Garbage (?) in output from BC3 in batch

[misi01]

I believe I'm suddenly seeing changes in the output from a batch job. I'll attach the 2 input files and maybe you can recreate it or point out what I've done wrong.

An example of the results from a comparison from before is something like

Left file: \\sthpv0013\misi01$\A_SHB_rexxcode\G01A1300_before .txt
Right file: \\sthpv0013\misi01$\A_SHB_rexxcode\G01A1300_after. txt
62c62
< DATUM_FROM 2027-05-11
---
> DATUM_FROM 2027-12-22
65c65
< RTESATS_SPEC_AVT -0.1500
---
> RTESATS_SPEC_AVT -1.0000
174,175c174,175
< DATUM_FROM 2027-04-20
< BELGRÄNS_RÄNTA_AVT -35000
---
etc etc

Now the new results

Ó…†£@†‰“…[email protected]¢£ˆ—¥ððñóq”‰¢‰ðñgqÁmâÈÂm™…§§ƒ–„…qÇðñÁ ñóððm‚…†–™…K£§£
%Ù‰‡ˆ£@†‰“…[email protected]¢£ˆ—¥ððñóq”‰¢‰ðñgqÁmâÈÂm™…§§ƒ–„…qÇð ñÁñóððm†£…™K£§£
%

The actual command I'm running from he DOS command is

"C:\Program Files\Beyond Compare 3\Bcompare.exe" "@H:\Excel examples\bc_batch.txt"

and the contents of bc_batch.txt are

Log normal "H:\Excel examples\Synclog.txt"
file-report layout:patch options:ignore-unimportant,patch-normal &
output-to:"H:\Excel examples\excel_diff.txt" &
          "\\sthpv0013\misi01$\A_SHB_rexxcode\G01A1300_before.txt" &
          "\\sthpv0013\misi01$\A_SHB_rexxcode\G01A1300_after.txt"

While I'm at it - how do I delete old attachments (I'm assuming I have to). I've uploaded 650 kb and seem (?) to be 2/3rd's full now.

[misi01]

Thought I'd added the info ...

This was using BC3 3.0.15 (9758). Have updated to the latest version, but still seeing the same garbage in the diff file.

[mdes]

G01A1300_before.txt = G01A1300_after.txt
These 2 files are identical: is it normal?

[misi01]

Normally, no, but I was forced to strip the files to be able to append them

By all means, download the files and change one or more lines. Run the comparison as I showed and see what you get. It might be something to do with my interactive BC settings. I just know that the garbage output shown is new - the old diff files looked similar to the example shown.

[mdes]

I have no problem... but I'm using BC3 Version 3.1.5 (build 10668).
Anyway, it seems that you are opening the result file with the wrong character encoding. Display it in hex to see the used encoding.

[misi01]

Character encoding

Anyway, it seems that you are opening the result file with the wrong character encoding. Display it in hex to see the used encoding.

Er - notepad ??????

As yo your version - how did you get hold of that ? (Opening BC this morning, I saw the info that a newer version was available, but downloading that only resulted in 3.1.4 (10554) being installed)

[misi01]

Aha !!!!

Found it (I think), though I don't understand why.

I start normal, interactive BC, and open the same 2 files. If I now select Tools, File format, and select Text format in the bottom left-hand pane, then the corresponding value for Miscellaneous group box (Encoding) was originally 01143 (EBCDIC Swedish/Finnish). Once I changed it to Detect instead, the resulting diff file was correct.

Don't really understand what the difference is between the two variations as shown in the attached screen. Can someone please explain. (Remember - the idea behind defining *.txt files as using 01143 was to be able to show special Swedish characters)

[mdes]

As yo your version - how did you get hold of that ? (Opening BC this morning, I saw the info that a newer version was available, but downloading that only resulted in 3.1.4 (10554) being installed)

See my signature

[mdes]

Don't really understand what the difference is between the two variations as shown in the attached screen. Can someone please explain.

I think (but this has to be confirmed) that the defaults ("Edit file format defaults:") initialize a new format (= when clicking the "New..." button, and selecting the format to use).

[mdes]

I start normal, interactive BC, and open the same 2 files. If I now select Tools, File format, and select Text format in the bottom left-hand pane, then the corresponding value for Miscellaneous group box (Encoding) was originally 01143 (EBCDIC Swedish/Finnish). Once I changed it to Detect instead, the resulting diff file was correct.

The 2 input files are in ANSI code page 1252 (probably) and not EBCDIC: this could be very seen by dumping them in hexadecimal (ASCII 1 = 0x31, EBCDIC 1 = 0xF1).

The output file looks to be UTF-8.

[Zoë]

The "Edit format defaults"->"Text format" values are used if there isn't an existing, enabled format in the list at the top that matches the file extensions.