Message 86180 - Python tracker

Message86180

Author	mark.dickinson
Recipients	benjamin.peterson, bob.ippolito, collinwinter, georg.brandl, giampaolo.rodola, jafo, mark.dickinson, nnorwitz, piman
Date	2009-04-19.20:50:16
SpamBayes Score	1.0215547e-08
Marked as misclassified	No
Message-id	<1240174219.18.0.437146040533.issue1530559@psf.upfronthosting.co.za>
In-reply-to

Content
The deprecated struct features (float coercion, overflow wrapping) have been removed for py3k in r70497, r70688, r71754. I don't plan to backport this to 2.7; I'll just try to fix the behaviour in a minimal way there. One thing that's not clear to me: what's the rationale for raising struct.error everywhere instead of more specific Python errors; e.g., TypeError for struct.pack('L', 'not an integer') and OverflowError for struct.pack('L', 10**100)? Is there a particular use-case for "except struct.error"?

Content

The deprecated struct features (float coercion, overflow wrapping) have 
been removed for py3k in r70497, r70688, r71754.  I don't plan to backport 
this to 2.7; I'll just try to fix the behaviour in a minimal way there.

One thing that's not clear to me:  what's the rationale for raising 
struct.error everywhere instead of more specific Python errors; e.g., 
TypeError for struct.pack('L', 'not an integer') and OverflowError for 
struct.pack('L', 10**100)?  Is there a particular use-case for "except 
struct.error"?

History
Date	User	Action	Args
2009-04-19 20:50:19	mark.dickinson	set	recipients: + mark.dickinson, nnorwitz, georg.brandl, collinwinter, jafo, bob.ippolito, piman, giampaolo.rodola, benjamin.peterson
2009-04-19 20:50:19	mark.dickinson	set	messageid: <1240174219.18.0.437146040533.issue1530559@psf.upfronthosting.co.za>
2009-04-19 20:50:18	mark.dickinson	link	issue1530559 messages
2009-04-19 20:50:16	mark.dickinson	create