languagetool-org/languagetool · error · IllegalArgumentException
JSON output format makes no sense for profiling
Error message
JSON output format makes no sense for profiling
What it means
parseOptions rejects -p/--profile together with -l/--json. Profiling mode prints a human-readable per-rule timing table, which cannot be represented in JSON, so the parser throws IllegalArgumentException for the combination.
Source
Thrown at languagetool-commandline/src/main/java/org/languagetool/commandline/CommandLineParser.java:161
}
if (options.isBitext()) {
throw new IllegalArgumentException("JSON output format is not implemented for Bitext");
}
if (options.isListUnknown()) {
throw new IllegalArgumentException("You cannot list unknown words in JSON output format");
}
} else if (args[i].equals("-a") || args[i].equals("--apply")) {
options.setApplySuggestions(true);
if (options.isTaggerOnly()) {
throw new IllegalArgumentException("You cannot apply suggestions when tagging only");
}
if (options.isJsonFormat()) {
throw new IllegalArgumentException("JSON output format makes no sense for automatic application of suggestions");
}
} else if (args[i].equals("-p") || args[i].equals("--profile")) {
options.setProfile(true);
if (options.isJsonFormat()) {
throw new IllegalArgumentException("JSON output format makes no sense for profiling");
}
if (options.isApplySuggestions()) {
throw new IllegalArgumentException("Applying suggestions makes no sense for profiling");
}
if (options.isTaggerOnly()) {
throw new IllegalArgumentException("Tagging makes no sense for profiling");
}
} else if (args[i].equals("--xmlfilter")) {
options.setXmlFiltering(true);
} else if (i == args.length - 1) {
options.setFilename(args[i]);
} else {
throw new UnknownParameterException("Unknown parameter: " + args[i]);
}
}
return options;
}
View on GitHub (pinned to 2e990059ce)
Solutions
- Remove --json when profiling; parse the tab-separated profiling table instead
- Profile without --json and redirect stdout to a file for later processing
- Do not attempt to combine the two modes; the parser will always reject it
Example fix
// before languagetool -l en -p --json file.txt // after languagetool -l en -p file.txt > profile.tsv
Defensive patterns
Strategy: validation
Validate before calling
if (hasFlag(args, "-p", "--profile") && hasFlag(args, "-l", "--json")) {
throw new IllegalArgumentException("Profiling output is a text table, not JSON");
} Type guard
static boolean hasFlag(String[] args, String... names) {
for (String a : args) for (String n : names) if (a.equals(n)) return true;
return false;
} Try / catch
try {
options = CommandLineParser.parseOptions(args);
} catch (IllegalArgumentException e) {
System.err.println(e.getMessage());
System.exit(2);
} Prevention
- Treat profiling output as tab-separated text, never JSON
- Keep benchmark flags in dedicated scripts without output-format flags
- Document in CI scripts that -p conflicts with -l
When it happens
Trigger: Command line includes both --profile (or -p) and --json (or -l), e.g. `languagetool -p --json file.txt`; checked inside the profile branch via options.isJsonFormat().
Common situations: Trying to capture rule-performance measurements as JSON for automated analysis; combining a benchmarking flag with an output-format flag in CI scripts.
Understand the failure class
Background: "mutually exclusive" flag errors: what "can't supply both nx and xx", "--raw is not compatible with -i" and "cannot be used with" mean, and how to fix them — this error's family across 29 libraries.
Related errors
- Applying suggestions makes no sense for profiling
- Tagging makes no sense for profiling
- You cannot list unknown words in JSON output format
- Profiling mode cannot be used with input from STDIN
- Unknown parameter: <arg>
AI-assisted analysis of languagetool-org/languagetool@2e990059ce (2026-09-06).
Data as JSON: /api/errors/50139027c37492b8.
Report an issue: GitHub.