8:04 p.m.
Hi Allin and others on the gretl team,
I tried to find a simple string function similar to OX’s strfind() function, or Excel’s
find() function, that gives me the position of a text string in another text string and
couldn’t seem to find one.
All the functions in the “Strings” section gives something else then what I want, it
seem.
Best,
Fred
9:22 p.m.
On Mon, 4 Apr 2022, Fred Engst wrote:
Hi Allin and others on the gretl team,
I tried to find a simple string function similar to OX’s strfind()
function, or Excel’s find() function, that gives me the position
of a text string in another text string and couldn’t seem to find
one.
All the functions in the “Strings” section gives something else
then what I want, it seem.
There isn't a built-in function that gives you exactly the match
position of one string in another, but the strstr() function can be
used for that purpose, as in
<hansl>
string s = "What's in a name?"
string p = strstr(s, "name")
if s != ""
printf "found at offset %d\n", strlen(s) - strlen(p)
endif
</hansl>
Allin
9:24 p.m.
On Sun, 3 Apr 2022, Allin Cottrell wrote:
There isn't a built-in function that gives you exactly the match
position of
one string in another, but the strstr() function can be used for that
purpose, as in
<hansl>
string s = "What's in a name?"
string p = strstr(s, "name")
if s != ""
printf "found at offset %d\n", strlen(s) - strlen(p)
endif
</hansl>
I meant,
if p != ""
in the test above.
Allin
9:22 a.m.
Am 04.04.2022 um 03:24 schrieb Allin Cottrell:
On Sun, 3 Apr 2022, Allin Cottrell wrote:
> There isn't a built-in function that gives you exactly the match
> position of one string in another, but the strstr() function can be
> used for that purpose, as in
>
> <hansl>
> string s = "What's in a name?"
In the spirit of Jack's famous one-liners I was going to suggest the
following: strlen(strsplit(s, "name", 1))
However, this gives me 11 instead of the expected 12. And the reason is
that trailing whitespace is chopped off by strsplit, as is duly documented.
So the one-liner based on Allin's suggestion is:
strlen(s) - strlen(strstr(s, "name"))
or in general strlen(s) - strlen(strstr(s, tobefound)) with tobefound
properly defined. I don't think the prior check s != "" is necessary.
Actually we could put this into the 'extra' addon; apparently there is
some demand.
But note that this is just based on the first occurrence that is found.
In principle we could generalize this to return a matrix with all positions.
cheers
sven
1:14 p.m.
Am 06.04.22 um 19:13 schrieb Sven Schreiber:
Am 04.04.2022 um 15:22 schrieb Sven Schreiber:
> Actually we could put this into the 'extra' addon; apparently there is
> some demand.
Or, as Jack reminded me, there's Artur's string_utils package which also
would be a natural place for that...
Hi,
I had the same idea, and would like to put the proposed function into
that package. Hope to find some time during the weekend.
Best,
Artur
911
days inactive
913
days old
5 comments
4 participants
participants (4)
-
Allin Cottrell -
Artur T. -
Fred Engst -
Sven Schreiber