REPLACE TEXT

[lazic67]

Hello everyone,

I have one problem that I do not know how to solve.
I have a file called eph.txt, plain ASCII file.
How do I change my name, in fact I would like to translate the text.
For example instead of 'filename' in the same spot to put the 'Ime fajla'
data behind ':' To stay where they are.

I'm using CompaqVisualFortran.


Thanks for your help

P.S. below is a look eph.txt size

filename: Baji029H.11o
file format: RINEX
file size (bytes): 4524915
start date & time: 2011-01-29 06:00:00.000
final date & time: 2011-01-29 06:59:59.000
sample interval: 1.0000
possible missing epochs: 0
4-char station code: Baji
station name: Baji
station ID number:
antenna ID number:
antenna type: TRM41249.00 TZGD
antenna latitude (deg): 43.968457
antenna longitude (deg): 19.570591
antenna elevation (m): 306.728
antenna height (m): 0.0000
receiver ID number: 0220368043
receiver type: TRIMBLE 5700
receiver firmware: Nav 2.24 / Boot 2
RINEX version: 2.10
RINEX translator: GPSNet 2.70 3641
trans date & time: 2011-01-29 06:59:00.000

 

[xwb]

It is only a problem if you are trying to read and write the same file. It is a lot easier with 2 files. The basic algorithm is

1) Copy the file to another file (say eph.old)
2) Open eph.txt for writing, eph.old for reading
3) Read a line from eph.old
4) Find out where the colon is
5) Extract text before the colon - compare and get translation. You can do this with a lookup table
6) Write translation, colon and text after the colon to eph.txt
7) Repeat from 3 until end of file.
8) Close both files

Which part are you having problems with?

 

[lazic67]

The problem is in step 4 and 5. I can not write code
for finding the column, Comparison and translation.

Thanks for your help

 

[GerritGroot]

Having CHARACTER(LEN=80) :: MyLine

Once you have read the MyLine variable you can use INDEX and then MyLine(i1:i2)

See:

http://publib.boulder.ibm.com/infoc...sp?topic=/com.ibm.xlf101a.doc/xlflr/index.htm

for an example

 

[mikrom]

Look at the example from this older thread

http://www.tek-tips.com/viewthread.cfm?qid=1602458&page=11

Although it's about parsing a config file, but maybe it helps you how to start.

 

[xwb]

You can have some very simple minded translations: phrase for phrase substitution. You have to make sure the number of space in between words is exactly the same

!  --------------------------------------------------
!  Silverfrost FTN95 for Microsoft Visual Studio
!  Free Format FTN95 Source File
!  --------------------------------------------------
subroutine xlat (engtext, maltext)
   character*(*), intent(in):: engtext
   character*(*), intent(out):: maltext
   integer, parameter:: english = 1, malay = 2, max = 4
   character*16:: xlatphrase(2,max) &
      / 'open door', 'buka pintu', &
        'close door', 'tutop pintu', &
        'eat', 'makan', &
        'get married', 'bersanding' /
   character*16:: eng
   
   ! so we don't need to trim
   eng = engtext
   maltext = '??' // engtext
   do ii = 1, max
      if (eng .eq. xlatphrase(english,ii)) then
          maltext = xlatphrase(malay,ii)
          return
      endif
   enddo
end subroutine


program main
   character*16 malay
   
   call xlat ('eat', malay)
   print *, malay
   call xlat ('close door', malay)
   print *, malay
   ! this will fail - no such word
   call xlat ('play', malay)
   print *, malay
   ! this will fail - too many spaces
   call xlat ('close  door', malay)
   print *, malay
   stop
end

 

[lazic67]

CIAO

I took silverfrost and tried to compile the code.

But we report a 572 error - String found where an operator was expected in the line of code 'get married', 'bersanding' /

Do you know what the problem was due to first use
silverfrost.

Greeting

 

[mikrom]

Trying to compile it with gfortran I got these errors

$ gfortran translation.f95 -o translation
translation.f95:12.35:

   character*16:: xlatphrase(2,max) &
                                  1
Error: Syntax error in data declaration at (1)
translation.f95:23.10:

      if (eng .eq. xlatphrase(english,ii)) then
         1
Error: Operands of comparison operator '.eq.' at (1) are CHARACTER(1)/REAL(4)
translation.f95:24.20:

          maltext = xlatphrase(malay,ii)
                   1
Error: Can't convert REAL(4) to CHARACTER(1) at (1)

I changed the declaration of xlatphrase to

   character*16:: xlatphrase(2,max) = reshape( &
     (/ 'open door', 'buka pintu', &
        'close door', 'tutop pintu', &
        'eat', 'makan', &
        'get married', 'bersanding' /), (/2, max/))

Now it compiles with gfortran and runs

$ translation
 makan           
 tutop pintu     
 ??play          
 ??close  door

 

[xwb]

That's strange - I wrote it using silverfrost. Anyway mikrom has come to the rescue.

 

[mikrom]

Btw, trying to compile the same source with g95 I got strange error:

$ g95 translation.f95 -o translation
In file translation.f95:13

     (/ 'open door', 'buka pintu', &
                     1
Error: Element in character array constructor at (1) has length 10 instead of 9

when I changed the array element to

     (/ 'open door', 'bukapintu', &

I got next error:

$ g95 translation.f95 -o translation
In file translation.f95:14

        'close door', 'tutop pintu', &
        1
Error: Element in character array constructor at (1) has length 10 instead of 9

Then I padded all array elements to the same length (i.e max length = 11)

   character*16:: xlatphrase(2,max) = reshape( &
     (/ 'open door  ', 'buka pintu ', &
        'close door ', 'tutop pintu', &
        'eat        ', 'makan      ', &
        'get married', 'bersanding ' /), (/2, max/))

...and the program compiled fine with g95.

I thought that it was bug of g95 compiler...

But, now I googled and found that:
"The F95 standard says that all of the elements in the array constructor have to be of the same length"

 

[xwb]

I get that when I use reshape on silverfrost.