Usage
Command-line reference for logfold 0.9.4.
logfold [-n] [-m N] [-c] [FILE ...]With no FILE, logfold reads standard input. With several files they are read one after another as a single stream, so a run can continue across a file boundary.
Options
| Option | Meaning |
|---|---|
-n, --normalize | Mask timestamps, hex values and numbers before comparing lines. The line that is printed is the original, unmasked one. |
-m N, --min N | Fold only runs of at least N lines. Shorter runs are printed as they are. Default 2; values below 2 are rejected. |
-c, --count-first | Print the count before the line, right-aligned, instead of (xN) after it. Easier to sort. |
-V, --version | Print the version and exit. |
-h, --help | Print a short help text and exit. |
What counts as the same line
Lines are compared exactly, except for the trailing newline. Without -n a single changed digit starts a new run. With -n the following parts are replaced before comparing, in this order:
| Pattern | Replaced by |
|---|---|
| ISO 8601 date and time, with optional fraction and zone | <ts> |
Clock time such as 10:00:01.250 | <time> |
Hexadecimal numbers with a 0x prefix | <hex> |
| Whole decimal numbers | <n> |
Only consecutive lines fold. If a line comes back after something else, it starts a new run; logfold never reorders or drops lines.
Exit status
| Code | Meaning |
|---|---|
| 0 | Input was read and written. A closed pipe, as with | head, is not an error. |
| 1 | A file could not be read. The message goes to standard error. |
| 2 | Invalid command line. |
Encoding
Input is read as UTF-8. Bytes that are not valid UTF-8 are replaced rather than rejected, because log files are rarely clean.