string stuff
in [Fortran]
|
From: Bart Vandewoestyne on 7 Oct 2009 12:05 On 2009-10-07, The Star King <jfb(a)npl.co.uk> wrote: > > I suggest > > character(30) :: data(10) > integer i > > read (pointsource_string,*) (data(i),i=1,10) Hmm... will this also work if the number of numbers after 'pulse' is variable? Regards, Bart -- "Share what you know. Learn what you don't."
From: Bart Vandewoestyne on 7 Oct 2009 12:18 On 2009-10-07, Richard Maine <nospam(a)see.signature> wrote: > > There isn't anything "magic". Just search for non-whitspace and then > search for whitespace. Loop the appropriate number of times. See the > index, scan, and verify intrinsics. OK. Thanks for the suggestion/input. So i think i need a subroutine that allows me to do stuff like the following: call split(inputstring, 7, outputstring1, outputstring2) Using this call, i would be splitting inputstring at the 7th whitespace-area between non-whitespace characters/words. The first part would end up in outputstring1, while the second part would end up in outputstring 2. Hmpf... looks like i'll have to do some 'annoying' programming here... :-( Unless somebody really sees a clever, short and elegant solution? Regards, Bart -- "Share what you know. Learn what you don't."
From: The Star King on 7 Oct 2009 12:35 On Oct 7, 5:18 pm, Bart Vandewoestyne <MyFirstName.MyLastN...(a)telenet.be> wrote: > On 2009-10-07, Richard Maine <nos...(a)see.signature> wrote: > > > > > There isn't anything "magic". Just search for non-whitspace and then > > search for whitespace. Loop the appropriate number of times. See the > > index, scan, and verify intrinsics. > > OK. Thanks for the suggestion/input. So i think i need a > subroutine that allows me to do stuff like the following: > > call split(inputstring, 7, outputstring1, outputstring2) > > Using this call, i would be splitting inputstring at the 7th > whitespace-area between non-whitespace characters/words. The > first part would end up in outputstring1, while the second > part would end up in outputstring 2. > > Hmpf... looks like i'll have to do some 'annoying' programming > here... :-( Unless somebody really sees a clever, short and > elegant solution? > > Regards, > Bart > > -- > "Share what you know. Learn what you don't." How about something like this (note I have "proved it correct not tried it" etc :-) i=1 do grabbed=grab(pointsource_string,data(i)) if (grabbed.eq.0) exit i=i+1 enddo function grab(string,res) integer grab character(*) string character(len(string)) res string=adjustl(string) do i=1,len(string)-1 if (string(i:i).eq.' '.and.string(i+1:i+1).ne.' ') then res=string(:i) string=string(i+1:) grab=1 return endif grab=0 end
From: The Star King on 7 Oct 2009 13:11 On Oct 7, 5:35 pm, The Star King <j...(a)npl.co.uk> wrote: > On Oct 7, 5:18 pm, Bart Vandewoestyne > > > > > > <MyFirstName.MyLastN...(a)telenet.be> wrote: > > On 2009-10-07, Richard Maine <nos...(a)see.signature> wrote: > > > > There isn't anything "magic". Just search for non-whitspace and then > > > search for whitespace. Loop the appropriate number of times. See the > > > index, scan, and verify intrinsics. > > > OK. Thanks for the suggestion/input. So i think i need a > > subroutine that allows me to do stuff like the following: > > > call split(inputstring, 7, outputstring1, outputstring2) > > > Using this call, i would be splitting inputstring at the 7th > > whitespace-area between non-whitespace characters/words. The > > first part would end up in outputstring1, while the second > > part would end up in outputstring 2. > > > Hmpf... looks like i'll have to do some 'annoying' programming > > here... :-( Unless somebody really sees a clever, short and > > elegant solution? > > > Regards, > > Bart > > > -- > > "Share what you know. Learn what you don't." > > How about something like this (note I have "proved it correct not > tried it" etc :-) > > i=1 > do > grabbed=grab(pointsource_string,data(i)) > if (grabbed.eq.0) exit > i=i+1 > enddo > > function grab(string,res) > > integer grab > character(*) string > character(len(string)) res > > string=adjustl(string) > do i=1,len(string)-1 > if (string(i:i).eq.' '.and.string(i+1:i+1).ne.' ') then > res=string(:i) > string=string(i+1:) > grab=1 > return > endif > > grab=0 > > end- Hide quoted text - > > - Show quoted text - Forgot the enddo! function grab(string,res) integer grab character(*) string character(len(string)) res string=adjustl(string) do i=1,len(string)-1 if (string(i:i).eq.' '.and.string(i+1:i+1).ne.' ') then res=string(:i) string=string(i+1:) grab=1 return endif enddo grab=0 end function grab
From: Richard Maine on 7 Oct 2009 14:26
The Star King <jfb(a)npl.co.uk> wrote: > > i=1 > > do > > grabbed=grab(pointsource_string,data(i)) > > if (grabbed.eq.0) exit > > i=i+1 > > enddo > > .... > Forgot the enddo! > > function grab(string,res) > > integer grab > character(*) string > character(len(string)) res > > string=adjustl(string) > do i=1,len(string)-1 > if (string(i:i).eq.' '.and.string(i+1:i+1).ne.' ') then > res=string(:i) > string=string(i+1:) > grab=1 > return > endif > enddo > > grab=0 > > > end function grab Not a style I favor in that it not only relies heavily on function side efects, but one of those side effects is to modify the "input" string. I'll not try to claim it is objectively bad style - just different from what I'd use. Looks a lot like C written with Fortran syntax (complete with using integer vaues 0 and 1 in place of logicals). But different stroke, etc. Somewhat more substantial is that this looks prone to exceeding array bounds (for the array data) without checking. (speaking of C-like :-). I am particularly careful about things like that when they can be triggered by oddities in input data because it is easy for that to slip by testing and show up only "in the field." If you have a fixed-size array that you are storing results in, then one might as well use that information, as in do i = 1 , size(data) grabbed=grab(pointsource_string,data(i)) if (grabbed.eq.0) exit end do if (i>size(data)) call error_halt('Some message') Or do whatever else you want for the error condition (including the possibility of ignoring it if that seems appropriate, for example if extra junk at the end of the string is ok). But most important, if this is supposed to be a general utility routine, is that it fails in a simple case. Always look for boundary conditions. In the particular case of the OP's sample, it wouldn't come up because the OP said he was looking to split after 7 fields I think it was. But if this is supposed to be a general utility routine, one might expect it to work for boundary cases such as 1 field. If the input string consists of a single field with no blanks, this function doesn't return that field. I'd consider that a bug. -- Richard Maine | Good judgment comes from experience; email: last name at domain . net | experience comes from bad judgment. domain: summertriangle | -- Mark Twain |