I deal with about 6GB+ worth of log files daily (firewall and proxy). And if your not familiar with the SQUID format, it uses the Unix Timestamp instead of the usual date and time format. I've written several "support apps" (perl scripts) that convert a large column (consisting of several hundred lines) of Unix timestamps to normal date and times.
Feature Request: I'd like to see an option to select a column of text, whether it be the left most column (timestamp) or a single column of text/numbers (such as the byte count in a directory listing, i.e. "ls -o").
And possibly an option to hide the menu and toolbar.
Thank you for a wonderful application which I use religiously.
Steve
Opps... found the Hide Toolbar option.
Thanks.
Right.... Options exists for hiding the toolbar, status bar and both scrollbars. The options exists under the Options->Properties->Appearance menu. Currently, there is no way to disable the menu, but I'll at least consider adding it at a future date.
As for your column selection, this really seems like a special application of 'rectangular' selection, which is not currently implemented in any form. For example, normal selection always copies the full width of the lines in the middle of the selected range. I've seen applications where you can hold down the 'alt' key while making the selection with the mouse, and the selected area will be a perfect rectangle.
It sounds like what you're asking for is a rectangular selection option that fits itself to a single column of data??? Something similar to the way a double-click will select a word??
perhaps if I modify the following text selection semantics:
mouse select = normal text selection (exists)
double-click = select word (exists)
triple-click = select line (exists)
quad-click = select page (exists)
alt+select = rectangular selection (proposed)
alt+doubleclick = column selection (proposed)
does this sound good?
Rectangular selection, or course, has to come first. The column detection will simply sit on top of that.
The detection of the full column, then, is what is left to define. How would you define a column? Any column of text, numbers, or symbols that is separated by spaces???
One more quick question about your date column.... Does it have embeded spaces like "January 2, 2002"???
If so, it would be difficult to detect the column based on simple text operations. The column would have to be recognized as a date field and treated accordingly, which is probably more comples than I want to get with this particular feature.
If, on the other hand, you convert the dates to: YYYYMMDD format (or any format without a space inside it), column detection can be accomplished by simply assuming that white spaces delimits the column from the other text around it.
Block select has been implemented in the new RC33. Can you please check it out?? Is this sufficient?
[url= http://www.celestialsoftware.net/telnet/beta_software.html ]http://www.celestialsoftware.net/telnet/beta_software.html[/url]
Further discussion should take place in the beta testing forum.
Great! The Beta does exactly what I need! I mostly only need to select a column that is displayed, not necesarily the entire column of the whole file... just whats on the screen.
Thanks.
Steve
Just to warn you.... RC34 has some extra debug code in the copy function that will pop up a few dialog boxes when you copy text. You can ignore these. They will be gone shortly.
Thanks.
