This is an archive of past discussions about Wikipedia:AutoWikiBrowser. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.
Latest comment: 17 years ago2 comments1 person in discussion
The best new feature for version 4 would be releasing it. Therefore I implore testers to give the alpha version a good going over this week, and as long as there's no serious bugs in article editing we should release v4.0 beta to the general user base asap. Trivial and cosmetic bugs, if any, can wait, as can feature requests. My tests with v4 (mostly with my plugin) have been pretty positive so I hope we can get this out soon! --kingboyk23:25, 24 July 2007 (UTC)
Is it possible to get a full list of articles in a category and its subcategories all the way down to the bottom of a category tree? ie. expand fully with one click, rather than expanding level by level? Carcharoth15:52, 27 March 2007 (UTC)
I think it would be good if AWB could make a list of articles in a given category and recursively add the articles of the subcategories of the given category. --ZeroOne (talk / @) 10:41, 12 May 2007 (UTC)
Added in revision
Im sure this had been added before, but it may have been removed (I havent seen it recently myself). I'll poke Mets about it for you Reedy Boy15:54, 27 March 2007 (UTC)
I don't know if it's already there or not; if it is I haven't seen it. Either way, I would strongly caution against an "all the way down to the bottom" approach. The category system isn't really a tree, there are loops and all sorts of peculiarities. A feature which did "select all categories in the list, get their contents, and remove the categories" in one click would be useful though. --kingboyk15:59, 27 March 2007 (UTC)
OK. A set distance downwards, then. That will at least improve matters. It is the need to expand the tree that is the killer. A recent example involved a definite tree structure (and many parts of the category system are trees), and involved 935 articles distributed over 661 categories, 9 of which were empty. It really shouldn't be necessary to click the "expand" toggle 661 time to get a list of 935 articles. Carcharoth16:52, 27 March 2007 (UTC)
You can just highlight them all, then right click, make list from category. Click the first item you want, then scroll down, press shift, and then click on the last category for multiple selections - There are workarounds!! Reedy Boy17:10, 27 March 2007 (UTC)
This feature was added a while back, but it's just too buggy with infinite loops and stuff. Use the workaround that Reedy Boy mentioned above instead. —METS501 (talk)18:03, 27 March 2007 (UTC)
OK, I'll use the workaround. But some limited functionality should be possible by cutting off loops and infinitely long expansions. Maybe something to come back to later? Carcharoth22:24, 27 March 2007 (UTC)
A semi-automated solution is what's needed here. The user must select which subcategories are trawled, but the software does the trawling, sorting, duplicates removal, etc. We definitely need to cut down on the number of mouse/key presses needed and make this process easier. (I had this planned but had pc problems, see User:Kingboyk/AWB) --kingboyk22:44, 9 April 2007 (UTC)
This feature is already implemented (including proper handling of circular references and duplicates), but after discussing possible strain on servers with Simetrical on WP:VPT it was restricted to debug builds only. I'm not ready to enable it unless we have constent from MW developers. MaxSem03:45, 12 April 2007 (UTC)
I don't see what difference it makes to the servers if it's done by the user manually, or by AWB programatically? --kingboyk11:15, 12 April 2007 (UTC)
This is what I had on my AWB to-do list. Would this be OK? Is there, then, code I can call or recycle to do this?
Make a semi-automated listbuilder feature, either in AWB or in my plugin. (Probably the former, as I think this could be widely useful; but if the latter, could inherit from the current listbuilder and then trick AWB into displaying ours).
Build list as normal
Display option to filter out non-main mainspaces
Display subcategories in a checklist, and prompt user
Trawl selected subcats, move others to a collection (so we can check we don't prompt about the same categories twice)
For plugin, convert to talk
Sort A-Z
Basically what I'm having to do manually, with a lot less keypresses and hassle.
I've commited a potential fix to stop it visiting categories it already has visited (ie, if the category is already in the list of categories visited, it wont load from that category... - rev 1398/rev 1399. Not sure which... Ideally needs testing... Wasnt sure of a category at the time to try it on... Reedy Boy16:42, 3 July 2007 (UTC)
de.wikipedia - persondata placement
Latest comment: 17 years ago4 comments3 people in discussion
On de.wikipedia (unlike en), the persondata is located behind the categories (and before the interwiki links). Is it possible to make AWB regarding this for de when sorting the metadata? Thanks, --Oxymoron8300:55, 5 August 2007 (UTC)
Hi, could you possible expand a bit more to as it is what you want. As "Append/prepend from text file." could mean anything, and therefore would be difficult for us to act upon it. Thanks, Reedy Boy11:28, 6 July 2007 (UTC)
Having it do it if there is 2 exactly the same is fine. Its if theres 2 for the same language under different names. I suppose this couldnt be done automatically... Reedy Boy13:58, 1 July 2007 (UTC)
This actually already exists where the Interwiki links are exactly the same... Removing them if they are the same language, but different pages is a no no. AWB cant tell which is right. See [1]Reedy Boy19:31, 10 August 2007 (UTC)
Don't show a message about {{inuse}} pages for pages which AWB is skipping anyway
Latest comment: 17 years ago2 comments2 people in discussion
Currently, if you're editing a list of pages, and set it to skip pages which do/don't have specific text (I had it skip any page which doesn't have the text [[WP:ANI# on pages that link to the shortcut WP:ANI), and one of the pages which it had been set to skip has the {{inuse}} template, it displays a message saying (this isn't a direct quote, but gives the idea) "This page has a {{inuse}} tag, consider skipping it". If AWB is skipping the page anyway, there is no reason to display the messgae. OdMishehu07:32, 9 July 2007 (UTC)
You do realise that these settings are (or at least are supposed to be) saved with the normal settings file..? Reedy Boy17:05, 23 May 2007 (UTC)
I mean save that the table in the text file, whether the "Find", "Replace with" and "Settings" (Multiline, Enabled ,....) --OsamaK08:17, 25 May 2007 (UTC)
i think osamak means he want to save the list of find and replace in a text file, so he dont reload the setting in each time --Mmustafa07:16, 4 June 2007 (UTC)
{{trivia}} (and its redirects) has been modified and should be auto-tagged with the date {{trivia|date=Month Year}} -- pb30<talk> 22:23, 31 May 2007 (UTC)
on de.wikipedia, the Personendaten template (the same as Template:Persondata on en) by convention is placed below the categories. I would like AWB to leave it there when fixing an article. Also links in the template should not count as duplicate links. --08-1508:06, 11 August 2007 (UTC)
Added in revision
De.wikipedia sorting order fixed for the next release. If by template you mean the infoboxes, then I think that they should count for duplicates, but that's just my opinion. If you mean something else by "templates", let us know :-) —METS501 (talk)21:43, 12 August 2007 (UTC)
Would it be possible to create up a list of all links on a page which go to disambig pages (something like the list of multiple links)? bd2412T21:37, 12 August 2007 (UTC)
Would be quite easy to add. Only requires a simple regex to look for and cases of [[ ]], so something like \[\[.*?\]\], do a regex.match, and if there is no success, add the tag... It may make it into v4, being simple, but the feature list is more a less stuck atm... Reedy Boy08:28, 20 July 2007 (UTC)
It would be nice to have a way to filter regular expressions by namespace, have certain regular expressions only match in certain namespaces. I figured out a way to do this but it is pretty ugly.
