<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Recent changes to feature-requests</title><link>https://sourceforge.net/p/zsh/feature-requests/</link><description>Recent changes to feature-requests</description><atom:link href="https://sourceforge.net/p/zsh/feature-requests/feed.rss" rel="self"/><language>en</language><lastBuildDate>Wed, 13 Feb 2013 19:49:19 -0000</lastBuildDate><atom:link href="https://sourceforge.net/p/zsh/feature-requests/feed.rss" rel="self" type="application/rss+xml"/><item><title>#1 position cursor to fix errors</title><link>https://sourceforge.net/p/zsh/feature-requests/1/?limit=25#de82</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;I classify this as unlikely to be implemented, because the &lt;span&gt;[ynae]&lt;/span&gt; prompts for correction et al. are part of the base shell and operate based on the command line after lexical analysis, whereas cursor placement is a function of the ZLE library which is not always available and which operates on the input text.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Bart Schaefer</dc:creator><pubDate>Wed, 13 Feb 2013 19:49:19 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net4c52656b658d2f1744a5c03147782bb29b666e86</guid></item><item><title>#12 Allow history expansion of elements to the right of cursor </title><link>https://sourceforge.net/p/zsh/feature-requests/12/?limit=25#6bb7</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Building in this sort of "knowledge" for selected commands is not scalable, and doing so in a general way would require adding semantic analysis to the spelling checker.  The completion system already provides this kind of analysis, so we recommend using the _correct completer instead of the CORRECT_ALL option when you want this kind of specific control.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Bart Schaefer</dc:creator><pubDate>Wed, 13 Feb 2013 19:09:15 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net0c095b1d156ed25e945a4fd31be4569da387dd44</guid></item><item><title>Allow history expansion of elements to the right of cursor </title><link>https://sourceforge.net/p/zsh/feature-requests/12/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Make history expansion of current line work with elements to the right of cursor.&lt;/p&gt;
&lt;p&gt;For example expand !#2 or !#1 to somefile in the following example:&lt;/p&gt;
&lt;p&gt;$ mv _ somefile&lt;/p&gt;
&lt;p&gt;Note cursor position is marked with an underscore.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Thor Andreassen</dc:creator><pubDate>Wed, 13 Feb 2013 13:21:36 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net5c0b8453a68bc40c3363c2bfde9132d438221c87</guid></item><item><title>Do not correct spelling of second argument of mv</title><link>https://sourceforge.net/p/zsh/feature-requests/11/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Renaming a file to something similar in zsh can be a bit annoying with the otherwise very useful spelling correction enabled. Take the following test case:&lt;/p&gt;
&lt;p&gt;&amp;gt; touch test1&lt;br /&gt;
&amp;gt; mv test1 test2&lt;br /&gt;
zsh: correct 'test2' to 'test1' [nyae]? &lt;/p&gt;
&lt;p&gt;It is pretty obvious, that I do *not* want to issue the command "mv test1 test1" as zsh suggests. Actually, I could very well live with losing spelling correction for the rare cases were I want to overwrite a file with another and mistyped the second argument. So I would very much have an option to disable the spelling checking of the second argument of mv altogether&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">mudd1</dc:creator><pubDate>Mon, 27 Aug 2012 12:53:15 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net2c9dd17193e766aa7b184629d37b87f6c1928deb</guid></item><item><title>add lualatex completion</title><link>https://sourceforge.net/p/zsh/feature-requests/10/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;I won't provide a patch because it's too trivial. Just add "luatex lualatex" to the #compdef line of zsh/Completion/Unix/Command/_tex&lt;/p&gt;
&lt;p&gt;I have to admit I don't understand the difference between adding these commands and having an entry such as for xelatex which is xelatex=latex. Somethinge like this might be appropriate, too, since lualatex is the envisioned successor of pdflatex. For more information on LuaLaTeX, see &lt;a href="http://www.luatex.org/." rel="nofollow"&gt;http://www.luatex.org/.&lt;/a&gt; The two main features in short: native UTF 8 support and OpenType support, including advanced OpenType features. lualatex is even more of a drop-in replacement for pdflatex than xelatex is, so this should not be a problem.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">mudd1</dc:creator><pubDate>Mon, 27 Aug 2012 12:38:37 -0000</pubDate><guid isPermaLink="false">https://sourceforge.neta5ad64970638945436369dca0f314e6b7977affd</guid></item><item><title>aptitude auto completion does not complete 'upgrade' command</title><link>https://sourceforge.net/p/zsh/feature-requests/9/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;aptitude upg&amp;lt;tab&amp;gt; gives:&lt;/p&gt;
