I can confirm this and add that it additionally breaks tag inheritance. I first noticed that a org-collector code block that I've been using forever was generating table that were way too long. My collector block includes =:conds ( not (string-match-p "noexport" ALLTAGS)= and the extra results I was getting all were from long trees tagged at the parent :no export:

I also have some Agenda custom views that use tag inheritance that broke at the same time.

I reverted =e0072f79137bbfabdf848da6865d8e4de776a549= and both behaviors corrected themselves.

Mike

On Monday, March 19, 2012, Charles Sebold <csebold@gmail.com> wrote:
> I think this patch may have broken org-map-entries for me:
>
> e0072f79137bbfabdf848da6865d8e4de776a549 is the first bad commit
> commit e0072f79137bbfabdf848da6865d8e4de776a549
> Author: David Maus <dmaus@ictsoc.de>
> Date:   Sun Mar 18 18:38:50 2012 +0100
>
>     Require one or more spaces (+) between keyword and headline
>
>     * org.el (org-scan-tags): Require one or more spaces (+) between
>     keyword and headline.
>
>     Otherwise the re will match a line like:
>
>     * TODO@ Foobar
>
>     And assumes the @ to be part of the headline.
>
>     This fixes a glitch reported by Simon Thum in
>     <4F53DEF7.8080604@gmx.de>.
>
>     > Hi all,
>     >
>     > I have found some irritating behaviour, potentially a bug. I have a
>     > block agenda which goes like:
>     >
>     > tags-todo "@home&TODO=\"TODO\"
>     >
>     > and it displays a certain org line that reads
>     >
>     > **** TODO_ state triggers
>     >
>     > Which is just a heading for dealing with TODO state triggers, and I
>     > appended the _ as I don't want it to be a TODO.
>     >
>     > For example, the global TODO list and syntax highlighting does not
>     > consider it a todo, but C-c a m TODO="TODO" does. TODO="T" does not,
>     > so it's not very grave.
>     >
>     > Most likely, it's simply an inconsistency arising from not having a
>     > real parser. I just wanted to report it here so it may get fixed.
>
> :040000 040000 8f5974f2dd7cf5f0ad10db56d223ba09a6dbca80
> 624cee5569de7ef8f240ad75943be215bb823ccc M      lisp
>
> ---------------------------------------------------------------------------------
>
> Try the following before and after this patch was applied, for a test result:
>
> ---------------------------------------------------------------------------------
> * test one
> * test two
> * test three
>
>  #+BEGIN_SRC emacs-lisp :results both
>    (org-map-entries 'org-get-heading nil nil)
>  #+END_SRC
> ---------------------------------------------------------------------------------
>
> The results ought to be the three test headings in a table output.
> After the patch it's blank.
> --
> Charles Sebold
> Ego delendus sum
>
>