How about this for a workaround: Let's say you want to replace the word vehicle with automobile only in the Wikipedia namespace. Set this as your rule:
In the "Edit box" give the user either a toolbar or keyboard shortcuts (like ctrl-B) to Bold, Italic, superscript/subscript, line-break, etc... —Dispenser05:06, 5 June 2007 (UTC)
One thing which AWB is good for is filtering out pages which do/don't have certain text or wiki-code, and then edit the remaining pages as one would directly using Wikipedia. I don't need this tool bar for anything, but for such things, this tool bar may be useful for some people. OdMishehu07:39, 9 July 2007 (UTC)
Is it possible to add an option of searching (including RegExp) in summary text? That would be helpful, e.g. for lists created from Recentchages page --213.170.94.12015:24, 12 March 2007 (UTC)
It would be useful to have a yes/no option somewhere to automatically skip a page if the edit made is blocked by the spam filters. Thanks Rjwilmsi17:15, 7 June 2007 (UTC)
I think he means an option to skip if the edit contains a keyword blocked by the mediawiki spam filter. --RParlateContribs@ (Let's Go Yankees!) 02:04, 8 June 2007 (UTC)
Yes, I mean the spam filter, not page protection. The spam filter parses the whole page, not just the changes made, so blocks edits. Rjwilmsi17:37, 8 June 2007 (UTC)
Doesn't 'skip article when first nudge doesn't help' accomplish effectively the same (at least in bot mode)? --Seed2.018:58, 8 June 2007 (UTC)
Just noticed this request again. I put some changes to the logging depending on the spam filter
730 else if (!BotMode && webBrowserEdit.Document.Body.InnerHtml.Contains("<A class=extiw title=m:spam_blacklist href=\"http://meta.wikimedia.org/wiki/spam_blacklist\">"))
731 {//check edit wasn't blocked due to spam filter
732 if (MessageBox.Show("Edit has been blocked by spam blacklist. Try and edit again?", "Spam blacklist", MessageBoxButtons.YesNo, MessageBoxIcon.Warning) == DialogResult.Yes)
733 Start();
734 else
735 SkipPage("Edit blocked by spam protection filter");
736
737 return;
738 }
It wouldnt be difficult to add on the else if, " && chkSkipIfBlockedBySpamFilter.Checked" - Causing it to automatically just skip it - not bugging the user with the message box...? Reedy Boy14:09, 21 August 2007 (UTC)
When running a hidden version in particular, or listening to music or games it would be nice to turn off the clicks. RichFarmbrough, 09:18 1 September2007 (GMT).
As far as I know, this cannot be done from within AWB. Do you know how to do it yourself? If not, I'll let you know. —METS501 (talk)16:45, 1 September 2007 (UTC)
A find dialog
Latest comment: 17 years ago3 comments2 people in discussion
I find the new method of displaying diffs much harder to read than the old way. The old red bracket highlighting really made the changes stand out. Could the colour scheme & font of the diffs box be made a customisable option? Colonies Chris13:36, 31 July 2007 (UTC)
I was going to suggest that it could be altered by using the custom style sheets feature in Internet Explorer; however, it doesn't seem to be working for me. Maybe this file should be externalize? —Dispenser21:39, 23 August 2007 (UTC)
Just how many programing languages am I suppose to learn? I got the idea after read a few bot request where people were combining frameworks in other languages to simple upload stuff to mediawiki. All it really needs to do is write to a file, call the program, when the program finishes, read the file back into the editor window. And it should probably be a plugin module for AWB. —Dispenser18:40, 31 July 2007 (UTC)
That was the idea of that code... Used as a plugin/module... Granted its VB, but AWB can and does compile VB mondules. Changed to C#, add a text reader, return to AWB and your done.... Reedy Boy18:51, 31 July 2007 (UTC)
Would you like me to write an AWB plugin to do this for you? Would need some input as to what would need to be processed from the text file... Reedy Boy22:18, 31 July 2007 (UTC)
So, working directory, is where the file you want to run is. File name is obvious, arguments, is probably how you want to pass it the article text or similar. If you want me to change the module to write to a text file the article text (something like input.txt), i can do. Reedy Boy11:12, 26 August 2007 (UTC)
Inclusion of Reference section
Latest comment: 17 years ago4 comments3 people in discussion
When there are <ref></ref> in the article but no {{reference}} with appropriate section header, is it possible to add the ==References==and {{reference}} section?SkierRMH20:38, 30 August 2007 (UTC)
There are loads of pointless templates that invoke <references/>, and people keep adding more. We must either have a 1000% reliable way to detect all such templates, or not fix this problem automatically at all, because too many fully-atomated bots run with general fixes nowadays, we cannot afford giving them a chance to screw articles. Besides, lack of <references/> is not that common mistake. MaxSem10:30, 31 August 2007 (UTC)
I've been trying to think of a "logical" manner to place the citation, and several test runs that gave some most unusual results, realize that although the section/heading is missing, there's simply no automated way to include the section. Guess I'll formally withdraw this and move to a simple 'locate and evaluate' mode! SkierRMH06:07, 3 September 2007 (UTC)
Click to add to list
Latest comment: 17 years ago2 comments2 people in discussion
It would be useful to have a button that allows you to populate a list based on clicks: For example, if I wanted to deliver a newsletter, and had a list on-wiki of all user's talk pages, I could just click each of the links and it would add them to the list. GrooveDog (talk) (Review) 20:19, 3 September 2007 (UTC)
Pages in template space do not receive general fixes. While it is understandable that pages with variables cause complication, the other page (such as documentation or navboxes) do not receive these general fixes. AWB should apply general fixes if the page do not contain {{{, to further exclude more pages possibly parser functions. —Dispenser05:23, 31 August 2007 (UTC)
Added in revision
It's because all general fixes are intended solely for mainspace, other pages either don't need such conversions, or these conversions would be harmful. General fixed would definitely break too many templates, that's why we use them only for articles. WONTFIX. MaxSem05:58, 31 August 2007 (UTC)
That's why I gave a trigger that would ensuring that templates wouldn't break. We have a lot of /doc subpages that could truly benefit from some general fixes. One of the general fixes that I've reimplemented is the underscore removal and capitalization of the namespace in wikilinks. Couldn't we enable fixes do not change the structure? —Dispenser13:22, 31 August 2007 (UTC)
Now that AWB 4.1.0.0 has the useful option to skip an edit if it's blocked by the spam filter, could we also have the option to skip articles tagged as 'in use'? Thanks Rjwilmsi17:45, 10 September 2007 (UTC)
The auto tag adds uncatogrized and wikify, would be good if the date parameter was preferred "date=" syntax rather than default parameter. RichFarmbrough, 09:29 16 April2007 (GMT).
Please, add "Make all text box from right to left in "User and project preferences". ar.wiki, he.wiki, fa.wiki and ur.wiki all this language write from right to left. Thanks--OsamaK10:51, 1 June 2007 (UTC)
Added in revision
This is feasible, well to an extent - Presumably for any and all text boxes everywhere...? As some would be a buggar to do, but ensuring that like the edit text box and such were done - ie the ones you would be typing in a lot. Reedy Boy11:48, 1 June 2007 (UTC)
Make the behavior of text prerpending feature customizable in terms of the amount of lines between the text prepended and other text. In some cases (usually when I insert hidden templates) users ask me to reduce the amount of newline characters down to one instead of two set by AWB by default. Mashiah21:45, 14 July 2007 (UTC)
An option to ignore template calls inside the find and replace menu would be useful to me when fixing capitalization errors. Right now I spend some time manually undoing replacements when a filename inside a template call is incorrectly capitalized. Jogers (talk) 13:13, 14 March 2007 (UTC)
Its done from transclusions of {{Lowercase and underscore}} and {{Underscore}}. So it doesnt have to be hard coded, and means, to add them to teh list of not being changed, you just need to transclude the templates, which also fixes the article header. Already exists! Reedy Boy08:21, 20 September 2007 (UTC)
Save edit toolbar show/hide preference
Latest comment: 17 years ago3 comments2 people in discussion
A good feature would be to add an "In parameter" SubRule. There is currently an "In template" SubRule which is convenient, but an "In parameter" would make life easier for trying to format specific parameters without effecting any others by mistake. Options could allow running a find and replace on the parameter, the equal sign, and the parameter value, or just the parameter value. Note: this would be used to add infoboxes to every city/town in the US, and add missing data to existing ones. Thanks, --CapitalR02:24, 5 March 2007 (UTC)
Only slightly annoying is that even if this isn't being used on a WM project wiki, it still appends "using [[Project:AutoWikiBrowser|AWB]]" to the end. Is there a way to disable this if the project being worked on isn't a WM wiki? --Izno05:52, 9 September 2007 (UTC)
Sometime I need to (remove or replace) more than one image or categorie At the same time. Now, AWB can't do this. Do you have any solution?--OsamaK07:00, 19 June 2007 (UTC)
Added in revision
The current work around, i suppose, is to do multiple runs through... Not the most efficient, i know. Other than that, we could change the image and category thing to like the FAR, so you can create multiple rules... (Sounds like the way to do it IMHO) Reedy Boy12:06, 19 June 2007 (UTC)
Sometimes there is an empty line under External links header and before the links. Here is an example. It would be nice if it was deleted. Moroever, I noticed that when this empty line exists AWB doesn't check if links above have "*". Obviously, because of an end condition.
When AWB checks links on a page, it seems to me that there is a chance to fix the redirected links. Perhaps all the links for redirect fixing could be displayed in a summary at the end of the page in a typical diff screen, ready for the selected ones to be changed. Perhaps also for the links checked to a certain point part way though a page by pressing a button. Snowman23:32, 18 October 2007 (UTC)
Would it be possible to automatically identify links on a page that point to disambiguation pages - perhaps by showing them in a different colour? Colonies Chris11:58, 19 October 2007 (UTC)
It is possible but would require loading all links on a page which means huge server load and excessive loading times. Jogers (talk) 14:47, 19 October 2007 (UTC)
Biographies of living people
Latest comment: 17 years ago3 comments2 people in discussion
But surely if we start this... We then do Dead people... I dont see the point of vanilla AWB doing it. Bot runs of tagging of pages based on the categories, usually clears up these missing talk pages.. I've started going through the deceased people again with my bot. See Special:Contributions/Reedy_Bot... Reedy Boy22:15, 20 October 2007 (UTC)
I think the feature is usufull for many new pages created. The vast majority of the users creates a page about a person but doesn't add a tag in the talk page. We can do that for dead people as well. I had the same idea for locations like if there is a geo-stub and no coord template exists and the talk page is black we can add LocateMe. :) -- Magioladitis 22:35, 20 October 2007 (UTC) ok, you are right. :) Magioladitis22:49, 20 October 2007 (UTC)
If there is a reference tag in the article (i.e. a </ref> somewhere) and there is not {{reflist}} then add ==References== and {{reflist}}. -- Magioladitis10:31, 21 October 2007 (UTC)
Added in revision
It already was discusseb before (somewhere in WT:AWB archives), and we concluded that this may cause more problems than it can fix. MaxSem11:18, 21 October 2007 (UTC)
References instead of Referenses
Latest comment: 17 years ago2 comments2 people in discussion
Using the Database Scanner to search the database is a (necessarily) very time consuming operation. It would be wonderful if the scanner could give some kind of feedback on it's progress. For example a percentage of the database search that is finished or an ETA. ʍαμ$ʏ504312:43, 14 September 2007 (UTC)
Added in revision
I do agree with you. With it being many GB it does indeed take ages... Im not sure how we'd implement it, but im sure its possible. Reedy Boy17:00, 14 September 2007 (UTC)
Added in rev 1693. It currently only updates the progress bar when an article is found. It would be better to run it in a separate thread and update the progress regularly instead. Jogers (talk) 15:00, 16 September 2007 (UTC)
You mean a category on a page..? Would require extra server load as it would have to do some form of query to check its presense... Reedy Boy18:16, 29 October 2007 (UTC)
I mean to check if a [[Category:Sample]] linked on a page exists of not. If not to display a warning.Well, I think, you are right. This requires extra server load. :( Magioladitis18:33, 29 October 2007 (UTC)
It is very useful to able to filter out non-mainspace pages from a list, but it would be great to have a full-scoped ability to filter the different namespaces. I envisage a popup box with a checkbox beside each of (main), talk, user, user talk, category, category talk, etc; just check the ones you want to keep and press go. There have been several recent occasions when I needed to filter a list in this way, but at the moment the only way to do it is to save the list to a text file, run a regex through it, and reload the list. It would be great not to have to do that. --BrownHairedGirl(talk) • (contribs) 23:37, 24 October 2007 (UTC)
Added in revision
There is the special filter where you can filter Main space, user, category, template, mediawiki and talk for each... Does this not satisfy? Click filter at the bottom of the list maker... If so, this already exists... Reedy Boy14:10, 25 October 2007 (UTC)
Ability to filter a list by whether or not the article name matches a particular regex. Particularly with long lists, I often find myself doing some filtering of this sort, but as with the filter out namespace facility suggested above, this is currently possible only by saving the list and importing it again once processed. The two facilities could perhaps be combined into one popup box. --BrownHairedGirl(talk) • (contribs) 23:56, 24 October 2007 (UTC)
Added in revision
The special filter (filter button at bottom of the list maker) has the ability to filter by some cases... I cant remember if it does regex, but if not, this should be add-able. Reedy Boy14:12, 25 October 2007 (UTC)
A feature which allows a user to create a list in Wikipedia of one language,and translate the list into an other language using the Interwiki (removing any page without an interwiki link into the requested langauge). OdMishehu07:30, 19 April 2007 (UTC)
Added in revision
Marking this as a no.. As its quite a lot of work on the servers - loading either the edit text, then parsing the IW links, or doing it via the actual display page... And i cant see it being that useful Reedy Boy11:44, 21 October 2007 (UTC)
Smart superscripts
Latest comment: 17 years ago4 comments3 people in discussion
Don't unicodify 123 => ¹²³ if there are other superscripts nearby (in the same article, probably). Reason, we can't achieve the same effect for (45)67890 (depending on font), let alone other superscripts. RichFarmbrough, 16:05 16 September2007 (GMT).
