reference Rule options

generic_engine

Value
spacegrep or aliengrep
Default
spacegrep
See also
languages, generic_multiline

A rule with languages: [generic] matches plain text, with one of two engines. spacegrep, the default, splits text into words and brackets, and an ellipsis ... spans at most 10 line breaks. aliengrep lets ... span any number of lines, and can be configured further.

Each engine reads its own options and ignores those of the other engine:

EngineOptions
spacegrepgeneric_ellipsis_max_span, generic_comment_style
aliengrepgeneric_multiline, generic_caseless, generic_braces, generic_extra_braces, generic_extra_word_characters

The option has no effect on rules for other languages.

Examples

A block longer than 10 lines

spacegrep.yaml
rules:  - id: block    pattern: begin ... end    message: a block    languages: [generic]    severity: INFO
aliengrep.yaml
rules:  - id: block    pattern: begin ... end    message: a block    languages: [generic]    severity: INFO    options:      generic_engine: aliengrep
Command and result
$ python3 -c "open('long.txt','w').write('begin\n' + 'x\n' * 11 + 'end\n')"
$ opengrep scan --config spacegrep.yaml --json long.txt 2>/dev/null | jq .
{
  "version": "X.Y.Z",
  "results": [],
  "errors": [],
  "paths": {
    "scanned": [
      "long.txt"
    ]
  },
  "interfile_languages_used": [],
  "skipped_rules": []
}
$ opengrep scan --config aliengrep.yaml --json long.txt 2>/dev/null | jq .
{
  "version": "X.Y.Z",
  "results": [
    {
      "check_id": "block",
      "path": "long.txt",
      "start": {
        "line": 1,
        "col": 1,
        "offset": 0
      },
      "end": {
        "line": 13,
        "col": 4,
        "offset": 31
      },
      "extra": {
        "metavars": {},
        "message": "a block",
        "metadata": {},
        "severity": "INFO",
        "fingerprint": "469c2010195f6b6494ec9b2c2fe48d1d88a70a644a35e637ede0e89e945d466f50cdb9d50a7436f43a10ff7ca084557e577b2d05be707e41f08512c8cb43626c_0",
        "lines": "begin\nx\nx\nx\nx\nx\nx\nx\nx\nx\nx\nx\nend",
        "is_ignored": false,
        "validation_state": "NO_VALIDATOR",
        "engine_kind": "Opengrep"
      }
    }
  ],
  "errors": [],
  "paths": {
    "scanned": [
      "long.txt"
    ]
  },
  "interfile_languages_used": [],
  "skipped_rules": []
}

The block has 12 line breaks between begin and end: too many for spacegrep, which finds nothing, but not for aliengrep, whose finding runs from line 1 to line 13.

An aliengrep option under spacegrep

caseless.yaml
rules:  - id: select-users    pattern: SELECT ... FROM users    message: query on users    languages: [generic]    severity: INFO    options:      generic_caseless: true
query.sql
select * from users
Command and result
$ opengrep scan -q --config caseless.yaml query.sql

The rule uses spacegrep, which ignores generic_caseless, so the lower-case query is not found and the command prints nothing.