;------------------------------------------------------------- ;+ ; NAME: ; WRITEGTE ; ; PURPOSE: ; Convert a data set (or data file) to NASA's GTE format. ; Allows for flexible handling of header information. ; ; CATEGORY: ; Data Input/Output ; ; CALLING SEQUENCE: ; WRITEGTE,{(DATA,HEADER,UNITS)|FILENAME=FILENAME }[,keywords] ; ; INPUTS: ; DATA, HEADER, UNITS --> data array, variable name list and unit ; name list containing the data to be converted. If DATA contains ; less than 2 elements, an input file must be provided with the ; keyword parameter FILENAME. ; ; KEYWORD PARAMETERS: ; FILENAME --> If no DATA is passed in the DATA parameter ; (togetehr with the appropriate HEADER and UNITS information), ; the data from this file will be read and converted into ; GTE format. The file format is not entirely flexible: Although ; the READDATA parameters SKP1, SKP2, AUTOSKIP and DELIM can ; be used, WRITEGTE expects the UNITS declaration always in ; the line immediately following the variable names (e.g. if ; the variable names (HEADER) are in line 2 (i.e. SKP1=1), UNITS ; must be in line 3). ; After program execution, the parameters DATA, HEADER, and UNITS ; contain the respective information of the data set read by FILENAME. ; ; OUTFILE --> Name of the output file to be generated. Default is ; 'mydata.gte'. ; ; SKP1, SKP2, AUTOSKIP, DELIM --> input parameters for the READDATA ; routine which is called if no DATA is passed. ; ; DESCFILE --> Name of a "descriptor" file which contains the ; structure information of the GTE header *and* the comments section. ; Default is 'gte_def.header'. This file is a simple ASCII file ; which contains 'tokens' identified with leading '$$' (token ; names must terminate with a blank or the end of a line). Each ; token may be repeated several times but must be unique in it's ; name and not contained in another token (e.g. $$VAR and $$VARIABLE ; will both be replaced with the cotents of VAR). Otherwise, the user ; may choose tokennames at will, and these will be replaced with ; the tokenvalues supplied in TOKENVALUES or prompted interactively. ; There are a few special tokens: ; $$FILENAME = adds the filename of the output file without path ; information ; $$SYSDATE = adds the system date in the form YYYY/MM/DD ; $$VARIABLE_NAMES = generates a comma seperated list of variable ; names (i.e. the entries of the HEADER field) ; $$VARIABLE_DEFINITIONS = generates the GTE variable definition ; block with NAME, UNIT, SCALE (always 1), OFFSET (always 0), ; MINIMUM, MAXIMUM, MISSING VALUE, DL-coding for each variable. ; $$NVARS = inserts the number of variables in the data set ; $$NCOMMENTS = inserts the number of comment lines in the data set ; $$NHEADERLINES = inserts the total number of header lines ; ; TOKENNAMES --> A string array with tokennames that shall be set ; to the corresponding values of TOKENVALUES. All tokens found in the ; description file that are neither special tokens nor passed in ; TOKENNAMES will be prompted interactively. ; ; TOKENVALUES --> A string array containing the values for the tokens ; passed in TOKENNAMES. ; ; /TEST --> set this flag to test the formatting of the header. Only ; 5 lines of data will be written. ; ; ; OUTPUTS: ; A data file that adheres to the NASA GTE specification (if the description ; file is correct). ; ; SUBROUTINES: ; pro add_variabledef,text,line,data,header,units ; constructs the variable definition block and inserts it into the ; current header ; ; pro replace_token,text,token,tokenresult,data,header,units ; replace the $$TOKENNAME in the header description with it's value. ; NOTE: One token can only be replaced once in each line of the header. ; ; function find_tokens,text ; compiles a list of all tokens in the description file. ; ; REQUIREMENTS: ; Uses READDATA and requires a header description file ; ; NOTES: ; ; EXAMPLE: ; WRITEGTE,file='this.data',skp1=1,skp2=1,delim=' ', $ ; out='thisdata.gte',desc='gtepemt.header', $ ; tokennames=['MISSIONNAME','MISSIONDATE', $ ; 'AIRCRAFT'], $ ; tokenvalues=['PEM-Tropics','1996/08/30', $ ; 'DC-8' ] ; ; produces a