A few months back, CSV import was improved by automatically inferring percent values in the data. That is whenever xlsgen would read a string such as "35%", xlsgen would actually store 0.35 in the current cell and associate a percent number format to it so it shows 35% in Excel and is a number.

With xlsgen 4.5, this automatic inference of percent values is extended to all other 3 input file types which are JSON, HTML and XML.

xlsgen 4.5 also introduces automatic inference for currencies. That is, if there is a string such as $340 or 340 €, xlsgen actually stores the corresponding number and asociates a currency number format to it. This works with all input file types, i.e. CSV, JSON, HTML and XML.

Overall, automatic inference differentiates numbers and strings, and disambiguates percent values and currency values. In other words, without any prior knowledge is the data being imported, xlsgen tries to understand what it imports and keep it usable in Excel.

Last but not least, CSV import now supports both files and memory buffers whereas before only files were accepted.