Direct xtrace output elsewhere than stderr in zsh

debuggingzsh

In zsh, is it possible to direct the trace output from set -x (set -o xtrace) to a file descriptor other than stderr?

I'm looking for an equivalent of $BASH_XTRACEFD or a way to emulate the same behaviour.

Best Answer

As of zsh 5.7, the answer is no. The trace output always goes to stderr.

Source: reading the source. The trace output is written to the file xtrerr, which looks promising, but the only assignments to xtrerr are to stderr, to a copy of it, or to NULL.

It should be possible to write a dynamically loadable module that sets xtrerr, but writing a module outside the zsh source tree is not easy.

A possible workaround is to emulate xtrace with a DEBUG trap. This provides the same basic information in most cases, but I'm sure there are plenty of corner cases where xtrace would be hard or impossible to emulate perfectly. One difference is that the inheritance of the xtrace option and the inheritance of traps follow different rules in some circumstances related to functions, subshells, emulate, etc. Proof-of-concept:

trap 'print -r -- "+$0:$LINENO>$ZSH_DEBUG_CMD" >>trace_file' DEBUG

Or maybe a bit more sophisticated (untested):

zmodload zsh/system
sysopen -a -o create -u xtrace_fd trace_file
trap 'syswrite -o $xtrace_fd "+$0:$LINENO>$ZSH_DEBUG_CMD"' DEBUG

Related Solutions

Zsh completion – show more options than just the alias

@Ilua's answer did not work, but it did give me some ideas of what to search for, and I solved the problem.

The style I needed was regular. From man zshcompsys:

regular
   This style is used by the _expand_alias completer and bindable command. 
   If set to ‘true’ (the default), regular aliases will be expanded but only in command position. 
   If it is set to ‘false’, regular aliases will never be expanded.
   If it is set to ‘always’, regular aliases will be expanded even if not in command position.

I used zstyle ':completion:*' regular 'false', and it works perfectly.

How to capture the xtrace output (only) in a file

Might not be the solution you are looking for, but having run into a similar scenario in the past, my go-to is to write a python parser to take the trace file, and output only the data I care about into a resulting file. As long as the patterns of the "garbage text" that you don't care about are predictable/differentiable from the xtrace output, the parsing should be extremely easy to do.

Example - grabbing a trace of all IOs across all submission queues for 20 min test consisting of reads and writes results in a 40GB-70GB file, but after parsing the data I care about out of it, my result file is only ~1-3GB in size, and actually usable for analysis and visualization. Granted, with such large file sizes, it takes a server-class system ~15-30 min to process through it all, but doing the same process (separate scenario, same idea - multiple logs each only about 24kb) - I can get all of the data from all logs into a single consumable file in under 2 seconds.

All that to say, it might be worth it to add a post-processing step that might add a few seconds/minutes to the overall capture, rather than waiting for a one-stop-shop.

Best of luck!

Best Answer

Related Solutions

Zsh completion – show more options than just the alias

How to capture the xtrace output (only) in a file

Related Question