Skip to main content
En Croissant supports multiple methods for importing games into your database. Import from local PGN files, download curated databases, or sync your online games directly from Lichess or Chess.com.

​
Importing from PGN files

PGN (Portable Game Notation) is the standard format for chess games. En Croissant can import PGN files with advanced compression and encoding support.
1

Open the Add Database dialog

Navigate to the Databases page and click the “Add Database” button in the top-right corner.
2

Select the Local tab

Switch to the “Local” tab to import from a file on your computer.
3

Choose your PGN file

Click “Click to select PGN file” and browse to your PGN file. Supported formats:
  • .pgn - Standard PGN text files
  • .pgn.zst - Zstandard compressed PGN (recommended for large databases)
  • .pgn.bz2 - Bzip2 compressed PGN
4

Configure database details

Enter a name and optional description. The name will be auto-filled based on the filename.
5

Convert to database

Click “Convert” to start the import process. Progress will be shown in real-time.

​
Import progress tracking

During import, En Croissant displays:
  • Number of games processed
  • Processing speed (games per second)
  • Estimated time remaining
  • Real-time progress updates every 1,000 games
The conversion process is implemented in src-tauri/src/db/mod.rs:convert_pgn().

​
Supported PGN features

En Croissant’s PGN parser supports:
  • Event - Tournament or match name
  • Site - Location or platform (e.g., “lichess.org”)
  • Date - Game date in YYYY.MM.DD format
  • Round - Round number in tournament
  • White/Black - Player names
  • WhiteElo/BlackElo - Player ratings
  • Result - Game outcome (1-0, 0-1, 1/2-1/2, *)
  • TimeControl - Time control string
  • ECO - Opening classification code

​
Batch import optimization

For optimal performance when importing large PGN files:
  • Disable indexes during import - Indexes are created after import completes
  • Journal mode OFF - Maximizes write speed (set automatically)
  • Foreign key checks disabled - Enabled only after completion
  • Transaction batching - All games imported in a single transaction
  • Parallel parsing - Uses multiple CPU cores for decompression
A compressed PGN file with 1 million games typically takes 5-15 minutes to import on modern hardware.

​
Importing from Lichess.org

En Croissant can import your games directly from Lichess using OAuth authentication.
1

Initiate OAuth login

In the import dialog, select Lichess as the source and enter your Lichess username.
2

Authorize the application

Your browser will open to Lichess.org. Click “Authorize” to grant En Croissant access to your games.Scopes requested:
  • preference:read - Read your account preferences
3

Automatic token handling

En Croissant receives an OAuth token and uses it to download your games. The token is stored securely and reused for future imports.
4

Download your games

Your complete game history is downloaded in PGN format and imported into a new database.

​
OAuth implementation details

The OAuth flow is implemented in src-tauri/src/oauth.rs: 
// OAuth configuration for Lichess
client_id: "org.encroissant.app"
auth_url: "https://lichess.org/oauth"
token_url: "https://lichess.org/api/token"

// Uses PKCE (Proof Key for Code Exchange) for security
// Local callback server on random port (127.0.0.1:RANDOM)
En Croissant uses a local callback server that automatically closes after authentication. No data is sent to external servers.

​
Importing from Chess.com

Chess.com game import follows a similar OAuth flow to Lichess:
  1. Enter your Chess.com username
  2. Authorize En Croissant in your browser
  3. Games are downloaded and imported automatically

​
Supported game types

Both Lichess and Chess.com imports include:
  • Rapid games
  • Blitz games
  • Bullet games
  • Classical games
  • Correspondence games
  • Tournament games
  • Casual games (optional)

​
Incremental updates

When re-importing from a source, En Croissant can skip games that already exist in your database:
  • Timestamp checking - Only imports games newer than the latest game in your database
  • Duplicate detection - Automatically skips identical games
  • Merge strategy - New games are appended to existing database
This is controlled by the optional timestamp parameter in convert_pgn(): 
// Skip games with timestamp before this Unix timestamp
timestamp?: number

​
Pre-built databases

En Croissant offers curated databases for download:

Master games database

High-level games from professional tournaments with full annotations.

Opening repertoire database

Specialized collections focused on specific openings and variations.

Historical games

Classic games from chess history, including world championship matches.

Computer games

High-quality engine vs. engine games for studying precise play.
Pre-built databases are downloaded in optimized .db3 format with indexes pre-generated for instant use.

​
Import error handling

En Croissant handles various PGN errors gracefully:
Skipped with warning:
  • Invalid FEN positions
  • Illegal moves in game notation
  • Missing required headers
  • Corrupted move sequences
Games with errors are skipped, and import continues.

​
Performance tips

1

Use compressed formats

.pgn.zst files are 5-10x smaller than uncompressed PGN and import just as fast due to parallel decompression.
2

Batch related imports

Import all related PGN files into a single database rather than creating multiple small databases.
3

Monitor disk space

The database will be approximately 30-40% the size of the uncompressed PGN file. Ensure adequate free space.
4

Use SSD storage

Database creation is I/O intensive. SSDs provide 10x faster import times compared to HDDs.

​
Import statistics

After import completes, the database Info table tracks:
  • GameCount - Total games in database
  • PlayerCount - Unique players
  • EventCount - Unique tournaments/events
  • SiteCount - Unique locations/platforms
  • Version - Database schema version
These counts are displayed in the database card and used for statistics and filtering.

Build docs developers (and LLMs) love

Get started for freeTalk to us