Skip to main content

A code analyzer without caring the C/C++ header files. It works with Java, C/C++, JavaScript, Python, Ruby, Swift, Objective C. Metrics includes cyclomatic complexity number etc.

Project description

Lizard is an extensible Cyclomatic Complexity Analyzer for many programming languages including C/C++ (doesn’t require all the header files or Java imports). It also does copy-paste detection (code clone detection/code duplicate detection) and many other forms of static code analysis.

A list of supported languages:

  • C/C++ (works with C++14)

  • Java

  • C# (C Sharp)

  • JavaScript (With ES6 and JSX)

  • TypeScript

  • Objective-C

  • Swift

  • Python

  • Ruby

  • TTCN-3

  • PHP

  • Scala

  • GDScript

  • Golang

  • Lua

  • Rust

  • Fortran

  • Kotlin

By default lizard will search for any source code that it knows and mix all the results together. This might not be what you want. You can use the “-l” option to select language(s).

It counts

  • the nloc (lines of code without comments),

  • CCN (cyclomatic complexity number),

  • token count of functions.

  • parameter count of functions.

You can set limitation for CCN (-C), the number of parameters (-a). Functions that exceed these limitations will generate warnings. The exit code of lizard will be none-Zero if there are warnings.

This tool actually calculates how complex the code ‘looks’ rather than how complex the code really ‘is’. People will need this tool because it’s often very hard to get all the included folders and files right when they are complicated. But we don’t really need that kind of accuracy for cyclomatic complexity.

It requires python2.7 or above (early versions are not verified).

Installation can be used as a stand alone Python script, most functionalities are there. You can always use it without any installation. To acquire all the functionalities of lizard, you will need a proper install.


If you want a proper install:

[sudo] pip install lizard

Or if you’ve got the source:

[sudo] python install --install-dir=/path/to/installation/directory/


lizard [options] [PATH or FILE] [PATH] ...

Run for the code under current folder (recursively):


Exclude anything in the tests folder:

lizard mySource/ -x"./tests/*"


-h, --help            show this help message and exit
--version             show program's version number and exit
                      List the programming languages you want to analyze. if left empty, it'll
                      search for all languages it knows. `lizard -l cpp -l java`searches for
                      C++ and Java code. The available languages are: cpp, java, csharp,
                      javascript, python, objectivec, ttcn, ruby, php, swift, scala, GDScript,
                      go, lua, rust, typescript
-V, --verbose         Output in verbose mode (long function name)
-C CCN, --CCN CCN     Threshold for cyclomatic complexity number warning. The default value is
                      15. Functions with CCN bigger than it will generate warning
-f INPUT_FILE, --input_file INPUT_FILE
                      get a list of filenames from the given file
-o OUTPUT_FILE, --output_file OUTPUT_FILE
                      Output file. The output format is inferred from the file extension (e.g.
                      .html), unless it is explicitly specified (e.g. using --xml).
-L LENGTH, --length LENGTH
                      Threshold for maximum function length warning. The default value is 1000.
                      Functions length bigger than it will generate warning
                      Limit for number of parameters
-w, --warnings_only   Show warnings only, using clang/gcc's warning format for printing
--warning-msvs        Show warnings only, using Visual Studio's warning format for printing
-i NUMBER, --ignore_warnings NUMBER
                      If the number of warnings is equal or less than the number, the tool will
                      exit normally; otherwise, it will generate error. If the number is
                      negative, the tool exits normally regardless of the number of warnings.
                      Useful in makefile for legacy code.
-x EXCLUDE, --exclude EXCLUDE
                      Exclude files that match the pattern. * matches everything, ? matches any
                      single character, "./folder/*" exclude everything in the folder
                      recursively. Multiple patterns can be specified. Don't forget to add ""
                      around the pattern.
                      number of working threads. The default value is 1. Using a bigger number
                      can fully utilize the CPU and often faster.
-X, --xml             Generate XML in cppncss style instead of the tabular output. Useful to
                      generate report in Jenkins server
--csv                 Generate CSV output as a transform of the default output
-H, --html            Output HTML report
-m, --modified        Calculate modified cyclomatic complexity number , which count a
                      switch/case with multiple cases as one CCN.
                      User the extensions. The available extensions are: -Ecpre: it will ignore
                      code in the #else branch. -Ewordcount: count word frequencies and
                      generate tag cloud. -Eoutside: include the global code as one function.
                      -EIgnoreAssert: to ignore all code in assert. -ENS: count nested control
                      Sort the warning with field. The field can be nloc,
                      cyclomatic_complexity, token_count, parameter_count, etc. Or an customized field.
                      Set the limit for a field. The field can be nloc, cyclomatic_complexity,
                      token_count, parameter_count, etc. Or an customized file. Lizard will
                      report warning if a function exceed the limit
                      The path and file name to the whitelist file. It's './whitelizard.txt' by
                      default. Find more information in README.

