YDBPython provides a Pythonic API for accessing YottaDB databases.
- A supported Linux Distribution: Ubuntu, Red Hat, or SUSE
- Python > 3.6 (f-string and type annotation used), including the
python3-dev
(orpython3-devel
on RHEL/SUSE) package that containsPython.h
- libffi
- YottaDB 1.34 or later
Before installing YDBPython:
- Install YottaDB per the Quick Start guide instructions or from source
- Install prerequisites:
- Ubuntu/Debian:
sudo apt install gcc python3 python3-setuptools python3-dev libffi-dev pkg-config python3-venv
- RHEL/CentOS:
yum install gcc python3 python3-setuptools python3-devel libffi-devel pkg-config python3-virtualenv
- Ubuntu/Debian:
- Set YottaDB environment variables:
source $(pkg-config --variable=prefix yottadb)/ydb_env_set
- Create venv:
- Enter directory where install is desired, e.g.
cd my-python-project
- Create venv:
python3 -m venv .venv
- Activate venv:
source .venv/bin/activate
- Enter directory where install is desired, e.g.
- Install into venv:
pip install yottadb
- Clone YDBPython source repository:
git clone https://gitlab.com/YottaDB/Lang/YDBPython.git /PATH/TO/YDBPYTHON
- Create venv:
- Enter directory where install is desired, e.g.
cd my-python-project
- Create venv:
python3 -m venv .venv
- Activate venv:
source .venv/bin/activate
- Install pre-requisites in
venv
:pip install setuptools pytest pytest-order psutil requests
- Enter directory where install is desired, e.g.
- Install into venv:
python -m pip install /PATH/TO/YDBPYTHON
Please install the pre-commit hook before committing changes:
ln -s ../../tools/git_hooks/pre-commit.sh .git/hooks/pre-commit
The pre-commit hook requires black
installed; you can install it with pip install black
.
To run YDBPython's test suite with a YDBPython source installation:
- Install YDBPython from source, per instructions above
- Run the tests in
venv
:python -m pytest
- Optional: Cleanup between tests:
- When making changes to code between test runs, some cleanup may be needed to prevent new changes being ignored due to Python caching. To clean up these files:
for artifact in $(cat .gitignore); do rm -rf $artifact; done
. Note that this will delete all files listed in.gitignore
, including core files. If these or any other such files need to be retained, move or rename them before running the aforementioned command.
- When making changes to code between test runs, some cleanup may be needed to prevent new changes being ignored due to Python caching. To clean up these files:
import yottadb
# Create Key objects for conveniently accessing and manipulating database nodes
key1 = yottadb.Key('^hello') # Create a key referencing the global variable '^hello'
print(f"{key1}: {key1.value}") # Display current value of '^hello'
key1.value = b'Hello world!' # Set '^hello' to 'Hello world!'
print(f"{key1}: {key1.value}")
key2 = yottadb.Key('^hello')['cowboy'] # Add a 'cowboy' subscript to the global variable '^hello', creating a new key
key2.value = 'Howdy partner!' # Set '^hello('cowboy') to 'Howdy partner!'
print(f"{key2}: {key2.value}")
key3 = yottadb.Key('^hello')['chinese'] # Add a second subscript to '^hello', creating a third key
key3.value = bytes('你好世界!', encoding="utf-8") # The value can be set to anything that can be encoded to `bytes`
print(key3, str(key3.value, encoding="utf-8")) # Returned values are `bytes` objects, and so may need to be encoded
for subscript in key1.subscripts: # Loop through all the subscripts of a key
sub_key = key1[subscript]
print(f"{sub_key}: {sub_key.value}")
key1.delete_node() # Delete the value of '^hello', but not any of its child nodes
print(f"{key1}: {key1.value}") # No value is printed
for subscript in key1.subscripts: # The values of the child nodes are still in the database
sub_key = key1[subscript]
print(f"{sub_key}: {sub_key.value}")
key1.value = 'Hello world!' # Reset the value of '^hello'
print(f"{key1}: {key1.value}") # Prints the value
key1.delete_tree() # Delete both the value at the '^hello' node and all of it's children
print(f"{key1}: {key1.value}") # Prints no value
for subscript in key1.subscripts: # Loop terminates immediately and displays no subscripts
sub_key = key1[subscript]
print(sub_key, sub_key.value)
# Database transactions are also available
@yottadb.transaction
def simple_transaction(value):
# Set values directly with the set() function
yottadb.set('test1', value=value) # Set the local variable 'test1' to the given value
yottadb.set('test2', value=value) # Set the local variable 'test2' to the given value
condition_a = False
condition_b = False
if condition_a:
# When a yottadb.YDBTPRollback exception is raised YottaDB will rollback the transaction
# and then propagate the exception to the calling code.
raise yottadb.YDBTPRollback("reason for the rollback")
elif condition_b:
# When a yottadb.YDBTPRestart exception is raised YottaDB will call the transaction again.
# Warning: This code is intentionally simplistic. An infinite loop will occur
# if yottadb.YDBTPRestart is continually raised
raise yottadb.YDBTPRestart()
else:
return yottadb.YDB_OK # Success, transaction will be committed
simple_transaction(b'test', db)
print(f"{db[b'test1']}: {db[b'test1'].value}")
print(f"{db[b'test2']}: {db[b'test2'].value}")
No, YDBPython does not support multithreading. This is due to the limitations of the Python Global Interpreter Lock for CPU-intensive multithreading. For background, see the following resources:
- Python documentation: Thread State and the Global Interpreter Lock
- Python's GIL - A Hurdle to Multithreaded Program
- Grok the GIL: How to write fast and thread-safe Python
- YDBPython GitLab discussion: Issue #7
Accordingly, the Python threading
and multithreading
should be avoided when developing applications with YDBPython. However, YDBPython does support multiprocessing and may be safely used with the Python multiprocessing
library for parallelism. For an example of multiprocessing
usage, see tests/test_threeenp1.py
.