Initing the atFile read/write code is always going to be tricky, but
perhaps not so tricky as it is at present.
A few days ago I rejoiced to eliminate the atAuto ivar, keyword args and
related code.
A few minutes ago, I was tempted to add it back. This is clearly a bad
idea, for two reasons:
1. *Switches should be as specific as possible*.
In this case, the switch will be called, allow_undefined_section_references.
General switches may also be reasonable. For example, a "toString" switch,
when passed to any AtFile write method, will never create a "huh?" when I
read it. Ditto for the at.toString ivar.
2. When possible, *switches should avoid polluting function signatures*.
This can be done using the following pattern:
ivar= 'allow_undefined_section_references'
setattr(self, ivar, True)
try:
<< call the code that uses this ivar >>
finally:
delattr(self, ivar)
The following code will test this switch:
if getattr(self, 'allow_undefined_section_references'):
I have never seen this Python pattern before, but here it makes sense. It
does complicate the calling and "using" code, but the complications are
minor. Crucially, *the complications are localized*. The new pattern
simplifies the rest of the AtFile class, which need not be aware of this
switch at all! Furthermore, this switch can be tested *anywhere* in the
AtFile class.
This is a big win. At present, various switches "infect" the signatures of
too many methods. The infection spreads to the at.initCommonIvars,
at.initReadIvars and at.initWriteIvars methods.
I suspect one or two other fairly mysterious switches pollute AtFile
signatures. If so, I'll remove them and the corresponding 'huh?' they evoke
in the reader.
Edward
--
You received this message because you are subscribed to the Google Groups
"leo-editor" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/leo-editor.
For more options, visit https://groups.google.com/d/optout.
