The Chess PDF Browser is an application programmed in Java that allows you to browse chess books in PDF format.

It also enables working with game files in .pgn format.

View a chessboard where you can browse through different games.

Allows you to open PDF chess books and extract games to save in .pgn format.

Allows you to edit the variants of the games stored in memory, whether they are read from .pgn, extracted from a PDF, or created directly by moving the pieces on the board.

There is a comprehensive manual that details how the application works.

Several bugs have been fixed, and new features have been added in the latest version of the application:

• New experimental game extractor.
• Mark moves as novelties. (Novelty).
• Language support in Russian has been added.

There is a comprehensive guide that explains how the application works.

The user guides have been updated with the new version of the application.

There is a comprehensive guide that explains how the application works.

With the latest version of the application, numerous new features have been added:

• The experimental line item extractor has been replaced by a new extractor that now takes into account brackets and square brackets.
• The system now has the capability to extract game data, including player names and ELO rankings.
• The ability to view the moves in a game in algebraic notation.
• An OCR has been implemented to recognize board positions and add them to games that do not start from the standard starting position.
• Now you can play games against another person or an engine or play two engines against each other.


• Enhanced connectivity to UCI engines:
  • Modify the engine settings, for example, by decreasing the level to play against a lower level than the maximum.
  • Analysis of positions.
  • Full game analysis.
  • You can use an engine as a player in a game.
• Dark mode option

There is a comprehensive guide that explains how the application works.

The application's new version can now extract game moves from PDFs using figurine algebraic notation.

What's new in this version:

• The software allows for extracting games in figurine algebraic notation, notating the movements of the pieces.
• Enhanced board position recognition using optical character recognition (OCR).
• Some bugs have been fixed.

There is a comprehensive guide that explains how the application works.

You can see a video of the new feature Demo video

The new version includes improvements in position detection (image -> FEN)

The new version includes a new binary that allows you to create PDFs from Pgns.

In this link you can find a demo video of the new functionality

The new version adds the option to work with scanned PDFs

When I implemented the code to handle PDFs, the option was to use pdfbox to work with the PDF details.

But I didn't like the idea of ​​the business logic explicitly using that dependency, so I decided to create a PDF interface that offered the functionality, and program an implementation that made use of the pdfbox library

The drawback was that for pure scanned PDFs (without anyone having added the text), the library couldn't return the text associated with the scanned pages.

Therefore, the application couldn't offer the functionality to extract entries from scanned PDFs.

This new version now allows you to try to extract entries from scanned PDFs.

This has been made possible by programming a new implementation of the PDFs interface, which attempts to offer functionality reasonably similar to that of "normal" PDFs with the existing pdfbox implementation.

This new library internally uses pdfbox to obtain the images of the scanned pages, and, for each page, invokes an external OCR located in the cloud.

Since working this way in real time would be extremely slow (invoking the OCR for each page takes between 3 and 10 seconds), the library invokes the OCR only once for each page.

Once the OCR has been invoked, the result is saved to the file system, and subsequent times the application needs it, it is retrieved from there.

The result is that the new library is equivalent and interchangeable with the pure pdfbox library, offering equivalent functionality for "normal" PDFs and scanned PDFs.

Although the drawback is that the text recognition offered by the OCR is not as perfect as the text extracted from "normal" PDFs.

For extracting games, the application already offered a mechanism that allows you to choose to change "l"s to "1"s and "S"s to "5"s.

For scanned PDFs, some new translations have also been added (changing "£"s to "f") and these translations are fixed, without user intervention.

These character transformations greatly improve the extraction of movements.

The result of extracting games is not perfect, but it is possible to extract some sequences of moves, and by editing the games and with a little patience, the sectioned games can be reassembled.

At this link you can find a demo video of the new functionality