Import Netlist Report.txt needs to be all ascii (lp:#1797449)

Original report created by Robert (robee)

When importing the netlist from schematic to layout, the log can be saved
to "report.txt".  This "text" file uses what I can only assume to be
unicode quotation marks ("), making the file challenging to read using a
standard text editor such as Vim.  Please fix this text file so that it is
really just a text file, I don't see any advantage to the way it is now.

Original tags: pcbnew

changed weight to 6

added pcbnew prioritylow statusconfirmed labels

Nick Østergaard (nickoe) wrote:

Please add version information.

What are you trying to do in vim with the report?

Robert (robee) wrote:

When the log window is open, none of the other Kicad windows work, so to
fix the errors listed, I have to close the window, in which case I can't
see the list of errors.  So the first thing I do is save the file and open
it so that I can fix schematic errors.  Also, I find myself swapping pins
in the layout often, and to sync up the schematic and layout, I have to
open the report.txt file.  Because that file has everything listed in it
instead of just the mismatches, I have to search for pin changes, and
readability makes this easier.  I can't recall exactly the problem I had,
but I was doing some copying and pasting.

Would something be lost if report.txt were simply a text file like
everything else in Kicad?

Thanks

Nick Østergaard (nickoe) wrote:

I don't see the problem

$ file report.txt
report.txt: ASCII text

The quotation marks are part ASCII. Please provide an example report if you
still think there is something to be fixed.

Robert (robee) wrote:

It most definitely is not completely readable with Vim and other text
editors.

Attachments:

• report.txt_file

Fabián Inostroza (fabianinostroza) wrote:

The file Robert attached doesn't use the ASCII quotation mark (checked
using an hex editor) but I'm still able to read it using geany, gedit,
nano, micro and vi.

In linux the report saved uses the ASCII quotation mark.

Application: kicad
Version: (6.0.0-rc1-dev-726-g82cb61b8a-dirty), release build
Libraries:
    wxWidgets 3.0.4
    libcurl/7.61.0 GnuTLS/3.5.19 zlib/1.2.11 libidn2/2.0.5 libpsl/0.20.2
(+libidn2/2.0.4) libssh2/1.8.0 nghttp2/1.32.0 librtmp/2.3
Platform: Linux 4.18.0-1-amd64 x86_64, 64 bit, Little endian, wxGTK
Build Info:
    wxWidgets: 3.0.4 (wchar_t,wx containers,compatible with 2.8) GTK+ 2.24
    Boost: 1.62.0
    OpenCASCADE Community Edition: 6.9.1
    Curl: 7.61.0
    Compiler: GCC 8.2.0 with C++ ABI 1013

Build settings:
    USE_WX_GRAPHICS_CONTEXT=OFF
    USE_WX_OVERLAY=OFF
    KICAD_SCRIPTING=ON
    KICAD_SCRIPTING_MODULES=ON
    KICAD_SCRIPTING_WXPYTHON=OFF
    KICAD_SCRIPTING_ACTION_MENU=ON
    BUILD_GITHUB_PLUGIN=ON
    KICAD_USE_OCE=ON
    KICAD_USE_OCC=OFF
    KICAD_SPICE=OFF

Robert (robee) wrote:

Sorry, I should have said that I'm using Kicad 5.0 on Windows.

Wayne Stambaugh (stambaughw) wrote:

If the report messages are translated, ascii would cause a loss of
information in translated report files.

Robert (robee) wrote:

I'm not familiar with unicode nor translating files, so I don't understand
why making this text file at least partly incompatible with text editors is
worth it.  I would think that translating ascii is something that can be
done.  I'm not 100% sure why the schematic and pcb files are also not held
to the same standard.  Shouldn't they also move to unicode then (not that I
actually want that)?  And finally, the Linux version appears to use ascii,
so the differing standards also seems strange.

jean-pierre charras (jp-charras) wrote:

Kicad uses Unicode.
But a text file is not necessary a ASCII7 text file. It is just not a
binary file.

All Kicad files use the UTF8 representation to code any Unicode char inside
a text file, to ensure compatibility between different OS and different
languages in different countries, because many users use localized names
for symbols, net names, comments... that are not ASCII7 codes.

On Windows, I am using a text editor(SciTE) that can accept not only the
current code page (but if a file uses a given code page and a non ASCII7
char, what I am seeing is no what a guy in a other country is seeing) or
UTF8 coding for Unicode chars.

Therefore, if the text file uses the UTF8 encoding, any user in any country
see the same text (as long the text editor accept UTF8 coding)

Robert (robee) wrote:

Thanks for the explanation, it finally prompted me to actually get Vim
working with unicode (set encoding=utf-8).  However, it's the inconsistency
that I don't understand.  Looking at the other files Kicad produces
(schematic, layout, project) they all use only ASCII characters, even for
the quotation mark (22), which I've verified with a hex editor.  If I had
the same problem with every other Kicad file, I would have figured it out
before now, but since report.txt is the only file I've had this problem
with, that's why I opened the bug report.  And according to Fabian, the
Linux version uses ASCII for the quotation mark.  I don't know what makes
this file special on Windows that it is treated differently.

Wayne Stambaugh (stambaughw) wrote:

I'm not sure why there is a unicode double quote in report files instead of
an ascii double quote but report files are for humans to read which entails
different languages so ascii is not adequate.  Board, schematic, project,
etc. files on the other hand are for computers to read so they always have
to be in the same other than embedded strings which are utf8 encoded.

Tomasz Wlostowski (twlostow) wrote:

Hi Wayne, probably because they're exported from a HTML view
(WX_HTML_REPORT_PANEL). I also agree that these files are intended to be
human-readable, not necessarily machine-readable. We might also remove the
quotes from these report files, I don't believe they are absolutely
necessary around component designators/values in the report file.

Wayne Stambaugh (stambaughw) wrote:

I'm fine with going to single quotes if that resolves the issue.  What I
don't understand is why double quotes would not be output as ascii unless
it has something to do with escaped characters in strings.  I'll go ahead
and make the change but technically it will still not make the report file
ascii.  Hopefully devs will see this and avoid using escaped characters in
report output strings in the future.

Wayne Stambaugh (stambaughw) wrote:

After doing some digging around, I found that this only effects the 5.0.0
package.  My development builds (including the 5.0 branch) create files
with ascii double quote characters.  It looks as though the problem has to
do with how our package builders are building KiCad.  Any ideas?  I'm going
to leave the double quotes until we understand why there is a difference
here.

Jeff Young (jeyjey) wrote:

Does this still affect the 5.1.x packages?

Robert (robee) wrote:

Yes, I'm running 5.1.4 and I still see UTF-8 characters used for double
quotes.

added import label

changed milestone to %5.1.7

added packaging label

mentioned in issue #5135 (closed)

assigned to @jeffyoung

mentioned in commit 621bfe36

added statusfix-committed label and removed statusconfirmed label

closed