
+            )
+            if result:
+                st.success(f"PDF unlocked successfully! Password: {password}")
+                with tempfile.NamedTemporaryFile(
+                    delete=False, suffix=".pdf"
+                ) as temp_output_file:
+                    result.save(temp_output_file.name)
+                    with open(temp_output_file.name, 'rb') as f:
+                        pdf_bytes = f.read()
+                    st.download_button(
+                        label="Download unlocked PDF",
+                        data=pdf_bytes,
+                        file_name=f"unlocked_{Path(input_pdf_path).name}",
+                        mime="application/octet-stream",
+                    )
+            else:
+                st.error("Failed to unlock PDF.")
+        else:
+            st.error("Password not found.")
+
+
+if __name__ == "__main__":
+    app()

src/unlock_pdf_password/unlock_pdf_password.py

@@ -0,0 +1,91 @@
+import itertools
+import logging
+import string
+from pathlib import Path
+from typing import Optional, Union
+
+import pikepdf
+from tqdm.auto import tqdm
+
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+
+def create_char_set(
+    numbers: bool, lowercase: bool, uppercase: bool, special_chars: bool
+) -> list[str]:
+    chars = []
+    if numbers:
+        chars.extend(string.digits)
+    if lowercase:
+        chars.extend(string.ascii_lowercase)
+    if uppercase:
+        chars.extend(string.ascii_uppercase)
+    if special_chars:
+        chars.extend(string.punctuation)
+    return chars
+
+
+def find_password(
+    input_pdf_path: Union[str, Path],
+    chars: list[str],
+    password_length: int,
+    password_start_length: int,
+) -> str:
+    for length in tqdm(range(password_start_length, password_length + 1)):
+        for password_tuple in itertools.product(chars, repeat=length):
+            password = ''.join(password_tuple)
+            try:
+                with pikepdf.open(input_pdf_path, password=password):
+                    logger.info(f'Password found: {password}')
+                    return password
+            except pikepdf.PasswordError:
+                logger.debug(f'Tried password: {password} - Incorrect')
+    logger.error('Password not found')
+    return ''
+
+
+def unlock_pdf(
+    input_pdf_path: Union[str, Path],
+    password: str,
+    output_pdf_path: Optional[Union[str, Path]] = None,
+) -> Union[pikepdf.Pdf, bool]:
+    try:
+        with pikepdf.open(input_pdf_path, password=password) as pdf:
+            pdf_unlocked = pikepdf.new()
+            pdf_unlocked.pages.extend(pdf.pages)
+            if output_pdf_path is not None:
+                pdf_unlocked.save(output_pdf_path)
+        return pdf_unlocked
+    except pikepdf.PasswordError:
+        return False
+
+
+def unlock_pdf_password(
+    input_pdf_path: Union[str, Path],
+    password_length: int = None,
+    password_start_length: int = None,
+    password_in_number: Optional[bool] = True,
+    password_in_lower: Optional[bool] = True,
+    password_in_upper: Optional[bool] = True,
+    password_in_sign: Optional[bool] = True,
+) -> Union[str, bool]:
+    input_pdf_path = Path(input_pdf_path)
+
+    if password_length is None or password_start_length is None:
+        logger.error('Password length and start length must be specified')
+        return False
+
+    chars = create_char_set(
+        numbers=password_in_number,
+        lowercase=password_in_lower,
+        uppercase=password_in_upper,
+        special_chars=password_in_sign,
+    )
+    password = find_password(
+        input_pdf_path=input_pdf_path,
+        chars=chars,
+        password_length=password_length,
+        password_start_length=password_start_length,
+    )
+    return password