csplit

Additional Notes

csplit splits a file into multiple output files based on line numbers, repeating counts, or context patterns such as a heading line. It is ideal for breaking one large document into chapters, sections, or records.

Unlike split, which cuts by fixed size, csplit cuts by content you describe with a pattern or line position.

Syntax

csplit [options] FILE PATTERN...

Parameters

• options: Flags that change naming and behavior.
• FILE: The file to split (use - for standard input).
• PATTERN: How to find each split point: a line number, a count, or a regular expression.

Pattern Forms

• N: Split before line number N.
• /REGEX/: Split before each line matching the regular expression.
• /REGEX/[OFFSET]: Apply an offset (such as +1 or -1) to the match line.
• {N}: Repeat the previous pattern N times, or {*} to repeat until the end.

Common Options

• -f PREFIX: Set the output filename prefix (default xx).
• -b FORMAT: Set the suffix format, such as %02d.
• -n N: Use N digists in the suffix.
• -k, --keep-files: Keep output files even on errors.
• -z: Use NUL instead of newline as the line delimiter.
• -s: Suppress counts that are normally printed.

Examples

csplit -z manual.txt /Chapter/ "{*}"

Split manual.txt before every line containing Chapter, repeating to the end.

csplit data.txt 20

Split before line 20, producing a first file with lines 1–19.

csplit -f part- -b "%03d.txt" log.txt /ERROR/ "{*}"

Split log.txt before each ERROR line into part-000.txt, part-001.txt, and so on.

csplit -z report.txt /SECTION/ +1

Split before each SECTION line, then shift the cut one line down.

csplit big.txt 100 {10}

Split before line 100, then repeat that every 100 lines 10 more times.

Practical Notes

• Patterns are tried in order; each splits the file at a point.
• {*} keeps splitting until the input ends, which is the usual case for documents.
• Use -f and -b to control the output filenames.
• csplit is content-aware, unlike split, which only cuts by size.