#1877: CSVWriter.java

Imports

java.io.PrintWriter / java.io.Writer — Text output
java.text.DateFormat / SimpleDateFormat — Date formatting for CSV date columns
java.util.Date — Date value type
java.util.TimeZone — UTC timezone for date formatting

Configuration

Setting	Default	Setter
CSV separator	`';'` (semicolon — European convention)	`setCsvSeparator(char)`
Line ending	`"\n"` (Unix newline)	`setCr(String)`
Date format	`"yyyy-MM-dd HH:mm:ss.SSS"`	`setDateFormat(DateFormat)`
Date timezone	UTC	Hardcoded in constructor

Fluent Builder API

All write() methods return this (the CSVWriter instance), enabling method chaining:

csvWriter.write("Name").write("Age").write(42L).writeEndOfLine();

Type-Specific Write Overloads

Method	Formatting	Quoting
`write(String s)`	Raw string with `"` escaped to `""` (doubling)	Always wrapped in `"..."` — follows RFC 4180 quoting rules
`write(long value)`	Direct numeric output via `PrintWriter.print()`	None — numbers are unquoted
`write(Date value)`	Formatted via `dateFormat.format()` in UTC	Wrapped in quotes — dates may contain spaces or special chars
`write(Object value)`	Fallback: `String.valueOf(value)`	None — generic objects are unquoted (risky for objects with commas/quotes)

Quoting Algorithm (String Writing)

The write(String) method implements proper CSV escaping per RFC 4180:

Output the separator if not the first entry in the row (managed by writeSeparator() with the firstEntry flag)
Output an opening double-quote (")
Iterate through each character of the string
If character is a double-quote ("), output it twice ("") — the standard CSV escape sequence
Output all other characters as-is
Output a closing double-quote (")

Line Writing

writeLine(Object[] values) — Writes all values in the array, then calls writeEndOfLine()
writeEndOfLine() — Outputs the configured line ending (\n by default) and resets firstEntry = true

Separator Logic (writeSeparator)

The firstEntry boolean flag tracks whether we're at the start of a line:

If firstEntry == true: set to false (first cell — no separator needed)
If firstEntry == false: output the separator character before the cell value

This is a classic state-tracking pattern for delimited output.

Integration with CSVParser

CSVWriter defines DEFAULT_CSV_SEPARATOR_CHAR = ';' as a public constant, which CSVParser references to ensure consistent default separator between reader and writer.

Comparison with Modern CSV Libraries

This implementation is intentionally minimal compared to Apache Commons CSV or OpenCSV:

No header mapping (that's handled by CSVParser.parseHeadCols())
No automatic type conversion (callers format values before writing)
No streaming/large-file optimizations (writes directly to a Writer)
No support for different quoting strategies (all strings are quoted; numbers never)

The simplicity reflects its origin as a purpose-built tool for ProjectForge data export, not a general-purpose CSV library.

The write(Object) method does NOT quote the output — it calls String.valueOf(value) and writes directly. If the object's toString() returns a string containing commas, semicolons, or newlines, the resulting CSV will be malformed. This is acceptable because callers are expected to use the type-specific overloads.

Date output uses UTC timezone — dates are normalized to UTC before formatting. This is a deliberate choice for data interchange: it avoids timezone ambiguity in exports. The format "yyyy-MM-dd HH:mm:ss.SSS" is ISO 8601-like but uses a space separator instead of 'T'.

#1877: `CSVWriter.java`

Architecture

Imports

Configuration

Fluent Builder API

Type-Specific Write Overloads

Quoting Algorithm (String Writing)

Line Writing

Separator Logic (writeSeparator)

Integration with CSVParser

Comparison with Modern CSV Libraries

Git History