Added in revision
It seems to me that AWB changes superscript to a smaller font, which is a bit too small too see. The wiki editing tool boxes have quite recently included superscripts and subscripts, which have become standard on the wiki. Could superscripts and subscripts be replaced with text of the same size font or just left unchanged? Why are they unicodified? Snowman17:08, 3 October 2007 (UTC)
Yes the Sup/sub is above the edit window and the unicode is below it. RichFarmbrough, 11:18 11 October2007 (GMT).
Not really. We have had extensive discussions about unicodifying, in the past and I think the agreement was to smhow avoid math pages(?). But AWB is doing what it says in the tin here, we want to change the tin. RichFarmbrough, 10:47 15 October2007 (GMT).
I had a user leave a message on my talk page asking that I stopped changing it to the smaller size, and I have to agree with them that it is much harder to read. Should this just be totally disabled in the next release? Tiddly-Tom18:43, 29 October 2007 (UTC)
The next version will already replace the old style references with the new. One step forward is to replace <div class="references-2column"><references/></div> with {{reflist|2}}. -- Magioladitis00:47, 3 November 2007 (UTC)
Add a new option under Categories (in More...) called Skip action if article contains ((box to add you text)). This will help in situations like the one described above. For example we can replace Year of birth missing with Year of birth missing (living people) only if Category:Living people exists. -- Magioladitis03:05, 4 November 2007 (UTC)
Some easy way of undoing the last x edits. Literally just going over those and reversing whatever was changed. Easy to do for top edits - just use what's given by the undo button. But if other people have edited since then, you'd have to search for the last diff made by the account and reverse that edit if possible. If an exact reversion isn't possible for whatever reason, output a list of pages for manual review. – Mike.lifeguard | @en.wb15:54, 6 November 2007 (UTC)
Added in revision
This isn't the type of thing that AWB is designed to do. Other tools (mostly vandal-fighting ones) allow you to revert edits in bulk. —METS501 (talk)19:01, 16 November 2007 (UTC)
Edit redirects
Latest comment: 16 years ago3 comments2 people in discussion
It would be very helpful, if one could edit redirects. For instance, changing a bunch of redirects to soft redirects, or, unmoving pages (cleanup after a pagemove vandal, for instance), etc. SQLQuery me!13:28, 13 November 2007 (UTC)
I brought this up with Lightmouse earlier. It would be nice if a bot could replace .28 and .29 inside links, left of the pipe, with the proper parens. Other symbols could be replaced as well. This would make editing articles more user-friendly and would be less confusing for number formatting scripts. TomTheHand14:04, 18 October 2007 (UTC)
I do not know how to explain it in terms of format names. I think the problem arises because people go to an article section and just copy the URL Horsepower#Shaft_horsepower_.28shp.29. It is perfectly understandable and have done this myself.
This may already be done by AWB's general fixes routine, which means that most AWB bots that hit an article for one reason or another will also fix this problem. — madmanbum and angel06:14, 20 October 2007 (UTC)
Answer, bot could be written to do it fairly/very easily - Its just a Far of .28 and .29 to ( and ). AWB could also be expanded/its general fixes be expanded to cater for this... Reedy Boy13:33, 2 November 2007 (UTC)
Will do as part of the general fixes - rev 1813. Probably not the best place to do it.. But it can be moved if needed. One good thing, is SmackBot does general fixes!! Will be there for next release. But i'll probably start going through getting it to change the .28 to .29 Reedy Boy14:16, 2 November 2007 (UTC)
Even if you do not get a list of pages to do one hit, it might come across it just by chance. As long as it the code is there for general fixes, we are better off. Lightmouse09:49, 3 November 2007 (UTC)
I have a lot of entries in my Find/Replace list and it would be very useful to be able to find a specific one so I can switch it off in particular cases. Colonies Chris16:39, 23 May 2007 (UTC)
I would recommend keeping the special pages thing the same, but adding Log (user) and Log (page). The only issue is on the user rights log, the user is either the account being created or the user creating a new account, on that log it would have to put the value of user into page if page isnt there. Changing the current special page method would cause problems, as it's only the log pages that contain both fields. Matt/TheFearow(Talk)(Contribs)(Bot)21:04, 12 August 2007 (UTC)
Change "<br>" and "<br/>" to "<br />" (with space and slash at end). It can just be another checkbox, or something like that, sort of like the RegexTypo Fix section bit. DREAMAFTER<TALK>00:15, 30 November 2007 (UTC)
The diff screen does not display changes in white space very well, so can white space changes be made more visible, perhaps emphasised in a different colour. Snowman20:30, 28 October 2007 (UTC)
AWB should automatically convert URLs pasted into the pagename box in the "(1) Make list" portion of the software. It is unnecessary to keep the URL intact as all options are sufficiently implemented in MediaWiki. —Dispenser18:33, 9 November 2007 (UTC)
Under the "Set options -->Find and replace -->Normal" tab, it would be very helpful to be able to change the row order of what is found first. Sometimes one doesn't realize that another row is needed before another that is already in place. If there is a way of doing this like in the "Advanced" tab, I am unaware of it. Also, it would be helpful to be able to insert a row. At present, you can only delete a row. —MJCdetroit17:41, 30 November 2007 (UTC)
Oh, and no, no current way of re-ordering. Should be fairly simple to do.. Just need to see if there is a better way than the way i just thought of. —ReedyBoy19:18, 2 December 2007 (UTC)
I have scanned the most recent database dump, I found and fixed about 250 instances on links containing a terminating comma, colon or semicolon. Unfortunately, there are a handful of false positives - instances of a wiki article that does end in such punctuation. In the case of a full-stop, there are so many many false positives as to make scanning inpractical. Consequently, IMHO, I think it would not be appropriate to try and make this a standard AWB feature, but individual editors are welcome to write their own regex rules. Gaius Cornelius (talk) 11:10, 16 December 2007 (UTC)
Are these cases just redirects? I participate in this project in which we correct punctuation error. I 've seen many cases that some users instead of correcting the error they were just creating a redirect. I tried to correct many of these cases. Most of these cases need to be deleted and all the articles linking to them must be fixed. Until now I haven't found any valid case for that. For the "point-at-the-end" I as aware. They are many companies, logos, etc. using it.