GTE conform data set of PEM-Tropics data from the file ; this.data . ; ; MODIFICATION HISTORY: ; mgs, 05 Jan 1998: VERSION 1.00 ; ;- ; Copyright (C) 1998, Martin Schultz, Harvard University ; This software is provided as is without any warranty ; whatsoever. It may be freely used, copied or distributed ; for non-commercial purposes. This copyright notice must be ; kept with any copy of this software. If this software shall ; be used commercially or sold as part of a larger package, ; please contact the author to arrange payment. ; Bugs and comments should be directed to mgs@io.harvard.edu ; with subject "IDL routine writegte" ;------------------------------------------------------------- pro add_variabledef,text,line,data,header,units nv = n_elements(header) vdef = strarr(nv) ; fake cfact -- should be 1.0 anyway cfact = fltarr(nv) + 1. for i=0,nv-1 do begin x = data(i,*) ind=where(x gt -666,c) if (c gt 0) then x = x(ind) vdef(i) = '"' +header(i)+ '","' +units(i)+ '",' + $ string(cfact(i),format='(f6.1)')+ $ ', 0.0, ' + $ string(min(x))+ $ ', ' + $ string(max(x))+ $ ', -999.99, 0' if(max(x) le -666) then print,header(i),' ONLY MISSING VALUES !!' if(units(i) eq 'X' OR units(i) eq 'unknown') then $ print,header(i),' : units = ',units(i) endfor nlines = n_elements(text) tmptext = [ text(0:line-1), vdef, text(line+1:nlines-1) ] text = tmptext return end pro replace_token,text,token,tokenresult,data,header,units ; limitation: each token can only be present once in each line ; (you can however have several tokens in one line, this procedure is ; called for every token !) ; while instead of for loop because number of lines can change ; common block holds number of lines after variable definition block ; (i.e. number of comment lines) common nlines,ncomment if (n_elements(ncomment) le 0) then ncomment = -1 i = 0 while (i lt n_elements(text)) do begin p = strpos(text(i),'$$'+token) if (p ge 0) then begin if (strmid(tokenresult,0,1) eq '$') then begin ; special tokens ; $1 : variable definition block if (tokenresult eq '$1') then begin ncomment = n_elements(text) - i - 1 add_variabledef,text,i,data,header,units tokenresult = '' endif ; $2 : number of comment lines if (tokenresult eq '$2') then $ tokenresult = string(ncomment,format='(i0)') ; $3 : total number of header lines if (tokenresult eq '$3') then $ tokenresult = string(n_elements(text),format='(i0)') endif if (tokenresult ne '') then begin ; normal tokens (replacement within line) ; split string into part before and after token s1 = strmid(text(i),0,p-1) p2 = strpos(text(i),' ',p+2) ; position of first blank after token if (p2 ge 0) then s2 = " "+strmid(text(i),p2+1,256) else s2 = "" ; re-compose text line text(i) = s1 + tokenresult + s2 endif endif i = i+1 endwhile return end function find_tokens,text tokens = '' for i=0,n_elements(text)-1 do begin p = 0 while (p ge 0) do begin p = strpos(text(i),'$$',p) if (p ge 0) then begin newtoken = str_sep(strupcase(strmid(text(i),p+2,25)),' ') if (newtoken(0) ne '') then $ tokens = [ tokens, newtoken(0) ] p = p+2 endif endwhile endfor if (n_elements(tokens) gt 1) then tokens = tokens(1:n_elements(tokens)-1) return,tokens end pro writegte,data,header,units,filename=filename,outfile=outfile, $ descfile=descfile, $ delim=delim,skp1=skp1,skp2=skp2,autoskip=autoskip, $ tokennames=tokennames,tokenvalues=tokenvalues, $ test=test on_error,2 ; default parameters if (not keyword_set(outfile)) then outfile = 'mydata.gte' if (not keyword_set(descfile)) then descfile = 'gte_def.header' ; check if tokenvalues were passed and add default values. ; Since default values will be appended and only the first entry ; for each token is used, defaults are overwritten by command line ; passed arguments if (n_elements(tokennames) le 0) then begin tokennames = '***' tokenvalues = '***' endif if (n_elements(tokennames) ne n_elements(tokenvalues)) then $ message,"Number of tokennames <> number of tokenvalues !" tokennames = [ strupcase(tokennames), $ "INVESTIGATOR", "PRODUCT", "MISSIONNAME", $ "MISSIONDATE", "FLIGHTNUMBER", "DATASETTYPE", $ "AVERAGING_TIME", "SAMPLING_FREQ" ] tokenvalues = [ tokenvalues, $ "Daniel J. Jacob and Martin G. Schultz", $ "Harvard University point model calculation", $ "PEM-Tropics (A)", $ '1996/08/30', $ '0', '0', '0.', '0.' ] ; check if data has been passed or must be read if (n_elements(data) lt 2 OR n_elements(header) lt 1 OR $ n_elements(units) lt 1) then begin if (not keyword_set(filename)) then $ message,"WRITEGTE: requires either input data or filename !" if (n_elements(delim) le 0) then delim = ' ' if (n_elements(skp1) le 0) then skp1=1 if (n_elements(skp2) le 0) then skp2=1 readdata,filename,data,header,comments=comments,skp1=skp1,skp2=skp2, $ delim=delim,autoskip=autoskip units = extract_comments(comments,skp1+1) ; units must be in line after ; variable names ; hope this crashes if attempt is made to read file without unit definitions ; here should be a chance to change the title line, for now we print it only if (n_elements(comments) ge 2) then title = comments(0) else title='' print,'Title line extracted from file '+filename print,title endif ; read description file with comment lines openr,dlun,descfile,/get_lun char = '' text = '' while (not eof(dlun)) do begin readf,dlun,char text = [ text, char ] endwhile close,dlun free_lun,dlun if (n_elements(text) gt 1) then text = text(1:n_elements(text)-1) $ else print,"No tokens in description file !!" ; extract tokens and try to fill them with values tokens = find_tokens(text) print,'Tokens found in description file : ' for i=0,n_elements(tokens)-1 do print,tokens(i) print tokenresults = strarr(n_elements(tokens)) for i=0,n_elements(tokens)-1 do begin ; first the easy ones if (tokens(i) eq 'FILENAME') then $ tokenresults(i) = extract_filename(outfile) if (tokens(i) eq 'NVARS') then $ tokenresults(i) = string(n_elements(header),format='(i0)') if (tokens(i) eq 'SYSDATE') then $ tokenresults(i) = strdate(/short,/sortable) if (tokens(i) eq 'VARIABLE_NAMES') then $ tokenresults(i) = string(header,format='(512(A,:,","))') ; use another level of templates for the more complicated ones ; the number determines the order in which they will be processed if (tokens(i) eq 'VARIABLE_DEFINITIONS') then $ tokenresults(i) = "$1" if (tokens(i) eq 'NCOMMENTS') then $ tokenresults(i) = "$2" if (tokens(i) eq 'NHEADERLINES') then $ tokenresults(i) = "$3" ; all others: check if entry exists in tokennames, else query if (tokenresults(i) eq "") then begin j = where(tokennames eq tokens(i)) if (j(0) ge 0) then tokenresults(i) = tokenvalues(j) $ else begin char = "" read,char,prompt="Enter value for token "+tokens(i)+" : " tokenresults(i) = char tokennames = [ tokennames, tokens(i) ] tokenvalues = [ tokenvalues, char ] endelse endif endfor ; now process token replacement into text array ; start with special tokens, then all the rest ind = where(tokenresults eq '$1') if (ind(0) ge 0) then $ replace_token,text,tokens(ind(0)),tokenresults(ind(0)),data,header,units ind = where(tokenresults eq '$2') if (ind(0) ge 0) then $ replace_token,text,tokens(ind(0)),tokenresults(ind(0)) ind = where(tokenresults eq '$3') if (ind(0) ge 0) then $ replace_token,text,tokens(ind(0)),tokenresults(ind(0)) for i=0,n_elements(tokens)-1 do $ replace_token,text,tokens(i),tokenresults(i) ; now we can write the new file print,'writing to file '+outfile openw,unit1,outfile,width=4096,/get_lun ; find index of first photolysis frequency for i=0,n_elements(header)-1 do begin p=strpos(strupcase(header(i)),'J') if (p ge 0) then begin ifrc=i goto,frcfound endif endfor print,' *** INDEX OF FIRST PHOTLOYSIS FREQ. NOT FOUND !!! ***' stop frcfound: for i=0,n_elements(text)-1 do $ printf,unit1,text(i) dummy = data(0,*) ndat = n_elements(dummy) if (keyword_set(test)) then ndat = 5 for i=0,ndat-1 do begin dataline = data(*,i) dats = string(dataline(0),dataline(1),dataline(2), $ format="(i4,2(',',f9.1))") for j=3,n_elements(header)-1 do $ if (header(j) eq 'mDENSITY' OR (j ge ifrc)) then $ dats = dats+','+string(dataline(j),format='(e11.4)') $ else dats = dats+','+string(dataline(j),format='(f10.2)') printf,unit1,dats endfor close,unit1 free_lun,unit1 print,' DONE.' return end