Example use

Analyze a folder recursively: lizard mahjong_game/src

  NLOC    CCN  token  param    function@line@file
    10      2     29      2    start_new_player@26@./html_game.c
     6      1      3      0    set_shutdown_flag@449@./httpd.c
    24      3     61      1    server_main@454@./httpd.c
2 file analyzed.
LOC    Avg.NLOC AvgCCN Avg.ttoken  function_cnt    file
    191     15      3        51        12     ./html_game.c
    363     24      4        86        15     ./httpd.c

!!!! Warnings (CCN > 15) !!!!
    66     19    247      1    accept_request@64@./httpd.c
Total NLOC  Avg.NLOC  Avg CCN  Avg token  Fun Cnt  Warning cnt   Fun Rt   NLOC Rt
       554        20     4.07      71.15       27            1      0.04    0.12

Warnings only (in clang/gcc formation):lizard -w mahjong_game

./src/html_ui/httpd.c:64: warning: accept_request has 19 CCN and 1 params (66 NLOC, 247 tokens)
./src/mahjong_game/mj_table.c:109: warning: mj_table_update_state has 20 CCN and 1 params (72 NLOC, 255 tokens)

Set warning threshold for any field:lizard -T nloc=25

The option -Tcyclomatic_complexity=10 is equal to -C10. The option -Tlength=10 is equal to -L10. The option -Tparameter_count=10 is equal to -a10.

You can also do -Tnloc=10 to set the limit of the NLOC. Any function that has NLOC greater than 10 will generate a warning.

Generated code

Lizard has a simple solution with generated code. Any code in a source file that is following a comment containing “GENERATED CODE” will be ignored completely. The ignored code will not generate any data, except the file counting.

Code Duplicate Detector

lizard -Eduplicate <path to your code>

Generate A Tag Cloud For Your Code

You can generate a “Tag cloud” of your code by the following command. It counts the identifiers in your code (ignoring the comments).

lizard -EWordCount <path to your code>

Using lizard as Python module

You can also use lizard as a Python module in your code:

>>> import lizard
>>> i = lizard.analyze_file("../cpputest/tests/AllTests.cpp")
>>> print i.__dict__
{'nloc': 9, 'function_list': [<lizard.FunctionInfo object at 0x10bf7af10>], 'filename': '../cpputest/tests/AllTests.cpp'}
>>> print i.function_list[0].__dict__
{'cyclomatic_complexity': 1, 'token_count': 22, 'name': 'main', 'parameter_count': 2, 'nloc': 3, 'long_name': 'main( int ac , const char ** av )', 'start_line': 30}

You can also use source code string instead of file. But you need to provide a file name (to identify the language).

>>> i = lizard.analyze_file.analyze_source_code("AllTests.cpp", "int foo(){}")


If for some reason you would like to ignore the warnings, you can use the whitelist. Add ‘whitelizard.txt’ to the current folder (or use -W to point to the whitelist file), then the functions defined in the file will be ignored. Please notice that if you assign the file pathname, it needs to be exactly the same relative path as Lizard to find the file. An easy way to get the file pathname is to copy it from the Lizard warning output. This is an example whitelist:

#The file name can only be whitelizard.txt and put it in the current folder.
#You may have commented lines begin with #.
function_name1, function_name2 # list function names in multiple lines or split with comma.
file/path/name:function1, function2  # you can also specify the filename

Options in Comments

You can use options in the comments of the source code to change the behavior of lizard. By putting “#lizard forgives” inside a function or before a function it will suppress the warning for that function.

int foo() {
    // #lizard forgives the complexity


Lizard requires syntactically correct code. Upon processing input with incorrect or unknown syntax:

  • Lizard guarantees to terminate eventually (i.e., no forever loops, hangs) without hard failures (e.g., exit, crash, exceptions).

  • There is a chance of a combination of the following soft failures:

    • omission

    • misinterpretation

    • improper analysis / tally

    • success (the code under consideration is not relevant, e.g., global macros in C)

This approach makes the Lizard implementation simpler and more focused with partial parsers for various languages. Developers of Lizard attempt to minimize the possibility of soft failures. Hard failures are bugs in Lizard code, while soft failures are trade-offs or potential bugs.

In addition to asserting the correct code, Lizard may choose not to deal with some advanced or complicated language features:

  • C/C++ digraphs and trigraphs are not recognized.

  • C/C++ preprocessing or macro expansion is not performed. For example, using macro instead of parentheses (or partial statements in macros) can confuse Lizard’s bracket stacks.

  • Some C++ complicated templates may cause confusion with matching angle brackets and processing less-than < or more-than > operators inside of template arguments.

Literatures Referring to Lizard

Lizard is often used in software related researches. If you used it to support your work, you may contact the lizard author to add your work in the following list.

Lizard is also used as a plugin for fastlane to help check code complexity and submit xml report to sonar.

Project details

Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

lizard-1.17.10.tar.gz (61.6 kB view hashes)

Uploaded Source

Built Distribution

lizard-1.17.10-py2.py3-none-any.whl (66.0 kB view hashes)

Uploaded Python 2 Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page