NAME
elpi - embeddable lambda-Prolog interpreter
SYNOPSIS
elpi [OPTION].. [FILE].. [-- ARGS..]
DESCRIPTION
Main options:
-test runs the query "main"
- -exec pred
-
runs the query "pred ARGS"
- -D var
-
Define variable (conditional compilation)
-document-builtins Print documentation for built-in predicates
-no-tc don't typecheck the program
- -delay-problems-outside-pattern-fragment (deprecated, for Teyjus
-
compatibility)
-version prints the version of Elpi
Parsing options:
- -I PATH
-
search for accumulated files in PATH
Tracing options:
- -trace-at FNAME START STOP
-
print trace between call START
-
and STOP of function FNAME (FNAME can be omitted, default is run)
-trace-on KIND FILE enable trace printing.
-
KIND is tty or json (default is tty). FILE is stdout or stderr (default) or host:port or /path or ./path
- -trace-skip REX
-
ignore trace items matching REX
- -trace-only REX
-
trace only items matching REX
- -trace-only-pred REX
-
trace only when the current predicate matches REX
- -trace-tty-maxbox NUM
-
Format max_boxes set to NUM
- -trace-tty-maxcols NUM
-
Format margin set to NUM
- -stats-on
-
Collect statistics
- -perf-on
-
Disable trace output, but keep perf
Tracing options can be used to debug your programs and the Elpi interpreter.
Tracing points for the user are prefixed with 'user:' while the ones for the Elpi developer with 'dev:'.
A sensible set of options to debug your programs is: -trace-on -trace-at 1 9999 -trace-only '\(run\|select\|user:\)'
Debug options (for debugging Elpi, not your program):
-print-accumulated-files prints files loaded via accumulate
-print-ast prints files as parsed, then exit
-print prints files after most compilation passes, then exit
-print-passes prints intermediate data during compilation, then exit
-print-units prints compilation units data, then exit