Module capture 

The job a runner took and the job’s log, read from inside its container while it runs. A small script, run with docker exec as root, waits for the runner’s _diag/Worker_*.log, prints the job’s ID (check_run_id) and run ID from the job message there, then follows the job’s log pages in _diag/pages, which the runner writes in 4 KiB blocks and deletes once uploaded: the script hard-links each page so none is lost, and prints what it has not printed before. These pages are what GitHub shows, with secrets already masked by the runner.

The script’s output: @@job <id> <run id>, then @@page <page> <offset> <length> followed by exactly that many bytes, and @@end <page> once a page is complete. Parser turns it into whole lines and the position to resume from, so a restarted daemon continues where the stored log ends.

Structs

Parser

Turns the script’s output, in chunks of any size, into Captured.

Enums

Captured

What the script reported.

Constants

MAX_LINE

The longest line kept: the rest of a longer one is dropped as it arrives, so one line never holds more memory.

SCRIPT

The script; bash -c SCRIPT capture <page> <offset> resumes at that page and byte offset.

Functions

command

The docker exec command line that captures a job from position on.