Skip to content

CLI

Jump to bottom
Ivan Kochurkin edited this page Mar 1, 2019 · 4 revisions

PT.PM CLI

PT.PM can be used in CLI (Command Line Interface) mode. In this case, the following parameters can be specified:

-f, -files

The path to the source code file or directory. PT.PM determines whether the path is a file or directory. Also, .zip archives and archives by link are supported (e.g. https://github.com/PositiveTechnologies/PT.PM/archive/dev.zip).

-l, --languages

Languages to be processed. CSharp, Java, Php, PlSql, TSql, Aspx, PlSql, TSql, MySql, and JavaScript are supported (case-insensitive). Commas (',') used as a separating character. If the parameter is not set, the language is determined automatically depending on the extension. If there is an ambiguity (the .sql extension can be both defined as T-SQL, PL/SQL and MySql), the algorithm try to determine language by file content (a file is parsed and the language with less parsing errors is chosen). If the language could not be determined for some reason after parsing a file, this file is ignored. If JavaScript is specified, JavaScript insertions will be processed inside PHP code.

--patterns

If the parameter value ends with .json, patterns are loaded from the passed file name. Also, DSL format is supported (files ends with .pattern). Otherwise, patterns are loaded from the string. The file and string patterns have the same format.

Each pattern is represented as follows:

  • Key — pattern ID (string or long)
  • Name — [optionally] pattern name
  • Languages — [optionally] target pattern languages (CSharp, Java, PHP, PLSQL, TSQL, MYSQL, JavaScript a combination thereof, for example, CSharp, Java). All languages are supported by default.
  • DataFormat — [optionally] format for describing a pattern (JSON, DSL). Default format: DSL. DSL description is located in document: DSL.
  • Value—pattern value in DataFormat.
  • Description — [optionally] pattern description.

An example of Hardcoded Password Pattern:

{
  "Key": "96",
  "Languages": "CSharp, Java, PHP, PLSQL, TSQL",
  "DataFormat": "Dsl",
  "Value": "<[(?i)password]> = <[ \"\\w*\" || null ]>"
}

JSON backslash (\) and double-quote mark (") characters should be escaped with a backslash \. Actually, Value looks more simple: <[(?i)password(?-i)]> = <[ "\w*" || null ]>.

If patterns are not set, the default built-in patterns are used for matching. It's equivalent to default option. If you want to run scan without patterns, use no option.

--pattern-ids

Patterns to be used. Commas are used as separators. By default all patterns are used.

-t, --threads

The maximum number of threads in use. 0 — at the discretion of the OS (it's default).

-s, --stage

Phase at which the parsing is stopped Default value: match.

  • file — file reading
  • language — detection of language. Not used if no ambiguity.
  • tokens tokenize text (only for ANTLR-based parsers).
  • parsetree — parsing.
  • ust — converting to UST.
  • match — matching with patterns
  • patterns — mode for checking the patterns

--not-fold-consts

Do not propagate and fold constants (integers and strings). If not defined, the expression 2 + 2 * 2 will be converted to 6.

-m, --memory

Approximate memory consumption (MB) of ANTLR parsers. Default value: 3000 MB. ANTLR сaches information on parsing; thus, memory consumption may increase significantly up to several gigabytes. Therefore, it's necessary to clear the cache from time to time.

--max-stack-size

The maximum stack size for the thread (in bytes). Used if the StackOverflow exception occurred during ANTLR parsing. You can first start processing without this parameter; then, if an exception occurred and the process is abruptly interrupted, you can increase this parameter (For example, specify the size of int.MaxValue / 8).

--timeout

Max spent time per file in seconds. By default: endless.

--logs-dir

The path to the directory to which the info, errors, and matching_result logs will be written while the engine is running. By default, the parameter is set to %LOCALAPPDATA%\PT.PM directory.

--temp-dir

The path to the directory for storing temp files.

--no-log

Do not log info and error messages.

--log-errors

Is error logging required? Default value: false.

--version

Shows the version of engine.

-d, --dump

Specify what intermediate structures should be dumped. Values from -s, --stage parameter are correct.

-r, --render

Stages to be rendered (with help of Graphviz library).

--serialize-format

Specify the format of UST serialization: json or msgpack.

Clone this wiki locally