v0.4.1 pre-release

Bug Fixes

• Similar to commit 9616984, fixed another bug in ParseUint32() whereby fields containing a number that exceeds 20 digits caused a panic.

Misc

• Added benchmark test for ParseUint32().

v0.4.0 pre-release

What's New

• Added Field.Bytes() method, which returns the field's backing []byte slice.
• Exported utility function ParseUint32([]byte) uint32.

Bug Fixes

• Fixed bug in Field.Uint32() whereby fields containing a number greater than 9,999,999,999 invoke a panic. The intended outcome is that an error is returned with a message indicating that this number overflows uint32 datatype.

v0.3.0 pre-release

What's New

• Added IsEmpty() to the Field class. This method returns true only if the field has no content.

v0.2.0 pre-release

What's Changed

Breaking change: The callback function for Read() and ReadFile() now returns an error, which gives the client the opportunity to halt reading the input (by returning a non-nil error).

v0.1.1 pre-release

Bug Fix

• Fixed bug in Field.Uint32() whereby a field containing an integer value that overflows a uint32 datatype did not result in an error, but instead just returned an incorrect value.

v0.1.0 pre-release

This is the first versioned release of hastycsv. Until a stable v1.0.0 version is reached, there may be more API-breaking changes in upcoming pre-releases.

Breaking Changes

• Reader.Row field is now private (i.e. changed to Reader.row).
• First record is no longer skipped by default.
• Renamed Delimiter to Comma on Reader struct in order to adopt same naming convention as Golang's csv reader.

Other Changes

• Added Go module support.
• Updated travis-ci config to build against Go 1.10, 1.11, and 1.12.
• Reader.ReadFile() now creates a 32K buffer (vs. relying on 4K default).