Python implementation of the package url spec. This project is sponsored by NLnet project https://nlnet.nl/project/vulnerabilitydatabase/ , the Google Summer of Code, nexB and other generous sponsors.
It would be nice to extend the current PackageURLFilter to accept multiple purl values in the form of `?purl=some_purl&purl=another_purl
String-based fields should not have null=True.
See: https://docs.djangoproject.com/en/3.0/ref/models/fields/#django.db.models.Field.null for details.
I am getting an error when I run url2purl on some sourceforge URLs:
http://master.dl.sourceforge.net/project/googleimagedown/project/v1.1/GoogleImageDownloader-v1.1-src.tar.bz2
ValueError('too many values to unpack (expected 2)')
The problem is where we split on the substring "/project/" (https://github.com/package-url/packageurl-python/blob/main/src/packageurl/contrib/url2purl.py#L350) and we get more than 2 elements returned by the split.
We should always create a generic purl otherwise if we cannot infer it as in pkg:generic/lxc-master?download_url=https://salsa.debian.org/lxc-team/lxc/-/archive/master/lxc-master.tar.gz ... see https://github.com/package-url/purl-spec/blob/master/PURL-TYPES.rst#generic
This helps a lot downstream users such as ScanCode to get always something and at the very least a generic purl.
url2purl does not make purls for the following PyPI URLs
Current implementation:
We need to support more npm package url schemes, such as:
With https://github.com/pypa/get-virtualenv/raw/20.0.31/public/virtualenv.pyz this purl is inferred:
pkg:github/pypa/get-virtualenv@raw#20.0.31/public/virtualenv.pyz and this is not correct.
It should be instead:
pkg:github/pypa/[email protected]#/public/virtualenv.pyz
>>> from packageurl import url2purl
>>> url2purl.get_purl('https://github.com/pypa/get-virtualenv/raw/20.0.31/public/virtualenv.pyz')
PackageURL(type='github', namespace='pypa', name='get-virtualenv', version='raw', qualifiers=OrderedDict(), subpath='20.0.31/public/virtualenv.pyz')The following code is duplicated many times in the url2purl module:
@purl_router.route(pattern)
def build_purl(uri):
return purl_from_pattern('type', pattern, uri)
We should add a simple registration function as a shortcut.
The current code is working but rather contrived.... We can do better!
For instance with this snippet:
>>> try:
... from urlparse import urlparse
... except ImportError:
... # Python 3
... from urllib.parse import urlparse
...
>>> urlparse('GO:google.golang.org/genproto@abcdedf?arch=i386#/googleapis/api/annotations/', allow_fragments=True)
ParseResult(scheme='go', netloc='', path='google.golang.org/genproto@abcdedf', params='', query='arch=i386', fragment='/googleapis/api/annotations/')
>>> urlparse('npm:%40angular/[email protected]')
ParseResult(scheme='npm', netloc='', path='%40angular/[email protected]', params='', query='', fragment='')
@ashcrow ping :)
dict is now the same as OrderedDict, and better, faster, cheaper in all respects.
setuptools install is deprecated and a warning regarding this is displayed during this project's current build process:
> python setup.py build sdist bdist_wheel
...
running bdist_wheel
/Users/phorton/Documents/GitHub/madpah/packageurl-python/lib/python3.10/site-packages/setuptools/command/install.py:34: SetuptoolsDeprecationWarning: setup.py install is deprecated. Use build and pip and other standards-based tools.
warnings.warn(
...
I would suggest that this project adopts a more modern package manager such as Poetry.
@pombredanne - do you have thoughts on this?
Probably in a new contrib/ module as a django_models.py:
from django.db import models
class PackageURLMixin(models.Model):
"""
Abstract Model for Package URL "purl" fields support.
"""
type = models.CharField(
max_length=16,
blank=True,
null=True,
help_text=_(
'A short code to identify the type of this package. '
'For example: gem for a Rubygem, docker for a container, '
'pypi for a Python Wheel or Egg, maven for a Maven Jar, '
'deb for a Debian package, etc.'
