Here is the tenhou replay
https://tenhou.net/0/?log=2020111100gm-0089-0000-d3c1f89e&tw=1

When I paste the link into the text box and try to right click > save link as.. it said "failed. No file"

I using chrome to run this, What did I do wrong?

Replay Link: https://tenhou.net/0/?log=2019112012gm-0009-0000-d020e5ec&tw=2

Analyzer:

It would be nice if you could claim a kan. Now, the only way to go on is to discard one of the tiles in the kan.

See: https://help.github.com/en/articles/setting-guidelines-for-repository-contributors

It would be good to add what your expectations are with regards to contributions. Do you expect mostly translation work or are you also open to some code contributions? Or whether you want a new issue created for each PR or not. A lot of the stuff currently in README.md could also be moved there, and maybe link the trello board.

I like the efficiency trainer, it is very helpful 😄

Hey there, I was trying to find out how to download majsoul game logs and found your comment here. However I tried this method on your replay analyzer and the link generated won't connect at all. Is there a new way to fetch the logs?

Also, I see that you're basically just using javascript and regular expressions to interpret the logs in a brute force way. I think it would be easier to compile the protobuf using javascript bindings and read it that way.

My ultimate goal is to cobble together code from various places and create a PR for Equim-chan/akochan-reviewer so that it can review Majsoul logs as well as Tenhou ones. I'll be using the protobuf approach once I figure out a way to download the game log reliably.

thanks!

❯ bun run build
$ react-scripts build
Creating an optimized production build...
Error: error:0308010C:digital envelope routines::unsupported
    at new Hash (node:internal/crypto/hash:67:19)
    at Object.createHash (node:crypto:133:10)
~~~
  opensslErrorStack: [ 'error:03000086:digital envelope routines::initialization error' ],
  library: 'digital envelope routines',
  reason: 'unsupported',
  code: 'ERR_OSSL_EVP_UNSUPPORTED'

here is avoidance. This project seems to have stopped, so I will develop it on my own fork!

export NODE_OPTIONS=--openssl-legacy-provider
bun run build

Initialized in #34

TODO:

• Trainer
• Analyzer
• All last trainer
• Folding
• Explorer
• Misc

Take the hand 123m12399p57s1122z .
The site says discarding 9p, 1z, 2z are equal with shanten 1, ukeire 8. and that "tiles that increase ukeire" is 0.
But after any of those discards drawing a 4s and discarding 7s increase ukeire because you can now advance the hand by drawing 3s and 6s instead of just 6s. Indeed if load 123m1239p457s1122z it does tell me that ukeire is now 12.

Thanks for developing this, it is a very nice tool to practice tile efficiency!

It would be nice if the final hand is indicated to be furiten yes/no.

2346678p3455577s -> discard 6p. Should be 3 outs instead of 2 outs.
2346678p4555677s -> discard 6p. Should be 7 outs instead of 6 outs.

Changing the suit from souzu to manzu recreated the error.

Discard: three of circles (4 tiles). That was the best choice! Your hand is now ready. Congratulations! Your efficiency was 297/298, or 99%.

99.6% gets rounded down to 99% and 98.8% gets rounded down to 98%. I suggest to keep the first decimal place for more accuracy.

Steps to reproduce:
After the first turn in the defense trainer, discard a tile that was discarded in the last turn, e.g.:

State:

Action: Discard 8s

Expected: 8s has safety rating of 15 (genbutsu).

Actual: Safety rating of 8s is not 15.

Missing space between exclamation mark and trainer.complete :triggered:

Currently you can only see it while discarding or not at all. It would be nice to be able to chose tiles without being influenced by knowledge of what the best move was, but still be able to see the optimal moves later.

In the following example, toimen declared riichi with 8p, which belongs to 2-5-8 suji line, thus 5p should fall into the category "9: suji 4/5/6". It is categorized into "1: non-suji 4/5/6" instead.