#1675 classic lineout()

[frank]

REXX-ooRexx_4.2.0(MT)_64-bit 6.04 22 Feb 2014
Microsoft Windows [Version 6.1.7601]

---

Sample code to reproduce the bug (?) with ooRexx vs. correct or better Regina behaviour:

   TST = 'deleteme.tmp'
   call SysFileDelete TST
   call lineout TST, 'line 1 of 2'
   call lineout TST, 'last line'
   call lineout TST

   do L = 2 until lines( TST ) = 0
      S.L = linein( TST )
   end L
   S.1 = S.2
   S.2 = 'inserted line'
   do L = 1 to L
      say L S.L
      call lineout TST, S.L, L
   end L

   say 'check' TST
   exit 0    

---

Regina result (deleteme.tmp, text/plain):

line 1 of 2
inserted line
last line

---

common STDOUT::

1 line 1 of 2
2 inserted line
3 last line
check deleteme.tmp

---

ooRexx result (deleteme.tmp, decoded hex.):

Offset(h) 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F

00000000  6C 69 6E 65 20 31 20 6F 66 20 32 0D 0A 6C 61 73  line 1 of 2..las
00000010  74 20 6C 69 6E 65 0D 0A 00 00 00 00 00 00 00 00  t line..........
00000020  00 00 00 00 00 69 6E 73 65 72 74 65 64 20 6C 69  .....inserted li
00000030  6E 65 0D 0A 6C 61 73 74 20 6C 69 6E 65 0D 0A     ne..last line..

---

The first written line by ooRexx at the original EOF should be line 1 of 2 (length 11 + 1 EOL),, but somehow ends up as 12 NUL bytes.

The code would work as expected with a call SysFileDelete TST before the output loop, but I was looking for a solution without any RexxUtil or RegUtil function, because static Regina doesn't have it. In other words, I expected trouble for L = 3 and got it at L = 1, the manual rexxref/bifLineout.html 7.4.40. LINEOUT (Line Output) mumbles something about...

Note that, unlike CHAROUT, you cannot specify a position beyond the end of the stream for non-binary streams.
Whatever that is supposed to mean, I got no error at L = 3.

[frank]

Counted again, 13 NUL bytes matching 11 + CrLf are less bizarre. And the question about Whatever that is supposed to mean wasn't a part of the quoted manual, I just like to know what cannot specify a position beyond the end actually means.

[frank]

That should be an ordinary open bug report with target 5.0.0 (or next) observed in 4.2.0., please tell me how to fix that.

[Erich]

unlike CHAROUT, you cannot specify a position beyond the end of the stream

this means that call lineout 'ttt', 'x', 100 works as expected (fills in NUL bytes),
while call lineout 'ttt', 'x', 100 doesn't write 99 emtpy lines
(this assumes ttt dones't exist)

[frank]

Thanks, reading charout for your first lineout, there is no such thing as a line x + y if x is the last line before EOF and y > 1 outside of a binary RECFM FIXED. However, my case is y = 1:

There were 2 lines in the file, and I tried to write line 1, 2, 3. Of course line 1 worked as expected. The new inserted line 2 ended up at a weird place instead of after line 1. After that line 3 worked again as expected, i.e., ending up after the weird line 2.