Tag: XLS

POI provides all-around functionalities for parsing and generating xls file. The problem is that it has a too low-level API that makes users write a lot of code from scratch even for the simple read/write actions. And to be all-around means that every detail needs to be taken care of, leading to complicated, heavy, hard-coded details handling process. Another issue is that POI invests too much in configuring too many appearance properties, such as font, color, thickness and alignment, yet its data processing capabilities are not as professional as expected when data parsing and generation should have been the focus

An indirect way is through a Java embedded database. We first load xls to it and then execute SQL. The advantage is that it makes full use of SQL’s computing ability, and disadvantages include extremely low real-time capability and a complicated loading process that involves checking if the same table name already exists, deleting and creating a table, parsing the file, importing data, creating the index, and so on. Moreover, data loading would fail if the source file has an irregular format. The direct way is to invoke the xls-based JDBC driver in Java, such as XLSJDBC, xlSQL and CDATA

POI does not have further query and computing capabilities after it parses xls. Programmers need to import data and hardcode a specific computing task. Embedded databases are exactly the opposite with their query and computing abilities and the inability to parse. Besides, they boast time-consuming data loading process and heavy frameworks. A fast and lightweight solution is to execute SQL directly on an xls file. But there are only a few open-source, rudimentary libraries for achieving the solution. Their practical uses are rather limited. DataFrame-style open-source libraries are more mature, having no need to interpret SQL and at the same