&lt;p&gt;aptitude -upgrade&lt;/p&gt;
&lt;p&gt;And it is trying to tell me to choose 'full-upgrade' or 'safe-upgrade'. However, aptitude does have a 'upgrade' command.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Hong Xu</dc:creator><pubDate>Sat, 31 Mar 2012 07:07:12 -0000</pubDate><guid isPermaLink="false">https://sourceforge.netf49d7c8e4dc818ef440f89301931dfa9fc2a48b8</guid></item><item><title>spelling / autocorrect feature</title><link>https://sourceforge.net/p/zsh/feature-requests/8/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;First, thanks so much for Zsh! Switching from Bash has made me a little bit more productive. That being said, there's one feature I would love to have:&lt;/p&gt;
&lt;p&gt;zsh: correct 'balh' to 'blah' [nyaNe]?&lt;/p&gt;
&lt;p&gt;The capital N would designate Never. I'm fully aware that I can alias "nocorrect" in my .zshrc, but this way would take much less time. What do you think?&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Elijah Chancey</dc:creator><pubDate>Sat, 10 Mar 2012 05:48:02 -0000</pubDate><guid isPermaLink="false">https://sourceforge.netab30a1d0f3bdc1d836a956c38362e77d8060e1ec</guid></item><item><title>warn/error for unquoted [:blank:], etc.</title><link>https://sourceforge.net/p/zsh/feature-requests/7/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;What do you think of adding a way to warn script writers/runners that&lt;br /&gt;
there is an unquoted glob like [:blank:] that is obviously intended&lt;br /&gt;
to be quoted, but that will work almost all of the time.  Until someone&lt;br /&gt;
expands it in a directory containing a file with one of these&lt;br /&gt;
single-letter file names: ":" "b" "l" "a" "n" "k".&lt;/p&gt;
&lt;p&gt;If there were an option I could set globally, somehow,&lt;br /&gt;
to make all such scripts emit a warning or even to fail,&lt;br /&gt;
I would use it.&lt;/p&gt;
&lt;p&gt;Why?  Many scripts use erroneous commands like tr -dc [:alnum:]&lt;br /&gt;
that work just fine during testing, but then fail&lt;br /&gt;
mysteriously when run in a different directory just&lt;br /&gt;
because that morning a file named "a" was created.&lt;br /&gt;
There, the [...] is being interpreted as a shell "glob",&lt;br /&gt;
rather than as the intended tr character class.  The per-application&lt;br /&gt;
fix is obviously to quote the argument: tr -dc '[:alnum:]'&lt;br /&gt;
But many don't ever notice the error until it's too late.&lt;/p&gt;
&lt;p&gt;I added a similar safety net in GNU grep:&lt;br /&gt;
There, the proper syntax is '[[:blank:]]', yet many people mistakenly&lt;br /&gt;
used '[:blank:]'.  Noticing the duplicate ":" in the character class,&lt;br /&gt;
with one at the beginning and one at the end, it's safe for grep to&lt;br /&gt;
diagnose the problem even in an otherwise-conforming implementation.&lt;br /&gt;
No one needs to include two ":" bytes in a grep [...] expression,&lt;br /&gt;
just as they never need to do that in a [...] glob.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Meyering</dc:creator><pubDate>Wed, 22 Feb 2012 12:27:15 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net89237b671dcdbf76c404fb7a0643de6b44ad7101</guid></item><item><title>Coloring stderr</title><link>https://sourceforge.net/p/zsh/feature-requests/6/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;It could be nice if zsh would color stderr natively (depending on an option for sure)&lt;/p&gt;
&lt;p&gt;There is a function to put in zshrc&lt;br /&gt;
&lt;a href="http://www.zsh.org/mla/users/2009/msg00498.html" rel="nofollow"&gt;http://www.zsh.org/mla/users/2009/msg00498.html&lt;/a&gt;&lt;br /&gt;
but due to race conditions sometimes the prompt comes back before stderr, and fuser (for example) gives output kind of shuffled&lt;br /&gt;
echo L1; echo L2 &amp;gt;&amp;amp;2; echo L3; echo L4 &amp;gt;&amp;amp;2 is a good test too&lt;/p&gt;
&lt;p&gt;I don't know if it's feasible, if not, why ? (i'm curious)&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">pir</dc:creator><pubDate>Wed, 28 Sep 2011 10:01:44 -0000</pubDate><guid isPermaLink="false">https://sourceforge.netf759123dd807aceec150a701d6299f360f612ee7</guid></item><item><title>modify command and run it.</title><link>https://sourceforge.net/p/zsh/feature-requests/5/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Hi, all&lt;/p&gt;
&lt;p&gt;If I run a command starts with  '_' , then It automatically convert into 'sudo '. For example:&lt;/p&gt;
&lt;p&gt;_vi file  #=&amp;gt; sudo vi file&lt;/p&gt;
&lt;p&gt;this is the best way to run sudo command. alias can do it, but I want all commands support this form, make every command as an alias is impossible. and preexec hook function can't do this.&lt;/p&gt;
&lt;p&gt;substitution the command and run the modified command.&lt;/p&gt;
&lt;p&gt;Guten,&lt;br /&gt;
thanks.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Guten</dc:creator><pubDate>Sun, 12 Jun 2011 16:01:26 -0000</pubDate><guid isPermaLink="false">https://sourceforge.nete63d873f5988959cde893bc831509cc7f14b09ef</guid></item></channel></rss>