Hi,
Just downloaded and built the latest from svn, and the "Find" and "Replace" dialogs really look great. :) I think the dialog that pops up to ask you if you want to wrap the search is good also, since, with other editors, sometimes you're clicking through too quickly and lose track if you've wrapped the search or not. The small dialog makes it noticable just the right amount, IMO.
But in the pursuit of finely-tuned perfection, here are some hopefully constructive comments:
In both the "Find" and "Replace" dialogs, you might change the text "Match only word start" to "Match only at start of word" and then provide a tooltip (is that the correct term? "popup help" maybe?) that says, "Only finds words that begin with the pattern you're searching for."
Actually, in both dialogs, you might consider disabling (greying out) "Match only a whole word", and "Match only at start of word"/"Match only word start" when the user selects "Use regular expressions", as those choices should make no difference when using regexes since the user would just use \b (erm... < and > I guess) if that's indeed what they wanted.
I see that you do provide the option of case insensitive regexes (which works nicely), though I suspect that the user will probably *expect* regex searches to be case sensitive by default -- thus you might consider making sure the "Case sensitive" checkbox gets flipped true (if not already selected) whenever the user selects "Use regular expressions". I think it would save the user some head-scratching. POLS (principle of least surprise) and all. :)
With regard to the "Replace control characters", I see that this means: if you put one of these 2-character patterns in your replace string (like \t), you'll actually get the control character when the replace happens (instead of the two separate '' and 't' characters). It may be an error that this option is in the "Find" dialog, correct?
Also regarding the "Find" dialog, it might be useful for the user to have a checkbox to "keep this dialog window open", since you might like to change one or more search options during a search. Maybe just my personal preference though. I see that I could just hit Ctrl-f to get that dialog up again if I need it.
Finally, with "Replace", a user might find it desirable to be able to replace a given bit of text, but then not automatically go on to the next occurrance (say you've got a large file with long stretches in between matches, not all of which you want to replace). It might be useful to have separate "replace" and "replace and find next" buttons. "find" could possibly even be changed to "find next" for extra clarity, but that's, again, maybe just personal preference.
Thanks for the great editor. Keep up the good work! :) ---John
On 04/08/06 06:52:51, John Gabriele wrote:
Hi,
Just downloaded and built the latest from svn, and the "Find" and "Replace" dialogs really look great. :) I think the dialog that pops up to ask you if you want to wrap the search is good also, since, with other editors, sometimes you're clicking through too quickly and lose track if you've wrapped the search or not. The small dialog makes it noticable just the right amount, IMO.
:)
But in the pursuit of finely-tuned perfection, here are some hopefully constructive comments:
In both the "Find" and "Replace" dialogs, you might change the text "Match only word start" to "Match only at start of word" and then provide a tooltip (is that the correct term? "popup help" maybe?) that says, "Only finds words that begin with the pattern you're searching for."
OK, I'll do this after the 0.8 release.
Actually, in both dialogs, you might consider disabling (greying out) "Match only a whole word", and "Match only at start of word"/"Match only word start" when the user selects "Use regular expressions", as those choices should make no difference when using regexes since the user would just use \b (erm... < and > I guess) if that's indeed what they wanted.
Just noticed that these options don't work when regular expressions are enabled (as well as being redundant), so I'll disable the options as you suggest.
I see that you do provide the option of case insensitive regexes (which works nicely), though I suspect that the user will probably *expect* regex searches to be case sensitive by default -- thus you might consider making sure the "Case sensitive" checkbox gets flipped true (if not already selected) whenever the user selects "Use regular expressions". I think it would save the user some head-scratching. POLS (principle of least surprise) and all. :)
I think this is not necessary - once the regex-irrelevant options are disabled, it will be more obvious that the case sensitive checkbox does apply to regexes.
With regard to the "Replace control characters", I see that this means: if you put one of these 2-character patterns in your replace string (like \t), you'll actually get the control character when the replace happens (instead of the two separate '' and 't' characters). It may be an error that this option is in the "Find" dialog, correct?
No, what if you want to find lines with a certain indentation? Also it can be used to find Unicode characters by their code.
Also regarding the "Find" dialog, it might be useful for the user to have a checkbox to "keep this dialog window open", since you might like to change one or more search options during a search. Maybe just my personal preference though. I see that I could just hit Ctrl-f to get that dialog up again if I need it.
By coincidence I thought about this yesterday, and now you mention it I think the Find dialog should not be hidden after clicking Find, so I'll leave it up to the user when to close the dialog.
Finally, with "Replace", a user might find it desirable to be able to replace a given bit of text, but then not automatically go on to the next occurrance (say you've got a large file with long stretches in between matches, not all of which you want to replace). It might be useful to have separate "replace" and "replace and find next" buttons. "find" could possibly even be changed to "find next" for extra clarity, but that's, again, maybe just personal preference.
I see the problem that the user might like to verify that the correct text was replaced. I will either have 2 replace buttons as you say or I'll add a checkbox to choose the behaviour - but this item may be done after 0.8.