)
)
namespace = models.CharField(
max_length=255,
blank=True,
null=True,
help_text=_(
'Package name prefix, such as Maven groupid, Docker image owner, '
'GitHub user or organization, etc.'
),
)
name = models.CharField(
max_length=100,
blank=True,
null=True,
help_text=_('Name of the package.'),
)
version = models.CharField(
max_length=50,
blank=True,
null=True,
help_text=_('Version of the package.'),
)
qualifiers = models.CharField(
max_length=1024,
blank=True,
null=True,
help_text=_(
'Extra qualifying data for a package such as the name of an OS, '
'architecture, distro, etc.'
),
)
subpath = models.CharField(
max_length=200,
blank=True,
null=True,
help_text=_(
'Extra subpath within a package, relative to the package root.'
),
)
class Meta:
abstract = True
@property
def package_url(self):
"""
Return a compact Package URL "purl" string.
"""
try:
purl = PackageURL(
self.type, self.namespace, self.name,
self.version, self.qualifiers, self.subpath
)
except ValueError:
return ''
return str(purl)
Signed-off-by: Thomas Druez [email protected]
Why is the colon in the name not translated?
would be
pkg:generic/code.google.com/android-notifier?download_url=https://storage.googleapis.com/google-code-archive-downloads/v2/code.google.com/android-notifier/android-notifier-desktop-0.5.1-1.i386.rpm
where the type is generic, namespace is code.google.com, name is the project name, and the download url is set as a qualifier.
We do not use code.google.com as a type because code.google.com repos has been shut down.
purl is an existing package module name. To avoid confusion, we could use packageurl instead.
@ashcrow ping
In the test_purl.py the from the statement in line 31 no path is set to find the file packageurl.py so its showing the error
`
Storing PackageURL instances in a set can be useful. Currently attempting this fails with the following error:
Python 3.8.2 (default, Feb 26 2020, 22:21:03)
[GCC 9.2.1 20200130] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from packageurl import PackageURL
>>> {PackageURL(name='test', type='pypi')}
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
TypeError: unhashable type: 'collections.OrderedDict'
I've run the latest version of url2purl and I've noticed the following issues:
{'download_url': 'current_purl_that_looks_wrong'}
{
'https://bitbucket.org/multicoreware/x265/downloads/x265_2.6.tar.gz': 'pkg:bitbucket/multicoreware/x265@downloads#x265_2.6.tar.gz',
'https://bitbucket.org/robeden/trove/downloads/trove-3.0.3.zip': 'pkg:bitbucket/robeden/trove@downloads#trove-3.0.3.zip',
'https://github.com/FasterXML/woodstox/archive/woodstox-core-5.0.2.zip': 'pkg:github/fasterxml/[email protected]',
'https://github.com/adobe-fonts/source-code-pro/archive/2.030R-ro/1.050R-it.tar.gz': 'pkg:github/adobe-fonts/[email protected]/1.050R-it',
'https://github.com/cassandra-rb/simple_uuid/archive/simple_uuid-0.3.0.zip': 'pkg:github/cassandra-rb/simple_uuid@simple_uuid-0.3.0',