The database dump is very big for me to download. If you give me the list of these links I'll try to fix them or check if they are any valid ones. -- Magioladitis (talk) 11:44, 16 December 2007 (UTC)
If there is any chance of false positives.. It cant be a standard feature. As Gaius said, someone could write their own FAR rules, or Regex's, and we'll happily code this into a custom module if needed/wanted... —ReedyBoy17:17, 17 December 2007 (UTC)
Do NOT skip if...
Latest comment: 16 years ago3 comments2 people in discussion
I think features "Do NOT skip if contains <fill in the blank box>" and "Do NOT skip if does NOT contain <fill in the blank box>" are useful. wjmt (talk) 01:04, 29 November 2007 (UTC)
Added in revision
You can do the "Do NOT skip if contains" by using skip if doesnt contain. Its reverse logic. Therefore, if its not there, it will skip. Doing esentially the same as not skipping if its the same. And surely "Do NOT skip if does NOT contain" is just double negative? —ReedyBoy10:44, 29 November 2007 (UTC)
Then I'm wondering if there could be a semi-automated module that would take images in this category, extracts the File Links, copies them in as "Article=X" and presents them to the user for review ans saving in AWB. I'm an accounting grad student, so I really have no idea how to write code or even if this is possible using the AWB interface, but I'd be willing to take the responsibility (read: complaints, blame, flames) for running such a feature, if another user could author it Mbisanz (talk) 06:43, 28 December 2007 (UTC)
Apparently, Category:Non-free images lacking article backlink already exists. So I'm just looking for a scripting extension that would copy the FUR to the # of times there is a file link, and insert an Article=x for each file link. I think this could be better done in AWB, as compared to a monobook script which lacks auto-page loads, scanning, etc. Mbisanz (talk) 17:38, 28 December 2007 (UTC)
When saving "over" an existing file, rename old file, save new file, delete old file. Tehn wehne AWB crashes in mid save, you don't lose hundreds of regexes. RichFarmbrough, 10:14 29 December2007 (GMT).
Added in revision
Added - rev 1940. Will create a file called (name of settings file).xml.old. Deletes it only if it hasnt encountered an error during saving settings —ReedyBoy17:10, 29 December 2007 (UTC)
Google List Searchability
Latest comment: 16 years ago3 comments2 people in discussion
If a user could create a text of excel list of words, that AWB could search through google and load a list of all pages with those words appearing, it would be useful. A user could then load a list from Wikipedia:Lists of common misspellings (I used excel and notepad to pull out the incorrect spellings in 5 mins) and search for all common misspellings faster than copying and pasting from IE to AWB Mbisanz (talk) 02:19, 13 December 2007 (UTC)
I'll give that a try. I wasnt to be able to take along list of misspellings and load it through AWB so taht it searches google for all articles that contain one or more of those words. Mbisanz (talk) 02:59, 18 December 2007 (UTC)
SSL support
Latest comment: 16 years ago10 comments6 people in discussion
I like AutoWikiBrowser and I wondered if there could be support for SSL (HTTPS), maybe using OpenSSL. As far as I see, there is only support für HTTP without SSL. (OK, I especially want this feature for my personal wiki... ;) but it could also be nice for paraniod folks on Wikipedia) --62.226.219.3519:22, 1 August 2007 (UTC)
Try to set up AWB to edit Wikipedia through the secure server. You will find that this is impossible because AWB assumes the URL starts with "http://" when in this case it starts with "https://". AWB should just ask the user to type out "http://en.wikipedia.org/w/" instead of "en.wikipedia.org/w/". Then, if the user types "en.wikipedia.org/w/" anyway it's safe to assume they meant HTTP. —Remember the dot(talk)21:11, 8 August 2007 (UTC)
Added in revision
Using OpenSSL shouldn't be necessary... It should be, in theory, fairly simple to have an option box to select http/https... Reedy Boy09:48, 2 August 2007 (UTC)
Dont really know. Depends how much re-working is needed, or it can just be changing the link.. If your bothered about security of your main account so much, how about having a Sock for AWB use...? —ReedyBoy22:35, 9 December 2007 (UTC)
Hi! I try the new tool in AWB (profiles). it's very good tool! but usually users use same nickname in multi wikis and need different profile at each wiki. I think you can add optional textbox (like password) to this.--OsamaK08:36, 12 July 2007 (UTC)
I've just told it to follow IE settings via the code... - rev 1912... I'll try it later (works fine with no proxy settings, i just need to setup a proxy to try it on :)) —ReedyBoy23:01, 18 December 2007 (UTC)
Just thought.. I dont think the actual web controls are following it... But im not sure.. I didnt tell them too, just the requests in code... Needs some more investigation tonight! —ReedyBoy11:55, 19 December 2007 (UTC)
Seems from googling it.. The web control just follows IE.. So fixing the code based webrequests to use the proxy.. So it should work fully... —ReedyBoy16:16, 19 December 2007 (UTC)
Oh, and with it following the IE settings.. I see no use for AWB to have a seperate way to change proxy settings too.. —ReedyBoy16:17, 19 December 2007 (UTC)
I want to be able to add a set of articles that belong to a particular WikiProject, but since the main category is only placed in the talk page, I am not sure if I can have AWB redirect it back to the main article and check it that way. So far I've had to use the Categories for stubs to check for typoes. If this feature already exists...oops I'm sorry... please let me know how to proceed in selecting the correct settings. Thanks for your help again! - Jameson L. Taitalk ♦ contribs05:50, 16 January 2008 (UTC)
This is not so much a feature request but a change in wording in the Advanced/Replace special. The tab worded now as Replace should be renamed with the word Find. and the With should be replaced with the wording Replace or Replace with. This would keep the wording standardized and not be so confusing to novices. As i was going over the manual I came upon it as being very confusing. Especially when trying to explain why the replace you type in what you want to find. Also note that the regex tester is named this way. Jeanenawhitney (talk) 17:37, 28 January 2008 (UTC)
What happened to this function. I seem to recall it being part of the "Skip articles when no changes made" option for a while. Could we have an option for this on its own? Harryboyles06:36, 3 July 2007 (UTC)
I'd also like to see this feature: I try to manually skip white-space-only changes, even though I sometimes do further cleanup on the articles, and I'd just as soon be able to not worry about them.--uɐɔlnʌɟoʞǝɹɐs15:29, 23 January 2008 (UTC)
Thanks :) Hmm... I didnt think it'd be overly difficult.. re.sub = "Regex Replace" ? And therefore re.sub(find, replace, whattolookin, re.UNICODE). Whats the old text and new text? old text being the original article text (before AWB started changing it?) and new text being what it is after AWB's moidfications?
