tips_and_tricks.md

Tips and Tricks

This notebook sums up the changes in syntax, compared to the 'raw' usage of the COM type library. Everything the API provides out of the box is still available and is working according to the docs, but the syntax of Python provides a few shortcuts to make coding easier.

Collections and Slicing

When accessing items of collection-like COM classes (like IAxisVMDomains, IAxisVMSurfaces, anything having an Item method), you can use the slicing mechanism of Python. Suppose that we have an IAxisVMModel instance called axm. The model has several domains, each of them having the property Weight. Let say we want to calculate the weight of all domains. The out of box solution for this would be something like

weights=[]
for i in range(axm.Surfaces.Count):
    weights.append(axm.Surfaces.Item[i+1].Weight)
weight = sum(weights)

or using a list comprehension

weight = sum([axm.Surfaces.Item[i+1].Weight for i in range(axm.Surfaces.Count)])

or maybe even this

surfaces = [axm.Surfaces.Item[i+1] for i in range(axm.Surfaces.Count)]
weight = sum(map(lambda s : s.Weight, surfaces))

Anyhow, although there is nothing inherently wrong with these approaches, they clearly doesn't measure up to

# if you are new to this, the colon means 'all indices in range' 
weight = sum(axm.Surfaces[:].Weight)

or

weight = sum(s.Weight for s in axm.Surfaces)

or maybe

weight = sum(map(lambda s : s.Weight, axm.Surfaces))

Notice how the loops here are carried out over the collection object itself. This is because collection types implement the so-called iterator protocol.

It is also possible to provide negative indices:

axm.Surfaces[-1].Weight # equivalent to axm.Surfaces[axm.Surfaces.Count].Weight
axm.Surfaces[-2].Weight # equivalent to axm.Surfaces[axm.Surfaces.Count - 1].Weight

Without further due, some other use cases that exploit Python's slicing mechanism:

axm.Surfaces[1, 5, 7].Weight
axm.Surfaces[1:4].Weight  # equivalent to axm.Surfaces[1, 2, 3].Weight
axm.Surfaces[1:8:2].Weight  # equivalent to axm.Surfaces[1, 3, 5, 7].Weight
axm.Surfaces[8:1:-2].Weight  # equivalent to axm.Surfaces[8, 6, 4, 2].Weight

Furthermore, instead of typing axm.Surfaces.Count, you can use len(axm.Surfaces) to get the number of surfaces in the model.

Be aware here, that the index of the first item in any iterable COM object is 1, opposed to the zero-indexed nature of Python.

Context Management

If you have some experience with AxisVM and COM, you know about the methods BeginUpdate and EndUpdate. With python, you don't need to care about this, instead you can simply use the with statement like this

with axm as model:
    # do some modification here on the model
    ...

Accessing the Type Library

When a new instace of IAxisVMApplication is created, the type library is generated on demand. After that, the type library can be accessed as

import axisvm.com.tlb as axtlb

Daemon

When creating a new interface, you can do it like

from axisvm.com.client import start_AxisVM

axvm = start_AxisVM(visible=True, daemon=True)

The keyword argument daemon=True is a simple shortcut, equivalent to

from axisvm.com.client import start_AxisVM
import axisvm.com.tlb as axtlb

axapp = start_AxisVM(visible=True, daemon=False)
axapp.CloseOnLastReleased = True
axapp.AskCloseOnLastReleased = False
axapp.AskSaveOnLastReleased = False
axapp.ApplicationClose = axtlb.acEnableNoWarning

As a result of these settings, if the COM server is shut down, AxisVM shuts down either, hence the term daemon. Shutting down the COM server can be done with typing

axapp.Quit()