Hash generator for HTML inline styles and scripts
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Gonçalo Valério 9fea295f24
Fix test workflow
1 year ago
.github/workflows Fix test workflow 1 year ago
inlinehashes Add line numbers and removed unneeded dependency 1 year ago
tests Add line numbers and removed unneeded dependency 1 year ago
.gitignore Add hash implementation and basic html parsing. The cli app should be usable. 1 year ago
LICENSE Initial commit 1 year ago
README.rst Add new target options to the readme 1 year ago
poetry.lock Add line numbers and removed unneeded dependency 1 year ago
pyproject.toml Add line numbers and removed unneeded dependency 1 year ago



A small tool and library to generate the hashes of inline content that needs to be whitelisted when serving an HTML document
with a `Content-Security-Policy <https://developer.mozilla.org/en-US/docs/Web/HTTP/CSP>`_ (because, as the name indicates,
using ``unsafe-inline`` is not recommended).

You provide the HTML content (directly or through a file path/URL) then ``inlinehashes`` will parse the document and provide
you with a list of elements that need to be explicitly added to the CSP header/tag.

The tool can be specially useful for scenarios where you use/include external software solutions in your website or application
(such as a 3rd party CMS, etc), since it will allow you to detect changes after updates and edit you CSP accordingly.

*Quick note: Always verify the content you are whitelisting and be careful when fetching live website data, since any existing
XSS code will be included in the results.*

**At the moment this package is still in a very early stage, so it still doesn't detect all possible items and the current API
might change with future releases.**

Inline content that is currently detected:

* ``<script></script>`` tags
* ``<style></style>`` tags
* Many event handlers defined in element/tag attributes
* Styles defined directly in the element/tag using the ``style`` attribute


Using pip you just need to ``pip install inlinehashes``


The package can be used through 2 different ways, either by using the CLI interface or programmatically in your python project.
Bellow you can find a quick summary of the available functionality.

CLI app

This is the available functionality:

.. code::

    usage: inlinehashes [-h] [-a {sha256,sha384,sha512}] [-f] [-t {all,scripts,styles}] source

    positional arguments:
    source                URL or local HTML file to check

    optional arguments:
    -h, --help            show this help message and exit
    -a {sha256,sha384,sha512}, --alg {sha256,sha384,sha512}
                            Hash algorithm to use (default: sha256)
    -f, --full            Include full content in the output
    -t {all,scripts,styles}, --target {all,scripts,styles}
                            Target inline content to look for

Here is an example of the output:

.. code::

    $inlinehashes https://ovalerio.net -a sha384
        "content": "\n      html {\n        height: 100%;\n      }\n      ",
        "hash": "sha384-Ku20lQH5qbr4EDPzXD2rf25rEHJNswNYRUNMPjYl7jCe0eHJYDe0gFdQpnKkFUTv",
        "line": 12,
        "position": 0


Here is the same example, but using the python shell:

.. code:: python

    >>> import requests
    >>> import inlinehashes
    >>> content = requests.get("https://ovalerio.net").text
    >>> inlines = inlinehashes.parse(content)
    >>> inlines
        html {
            height: 100%;
    >>> first = inlines[0]
    >>> first.short_content
    '\n      html {\n        height: 100%;\n      }\n      '
    >>> first.sha256
    >>> first.sha384
    >>> first.sha512
    >>> first.content
    '\n      html {\n        height: 100%;\n      }\n      body {\n        background-image: url("data:image/png;base64,iVBORw0KGgoAAAANS...'


All contributions and improvements are welcome.