r (as in raw) outside the quotes tell the compiler how to read the string. This avoid escaping the \ with \\, C# uses @"" I believe. re.UNICODE flag is enabled as to capture the unicode set of spaces ( ,  , etc.). — Dispenser21:03, 20 February 2008 (UTC)
Right. I call that "literal", so thats the same thing
rev 2095 - I've enabled it by default. Thanks Dispenser! Its a simple check, only catering for the added/removed whitespace, but its certainly a start! —ReedyBoy21:45, 20 February 2008 (UTC)
refs and punctuation
Latest comment: 16 years ago3 comments3 people in discussion
Currently, 'unlink' and 'unpipe' are available as options to use when choosing how to disambig a link. Can a 'reverse piping' option be added for when the piping is backwards and simply needs to be reversed? Example: a link is set up as [[USC|University of Southern California]] when what is intended is [[University of Southern California|USC]] -Gwguffey (talk) 04:48, 4 February 2008 (UTC)
A modified regexHeadings1, and removal of regexHeadings2 completely (as a regex and as a replace), should "fix"/sort this? —ReedyBoy21:51, 23 February 2008 (UTC)
Option to add/remove/change multiple categories without creating your own regex. So on the 2nd tab (More...) there'd by instead a dialog box for categorization options where multiple add/remove/change rules can be entered. —Mike.lifeguard|@en.wb14:33, 12 February 2008 (UTC)
In the advanced FaR (if that is what you mean), the items can be dragged and dropped to change position, and also highlighted, and the up and down arrow move the rules up and down... So it exists... —ReedyBoy17:46, 21 February 2008 (UTC)
Removing "Insert * here" generated by wiki toolbar
Latest comment: 16 years ago1 comment1 person in discussion
'''Bold text'''list ''Italic text''list [[Link title]]list == Headline text ==list <math>Insert formula here</math>list <nowiki>Insert non-formatted text here</nowiki>list #REDIRECT [[Insert text]] <s>Strike-through text</s>list <sup>Superscript text</sup>list <sub>Subscript text</sub>list <small>Small Text</small>list <!-- Comment --> <ref>Insert footnote text here</ref>list
Some of these are a bit more error-prone when the user does it recursively, like the headers will insert newlines. But I would expect the number of false positives to be the same as any of other General fixes.
Wait aren't editors responsible for changes made with AWB? I'm not proposing that AWB try to fix vandalism, I'm asking AWB to fix a common mistake where people try to click the edit field before the page loads an inadvertently click the toolbar which pops up and moves the page down. And since such text is almost always inappropriate for main article space I suggest we remove these instances. And if you wanted to limited it to just accidental clicking then only remove at the begging or end of section/articles.
Sort bulleted lists (alphabetically)
Latest comment: 16 years ago2 comments2 people in discussion
Improper list sorting is a major problem on Wikipedia, on stand-alone lists, in see also sections, etc. Thousands of lists on Wikipedia have sorting errors in them. An option to sort bulleted lists would be very helpful. The sort feature would have to take into account sub-bullets, and keep them under the appropriate list entry. Columnization would also need to be taken into account, as well as link brackets. Concerning sorting, when an item begins with a link that has a pipe, the pipe part should be what that item is sorted by. To make the sorting truly great, the sorting feature should recognize persons' names and sort these by surname (followed by the first name for continued sort order). The Transhumanist15:11, 2 March 2008 (UTC)
I think that may be good idea to edit just one section if the other sections in the article are not touched, but I don't know whether it is possible--84.234.42.68 (talk) 15:02, 28 January 2008 (UTC)
Added in revision
How are we supposed to get the number of the needed section? The only way is to retrieve whole page content. And if we have it all - why change only one particular section? This proposal makes no sense. MaxSem(Han shot first!)21:16, 3 March 2008 (UTC)
I think this is a good idea. The media wiki has a capacity up upload sections only, so it is possible. How does mediawiki know what sections to upload? The software could count the number of headings and work out what sections need updating. Or it could work out the line numbers of the headings and work out the sections that need uploading with the line number of the edits made on AWB. Surely, this will made the AWB uploads quicker, and use up less media wiki capacity. Snowman (talk) 22:41, 3 March 2008 (UTC)
You're suggesting to replace "load page→modify it→save page" with "get page text→determine which section we should modify→open that section for editing→modify it→save it". That is, 1 full-size request and two potentially smaller ones instead of two full-sized. This is not going to improve performance. MaxSem(Han shot first!)09:33, 4 March 2008 (UTC)
Add a button to copy the regex and the associated settings in the regex tester into the normal or advanced find and replace settings. mattbr17:15, 26 January 2008 (UTC)
There's a [Wikipedia:Disambiguation pages with links#Progress|big] disambig repair backlog still to be done on Wikipedia. Imagine if AWB had a feature like this:
-----------------------------------------------------------------
AI-assisted Disambig Repair Dialog Box [_][X]
-----------------------------------------------------------------
AWB found 3 links to race which should probably be modified
to point to Race (classification of human beings). Please
select the links to modify.
[ ] "...these flows -- such as race, gender, socioeconomic status..."
from article Anatol Rapoport
[ ] "...a cognizable racial group, and that the..."
from article Batson v. Kentucky
[ ] "...part, a national, ethnical, racial or religious group, as..."
from article Genocides in history
[ Modify checked links ] [ Cancel ]
The artificial intelligence could look at one or many things:
the source and suggested target pages' places in the category tree,
nearby links (when someone writes "race or ethnic group" you know what they meant by "race"),
nearby words in the paragraph,
popularity of the target page (AWB should think twice before repointing a link to point to an obscure page with few inbound links like Colonel Race instead of to a popular page like Race (classification of human beings)
context of the link (is the link in italics at the top of a page? or in the middle of an article? or a See Also list? Italic links at the top of a page should probably not be changed.)
or other things (suggestions welcome; perhaps a computer science or linguistics professor could help us out more here).
It does sounds like a good idea, in principle, applying it wouldnt be so easy though. Implementing more than very basic AI isnt a simple process... AWB does have its basic disambiguation program... I dont know if this is just outside the scope for AWB. As the ruleset could/would have to be VERY large to be able to get it to do more than a few cases... —ReedyBoy20:48, 22 January 2008 (UTC)
Unfortunately, this task is extremely hard to accomplish, and even when we make it, there will always be mor... id... lazy as... people who will hit 'Save' w/o actually checking what machine suggests them to do. I'm inclined to WONTFIX this. MaxSem(Han shot first!)20:53, 3 March 2008 (UTC)
I'd like to see a strong consensus that EACH AND EVERY such link are to be removed, no matter what the editors of a particular page may agree upon before implementing this feature. MaxSem(Han shot first!)14:00, 3 March 2008 (UTC)
It is not just solitary years, there are lots of date fragments (e.g. months only, days of the week) that have unnecessary links. Code to delink date fragments is available at User:Lightmouse/javascript conversion. It would be easy to build an AWB option using the code. In the meantime, feel free to use the code yourself.
