Message193986
| Author |
steven.daprano |
| Recipients |
belopolsky, cben, eric.araujo, ezio.melotti, flox, giampaolo.rodola, larry, lesmana, loewis, mark.dickinson, ned.deily, pitrou, python-dev, r.david.murray, ronaldoussoren, serhiy.storchaka, steven.daprano, techtonik, tshepang |
| Date |
2013年07月31日.08:04:14 |
| SpamBayes Score |
-1.0 |
| Marked as misclassified |
Yes |
| Message-id |
<51F8C4F5.6020506@pearwood.info> |
| In-reply-to |
<1375254853.62.0.149911365749.issue5845@psf.upfronthosting.co.za> |
| Content |
On 31/07/13 17:14, Larry Hastings wrote:
> IMO the optimal solution is that tab preceded by only whitespace indents, and tab preceded by any non-whitespace character attempts to complete. Can we goad readline into behaving this way?
Yes we can. Attached are a pair of lightweight modules I've used for this for the last 3-4 years, and it works fine in Python 2.4 through 3.3. I put something like this in my startup.py:
import completer
import history
history = history.History()
completer = completer.Completer(
bindings=(r'"\C-xo": overwrite-mode',
r'"\C-xd": dump-functions',
))
Originally they were in a single class, but I was persuaded that it was cleaner to separate them. |
| Files |
|---|
| File name |
Uploaded |
|
completer.py
|
steven.daprano,
2013年07月31日.08:04:13
|
|
history.py
|
steven.daprano,
2013年07月31日.08:04:13
|
|
History
|
|---|
| Date |
User |
Action |
Args |
| 2013年07月31日 08:04:15 | steven.daprano | set | recipients:
+ steven.daprano, loewis, ronaldoussoren, mark.dickinson, cben, belopolsky, pitrou, larry, techtonik, giampaolo.rodola, ned.deily, ezio.melotti, eric.araujo, r.david.murray, flox, lesmana, tshepang, python-dev, serhiy.storchaka |
| 2013年07月31日 08:04:14 | steven.daprano | link | issue5845 messages |
| 2013年07月31日 08:04:14 | steven.daprano | create |
|