Vimgrep before any empty line - regex

Question

I have a lot of files which starts with some tags I defined.
Example:
=Title
#context
!todo
#topic
#subject
#etc
And some text (notice the blank line just before this text).
Foo
Bar
I'd like to write a Vim search command (with vimgrep) to match something before an empty line.
How do I grep only in the lines before the first blank line? Will it make quicker grep action? Please, no need to mention :grep and binary like Ag - silver search.
I know \_.* to match everything including EOL. I know the negation [^foo]. I succeed to match everything but empty lines with /[^^$]. But I didn't manage to compose my :vimgrep command. Thank you for your help!

Answer 1

If you want a general solution which works for any content of file let me tell you that AFAK, you can't with that form of text. You may ask why ?
Explanation:
vimgrep requires a pattern argument to do the search line by line which behaves exactly as the :global cmd.
For your case we need to get the first part preceding the first blank line. (It can be extended to: Get the first non blank text)
Let's call:
A :Every block of text not containing any single blank line inside
x :Blank lines
With these only 5 forms of content file you can get the first A block with vimgrep(case 2,4,5 are obvious):
1 | 2 | 3 | 4 | 5
x | x | A | x | A
A | A | x | A | x
x | x | A
A |
Looking to your file, it is having this form:
A
x
A
x
A
the middle block causes a problem that's why you cannot split the first A unless you delimit it by some known UNIQUE text.
So the only solution that I can come up for the only 5 cases is:
:vimgrep /\_.\{-}\(\(\n\s*\n\)\+\)\#=/ %

Answer 2

AFAIK the most you can do with :vimgrep is use the \%<XXl atom to search below a specific line number:
:vim /\%<20lfunction/ *.vim
That command will find all instances of function above line 20 in the given files.
See :help \%l.

Answer 3

[...] always matches a single character. [^^$] matches a character that is not ^ or $. This is not what you want.
One of the things you can do is:
/\%^\%(.\+\n\)\{-}.\{-}\zsfoo/
This matches
\%^ - the beginning of the file
\%( \) - a non-capturing group
\{-} - ... repeated 0 or more times (as few as possible)
.\+ - 1 or more non-newline characters
\n - a newline
.\{-} - 0 or more non-newline characters (as few as possible)
\zs - the official start of the match
This will find the first occurrence of foo, starting from the beginning of the file, searching only non-empty lines. But that's all it does: You can't use it to find multiple matches.
Alternatively:
/\%(^\n\_.*\)\#<!foo/
\%( \) - a non-capturing group
\#<! - not-preceded-by modifier
^ - beginning of line
\n - newline
\_.* - 0 or more of any character
This matches every occurrence of foo that is not preceded anywhere by an empty line (i.e. a beginning-of-line / newline combo).