For example, here is a copy of the regex that unlinks solitary years:
i'm sorry, MaxSem i'm not sure i understand your objections. consensus for awb inclusion? or consensus in article changes? what are you arguing? --emerson720:35, 3 March 2008 (UTC)
Was just reporting what was asked... The Regex tester, is a good idea in AWB, but agreed, this is too specific... With other tools being out there! —ReedyBoy10:15, 5 March 2008 (UTC)
I've added a help feature to regex tester, which opens MSDN in browser. Don't think that anything else needs to be done. Except for tighter integration of tester with other parts, maybe. MaxSem(Han shot first!)10:59, 5 March 2008 (UTC)
That would sound better. I believe there is a feature request, asking to be able to open the regex tester from a few various places, transferring certain text accross, allowing the user to edit it, and it be copied back automatically... —ReedyBoy14:36, 5 March 2008 (UTC)
Please can AWB remove leading and trailing spaces in wikilinks. I cannot see any circumstance when these are valid. They are hard for the human eye to detect but easy for a machine. Lightmouse13:35, 2 December 2007 (UTC)
Suggested regex for spaces in wikilink problem (I am sure it can be more efficient:
//remove undesirable space from beginning of wikilink and move it outside link (up to 30 characters in wikilink)ArticleText=Regex.Replace(ArticleText,@"\[\[ ([^\]]{1,30})\]\]"," [[$1]]");//remove undesirable space from end of wikilink and move it outside link (up to 30 characters in wikilink)ArticleText=Regex.Replace(ArticleText,@"\[\[([^\]]{1,30}) \]\]","[[$1]] ");//remove undesirable double space from middle of wikilink (up to 61 characters in wikilink)ArticleText=Regex.Replace(ArticleText,@"\[\[([^\]]{1,30}) ([^\]]{1,30})\]\]","[[$1 $2]] ");
Suggested regex for double spaces in dates problem:
//remove undesirable double space between links in date (day first)ArticleText=Regex.Replace(ArticleText,@"(?i)(\[\[\d\d? (?:January|February|March|April|May|June|July|August|September|October|November|December)\]\]),? ? ?(\[\[\d{1,4}\]\])","$1 $2");//remove undesirable double space between links in date (day second)ArticleText=Regex.Replace(ArticleText,@"(?i)(\[\[(?:January|February|March|April|May|June|July|August|September|October|November|December) \d\d?\]\]),? ? ?(\[\[\d{1,4}\]\])","$1 $2");
I'll just smack it in the sandbox and do some tests. Thanks for the actual code comments too, means we can put them in with it. I'll try and work out where to put them! —ReedyBoy10:01, 22 February 2008 (UTC)
rev 2098 - Added it, in a very basic form, im doing it just quickly before i go out to Uni. I will change it later to use compiled regex's, and also look for matches, and only replace if there is matches - A bigger improvement to preformance than just by improving the regex alone (dont know if they actually can be improved. ;)). Also uploaded it to my SVN snapshots, so feel free to help out with testing. Thanks!! —ReedyBoy10:12, 22 February 2008 (UTC)
I have just spotted a small error. It adds an unwanted space after wikilink, not a big deal but worth correcting
//remove undesirable double space from middle of wikilink (up to 61 characters in wikilink)//do not use this: unwanted space at end of replace text: ArticleText = Regex.Replace(ArticleText, @"\[\[([^\]]{1,30}) ([^\]]{1,30})\]\]", "[[$1 $2]] ");//use this: unwanted space removed: ArticleText = Regex.Replace(ArticleText, @"\[\[([^\]]{1,30}) ([^\]]{1,30})\]\]", "[[$1 $2]]");
More errors spotted. The date '2 October' has a double space F-105 Thunderchief, it correctly changes that to a single space but it delinks it incorrectly. I think you may have taken the regex from this page in view mode. There are square brackets in the regex that can only be seen in edit mode. Let me know if that is not clear. Lightmouse (talk) 16:33, 23 February 2008 (UTC)
//remove undesirable double space between links in date (day first)if(DateLinkWhitespace1.Match(ArticleText).Success)ArticleText=DateLinkWhitespace1.Replace(ArticleText,"$1 $2");//remove undesirable double space between links in date (day second)if(DateLinkWhitespace2.Match(ArticleText).Success)ArticleText=DateLinkWhitespace2.Replace(ArticleText,"$1 $2");
Presumably, its due to the ([^\]]{1,30}) on the other links - They are not matching the closing brackets at the start.. —ReedyBoy16:46, 24 February 2008 (UTC)
Whatever it is, I can't get my head round it. Can anyone else help?
Incidentally, I think I saw an error in "LinkWhitespace1", you have "$1"); instead of " $1");
On your date Regex's above, you dont have the date ones with wikilinks on the replace, but i did add it when i added the others above because i had forgotten them. Tidied up - rev 2107 - [7]
Only problem i can see Space at end: [[wiki link]] blah. - We've got a double space. Probably due to the space already there? —ReedyBoy18:22, 24 February 2008 (UTC)
Yes. Moving the space outside the wikilink will create a double space in most cases. We can do better but it will take two lines for the space at beginning of the wikilink and two lines for the space at the end of the wikilink.