Thanks for the great editor. Keep up the good work! :)
Thanks for the suggestions :)
Regards, Nick
On 8/4/06, Nick Treleaven nick.treleaven@btinternet.com wrote:
On 04/08/06 06:52:51, John Gabriele wrote: [snip]
With regard to the "Replace control characters", I see that this means: if you put one of these 2-character patterns in your replace string (like \t), you'll actually get the control character when the replace happens (instead of the two separate '' and 't' characters). It may be an error that this option is in the "Find" dialog, correct?
No, what if you want to find lines with a certain indentation? Also it can be used to find Unicode characters by their code.
Ah. Ok. I was confused because the overloading of the word "replace" (in connection with "find and replace"). In both the "Find" and "Replace" dialogs, it might make it more clear to change this checkboxe's name from "Replace control characters" to "Interpret control characters", since the editor is "interpreting" what you really mean when you type \t or \uXXXX. Also, in the "Replace" dialog, you might add to the popup help for this item that the replacement happens for *both* the find text *and* the replace text. (That is, I think you can search for foo_\uXXXX and replace it with foo_\uYYYY.)
BTW, there seems to be a small bug in the "Replace" dialog: when the search text is the same as the replace text, and you don't have the case sensitive box checked, the editor just beeps at you when you hit "Find".
Also regarding the "Find" dialog, it might be useful for the user to have a checkbox to "keep this dialog window open", since you might like to change one or more search options during a search. Maybe just my personal preference though. I see that I could just hit Ctrl-f to get that dialog up again if I need it.
By coincidence I thought about this yesterday, and now you mention it I think the Find dialog should not be hidden after clicking Find, so I'll leave it up to the user when to close the dialog.
Just grabbed and built today's svn. Works great. :)
Thanks, ---John
On 8/4/06, John Gabriele jmg3000@gmail.com wrote:
On 8/4/06, Nick Treleaven nick.treleaven@btinternet.com wrote:
On 04/08/06 06:52:51, John Gabriele wrote: [snip]
With regard to the "Replace control characters", I see that this means: if you put one of these 2-character patterns in your replace string (like \t), you'll actually get the control character when the replace happens (instead of the two separate '' and 't' characters). It may be an error that this option is in the "Find" dialog, correct?
No, what if you want to find lines with a certain indentation? Also it can be used to find Unicode characters by their code.
Ah. Ok. I was confused because the overloading of the word "replace" (in connection with "find and replace"). In both the "Find" and "Replace" dialogs, it might make it more clear to change this checkboxe's name from "Replace control characters" to "Interpret control characters", since the editor is "interpreting" what you really mean when you type \t or \uXXXX. [snip]
You know, it just occurs to me now (since I don't think I've ever explicitly used unicode) that this feature is probably much more widely used for the unicode than for the ascii control chars (the only ascii one I can think of that I might come across is \t, and for that I always think "regex"). So, the checkbox item could say "Interpret unicode" or else maybe the possibly-too-verbose "Interpret unicode and control characters".
Hmm... Wait a sec. Have you ever considered using that checkbox explicitly and only for unicode interpretation, and leaving the ascii control codes only to be specified as escapes in regexes?
My guess is that it's conceivable that you might one day either allow extended regexes (another checkbox in the Find and Replace dialogs?), or maybe even migrate entirely to a different regex engine. Dunno how regexes and unicode chars are connected wrt the regex code though...
---John
On 04/08/06 17:40:27, John Gabriele wrote:
On 8/4/06, John Gabriele jmg3000@gmail.com wrote:
On 8/4/06, Nick Treleaven nick.treleaven@btinternet.com wrote:
On 04/08/06 06:52:51, John Gabriele wrote: [snip]
With regard to the "Replace control characters", I see that this means: if you put one of these 2-character patterns in your
replace
string (like \t), you'll actually get the control character when
the
replace happens (instead of the two separate '' and 't'
characters).
It may be an error that this option is in the "Find" dialog,
correct?
No, what if you want to find lines with a certain indentation?
Also it
can be used to find Unicode characters by their code.
Ah. Ok. I was confused because the overloading of the word "replace" (in connection with "find and replace"). In both the "Find" and "Replace" dialogs, it might make it more clear to change this checkboxe's name from "Replace control characters" to "Interpret control characters", since the editor is "interpreting" what you really mean when you type \t or \uXXXX. [snip]
You know, it just occurs to me now (since I don't think I've ever explicitly used unicode) that this feature is probably much more widely used for the unicode than for the ascii control chars (the only ascii one I can think of that I might come across is \t, and for that I always think "regex"). So, the checkbox item could say "Interpret unicode" or else maybe the possibly-too-verbose "Interpret unicode and control characters".
I use \t quite a lot when I'm replacing. I think just "Interpret control characters" is best.
Hmm... Wait a sec. Have you ever considered using that checkbox explicitly and only for unicode interpretation, and leaving the ascii control codes only to be specified as escapes in regexes?
But as we already have the control character feature it would be overkill to have to use a regex for doing this.
My guess is that it's conceivable that you might one day either allow extended regexes (another checkbox in the Find and Replace dialogs?), or maybe even migrate entirely to a different regex engine. Dunno how regexes and unicode chars are connected wrt the regex code though...