'https://github.com/djberg96/sys-filesystem/archive/sys-filesystem-1.1.4.zip': 'pkg:github/djberg96/[email protected]',
'https://github.com/freedesktop/xorg-intel-gpu-tools/archive/igt-gpu-tools-1.23.tar.gz': 'pkg:github/freedesktop/[email protected]',
'https://github.com/grnet/synnefo/archive/synnefo/v0.12.3.zip': 'pkg:github/grnet/synnefo@synnefo/v0.12.3',
'https://github.com/n8n-io/n8n/archive/[email protected]': 'pkg:github/n8n-io/n8n@n8n%400.23.0',
'https://github.com/nginx/nginx/archive/branches/stable-0.7.zip': 'pkg:github/nginx/nginx@branches/stable-0.7',
'https://github.com/fanf2/unifdef/blob/master/unifdef.c': 'pkg:github/fanf2/unifdef@blob#master/unifdef.c',
'https://github.com/joebeeson/amazon/blob/master/vendors/aws-sdk/sdk.class.php': 'pkg:github/joebeeson/amazon@blob#master/vendors/aws-sdk/sdk.class.php',
'https://github.com/modelfabric/yowl/blob/master/bin/yowl': 'pkg:github/modelfabric/yowl@blob#master/bin/yowl',
'https://github.com/downloads/mozilla/rhino/rhino1_7R4.zip': 'pkg:github/downloads/mozilla@rhino#rhino1_7R4.zip',
'https://github.com/pombredanne/schematics.git': 'pkg:github/pombredanne/schematics.git',
'https://github.com/syncthing/syncthing/releases/download/v0.14.36/syncthing-source-v0.14.36.tar.gz': 'pkg:github/syncthing/syncthing@releases#download/v0.14.36/syncthing-source-v0.14.36.tar.gz',
'https://github.com/torakiki/pdfsam/releases/download/v3.3.2/pdfsam-3.3.2-bin.zip': 'pkg:github/torakiki/pdfsam@releases#download/v3.3.2/pdfsam-3.3.2-bin.zip',
'https://github.com/yarnpkg/yarn/releases/download/v1.3.2/yarn-v1.3.2.tar.gz': 'pkg:github/yarnpkg/yarn@releases#download/v1.3.2/yarn-v1.3.2.tar.gz',
'https://github.com/z3APA3A/3proxy/releases/download/0.8.11/3proxy-0.8.11.zip': 'pkg:github/z3apa3a/3proxy@releases#download/0.8.11/3proxy-0.8.11.zip',
'https://raw.githubusercontent.com/LeZuse/flex-sdk/master/frameworks/projects/mx/src/mx/containers/accordionClasses/AccordionHeader.as': 'pkg:github/lezuse/flex-sdk/master/frameworks/projects/mx/src/mx/containers@accordionClasses#AccordionHeader.as',
'https://raw.githubusercontent.com/NCIP/lexevs/master/lgSharedLibraries/jettison/jettison-1.1.jar': 'pkg:github/ncip/lexevs/master/lgsharedlibraries@jettison#jettison-1.1.jar',
'https://raw.githubusercontent.com/jgallen23/routie/master/dist/routie.min.js': 'pkg:github/jgallen23/routie/master@dist#routie.min.js',
'https://raw.githubusercontent.com/jquery/jquery-ui/1.8.14/ui/jquery.ui.mouse.js': 'pkg:github/jquery/jquery-ui/1.8.14@ui#jquery.ui.mouse.js',
'https://raw.githubusercontent.com/mariopacio/projeto.dprf.info/master/assets/plugins/xbreadcrumbs/xbreadcrumbs.js': 'pkg:github/mariopacio/projeto.dprf.info/master/assets/plugins@xbreadcrumbs#xbreadcrumbs.js'
}
I have a django webapp that uses packageurl.contrib.django.models.PackageURLMixin and I get the following error when I try to save a package with a very long name.