//remove undesirable space from beginning of wikilink (space before wikilink) - parse this line first
ArticleText = Regex.Replace(ArticleText, @" \[\[ ([^\]]{1,30})\]\]", " $1");
//remove undesirable space from beginning of wikilink and move it outside link - parse this line second
ArticleText = Regex.Replace(ArticleText, @"\[\[ ([^\]]{1,30})\]\]", " $1");
//remove undesirable space from end of wikilink (space after wikilink) - parse this line first
ArticleText = Regex.Replace(ArticleText, @"\[\[([^\]]{1,30}) \]\] ", "$1 ");
//remove undesirable space from end of wikilink and move it outside link - parse this line second
ArticleText = Regex.Replace(ArticleText, @"\[\[([^\]]{1,30}) \]\]", "$1 ");
rev 2108 - [8], looks good. Double spaces should be eliminated, but it depends if the code is run before or after this block.. Could be moved position in general fixes to sort it —ReedyBoy19:11, 24 February 2008 (UTC)
If you want fewer lines of code, use just the 'parse this second' lines and eliminate the double spaces by other means. If you want stand-alone modules, use the four lines. I do not know whether performance is an issue. I have not spotted any more errors, perhaps we are there. Lightmouse (talk) 18:33, 25 February 2008 (UTC)
Its fine really.. Its checking for matches, then only replacing if there are matches.. Which is better than just trying to replace... =) —ReedyBoy20:07, 25 February 2008 (UTC)
readonlyRegexLinkWhitespace1=newRegex(@" \[\[ ([^\]]{1,30})\]\]",RegexOptions.IgnoreCase|RegexOptions.Compiled);readonlyRegexLinkWhitespace2=newRegex(@"\[\[ ([^\]]{1,30})\]\]",RegexOptions.IgnoreCase|RegexOptions.Compiled);readonlyRegexLinkWhitespace3=newRegex(@"\[\[([^\]]{1,30}) ([^\]]{1,30})\]\]",RegexOptions.IgnoreCase|RegexOptions.Compiled);readonlyRegexLinkWhitespace4=newRegex(@"\[\[([^\]]{1,30}) \]\] ",RegexOptions.IgnoreCase|RegexOptions.Compiled);readonlyRegexLinkWhitespace5=newRegex(@"\[\[([^\]]{1,30}) \]\]",RegexOptions.IgnoreCase|RegexOptions.Compiled);readonlyRegexDateLinkWhitespace1=newRegex(@"(?i)(\[\[\d\d? (?:January|February|March|April|May|June|July|August|September|October|November|December)\]\]),? ? ?(\[\[\d{1,4}\]\])",RegexOptions.IgnoreCase|RegexOptions.Compiled);readonlyRegexDateLinkWhitespace2=newRegex(@"(?i)(\[\[(?:January|February|March|April|May|June|July|August|September|October|November|December) \d\d?\]\]),? ? ?(\[\[\d{1,4}\]\])",RegexOptions.IgnoreCase|RegexOptions.Compiled);/// <summary>/// Fix leading, trailing and middle spaces in Wikilinks/// </summary>/// <param name="ArticleText">The wiki text of the article</param>/// <returns>The modified article text.</returns>publicstringFixLinkWhitespace(stringArticleText){//remove undesirable space from beginning of wikilink (space before wikilink) - parse this line firstif(LinkWhitespace1.Match(ArticleText).Success)ArticleText=LinkWhitespace1.Replace(ArticleText," [[$1]]");//remove undesirable space from beginning of wikilink and move it outside link - parse this line secondif(LinkWhitespace2.Match(ArticleText).Success)ArticleText=LinkWhitespace2.Replace(ArticleText," [[$1]]");//remove undesirable double space from middle of wikilink (up to 61 characters in wikilink)if(LinkWhitespace3.Match(ArticleText).Success)ArticleText=LinkWhitespace3.Replace(ArticleText,"[[$1 $2]]");//remove undesirable space from end of wikilink (space after wikilink) - parse this line firstif(LinkWhitespace4.Match(ArticleText).Success)ArticleText=LinkWhitespace4.Replace(ArticleText,"[[$1]] ");//remove undesirable space from end of wikilink and move it outside link - parse this line secondif(LinkWhitespace5.Match(ArticleText).Success)ArticleText=LinkWhitespace5.Replace(ArticleText,"[[$1]] ");//remove undesirable double space between links in date (day first)if(DateLinkWhitespace1.Match(ArticleText).Success)ArticleText=DateLinkWhitespace1.Replace(ArticleText,"$1 $2");//remove undesirable double space between links in date (day second)if(DateLinkWhitespace2.Match(ArticleText).Success)ArticleText=DateLinkWhitespace2.Replace(ArticleText,"$1 $2");returnArticleText;}
Hmm. I am not sure what the bug is. Moving the space outside the wikilink but inside the italics is inelegant but I cannot detect a fault. It is better than a space at the end of a wikilink. Have I misunderstood?
A fix will need a few goes to get right but testing for a few special characters may look something like:
readonly Regex LinkWhitespace2 = new Regex(@"(['])\[\[ ([^\]]{1,30})\]\]", RegexOptions.IgnoreCase | RegexOptions.Compiled);
if (LinkWhitespace2.Match(ArticleText).Success)
ArticleText = LinkWhitespace2.Replace(ArticleText, "$1$2");
Testing for non-word characters may look something like:
readonly Regex LinkWhitespace2 = new Regex(@"(\W)\[\[ ([^\]]{1,30})\]\]", RegexOptions.IgnoreCase | RegexOptions.Compiled);
if (LinkWhitespace2.Match(ArticleText).Success)
ArticleText = LinkWhitespace2.Replace(ArticleText, "$1$2");
I am also intrigued by the suggestion that such spaces may be desirable sometimes. I would like to know more.
Add a button to copy a regex and the associated settings in the find and replace settings into the regex tester, make some changes and 'save' them back into the settings. mattbr17:15, 26 January 2008 (UTC)
Jump to bottom, when working on nav templates and categories that need a lot of manual tweaking it would be handy if AWB would should the end of the page not the top for speedier manual changes (Gnevin (talk) 15:31, 28 January 2008 (UTC))
Added in revision
Presumably you mean of the edit box...? Could probably do this as a menu option fairly easily - If its checked, jump to bottom? —ReedyBoy16:34, 24 February 2008 (UTC)
1: In the Make list section, change the dimensions of the Filter button so that the whole text fits on it
2: In the filter dialog box, switch the buttons for talk only and non-talk only namespaces so that they are more logically associated. Harryboyles08:17, 10 March 2008 (UTC)
Added in revision
For no 2 - Errm... The "Talk only" button is under talk... And "Non-talk only" under non talk pages...? And for no 1 - Thats all that is on the filter button... Is "filter" —ReedyBoy21:08, 11 March 2008 (UTC)
Regarding #2. maxsem fixed it in rev 2264. As for #1, when I open ListMaker.cs in VS, the filter button is indeed showing "Filter". However in Main.cs, and when the program runs, it only shows "Filte", truncating the "r". Same thing for the Remove Duplicates button, truncating " duplicates" and leaving "Remove". I fiddled with the Width parameter of the listMaker1 control in Main.cs, increasing it by 4 from 197 to 201 and that fixed the problem. Harryboyles10:46, 12 March 2008 (UTC)
Option to prevent duplicate entries in article list: when I create a list of articles from various searches duplicate entries may occur, which I can then remove using the duplicate entry filter. However, I'd find it helpful if there were an option to prevent/automatically remove duplicate entires in the list. Thanks Rjwilmsi (talk) 20:45, 11 March 2008 (UTC)
Small request from new user, so sorry if already old news. I just find it a niggle to have to click on the arrow right/left buttons in order to move between the "(2) Set options" and "(3) Start" tabs -- could the number of tabs between them be reduced somehow (say by hiding the "Bots" tab by default?) or maybe make the widths of the three areas ("(1) Make list", tabs, Edit box) adjustable? Otherwise looks like a nice piece of software so far. Sardanaphalus (talk) 19:13, 11 March 2008 (UTC)
The resizing of the 3 area's, i am going to say no for. When you start resizing stuff, it can cause problems, and we either get things overlapping, or cant get access to them, etc. Please try a svn snapshot - Select the one with the highest number, and let us know if that is acceptable. As for the "show/hide bot tab" - that is still feasible. :) —ReedyBoy20:42, 11 March 2008 (UTC)
Yes, understood about the three areas. Looks like an option to show/hide one or more of the tabs (not just or only the bot tab?) is neatest solution. Thanks for your reply. Sardanaphalus (talk) 14:49, 12 March 2008 (UTC)
Feature added in next release
