Hi @seanbudd, this is a small fix, can it be merged?

HI @seanbudd
I can't seem to successfully add a translator's note with Chinese characters, otherwise Gettext will throw an error.
So, even though I applied your suggested changes, translators using Poedit cannot see the comments.

You may give a try be defining the coding as first line of your file:

# -*- coding: utf-8 -*-

It is written in NVDA's codingStandards.md, in the 'Encoding' paragraph.

Even if it is not required for Python 3, it seems it is required in some situations for gettext. See here.

• PASS: Unit tests.
• PASS: System tests.
• PASS: Lint check.
• Build (for testing PR): https://ci.appveyor.com/api/buildjobs/3kbl0q3w53p8wai8/artifacts/output/nvda_snapshot_pr13675-25389,50e31124.exe
• FAIL: Translation comments check. Translation comments missing or unexpectedly included. See build log for more information.

See test results for failed build of commit 50e311247a

Thanks @CyrilleB79

I define the coding on the first line, but when I run scons pot it still throws an error:

cd source && D:\git\nvda\miscDeps\tools\xgettext.exe -o D:\git\nvda\output\nvda_snapshot_source-CorrectedChineseBrailleT
ableNames-4aa8735.pot --package-name NVDA --package-version source-CorrectedChineseBrailleTableNames-4aa8735 --foreign-u
ser --add-comments=Translators: --keyword=pgettext:1c,2 --from-code utf-8 --language=python --files-from=D:\git\nvda\out
put\potSourceFileList.txt
scons: *** [output\nvda_snapshot_source-CorrectedChineseBrailleTableNames-4aa8735.pot] UnicodeDecodeError : 'gbk' codec
can't decode byte 0xad in position 2154: illegal multibyte sequence
Traceback (most recent call last):
  File "D:\git\nvda\.venv\lib\site-packages\SCons\Action.py", line 1279, in execute
    result = self.execfunction(target=target, source=rsources, env=env)
  File "D:\git\nvda\SConstruct", line 474, in makePot
    for lineNum, line in enumerate(inp):
UnicodeDecodeError: 'gbk' codec can't decode byte 0xad in position 2154: illegal multibyte sequence
scons: building terminated because of errors.
Deactivating NVDA Python virtual environment

• FAIL: Translation comments check. Translation comments missing or unexpectedly included. See build log for more information.
• PASS: Unit tests.
• PASS: Lint check.
• PASS: System tests.
• Build (for testing PR): https://ci.appveyor.com/api/buildjobs/xq6abp6b7p30mqun/artifacts/output/nvda_snapshot_pr13675-25390,b5b30f0b.exe

See test results for failed build of commit b5b30f0ba8

Thanks @CyrilleB79

I define the coding on the first line, but when I run scons pot it still throws an error:

cd source && D:\git\nvda\miscDeps\tools\xgettext.exe -o D:\git\nvda\output\nvda_snapshot_source-CorrectedChineseBrailleT
ableNames-4aa8735.pot --package-name NVDA --package-version source-CorrectedChineseBrailleTableNames-4aa8735 --foreign-u
ser --add-comments=Translators: --keyword=pgettext:1c,2 --from-code utf-8 --language=python --files-from=D:\git\nvda\out
put\potSourceFileList.txt
scons: *** [output\nvda_snapshot_source-CorrectedChineseBrailleTableNames-4aa8735.pot] UnicodeDecodeError : 'gbk' codec
can't decode byte 0xad in position 2154: illegal multibyte sequence
Traceback (most recent call last):
  File "D:\git\nvda\.venv\lib\site-packages\SCons\Action.py", line 1279, in execute
    result = self.execfunction(target=target, source=rsources, env=env)
  File "D:\git\nvda\SConstruct", line 474, in makePot
    for lineNum, line in enumerate(inp):
UnicodeDecodeError: 'gbk' codec can't decode byte 0xad in position 2154: illegal multibyte sequence
scons: building terminated because of errors.
Deactivating NVDA Python virtual environment

Sorry, I do not have any other idea on the issue here.

Hello @seanbudd and @cary-rowen

After thinking again to this PR, I think that it is not a good idea to change the name of Chinese Braille tables in English. Indeed it seems to me that it will be very difficult for non-Chinese translators to find a correct translation of the translatable strings introduced in this PR.
Wouldn't it rather be possible to keep the table name as is (i.e. close this PR) and contact Chinese translators so that they translate the table name correctly in Chinese?

