Looking for PGN file tools

[mjlef]

I am looking for a good set of PGN file reading tools. Ideally, it would parse the file, not the win/loss/draw for each opponent by name, and have enough flexibility to parse embedded comment sin the file, like search depth or eval. I assume I will have to write the last part.

An example of the kind of thing I wold like to do is to scan a huge file of computer-computer games, and see how the win percentage of each program matches the eval, with output like this:

eval
0-10 cp: 51.2%
11-20 cp: 53%
21-30 cp 54%

and so on. It would be interesting to see how well programs eval correlate to the static eval in each searched position.

If you know of public domain PGN parsing tools that might help to this, let me know. If not, then I will write my own.

Mark

[Sven]

You might want to check the PGN tool collection by Norm Pollock.

[Laskos]

I am looking for a good set of PGN file reading tools. Ideally, it would parse the file, not the win/loss/draw for each opponent by name, and have enough flexibility to parse embedded comment sin the file, like search depth or eval. I assume I will have to write the last part.

An example of the kind of thing I wold like to do is to scan a huge file of computer-computer games, and see how the win percentage of each program matches the eval, with output like this:

eval
0-10 cp: 51.2%
11-20 cp: 53%
21-30 cp 54%

and so on. It would be interesting to see how well programs eval correlate to the static eval in each searched position.

If you know of public domain PGN parsing tools that might help to this, let me know. If not, then I will write my own.

Mark

Maybe you find of interest the thread:
http://www.talkchess.com/forum/viewtopic.php?t=55210

[G.B. Harms]

Hello Mark,

For your goal you can probably use the PGNFileReader class in
https://github.com/Bobcat/bobcat/blob/master/src/Pgn.h

It can be compiled independently from the other files in the project.

Example how to use it is in
https://github.com/Bobcat/bobcat/blob/m ... gnPlayer.h and
https://github.com/Bobcat/bobcat/blob/master/src/Tune.h

[Steve Maughan]

Hi Mark,

Did you take a look at Python Chess

https://github.com/niklasf/python-chess

Cheers - Steve

[Ozymandias]

I am looking for a good set of PGN file reading tools. Ideally, it would parse the file, not the win/loss/draw for each opponent by name, and have enough flexibility to parse embedded comment sin the file, like search depth or eval. I assume I will have to write the last part.

An example of the kind of thing I wold like to do is to scan a huge file of computer-computer games, and see how the win percentage of each program matches the eval, with output like this:

eval
0-10 cp: 51.2%
11-20 cp: 53%
21-30 cp 54%

and so on. It would be interesting to see how well programs eval correlate to the static eval in each searched position.

If you know of public domain PGN parsing tools that might help to this, let me know. If not, then I will write my own.

Mark

Maybe you find of interest the thread:
http://www.talkchess.com/forum/viewtopic.php?t=55210

I was actually going to point him to it. Did you find a1 and b1 for SF6?

[Laskos]

I am looking for a good set of PGN file reading tools. Ideally, it would parse the file, not the win/loss/draw for each opponent by name, and have enough flexibility to parse embedded comment sin the file, like search depth or eval. I assume I will have to write the last part.

An example of the kind of thing I wold like to do is to scan a huge file of computer-computer games, and see how the win percentage of each program matches the eval, with output like this:

eval
0-10 cp: 51.2%
11-20 cp: 53%
21-30 cp 54%

and so on. It would be interesting to see how well programs eval correlate to the static eval in each searched position.

If you know of public domain PGN parsing tools that might help to this, let me know. If not, then I will write my own.

Mark

Maybe you find of interest the thread:
http://www.talkchess.com/forum/viewtopic.php?t=55210

I was actually going to point him to it. Did you find a1 and b1 for SF6?

Ah, those fits, no, it was too tedious.

[Rebel]

I am looking for a good set of PGN file reading tools. Ideally, it would parse the file, not the win/loss/draw for each opponent by name, and have enough flexibility to parse embedded comment sin the file, like search depth or eval. I assume I will have to write the last part.

An example of the kind of thing I wold like to do is to scan a huge file of computer-computer games, and see how the win percentage of each program matches the eval, with output like this:

eval
0-10 cp: 51.2%
11-20 cp: 53%
21-30 cp 54%

and so on. It would be interesting to see how well programs eval correlate to the static eval in each searched position.

If you know of public domain PGN parsing tools that might help to this, let me know. If not, then I will write my own.

Mark

Something like this ?

[Ozymandias]

I am looking for a good set of PGN file reading tools. Ideally, it would parse the file, not the win/loss/draw for each opponent by name, and have enough flexibility to parse embedded comment sin the file, like search depth or eval. I assume I will have to write the last part.

An example of the kind of thing I wold like to do is to scan a huge file of computer-computer games, and see how the win percentage of each program matches the eval, with output like this:

eval
0-10 cp: 51.2%
11-20 cp: 53%
21-30 cp 54%

and so on. It would be interesting to see how well programs eval correlate to the static eval in each searched position.

If you know of public domain PGN parsing tools that might help to this, let me know. If not, then I will write my own.

Mark

Maybe you find of interest the thread:
http://www.talkchess.com/forum/viewtopic.php?t=55210

I was actually going to point him to it. Did you find a1 and b1 for SF6?

Ah, those fits, no, it was too tedious.

Them's the breaks, I guess. Did you at least get enough data for SF6, at move 70? I'm interested in the expected score, at an eval of 350cp.