;-------------------------------------------------------------
;+
; NAME:
;       GETWRD
; PURPOSE:
;       Return the n'th word from a text string.
; CATEGORY:
; CALLING SEQUENCE:
;       wrd = getwrd(txt, n, [m])
; INPUTS:
;       txt = text string to extract from.         in
;         The first element is used if txt is an array.
;       n = word number to get (first = 0 = def).  in
;       m = optional last word number to get.      in
; KEYWORD PARAMETERS:
;       Keywords:
;         LOCATION = l.  Return word n string location.
;         DELIMITER = d. Set word delimiter (def = space & tab).
;         /LAST means n is offset from last word.  So n=0 gives
;           last word, n=-1 gives next to last, ...
;           If n=-2 and m=0 then last 3 words are returned.
;         /NOTRIM suppresses whitespace trimming on ends.
;         NWORDS=n.  Returns number of words in string.
; OUTPUTS:
;       wrd = returned word or words.              out
; COMMON BLOCKS:
;       getwrd_com
; NOTES:
;       Note: If a NULL string is given (txt="") then the last string
;             given is used.  This saves finding the words again.
;             If m > n wrd will be a string of words from word n to
;             word m.  If no m is given wrd will be a single word.
;             n<0 returns text starting at word abs(n) to string end
;             If n is out of range then a null string is returned.
;             See also nwrds.
; MODIFICATION HISTORY:
;       Ray Sterner,  6 Jan, 1985.
;       R. Sterner, Fall 1989 --- converted to SUN.
;       R. Sterner, Jan 1990 --- added delimiter.
;       R. Sterner, 18 Mar, 1990 --- added /LAST.
;       R. Sterner, 31 Jan, 1991 --- added /NOTRIM.
;       R. Sterner, 20 May, 1991 --- Added common and NULL string.
;       R. Sterner, 13 Dec, 1992 --- Made tabs equivalent to spaces.
;       R. Sterner,  4 Jan, 1993 --- Added NWORDS keyword.
;       R. Sterner, 2001 Jan 15 --- Fixed to use first element if not a scalar.
;       Johns Hopkins University Applied Physics Laboratory.
;
; Copyright (C) 1985, Johns Hopkins University/Applied Physics Laboratory
; This software may be used, copied, or redistributed as long as it is not
; sold and this copyright notice is reproduced on each copy made.  This
; routine is provided as is without any express or implied warranties
; whatsoever.  Other limitations apply as described in the file disclaimer.txt.
;-
;-------------------------------------------------------------
 
 
	FUNCTION GETWRD, TXTSTR, NTH, MTH, help=hlp, location=ll,$
	   delimiter=delim, notrim=notrim, last=last, nwords=nwords
 
	common getwrd_com, txtstr0, nwds, loc, len
 
	if (n_params(0) lt 1) or keyword_set(hlp) then begin
	  print," Return the n'th word from a text string."
	  print,' wrd = getwrd(txt, n, [m])'
	  print,'   txt = text string to extract from.         in'
	  print,'     The first element is used if txt is an array.'
	  print,'   n = word number to get (first = 0 = def).  in'
	  print,'   m = optional last word number to get.      in'
	  print,'   wrd = returned word or words.              out'
	  print,' Keywords:'
	  print,'   LOCATION = l.  Return word n string location.'
	  print,'   DELIMITER = d. Set word delimiter (def = space & tab).'
	  print,'   /LAST means n is offset from last word.  So n=0 gives'
	  print,'     last word, n=-1 gives next to last, ...'
	  print,'     If n=-2 and m=0 then last 3 words are returned.'
	  print,'   /NOTRIM suppresses whitespace trimming on ends.'
	  print,'   NWORDS=n.  Returns number of words in string.'
	  print,'Note: If a NULL string is given (txt="") then the last string'
	  print,'      given is used.  This saves finding the words again.'
	  print,'      If m > n wrd will be a string of words from word n to'
	  print,'      word m.  If no m is given wrd will be a single word.'
	  print,'      n<0 returns text starting at word abs(n) to string end'
	  print,'      If n is out of range then a null string is returned.'
	  print,'      See also nwrds.'
	  return, -1
	endif
 
	if n_params(0) lt 2 then nth = 0		; Def is first word.
	IF N_PARAMS(0) LT 3 THEN MTH = NTH		; Def is one word.
 
	if strlen(txtstr[0]) gt 0 then begin
	  ddel = ' '					; Def del is a space.
	  if n_elements(delim) ne 0 then ddel = delim	; Use given delimiter.
	  TST = (byte(ddel))(0)				; Del to byte value.
	  tb = byte(txtstr[0])				; String to bytes.
	  if ddel eq ' ' then begin		        ; Check for tabs?
	    w = where(tb eq 9B, cnt)			; Yes.
	    if cnt gt 0 then tb[w] = 32B		; Convert any to space.
	  endif
	  X = tb NE TST					; Non-delchar (=words).
	  X = [0,X,0]					; 0s at ends.
 
	  Y = (X-SHIFT(X,1)) EQ 1			; Diff=1: word start.
	  Z = WHERE(SHIFT(Y,-1) EQ 1)			; Word start locations.
	  Y2 = (X-SHIFT(X,-1)) EQ 1			; Diff=1: word end.
	  Z2 = WHERE(SHIFT(Y2,1) EQ 1)			; Word end locations.
 
	  txtstr0 = txtstr[0]				; Move string to common.
	  NWDS = long(TOTAL(Y))				; Number of words.
	  LOC = Z					; Word start locations.
	  LEN = Z2 - Z - 1				; Word lengths.
	endif else begin
	  if n_elements(nwds) eq 0 then begin		; Check if first call.
	    print,' Error in getwrd: must give a '+$
	      'non-NULL string on the first call.'
	    return, -1					; -1 = error flag.
	  endif
	endelse
 
	nwords = nwds					; Set nwords
 
	if keyword_set(last) then begin			; Offset from last.
	  lst = nwds - 1
	  in = lst + nth				; Nth word.
	  im = lst + mth				; Mth word.
	  if (in lt 0) and (im lt 0) then return, ''	; Out of range.
	  in = in > 0					; Smaller of in and im
	  im = im > 0					;  to zero.
	  if (in gt lst) and (im gt lst) then return,'' ; Out of range.
	  in = in < lst					; Larger of in and im
	  im = im < lst					;  to be last.
	  ll = loc[in]					; Nth word start.
	  return, strtrim(strmid(txtstr0,ll,loc[im]-loc[in]+len[im]), 2) 
	endif
 
	N = ABS(NTH)					; Allow nth<0.
	IF N GT NWDS-1 THEN RETURN,''			; out of range, null.
	ll = loc[n]					; N'th word position.
	IF NTH LT 0 THEN GOTO, NEG			; Handle nth<0.
	IF MTH GT NWDS-1 THEN MTH = NWDS-1		; Words to end.
 
	if keyword_set(notrim) then begin
	  RETURN, STRMID(TXTSTR0,ll,LOC[MTH]-LOC[NTH]+LEN[MTH])
	endif else begin
	  RETURN, strtrim(STRMID(TXTSTR0,ll,LOC[MTH]-LOC[NTH]+LEN[MTH]), 2)
 	endelse
 
NEG:	if keyword_set(notrim) then begin
	  RETURN, STRMID(TXTSTR0,ll,9999)
	endif else begin
	  RETURN, strtrim(STRMID(TXTSTR0,ll,9999), 2)
	endelse
 
	END