Python Commands
Table of contents
Python Command
The python command is the basic unit of work of PythonBridge. It allows developers to run arbitrary python code from Pharo. It is composed by a sequence of python statements, a collection of variable bindings, pharo observers callbacks and a transformation block.
Python statements
The python statements are created in Pharo by using Python3Generator.
Attribute accessing
Python objects instance variables and methods are accessible in Python by operator .. From Pharo, this operator is defined using => and allow accessing instance variables and invoking methods.
#foo asP3GI => #a "foo.a"
Method and function invoking
Python functions and methods are called by using callWith: and callWith:with:.
The callWith: method receive as argument a collection of objects which corresponds to the positional arguments of the call.
#foo asP3GI callWith: #(1 2 3) "foo(1,2,3)"
The callWith:with: method receives as first agument a collection of positional arguments and, as second argument, a dictionary of named arguments.
#foo asP3GI
callWith: #(1 2 3)
with: {
#nameArg1 -> 'bar'.
#nameArg2 -> 5 } asDictionary
"foo(1, 2, 3, nameArg1='bar', nameArg2=5)"
Variable assignment
Assignments in python allow you to store objects in globals, temporary variables and instance variables. Python3Generator uses the message <- to assign the value at the right of the message to the receiver.
Assigning the temporary variable w with the array [1,2,3].
#foo asP3GI <- #(1 2 3) "foo = [1,2,3]"
Assigning instance variable a from object in variable foo with the string 'bar'.
(#foo asP3GI => #a) <- 'bar' "foo.a = 'bar'"
Importing an External Module
PythonBridge can interact with any python modules. Consider the numpy Python module, which essential in scientific computing. We will perform the following Python instruction within Smalltalk: numpy.arange(15).reshape(3, 5)
First, we need to augment the Pipenv environment with the numpy module. You can simply do it by modifying the Pipfile by adding the line numpy = "*". The complete Pipfile should be:
[[source]]
name = "pypi"
url = "https://pypi.org/simple"
verify_ssl = true
[dev-packages]
[packages]
flask = "*"
requests = "*"
msgpack = "*"
numpy = "*"
[requires]
python_version = "3"
Once the Pipfile file modified, the pipenv environment must be upgraded, which can be done using the command line:
pipenv update
You should see something like:
➜ PythonBridge git:(master) ✗ pipenv update
Running $ pipenv lock then $ pipenv sync.
Locking [dev-packages] dependencies…
Locking [packages] dependencies…
✔ Success!
Updated Pipfile.lock (deca6d)!
Installing dependencies from Pipfile.lock (deca6d)…
🐍 ▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉▉ 14/14 — 00:00:07
To activate this project's virtualenv, run pipenv shell.
Alternatively, run a command inside the virtualenv with pipenv run.
All dependencies are now up-to-date!
Numpy is now installed in your Pipenv environment, and it is ready to be used from Smalltalk. You can do
PBApplication do: [
"We import the numpy module"
PBCF << (P3GImport moduleIdentifier: #numpy asP3GI).
"We construct the Python expression numpy.arange(15).reshape(3, 5)"
r := P3GCall target: (#numpy asP3GI => #arange) positionalArguments: #(15).
PBCF << (P3GCall target: r => #reshape positionalArguments: #(3 5)).
s := PBCF send waitForValue.
"The variable s points to a Smalltalk wrapper of the Python object"
"We simply execute the Python expression str(s) to obtain a string that represents the numpy object"
PBCF << (P3GCall target: #str asP3GI positionalArguments: (Array with: s)).
PBCF send waitForValue.
].
Printing this expression will result in:
[[ 0 1 2 3 4]
[ 5 6 7 8 9]
[10 11 12 13 14]]
This small example shows how you can augment the Pipenv environment with a new Python module, and how you can access it from Smalltalk.
Command Factory
Each PythonBridge extension defines its own CommandFactory as a global object. In the case of PythonBridge it’s called PBCF and in KerasBridge is called KCF.