Pyalma

A Python library for SSH connections and remote command execution using Paramiko.

To Install pyalma using pip (no need to download the code)

Run the following in your Python environment:

pip install "pyalma @ git+https://github.com/ICR-RSE-Group/pyalma.git@main"

To build and install the package locally:

1. Clone this repo locally to your machine: Make sure to navigate to the package directory before building

git clone [email protected]:ICR-RSE-Group/pyalma.git
cd pyalma

2. Create a Python Environment and activate it: Before building and installing the package, it is recommended to create a dedicated Python environment for your project.

python -m venv .env-py
source .env-py/bin/activate

3. Install Build Dependencies

python -m pip install --upgrade pip build

4. Build the Package Locally:

python -m build

5. Install locally using pip:

pip install -e .

For command line,

pyalma-cli --cmd "ls -l"

For remote access, from within python

from pyalma import SshClient

ssh = SshClient(server='your_server', username='your_username', password='your_password')
result = ssh.run_cmd('ls -l')
print(result["output"])
print(result["err"])

For remote access using Ssh Keys, from within python or jupyter notebooks

from pyalma import SecureSshClient

secure_ssh = SecureSshClient(server='your_server', username='your_username')
result = ssh.run_cmd('ls -l')
print(result["output"])
print(result["err"])

For local access, from within python

from pyalma import LocalFileReader

local = LocalFileReader()
result = local.run_cmd('ls -l')
print(result["output"])
print(result["err"])

To read remote anndata files, from within python:

remote_path = "/full/remote/path/file.h5ad"
local_path = "local_copy.h5ad"
ssh = SshClient(server='your_server', username='your_username', password='your_password')
ssh.load_h5ad_file(path, local_path)
adata = ssh.read_h5ad(local_path)
print(adata)

To read pdf files, from within python:

path = "file.pdf"
ssh = SshClient(server='your_server', username='your_username', password='your_password')
# Read PDF into a DataFrame using pyalma
pdf_df = ssh.read_file_into_df(path,'pdf')
# returned dataframe contains three columns : "Page", "Content" and "Images"

Setting Up SSH Keys for SecureSshClient

Generate a new SSH key:

ssh-keygen -t rsa -b 4096 -C "[email protected]"

Add Your SSH Key to the Remote Server:

Access and copy the newly generated public key via:

cat ~/.ssh/id_rsa.pub

Then paste it into the remote server’s ~/.ssh/authorized_keys file.

Test the setup

Before using the Python client, confirm the connection works:

ssh [email protected]

You should connect without a password prompt