CsvObject

📖 Description

CsvObject package allows to dump structured data to CSV file string, and vise versa - parse CSV data from string to slice of structures.

In order to minimize dependencies, this package does not do direct write to file (filesystem). It only deals with (CSV) strings.

Limitations:

CSV data can be represented as an arbitrary structure, where fields can represent other structures (with unlimited nesting), but every "row" should be "flat", or in other words every last field in nested structure chain should be scalar and have standard type:

• string
• int8
• int
• int32
• int64
• float32
• float64
• bool
• *big.Int (this type added 11.12.204 to expand the application of the module to crypto projects)

📦 Installation

To make CsvObject available in your project, you can run the following command. Make sure to run this command inside your project, when you're using go modules 😉

go get github.com/anxp/csvobject

👀 Example

For example, if we have in our code some structured data like:

type Student struct {
    StudentID      int64
    FirstName      string
    LastName       string
    EntryYear      int
    Speciality     string
    AdditionalData AdditionalData
}

type AdditionalData struct {
    IsCovidVaccinated bool
    HasDriveLicense   bool
    IsPaidStudent     bool
    Stipend           *big.Int // Yeah, our students are really rich
}

And we have series of such data (slice):

And we want to have the ability to easily store such data on disk and then read back to structured format, csvobject library can be used for that purpose.

Data will be converted to CSV format, with a specially formatted header, where original types of data are preserved:

StudentID|string	FirstName|string	LastName|string	EntryYear|int	Speciality|string	AdditionalData.IsCovidVaccinated|bool	AdditionalData.HasDriveLicense|bool	AdditionalData.IsPaidStudent|bool	AdditionalData.Stipend|*big.Int
1	John	Smith	2010	Physics	true	true	false	1000000000000000000
2	Max	Mustermann	2011	Electronics	false	true	true	2000000000000000000
3	John	Davis	2012	Mathematics	true	false	true	3000000000000000000

When we read such data back from disk and want to parse to original structured format, column headers are used to get original structure field names and their type, what allows to completely reconstruct objects in runtime.