Example in Readme has < instead of <

for (int i=0; i<csvReader.FieldsCount; i++) {

for (int i=0; i<csvReader.FieldsCount; i++) {

File contains 3 columns:
AddressId, JsonData, UpdatedAt
1,"{"address": {"zip": null, "city": null, "state": null, "country": "USA", "street1": null, "street2": null, "street3": null}}","2021-03-31T23:36:04"
2,"{"guardian": {"name": "Abhishek", "phone": true, "address": {"zip": "6204", "city": "York", "state": "PA", "country": "USA", "street1": "62115B Greenwood Rd", "street2": null, "street3": null}, "relationship": "Son"}}","2021-03-31T23:36:04"

I am not able to find "UpdatedAt" field at csvReader[2]. Please add json support also. thanks

Introduce 2 new properties - SkippedLines and ReadLines or similarly named that gets the point across.

• SkippedLines - returns the number of skipped lines of the file.
• ReadLines - returns the number of read lines of the file.

I'll attempt to tackle this now as I'm in need of it. Let me know if this is something you'd like to see implemented in the library!

Why there is no possibility to accept a different quote?

Something like this
var csvReader = new CsvReader(stream, delimiter, quote); (eventually also comment char....)

This lib is very fast but also incomplete with only delimiter. I hope you add this feature.

In .NET Core 2.1 and upcoming netstandard 2.1 ReadOnlySpan can be used for zero-allocation handling of string values - without need to create 'string' object from the buffer.

To support this it is enough to add ProcessValueInBuffer overload or maybe add method ReadOnlySpan<char> GetSpanValue(int idx).

It seems the library is capable of reading not fully correct CSV files. if you can expand on the differences, as I'm guessing your model is specified in RFC4180?

With that said, does the library itself produce CSV files conforming to the above RFC standard?

In some cases it might be useful to read CSV stream as System.Data IDataReader.
This unlocks simple way to read CSV to DataTable or to list of POCO-models - say, this is possible with DataReaderResult class from NReco.Data library.

In the initial implementation, lets expose all CSV line fields as 'string' FieldType, and add option to skip first (headers) line.

Stack trace:
at NReco.Csv.CsvReader.GetOrAddField(Int32 startIdx)
at NReco.Csv.CsvReader.Read()

Using version 1.0.1

The Read() method calls itself when an empty line is encountered. If there are sufficient empty lines in a file, this has the possibility of causing a stack overflow exception, killing the entire process. Could you please make it so that the Read() method works without recursion?

Currently CsvWriter always wraps with double quotes values that start and/or end with space. This is allowed by RFC4180, however double quotes are not required in that case.

Let's add an option to control this behaviour.