view tests/test-hgweb-auth.py @ 47802:de2e04fe4897

hgwebdir: avoid systematic full garbage collection

Forcing a systematic full garbage collection upon each request
can serioulsy harm performance. This is reported as
https://bz.mercurial-scm.org/show_bug.cgi?id=6075

With this change we're performing the full collection according
to a new setting, `experimental.web.full-garbage-collection-rate`.
The default value is 1, which doesn't change the behavior and will
allow us to test on real use cases. If the value is 0, no full garbage
collection occurs.

Regardless of the value of the setting, a partial garbage collection
still occurs upon each request (not attempting to collect objects from
the oldest generation). This should be enough to take care of
reference cycles that have been created by the last request
(assessment of this requires changing the setting, not to be 1).

In my experience chasing memory leaks in Mercurial servers,
the full collection never reclaimed any memory, but this is with
Python 3 and biased towards small repositories.

On the other hand, as explained in the Python developer docs [1],
frequent full collections are very harmful in terms of performance if
lots of objects survive the collection, and hence stay in the
oldest generation. Note that `gc.collect()` is indeed trying to
collect the oldest generation [2]. This happens usually in two cases:

- unwanted lingering objects (i.e., an actual memory leak that
  the GC cannot do anything about). Sadly, we have lots of those
  these days.
- desireable long-term objects, typically in caches (not inner caches
  carried by repositories, which should be collected with them). This
  is a subject of interest for the Heptapod project.

In short, the flat rate that this change still permits is
probably a bad idea in most cases, and the default value can
be tweaked later on (or even be set to 0) according to experiments
in the wild.

The test is inspired from test-hgwebdir-paths.py

[1] https://devguide.python.org/garbage_collector/#collecting-the-oldest-generation
[2] https://docs.python.org/3/library/gc.html#gc.collect

Differential Revision: https://phab.mercurial-scm.org/D11204

from __future__ import absolute_import, print_function

from mercurial import demandimport

demandimport.enable()
from mercurial import (
    error,
    pycompat,
    ui as uimod,
    url,
    util,
)
from mercurial.utils import (
    stringutil,
    urlutil,
)

urlerr = util.urlerr
urlreq = util.urlreq


class myui(uimod.ui):
    def interactive(self):
        return False


origui = myui.load()


def writeauth(items):
    ui = origui.copy()
    for name, value in items.items():
        ui.setconfig(b'auth', name, value)
    return ui


def _stringifyauthinfo(ai):
    if ai is None:
        return ai
    realm, authuris, user, passwd = ai
    return (
        pycompat.strurl(realm),
        [pycompat.strurl(u) for u in authuris],
        pycompat.strurl(user),
        pycompat.strurl(passwd),
    )


def test(auth, urls=None):
    print('CFG:', pycompat.sysstr(stringutil.pprint(auth, bprefix=True)))
    prefixes = set()
    for k in auth:
        prefixes.add(k.split(b'.', 1)[0])
    for p in prefixes:
        for name in (b'.username', b'.password'):
            if (p + name) not in auth:
                auth[p + name] = p
    auth = {k: v for k, v in auth.items() if v is not None}

    ui = writeauth(auth)

    def _test(uri):
        print('URI:', pycompat.strurl(uri))
        try:
            pm = url.passwordmgr(ui, urlreq.httppasswordmgrwithdefaultrealm())
            u, authinfo = urlutil.url(uri).authinfo()
            if authinfo is not None:
                pm.add_password(*_stringifyauthinfo(authinfo))
            print(
                '    ',
                tuple(
                    pycompat.strurl(a)
                    for a in pm.find_user_password('test', pycompat.strurl(u))
                ),
            )
        except error.Abort:
            print('    ', 'abort')

    if not urls:
        urls = [
            b'http://example.org/foo',
            b'http://example.org/foo/bar',
            b'http://example.org/bar',
            b'https://example.org/foo',
            b'https://example.org/foo/bar',
            b'https://example.org/bar',
            b'https://[e-mail]/bar',
            b'https://[e-mail]/bar',
        ]
    for u in urls:
        _test(u)


