Find & Extract hashtags in text - lazarus

Question

I'm looking for an easy/quick way to identify and extract hashtags from a string, and temporarily store them separately - e.g.:
If I have the following string:
2017-08-31 This is a useless sentence being used as an example. #Example #Date:2017-09-01 #NothingWow (and then some more text for good measure).
Then I want to be able to get this:
#Example
#Date:2017-09-01
#NothingWow
I figured storing it in a TStringList should be sufficient until I'm done. I just need to store them outside of the original string for easier cross referencing, then if the original string changes, add them back at the end.
(but that's easy - its the extracting part I'm having trouble with)
It should start at the # and end/break when it encounters a [space].
The way I initially planned it was to use Boolean flags (defaulted to False), then check for the different hashtags, set them to true if found, and extract anything after a [:] separately.
(but I'm sure there is a better way of doing it)
Any advice will be greatly appreciated.

Answer 1

The following shows a simple console application which you could use as the basis
for a solution. It works because assigning your input string to the DelimitedText property of a StringList causes the StringList to parse the input into a series of space-limited lines. It is then a simple matter to look for the ones which start with a #.
The code is written as a Delphi console application but should be trivial to convert to Lazarus/FPC.
Code:
program HashTags;
{$APPTYPE CONSOLE}
uses
Classes, SysUtils;
procedure TestHashTags;
var
TL : TStringList;
S : String;
i : Integer;
begin
TL := TStringList.Create;
try
S := '2017-08-31 This is a useless sentence being used as an example. #Example #Date:2017-09-01 #NothingWow (and then some more text for good measure)';
TL.DelimitedText := S;
for i := 0 to TL.Count - 1 do begin
if Pos('#', TL[i]) = 1 then
writeln(i, ' ', TL[i]);
end;
finally
TL.Free;
end;
readln;
end;
begin
TestHashTags;
end.