Traceback (most recent call last):
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/backends/utils.py", line 84, in _execute
return self.cursor.execute(sql, params)
psycopg2.errors.StringDataRightTruncation: value too long for type character varying(100)
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "./manage.py", line 10, in <module>
execute_from_command_line(sys.argv)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/core/management/__init__.py", line 419, in execute_from_command_line
utility.execute()
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/core/management/__init__.py", line 413, in execute
self.fetch_command(subcommand).run_from_argv(self.argv)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/core/management/base.py", line 354, in run_from_argv
self.execute(*args, **cmd_options)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/core/management/base.py", line 398, in execute
output = self.handle(*args, **options)
File "/home/jono/nexb/src/dejacode/component_catalog/management/commands/setpurls.py", line 88, in handle
Package.objects.filter(pk=package.pk).update(**package_url_dict)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/models/query.py", line 783, in update
rows = query.get_compiler(self.db).execute_sql(CURSOR)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/models/sql/compiler.py", line 1559, in execute_sql
cursor = super().execute_sql(result_type)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/models/sql/compiler.py", line 1175, in execute_sql
cursor.execute(sql, params)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/backends/utils.py", line 98, in execute
return super().execute(sql, params)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/backends/utils.py", line 66, in execute
return self._execute_with_wrappers(sql, params, many=False, executor=self._execute)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/backends/utils.py", line 75, in _execute_with_wrappers
return executor(sql, params, many, context)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/backends/utils.py", line 84, in _execute
return self.cursor.execute(sql, params)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/utils.py", line 90, in __exit__
raise dj_exc_value.with_traceback(traceback) from exc_value
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/db/backends/utils.py", line 84, in _execute
return self.cursor.execute(sql, params)
django.db.utils.DataError: value too long for type character varying(100)
This isn't valid python2:
normalized_purl = f'pkg:{purl}' if not purl.startswith('pkg:') else purl
Python2 doesn't support f-strings and I get an error when building the module:
byte-compiling build/bdist.linux-x86_64/egg/packageurl/contrib/django/models.py to models.pyc
File "build/bdist.linux-x86_64/egg/packageurl/contrib/django/models.py", line 60
normalized_purl = f'pkg:{purl}' if not purl.startswith('pkg:') else purl
^
SyntaxError: invalid syntax
This no longer works
https://rubygems.org/gems/i18n-js-3.0.11.gem should be detected as a gem but instead is detected as a generic purl
In NixOS a special syntax is used in the package configurations that allow downloading from mirrors, which are then locally configured and can be changed or searched automatically, and allow a user to pick a mirror closest, or do some very basic failover.
An example is this package: https://github.com/NixOS/nixpkgs/blob/master/pkgs/tools/graphics/barcode/default.nix
url = "mirror://gnu/${pname}/${pname}-${version}.tar.xz";
This then tells the download script to substitute the package and use the configured mirror. It might be useful for purl2url.py to either use something similar or to return something similar to what NixOS does, so the actual fetcher can then replace it with the correct mirror.
hey 😄
Thanks for the great project 🚀 We are using both your project and CycloneDX to create SBOMs.
I would like to know, if you are interested in adding type hints to your project? I'm a big supporter of improving Python projects by adding type hints to them and reduce the amount of errors, which come from passing the wrong types to a functions or expecting a different result to come back.
I'm willing to do the change, just need to know if you prefer stub files or inline type hints, I usually prefer the latter one? I would also add mypy to your CI to check, if the added type hints make sense or not 🙂
Currently their is no support for Rust Packages (crates) in package-url python repo
Qualifiers are currently stored as strings in the database, which implies that the order of each qualifier matters. This contradicts the Package URL specification ( check for more details package-url/purl-spec#51 (comment) )
django/django#12392 adds JSONField in a db agnostic way to the django orm.
This field will be implemented for Django 3.1+ is not planned to be backported.
We are using PackageURLMixin in nexB/vulnerablecode.
While trying to import security data from Debian, I've come across a version "number" that exceeds the limit of 50 characters on the version field of the model mixin: 1:0.0~git20170407.0.55a552f+REALLY.0.0~git20161012.0.5f31782-1.
Can we bump max_length to 255? Should I open a PR for it?
Currently no code formatter is not used in packageurl-python, so a suggestion to add black for code formatting .
This repo currently has a test suite file within the repo. As part of the test process it seems it would be better to grab the current one and test off that.
Is it possible to set attributes (such as version) in a packageurl object?
>>> import packageurl
>>> purl = packageurl.PackageURL.from_string('pkg:generic/busybox')
>>> purl.version
>>> purl.version = 3
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
AttributeError: can't set attribute
From gitter chat https://gitter.im/package-url/Lobby
@tclasen :
Anyone know why the python package isn't letting me get a url from a purl?