For French we have modified the translations so that it corresponds to what is known in French, but only at the translation level.; the English name has not been modified. For example:

• In English: "French (unified) grade 2"
• In French: "Français (unifié) abrégé"

In French, this notion of Braille grade is not used and we use "abrégé" (contracted) to describe it.

What do you think?

Hi @CyrilleB79
I think it makes sense to use the names suggested in the Chinese national standard documents in NVDA.
Also, I'm a member of the Chinese translation team, and the note to the translator doesn't seem necessary because I'm well aware of what this change represents.
Seriously, as a Chinese translator, I'm confused by the original display name using "Grade1","Grade2" to distinguish the two tables, because no one in China knows such a name.
The NVDA Chinese team also intends to submit a new Chinese Braille table to Liblouis. Obviously this new Braille table should not be named "Grade3".
Anyway, thanks for your help.

Hi,

I'm raising a possible red flag here: even if native speakers know about what the original string and the translated text says, you should still provide translations for benefit of non-native speakers or folks who may not know about national standards. Therefore, I strongly urge you to add translator comments.

A possible compromise would be something I did when adding the 2006 edition of Korean braille to Libluis and NVDA: add the names to be used in China and elsewhere is parentheses. If not, you should still provide translator comments, as without that, this pull request will not proceed.

Thanks.

Hi, @josephsl
Since the translation contains Chinese strings, I defined utf-8 encoding at the beginning of the file, but the translation check still throws an error, and there is no way to go to the next step.
I might need some help, do you have any possible hints?

Thanks

Hi, I did add translation comments in brailleTables.py.

Hi, I did add translation comments in brailleTables.py.

@josephsl, to clarify things regarding the technical encoding issue:
The issue is the following:
In the source file brailleTables.py, @cary-rowen has added non-Ascii characters in one of the translator comment. The file is correctly encoded in UTF8 (without BOM) and the first line is the encoding specification as defined at Python 2 time. Though, the Translation comments check fails.
Could you help here?

@cary-rowen:
Regarding the translation of braille table names, I understand that grade 1/2 does not make sense for Chinese people. I will let people more familiar with Braille in various languages comment here, in case they have an idea of translator comment that could help clarifying things for non-Chinese translators.

As to why the checkPot fails it is because Python open function opens the text files using the default ANSI code page on Windows and the ANSI code page on your system (and on AppVeyor) cannot correctly decode utf-8 text.
To fix this you need to specify explicit encoding for the open calls in sconstruct i.e. replace:
with open(potFn, "rt") as inp, open(tmpFn, "wt") as out:
with:
with open(potFn, "rt", encoding="utf-8") as inp, open(tmpFn, "wt", encoding="utf-8") as out:
and do the same for the open call in the checkPot file in the tests directory.

As to why the checkPot fails it is because Python function opens the text files using the default ANSI code page on Windows and the ANSI code page on your system (and on AppVeyor) cannot correctly decode utf-8 text.
Hi, @lukaszgo1

Thank you very much for your help, I followed your instructions, made these changes and it works perfectly.
However, this change seems to be outside the scope of this PR, do I need to open a separate PR to fix this?

Thanks Again

@cary-rowen I'm not sure if this deserves a separate PR. @seanbudd What do you think?

As to why the checkPot fails it is because Python open function opens the text files using the default ANSI code page on Windows and the ANSI code page on your system (and on AppVeyor) cannot correctly decode utf-8 text. To fix this you need to specify explicit encoding for the open calls in sconstruct i.e. replace: with open(potFn, "rt") as inp, open(tmpFn, "wt") as out: with: with open(potFn, "rt", encoding="utf-8") as inp, open(tmpFn, "wt", encoding="utf-8") as out: and do the same for the open call in the checkPot file in the tests directory.

While this is a small change, it is a separate concern and would be better fixed in a separate PR.

• FAIL: Translation comments check. Translation comments missing or unexpectedly included. See build log for more information.
• PASS: Unit tests.
• PASS: Lint check.
• PASS: System tests.
• Build (for testing PR): https://ci.appveyor.com/api/buildjobs/funqkpbk74pvqx6r/artifacts/output/nvda_snapshot_pr13675-25400,a31c9f60.exe

See test results for failed build of commit a31c9f6019

Hi, @seanbudd
I applied the changes suggested by @lukaszgo1 and created #13698
Can you take a look?