Importing games¶

Everything arrives as PGN in the end. What differs is where it comes from and how it is encoded.

PGN files and clipboard¶

Import games → PGN files & clipboard. Choose one or more .pgn files, or paste the text. Name the collection first; the name is required, and the file names itself if you have not typed one.

Duplicates are detected two ways: by source id where the file carries one (a lichess or chess.com game id), and otherwise by a signature over the tags and the moves. Re-importing the same file adds nothing and says so.

ChessBase exports¶

ChessBase cannot be read natively — .cbh, .cbv and .si4 are undocumented binary formats, so export to PGN from ChessBase itself (File → Export → PGN, or right-click a database and choose the PGN format). Everything after that is handled.

Encoding¶

Why your accented names used to break

ChessBase writes PGN in the Windows code page, not UTF-8. Read as UTF-8, Réti becomes R<?>ti and the original letter is gone for good, because the replacement happens before anything is stored.

Files are now decoded by sniffing: UTF-8 first, and where that is genuinely impossible, Windows-1252 and then Latin-1, applied to the whole file rather than guessed per chunk. UTF-16 files are detected from their byte-order mark. The same rule applies to files you drop in through the browser, to files read from disk, and to PGNs inside a ZIP.

The tags ChessBase writes¶

ChessBase’s tag set is richer than lichess’s, and these are read, stored and searchable:

Tag	Used for
`EventDate`	sorting and filtering by the tournament rather than the game
`EventType`	filtering by format, e.g. `swiss`, `team-swiss`, `k.o.`
`WhiteTeam`, `BlackTeam`	club and league play — one filter matches either side
`WhiteTitle`, `BlackTitle`	shown beside the name, and filterable (`GM`, `IM`, …)
`WhiteFideId`, `BlackFideId`	telling apart players who share a name
`SourceTitle` / `Source`	the publication a game came from
`Variation`, `SubVariation`	shown next to the opening name
`Annotator`, `Termination`, `PlyCount`	as elsewhere

A library imported before these columns existed re-reads its own PGN once on the next launch and fills them in — you do not need to import anything again.

What is still missing¶

For completeness, the things ChessBase does that this does not:

Native database files. .cbh/.cbv/.si4 need an export step.
Medals and training annotations. ChessBase’s [%mdl …] and [%tqu …] commands survive a round trip inside the PGN, but nothing filters on them.
Player and tournament entities. Players are matched by name and FIDE ID, not by a curated player database with photographs and biographies.

Archives, folders and URLs¶

Archives, folders & URLs streams .pgn, .zip, .gz, .bz2 and — with the optional zstandard package — .zst, from a local path or an HTTPS URL. A folder is scanned recursively. Nothing is extracted to disk: archives are read as streams, so a multi-gigabyte dump does not need a multi-gigabyte temporary file.

C:\Chess\TWIC                      a folder, scanned recursively
C:\Chess\twic1500g.zip             a local archive
https://theweekinchess.com/...zip  downloaded and streamed

Your online games¶

Your online games imports from lichess or chess.com by username. A lichess token is optional here but makes rate limits kinder, and if you have connected your account in Settings the stored token is used automatically.

lichess studies¶

lichess studies imports whole studies, including private and unlisted ones once your account is connected. See Your lichess account.

Master games¶

The Master games module points at free collections — The Week in Chess by issue number, and PGN Mentor’s player and tournament files — and imports them the same way.

Naming openings¶

Games imported without an Opening tag can be named from the bundled ECO data: Database → Name openings. It never overwrites a name a file already carries.