@app.command()
def get(purl: str):
parsed = PackageURL.from_string(purl)
typer.echo(f"Grabbing {parsed}")
typer.echo("A", purl2url.get_url(purl))
typer.echo("B", purl2url.get_url(parsed.to_string()))
typer.echo("C", purl2url.get_url(str(parsed)))
typer.echo("D", purl2url.get_url(parsed.to_string()))
For this I've tried the two following PURLs:
pkg:maven/org.apache.commons/[email protected]
pkg:pypi/[email protected]
And for both of them I successfully parse and can print the purl, but the A, B, C, D echos after that are all empty (but don't crash).
@tclasen because it has not been implemented yet : https://github.com/package-url/packageurl-python/blob/main/src/packageurl/contrib/purl2url.py :)
It should be reasonably straightforward using tidbits of code from https://github.com/nexB/fetchcode/blob/master/src/fetchcode/package.py and the expansive set in https://github.com/nexB/scancode-toolkit/tree/develop/src/packagedcode
Currently it seems the packageurl-python library is not typed according to PEP 561.
Raising for awareness.
FYI @jkowalleck
When we create a generic sourceforge purl for sourceforge URLs that do not fit the pattern we use to parse sourceforge URLs, we should use the project name for the name field when creating the purl rather than using the filename.
For example, url2purl should process this:
into this:
pkg:sourceforge/aloyscore?download_url=http://master.dl.sourceforge.net/project/aloyscore/aloyscore/0.1a1%2520stable/0.1a1_stable_AloysCore.zip
instead of this:
pkg:sourceforge/0.1a1_stable_AloysCore.zip?download_url=http://master.dl.sourceforge.net/project/aloyscore/aloyscore/0.1a1%2520stable/0.1a1_stable_AloysCore.zip
PackageURL.from_string("pkg:npm/@babel/core")
The purl is identified as a wrong one.
The is a common requirement on apps that implement the PackageURLMixin class.
The current implementation uses a namedtuple for PackageURL but this is an implementation detail that should be relied uppon. We should have a method to create a new purl from an existing purl passing new field values as arguments.
See https://github.com/nexB/skeleton and https://github.com/nexB/fetchcode/ for examples
>>> purl = "pkg:alpine/openssl@0?arch=aarch64&distroversion=edge&reponame=main"
>>> purl_to_lookups(purl_str=purl)
{'type': 'alpine', 'name': 'openssl', 'version': '0', 'qualifiers': 'arch=aarch64&distroversion=edge&reponame=main'}
We should consider creating Package URLs for googlesource URLs. For example:
https://chromium.googlesource.com/v8/v8.git/%2Barchive/refs/heads/6.4.388.tar.gz
would be
pkg:googlesource/chromium/v8/[email protected]
Where googlesource is the type, chromium is the namespace, v8/v8 the name, and 6.4.388 the version.
Likewise:
https://android.googlesource.com/platform/system/bt.git/+archive/67f5f372.tar.gz
would be
pkg:googlesource/android/platform/system/bt@67f5f372
>>> from packageurl.contrib import purl2url
>>> purl2url.purl2url('pkg:github/package-url/[email protected]')
'https://github.com/package-url/packageurl-python/tree/0.9.1'
The returned URL https://github.com/package-url/packageurl-python/tree/0.9.1 leads to a 404.
The proper URL is actually https://github.com/package-url/packageurl-python/tree/v0.9.1 with the "v" prefix.
Note that the "v" prefix in a tag name is a convention and optional.