We use the integrated Scintilla regex engine because it's best for searching Scintilla documents - otherwise we'd probably have to read the current document into one long string to pass to another regex parser. Find in files uses the external grep tool because it parses files that are not necessarily open in Geany.
Regards, Nick
On 04/08/06 17:11:32, John Gabriele wrote:
On 8/4/06, Nick Treleaven nick.treleaven@btinternet.com wrote:
On 04/08/06 06:52:51, John Gabriele wrote: [snip]
With regard to the "Replace control characters", I see that this means: if you put one of these 2-character patterns in your replace string (like \t), you'll actually get the control character when
the
replace happens (instead of the two separate '' and 't'
characters).
It may be an error that this option is in the "Find" dialog,
correct?
No, what if you want to find lines with a certain indentation? Also it can be used to find Unicode characters by their code.
Ah. Ok. I was confused because the overloading of the word "replace" (in connection with "find and replace"). In both the "Find" and "Replace" dialogs, it might make it more clear to change this checkboxe's name from "Replace control characters" to "Interpret control characters", since the editor is "interpreting" what you really mean when you type \t or \uXXXX. Also, in the "Replace" dialog, you might add to the popup help for this item that the replacement happens for *both* the find text *and* the replace text. (That is, I think you can search for foo_\uXXXX and replace it with foo_\uYYYY.)
Correct. Will do after 0.8.
BTW, there seems to be a small bug in the "Replace" dialog: when the search text is the same as the replace text, and you don't have the case sensitive box checked, the editor just beeps at you when you hit "Find".
Fixed now.
Regards, Nick
2006/8/4, Nick Treleaven nick.treleaven@btinternet.com:
On 04/08/06 06:52:51, John Gabriele wrote:
Hi,
Just downloaded and built the latest from svn, and the "Find" and "Replace" dialogs really look great. :) I think the dialog that pops up to ask you if you want to wrap the search is good also, since, with other editors, sometimes you're clicking through too quickly and lose track if you've wrapped the search or not. The small dialog makes it noticable just the right amount, IMO.
:)
But in the pursuit of finely-tuned perfection, here are some hopefully constructive comments:
In both the "Find" and "Replace" dialogs, you might change the text "Match only word start" to "Match only at start of word" and then provide a tooltip (is that the correct term? "popup help" maybe?) that says, "Only finds words that begin with the pattern you're searching for."
OK, I'll do this after the 0.8 release.
Actually, in both dialogs, you might consider disabling (greying out) "Match only a whole word", and "Match only at start of word"/"Match only word start" when the user selects "Use regular expressions", as those choices should make no difference when using regexes since the user would just use \b (erm... < and > I guess) if that's indeed what they wanted.
Just noticed that these options don't work when regular expressions are enabled (as well as being redundant), so I'll disable the options as you suggest.
I see that you do provide the option of case insensitive regexes (which works nicely), though I suspect that the user will probably *expect* regex searches to be case sensitive by default -- thus you might consider making sure the "Case sensitive" checkbox gets flipped true (if not already selected) whenever the user selects "Use regular expressions". I think it would save the user some head-scratching. POLS (principle of least surprise) and all. :)
I think this is not necessary - once the regex-irrelevant options are disabled, it will be more obvious that the case sensitive checkbox does apply to regexes.
With regard to the "Replace control characters", I see that this means: if you put one of these 2-character patterns in your replace string (like \t), you'll actually get the control character when the replace happens (instead of the two separate '' and 't' characters). It may be an error that this option is in the "Find" dialog, correct?
No, what if you want to find lines with a certain indentation? Also it can be used to find Unicode characters by their code.
Also regarding the "Find" dialog, it might be useful for the user to have a checkbox to "keep this dialog window open", since you might like to change one or more search options during a search. Maybe just my personal preference though. I see that I could just hit Ctrl-f to get that dialog up again if I need it.
By coincidence I thought about this yesterday, and now you mention it I think the Find dialog should not be hidden after clicking Find, so I'll leave it up to the user when to close the dialog.
Just thought it would be nice to mention, that one of the personal patches I will try to do with my own Geany (whenever I have the time too, which is short these days) is to create alternative find/repleace dialogs based on a bottom bar (much like Firefox's)... to match the confort I have with gvim's search/replace (my editor-of-choice)
Finally, with "Replace", a user might find it desirable to be able to replace a given bit of text, but then not automatically go on to the next occurrance (say you've got a large file with long stretches in between matches, not all of which you want to replace). It might be useful to have separate "replace" and "replace and find next" buttons. "find" could possibly even be changed to "find next" for extra clarity, but that's, again, maybe just personal preference.
I see the problem that the user might like to verify that the correct text was replaced. I will either have 2 replace buttons as you say or I'll add a checkbox to choose the behaviour - but this item may be done after 0.8.
Thanks for the great editor. Keep up the good work! :)
Thanks for the suggestions :)
Regards, Nick _______________________________________________ Geany mailing list Geany@uvena.de http://uvena.de/cgi-bin/mailman/listinfo/geany
Regards, Alexandre Moreira