print('\n*** Test in-uri schemes\n')
test({b'x.prefix': b'http://example.org'})
test({b'x.prefix': b'https://example.org'})
test({b'x.prefix': b'http://example.org', b'x.schemes': b'https'})
test({b'x.prefix': b'https://example.org', b'x.schemes': b'http'})

print('\n*** Test separately configured schemes\n')
test({b'x.prefix': b'example.org', b'x.schemes': b'http'})
test({b'x.prefix': b'example.org', b'x.schemes': b'https'})
test({b'x.prefix': b'example.org', b'x.schemes': b'http https'})

print('\n*** Test prefix matching\n')
test(
    {
        b'x.prefix': b'http://example.org/foo',
        b'y.prefix': b'http://example.org/bar',
    }
)
test(
    {
        b'x.prefix': b'http://example.org/foo',
        b'y.prefix': b'http://example.org/foo/bar',
    }
)
test({b'x.prefix': b'*', b'y.prefix': b'https://example.org/bar'})

print('\n*** Test user matching\n')
test(
    {
        b'x.prefix': b'http://example.org/foo',
        b'x.username': None,
        b'x.password': b'xpassword',
    },
    urls=[b'http://[e-mail]/foo'],
)
test(
    {
        b'x.prefix': b'http://example.org/foo',
        b'x.username': None,
        b'x.password': b'xpassword',
        b'y.prefix': b'http://example.org/foo',
        b'y.username': b'y',
        b'y.password': b'ypassword',
    },
    urls=[b'http://[e-mail]/foo'],
)
test(
    {
        b'x.prefix': b'http://example.org/foo/bar',
        b'x.username': None,
        b'x.password': b'xpassword',
        b'y.prefix': b'http://example.org/foo',
        b'y.username': b'y',
        b'y.password': b'ypassword',
    },
    urls=[b'http://[e-mail]/foo/bar'],
)

print('\n*** Test user matching with name in prefix\n')

# prefix, username and URL have the same user
test(
    {
        b'x.prefix': b'https://example.org/foo',
        b'x.username': None,
        b'x.password': b'xpassword',
        b'y.prefix': b'http://[e-mail]/foo',
        b'y.username': b'y',
        b'y.password': b'ypassword',
    },
    urls=[b'http://[e-mail]/foo'],
)
# Prefix has a different user from username and URL
test(
    {
        b'y.prefix': b'http://[e-mail]/foo',
        b'y.username': b'y',
        b'y.password': b'ypassword',
    },
    urls=[b'http://[e-mail]/foo'],
)
# Prefix has a different user from URL; no username
test(
    {b'y.prefix': b'http://[e-mail]/foo', b'y.password': b'ypassword'},
    urls=[b'http://[e-mail]/foo'],
)
# Prefix and URL have same user, but doesn't match username
test(
    {
        b'y.prefix': b'http://[e-mail]/foo',
        b'y.username': b'z',
        b'y.password': b'ypassword',
    },
    urls=[b'http://[e-mail]/foo'],
)
# Prefix and URL have the same user; no username
test(
    {b'y.prefix': b'http://[e-mail]/foo', b'y.password': b'ypassword'},
    urls=[b'http://[e-mail]/foo'],
)
# Prefix user, but no URL user or username
test(
    {b'y.prefix': b'http://[e-mail]/foo', b'y.password': b'ypassword'},
    urls=[b'http://example.org/foo'],
)


def testauthinfo(fullurl, authurl):
    print('URIs:', fullurl, authurl)
    pm = urlreq.httppasswordmgrwithdefaultrealm()
    ai = _stringifyauthinfo(
        urlutil.url(pycompat.bytesurl(fullurl)).authinfo()[1]
    )
    pm.add_password(*ai)
    print(pm.find_user_password('test', authurl))


print('\n*** Test urllib2 and urlutil.url\n')
testauthinfo('http://[e-mail]:8080/foo', 'http://example.com:8080/foo')