I was running some code that updates purls based on download URLs and url2purl crashed on the URL http://apt-rpm.org/
Traceback (most recent call last):
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/packageurl/contrib/url2purl.py", line 54, in url2purl
return purl_router.process(url)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/packageurl/contrib/route.py", line 174, in process
endpoint = self.resolve(string)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/packageurl/contrib/route.py", line 197, in resolve
raise NoRouteAvailable(string)
packageurl.contrib.route.NoRouteAvailable: http://apt-rpm.org/
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "./manage.py", line 10, in <module>
execute_from_command_line(sys.argv)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/core/management/__init__.py", line 419, in execute_from_command_line
utility.execute()
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/core/management/__init__.py", line 413, in execute
self.fetch_command(subcommand).run_from_argv(self.argv)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/core/management/base.py", line 354, in run_from_argv
self.execute(*args, **cmd_options)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/django/core/management/base.py", line 398, in execute
output = self.handle(*args, **options)
File "/home/jono/nexb/src/dejacode/component_catalog/management/commands/setpurls.py", line 79, in handle
package_url = url2purl.get_purl(package.download_url)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/packageurl/contrib/url2purl.py", line 58, in url2purl
return build_generic_purl(url)
File "/home/jono/nexb/src/dejacode/lib/python3.6/site-packages/packageurl/contrib/url2purl.py", line 105, in build_generic_purl
file_name = uri_path_segments[-1]
IndexError: list index out of range
We should check after https://github.com/package-url/packageurl-python/blob/main/src/packageurl/contrib/url2purl.py#L104 to see if we have URL path segments before creating a purl.
travis.com is dead but still listed. We could enable GH actions and/or Azure in addition to appveyor
It would be very useful if it would be possible to supply a name of a package and then get a purl object.
For example. I would like to do something like:
>>> purl_object = packageurl.PackageURL.from_package_name('vim_8.1.0875-5_amd64.deb')
and get a correct purl object.
Currently there is a method to_dict() but there is no equivalent from_dict(). It might be useful to have this as well (use case: templating).
Working with regex syntax is always hard but should not be necessary for most of the simple routes.
For example, a common pattern '[^/]+' in path segment should be abstracted for better readability and new route addition.
We could re-use some ideas from the recent Django's URL route system that now replaces the old regex system: https://docs.djangoproject.com/en/3.1/topics/http/urls/#url-dispatcher
This system abstracts the regex complexity into "converters", for example r'^articles/(?P<year>[0-9]{4})/$' becomes articles/<yyyy:year>/
Using a current url2purl example:
pattern = r"https?://raw.githubusercontent.com/(?P<namespace>[^/]+)/(?P<name>[^/]+)/(?P<version>[^/]+)/(?P<subpath>.*)$"Could become:
route = "https://raw.githubusercontent.com/<str:namespace>/<str:name>/<str:version>/<path:subpath>"Much easier to write and to read.
Playing around with the Django's _route_to_regex
from django.urls.resolvers import _route_to_regex
route = "https://raw.githubusercontent.com/<str:namespace>/<str:name>/<str:version>/<path:subpath>"
pattern = _route_to_regex(route, is_endpoint=True)[0]
# -> "^https\\:\\/\\/raw\\.githubusercontent\\.com\\/(?P<namespace>[^/]+)\\/(?P<name>[^/]+)\\/(?P<version>[^/]+)\\/(?P<subpath>.+)$"
url = "https://raw.githubusercontent.com/LeZuse/flex-sdk/master/frameworks/projects/mx/src/mx/containers/accordionClasses/AccordionHeader.as"
re.compile(pattern, re.VERBOSE).match(url).groupdict()
# -> {'namespace': 'LeZuse', 'name': 'flex-sdk', 'version': 'master', 'subpath': 'frameworks/projects/mx/src/mx/containers/accordionClasses/AccordionHeader.as'}
We could add custom converter for the specific needs of purl https://docs.djangoproject.com/en/3.1/topics/http/urls/#registering-custom-path-converters
Some parts like the (http|https) will need support as well as the domain section is not part of the Django system:
from django.urls.resolvers import _route_to_regex
from django.urls.converters import register_converter
from django.urls.converters import StringConverter
class ProtocolConverter(StringConverter):
regex = '(http|https|ftp)'
register_converter(ProtocolConverter, 'protocol')
route = "<protocol:protocol>://raw.githubusercontent.com/<str:namespace>/<str:name>/<str:version>/<path:subpath>"
_route_to_regex(route, is_endpoint=True)[0]
'^(?P<protocol>(http|https|ftp))\\:\\/\\/raw\\.githubusercontent\\.com\\/(?P<namespace>[^/]+)\\/(?P<name>[^/]+)\\/(?P<version>[^/]+)\\/(?P<subpath>.+)$'
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.