Chinese Encyclopedias (Leishu) are reference books consisting of extracts from sources extant at the time of compilation which were put under different categories. By the time of the Tang and Song dynasties, the compilation of Leishu became fashionable, and well-known examples of them survive in the present day, for instance, Beitong Shuchao, Yiwen Leiju, Taiping Yulan, Chuxueji. The CHANT Database contains the entire corpus of traditional Chinese texts down to the sixth century, but some missing titles/passages are still to be found in the form of citations in various Leishu. The Research Grants Council of the Hong Kong Government funded this project in 2000. The objectives of this study are: i) to build up a computerized database of extant Leishu. ii) to make a comparative study of citations appearing in Leishu and the CHANT Database. Regarding the comparative study, there are three tasks we set ourselves. First, by comparing the earliest citations with extant versions, we hope to be able to correct textual mistakes wherever they occur. Second, we would make an analysis of the variant readings through textual comparison. In this way, we can get closer to the texts as they were in their original form, which may differ significantly from extant versions. Third, we are able to identify all texts which are not found in extant editions so that such missing parts can be culled and re-constituted for purposes of textual studies.

1. The database contains all major Leishu from the Weijin period down to the Qing Dynasty, including Qunshuzhiyao, Taiping Yulan, and Cefu Yuangui. It is estimated that the database will exceed 60 million words.
2. All inputted texts were based on the best available editions with the least subsequent tampering. Textual comparisons and punctuation were carried out for all texts. Variant readings were given in footnotes for reference.
3. It provides a comprehensive database of Chinese Leishu which helps scholars to focus on high-level studies of textual comparison.