A New Breed of CSV Parser (+ a lot more)
TFieldedText takes a new approach to reading CSV files (and other text files containing tables of values). It treats them like a database!
The approach is quite straight forward. You provide a small XML file which defines the Meta structure of the information in the text file. TFieldedText uses that to read from and write to the text file as if it were a database. Typically you would use a Fielded Text Editor to create these Meta XML files.
The format (schema) of this Meta XML file is specified by a standard defined at http://www.fieldedtext.org. In addition to using these Meta XML files, TFieldedText can also be used to create them. These XML files can then also be used by any other software component/library which support this standard. Refer to the http://www.fieldedtext.org website for more information about this standard.
TFieldedText can be used to parse or generate nearly any type of CSV and Fielded Text file. This includes files which have:
- fields are separated by delimiters (eg CSV files).
- fields which are fixed length.
- fields that are quoted.
- fields that have new lines embedded in them.
- headings.
- substitutions
- lines with different sets of fields based on key fields.
- much more ...
A more comprehensive list of features can be viewed here.
TFieldedText is a .NET 2 component (written in Delphi.Net) and can be used with C#, Delphi.Net and other .NET languages. (A .NET 1.1 version is also available). Examples on how to use TFieldedText can be viewed here.
The code for this component has been placed in Public Domain - so you are free to use it however you like. The source code can be accessed from the SourceForge repository.
If you would like to join this project and help improve this component, please use the SourceForge project page to contact me.