You should see this message only if your browser is not handling the DjVu format, but sometimes it is superflous:

Please install a DjVu plugin for your browser (Firefox is the recommended one) or use the djview for Poliqarp client program.

This is another retrodigitization of the second edition of Linde's dictionary in a form of a two-layer (texts and scans) DjVu corpus and the two-level regular expressions Poliqarp search engine. It is supplemented by a preliminary index which allows to find entries using *a fronte* and *a tergo* lists.

This version of the corpus is available since September 2016. It contains about 7 millions tokens representing about 5000 pages of the dictionary.

The recommended form of acknowledging in the scientific publications the use of this retrodigitization and the search engine is to reference the paper Efficient search in hidden text of large DjVu documents or Skanowane teksty jako korpusy. There are also some relevant presentations, such as An incremental approach to retrodigitization and Przyrostowa metoda dygitalizacji słowników.

The recommended form of acknowledging the use of index is referencing the presentation Elektroniczny indeks do słownika Lindego.

The recommended way to use the search engine and the index is the djview4poliqarp remote client available on all basic platforms, preferably together with the djview4 viewer which supports outlines included in the dictionary files.

Scans were made with Fujitsu fi-6130 scanner and scanhelper software by Joanna A. Bilińska, who also improved the scans using Scan Tailor. They were converted to the DjVu format by Janusz S. Bień with the didjvu program. OCR was prepared by Janusz S. Bień with ocrodjvu using the Tesseract engine. Various adjustments of the OCR results has been done by Janusz S. Bień and Michał Rudolf. The corpus was created and published online by Michał Rudolf.

Search can be limited to a specific volume using metadata, e.g. meta volume=6 .

The meta clause can be also used to limit the search using other metadata fields, but it's not practical.

Every token has the following attributes: