k6
k6 copied to clipboard
grafana
Stabilize Streams module
Feature Description
k6 exposes a k6/experimental/streams module to users. It implements the parts of the Streams API specification the team has deemed worth supporting.
The module is currently experimental and considered incomplete. To make it stable and strip it from its "experimental" label, we hereby define a list of prerequisite tasks to be completed.
Definition of done
- [ ] #3919
- [ ] #3918
- [ ] #3921
Suggested Solution (optional)
No response
Already existing or connected issues / PRs (optional)
#2974 #2978
Since the critical apparatus is used in many fields of study, it seemed to make sense to provide a common characters file for the signs found in texts where a Critical Apparatus is being used.
Further, since the project where a common character file for Hebrew has been abandoned, I created another that also includes support for back translation. This includes support for the tricky "final" characters. Finally, I updated the HBO table to reflect these changes.
We will upload tests for these next week.
Greetings, The Hebrew Common character file is ready. It follows the standard set by IHBO. This file includes support for forward and back translation. This file includes the following characters: Consonants, Consonants with dagesh, “final” characters, and vowels. Tests are included, and both the common characters and the HBO appear to pass.
Special Notes: There are currently 2 Hebrew tables on Lib Louis. HBO which is classical Hebrew and the Modern Hebrew- He.Il table which is a computer braille table. The HBO table which includes the Hebrew common character file, produces correct braille for All letters, all vowels, and all Cantillation marks. This includes notes for the Masorah both in the text of the Hebrew Bible and for the Masorah itself. You can obtain access to the Masorah through HBHS Quinta.
Additional notes: About the Modern Hebrew Table: The current Hebrew Il table is a computer braille table. It does contain some of the characters in this Hebrew common character file
If you read the Hebrew Computer Braille table documentation, the table does attempt to engage with dagesh characters that are known to be problematic. However, in my testing, I have discovered that the Hebrew Computer Braille does not effectively provide correct braille for dagesh characters, in every circumstance; especially where there are vowels.
The reason for this is because of the nature of the Unicode characters and how they are written. Even though the Unicode characters look correct visually, they are written in the Unicode stanza in a different order. For example: Bet dagesh hiriq might look correct visually, but in the Unicode it can be written, bet hiriq dagesh.
For Lib Louis to produce the correct braille, both possibilities must be presented in the table. The Hebrew common character table has been updated to provide all possibilities to ensure correct braille for both forward and back translation.
Special Note about the final characters:
The final characters are problematic because the braille for the regular characters and the final characters are the same.
To provide accurate back translation, we did 2 things: 1 we included the opt code to tell it not to back translate just the regular letter. 2 Then we also included a space and punctuation according the UEB to the final characters where you would typically see them.
This way when the back translation sees the character with the space or punctuation, it will back translate to the desired final character.
This should allow one to input a character and punctuation or space, while typing with a braille keyboard, and get the expected final character.
Final Note:
I need to do a test for the critical apparatus table. This will be done with in the next 2 days.
Finally, @egli and @bertfrees I have been in contact with the guys who developed the current Hebrew Computer Braille table, and we have some exciting updates coming .
Hi @yeatersink I fixed a few problems. Can you look at the test reports and see what is wrong?
FAIL: braille-specs/hbo
=======================
Tables have not been indexed yet. Indexing LOUIS_TABLEPATH.
