This article covers details about the file formats supported by the Peak platform. It also gives users an insight into the preferred file naming conventions and some guidelines to design data extracts.
Contents
Supported file types and formatting
Peak accepts files in the following formats:
- CSV
- TXT
- JSON
- XML
Formatting
- Use comma(,) or tab or pipe(|) as a column delimiter for CSV / TXT files.
- All fields must be enclosed by double-inverted commas ("").
- Replace delimiters from data where possible.
- The first row must always be the header.
- The file should be compressed (GZIP with .gz or .gzip extension) when possible before being transferred to reduce file size.
- Peak supports plain text using a character set such as ASCII, or Unicode character sets (UTF-8).
- Use CRLF for the new line.
- Remove any LF, or CR characters from the data other than the new line.
File naming convention
- The file name must contain at least one alphanumeric character.
- The file name should follow either of the formats stated below.
List of Valid date formats
Guidelines for designing your data extracts
- Preferably all rows in the data file should be uniquely identifiable using a single or composite unique identifier. This is mandatory if data files are not insert only.
- Consistent format for date, date time, and timestamp across data attributes and files.
- Remove trailing and lead space for text.
- After the initial load, all transaction data should be sent as incremental files (new and updated records) with a unique key shared with Peak.
- Peak’s platform doesn’t handle record deletion by default.