Best match: /home/runner/work/liblouis/liblouis/tables/hbo.utb (15)
Input: '֯֯אֱלֹהִ֑ים '
Expected: '⠸⠍⠸⠍⠁⠢⠇⠕⠓⠊⠆⠚⠍ ' (length 14)
Received: '⠸⠍⠸⠍⠁⠢⠇⠕⠓⠊⠆⠚⠍⠀' (length 14)
Diff: Expected ' ' but received '⠀' in index 13
lou_checkyaml:./braille-specs/hbo.yaml:16: Failure
Table: /home/runner/work/liblouis/liblouis/tables/hbo.utb
Display table: unicode.dis
Input: 'הָאָֽרֶץ׃ '
Expected: '⠓⠣⠁⠣⠈⠗⠑⠮⠲ ' (length 10)
Received: '⠓⠣⠁⠣⠈⠗⠑⠮⠲⠀' (length 10)
Diff: Expected ' ' but received '⠀' in index 9
lou_checkyaml:./braille-specs/hbo.yaml:17: Failure
Table: /home/runner/work/liblouis/liblouis/tables/hbo.utb
Display table: unicode.dis
Input: 'וְהָאָ֗רֶ֯ץ '
Expected: '⠺⠄⠓⠣⠁⠣⠨⠗⠗⠑⠸⠍⠮ ' (length 15)
Received: '⠺⠄⠓⠣⠁⠣⠨⠗⠗⠑⠄⡳⠭⠆⠴⠴⠉⠄⠸⠍⠮⠀' (length 22)
Diff: Expected '' but received '⠄' in index 10
lou_checkyaml:./braille-specs/hbo.yaml:18: Failure
Table: /home/runner/work/liblouis/liblouis/tables/hbo.utb
Display table: unicode.dis
Input: 'הָיְתָ֥ה תֹ֨ה. '
Expected: '⠓⠣⠚⠄⠹⠣⠨⠍⠓ ⠹⠕⠨⠏⠓⠲ ' (length 17)
Received: '⠓⠣⠚⠄⠹⠣⠨⠍⠓⠀⠹⠕⠨⠏⠓⠲⠀' (length 17)
Diff: Expected ' ' but received '⠀' in index 9
lou_checkyaml:./braille-specs/hbo.yaml:19: Failure
Table: /home/runner/work/liblouis/liblouis/tables/hbo.utb
Display table: unicode.dis
FAILURE (6 tests, 4 failures)
FAIL braille-specs/hbo.yaml (exit status: 1)
@egli I got in a bit of a hurry and overlooked problematic spacing. Here is an updated YAML. For others who want to look, this is the first 2 verses from Genesis 1. These verses are copied from the BHS, which should explain the references to the Masorah.
It is possible that I missed some formatting from Logos Bible Software which always is fun for braille conversion software. Hopefully this should produce a passing test.
@yeatersink the problem is (I presume) in part that you generate your yaml test and you undo my changes to the yaml file. There is no display: both and I had this corrected in my changes. It needs to say display: unicode.dis.
Make sure you do not overwrite my changes. It is not fun to correct the same thing over and over again.
@egli Sorry, didn't realize that i was overwriting your change. Thanks for the feedback, will make sure to be more careful in the future.
This push includes the yaml file for the criticle apparatus as well.
@egli There is a wierd character from Logos bible software that was creating the problem in this document. The character comes from formatting that Logos uses in their text files. Usually this is removed with a "paste special" in MS Word. I believe that I have removed it.
However, there is a conflict in the fuzzing file that I am not sure that I can repair. Can you do this, please?
Hopefully this fixes the yaml test.
Sure. Thank you for all your help brother.
@bertfrees commented on this pull request.
Ok, so I'll drop everything except display-name and index-name?
If it's possible to send me the ones that are failing, that would be a huge help. It'll help me to know which animal test to look at. We've got two different ones.
@bertfrees The YAML file i for the He-common characters.yml had several unnecessary quotes and spaces that I believe were causing the errors. I make the necessary changes, I hope this fixes the problem?
@bertfrees I see that i need to add some characters to the table. I am missing a few consonants with vowels and dagesh fortays. I will update this and push the changes today.
@bertfrees and @egli I spent the past couple of hours going through the table and updating it to be sure it included the letters with dagesh fortay. I also re- did the tests making sure not to over write your changes. I have looked at the expected braille carefully, and do believe that it is correct. I hope this is ready and passes. lol
@yeatersink sorry I have to start making the release and there isn't enough time to work this out. Hopefully we can include it in the December release
If it's possible, I will have this pushed in less than an hour
After we merge all PRs it takes time to build all the artifacts (tarballs, windows binaries, etc) and I really have to leave at 5 today. So I think it is just too tight. There is always another release.
Us too. We cannot get Lib Luis to run a test on windowMattUs too. We cannot get Lib Louis to run a test on windows for some reason. The table is finished in it seems to be perfect. It just a matter of getting a test completed. Yeater- Yeshurun
You're probably best of to run tests within the windows subsystem for linux if you want to of to develop on Windows.
That's what we are working on. We started the process but ran into errors and just have to work through the process. I appreciate the advice Matt Yeater- Yeshurun
@LeonarddeR I was actually trying to create a dev container based on the existing dockerfile but I was getting a lot of build errors.
@yeatersink wrote:
The table is finished in it seems to be perfect.
You might have missed my comment regarding the inline comments wit the character names, which seem to be not matching the actual characters. For example there is:
always בְּ 12-3 # gimel with dagesh
This is definitely a bet, not a gimel.
Thanks for fixing @yeatersink. Is there any reason why some lines end with an asterisk, e.g. letter ֻ 346 # qubuts*
Yes sir, the reason why those Astria are there is because I'm an idiot and did not remove them. LolI'll remove them and update the changes tomorrow.Matt Yeater- Yeshurun
Replaced by https://github.com/liblouis/liblouis/pull/1658 and https://github.com/liblouis/liblouis/pull/1659
