hgext/largefiles/basestore.py

httppeer: detect redirect to URL without query string (issue5860)

197d10e157ce subtly changed the HTTP peer's handling of HTTP redirects.

Before that changeset, we instantiated an HTTP peer instance and
performed the capabilities lookup with that instance. The old code had
the following relevant properties:

1) The HTTP request layer would automatically follow HTTP redirects.
2) An encountered HTTP redirect would update a peer instance variable
   pointing to the repo URL.
3) The peer would automagically perform a "capabilities" command
   request if a caller requested capabilities but capabilities were
   not yet defined.

The first HTTP request issued by a peer is for ?cmd=capabilities. If
the server responds with an HTTP redirect to a ?cmd=capabilities URL,
the HTTP request layer automatically followed it, retrieved a valid
capabilities response, and the peer's base URL was updated
automatically so subsequent requests used the proper URL. In other
words, things "just worked."

In the case where the server redirected to a URL without the
?cmd=capabilities query string, the HTTP request layer would follow
the redirect and likely encounter HTML. The peer's base URL would be
updated and the unexpected Content-Type would raise a RepoError. We
would catch RepoError and immediately call between() (testing the case
for pre 0.9.1 servers not supporting the "capabilities" command). e.g.

    try:
        inst._fetchcaps()
    except error.RepoError:
        inst.between([(nullid, nullid)])

between() would eventually call into _callstream(). And _callstream()
made a call to self.capable('httpheader'). capable() would call
self.capabilities(), which would see that no capabilities were set
(because HTML was returned for that request) and call the "capabilities"
command to fetch capabilities. Because the base URL had been updated
from the redirect, this 2nd "capabilities" command would succeed and
the client would immediately call "between," which would also succeed.
The legacy handshake succeeded. Only because "capabilities" was
successfully executed as a side effect did the peer recognize that it
was talking to a modern server. In other words, this all appeared to
work accidentally.

After 197d10e157ce, we stopped calling the "capabilities" command on
the peer instance. Instead, we made the request via a low-level opener,
detected the redirect as part of response handling code, and passed the
redirected URL into the constructed peer instance.

For cases where the redirected URL included the query string, this
"just worked." But for cases where the redirected URL stripped the query
string, we threw RepoError and because we removed the "between" handshake
fallback, we fell through to the "is a static HTTP repo" check and
performed an HTTP request for .hg/requires.

While 197d10e157ce was marked as backwards incompatible, the only
intended backwards incompatible behavior was not performing the
"between" fallback. It was not realized that the "between" command
had the side-effect of recovering from an errant redirect that
dropped the query string.

This commit restores the previous behavior and allows clients to
handle a redirect that drops the query string. In the case where
the request is redirected and the query string is dropped, we raise
a special case of RepoError. We then catch this special exception in
the handshake code and perform another "capabilities" request against
the redirected URL. If that works, all is well. Otherwise, we fall back
to the "is a static HTTP repo" check.

The new code is arguably better than before 197d10e157ce, as it is
explicit about the expected behavior and we avoid performing a
"between" request, saving a server round trip.

Differential Revision: https://phab.mercurial-scm.org/D3433

# Copyright 2009-2010 Gregory P. Ward
# Copyright 2009-2010 Intelerad Medical Systems Incorporated
# Copyright 2010-2011 Fog Creek Software
# Copyright 2010-2011 Unity Technologies
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.

'''base class for store implementations and store-related utility code'''
from __future__ import absolute_import

from mercurial.i18n import _

from mercurial import node, util

from . import lfutil

class StoreError(Exception):
    '''Raised when there is a problem getting files from or putting
    files to a central store.'''
    def __init__(self, filename, hash, url, detail):
        self.filename = filename
        self.hash = hash
        self.url = url
        self.detail = detail

    def longmessage(self):
        return (_("error getting id %s from url %s for file %s: %s\n") %
                 (self.hash, util.hidepassword(self.url), self.filename,
                  self.detail))

    def __str__(self):
        return "%s: %s" % (util.hidepassword(self.url), self.detail)

class basestore(object):
    def __init__(self, ui, repo, url):
        self.ui = ui
        self.repo = repo
        self.url = url

    def put(self, source, hash):
        '''Put source file into the store so it can be retrieved by hash.'''
        raise NotImplementedError('abstract method')

    def exists(self, hashes):
        '''Check to see if the store contains the given hashes. Given an
        iterable of hashes it returns a mapping from hash to bool.'''
        raise NotImplementedError('abstract method')

    def get(self, files):
        '''Get the specified largefiles from the store and write to local
        files under repo.root.  files is a list of (filename, hash)
        tuples.  Return (success, missing), lists of files successfully
        downloaded and those not found in the store.  success is a list
        of (filename, hash) tuples; missing is a list of filenames that
        we could not get.  (The detailed error message will already have
        been presented to the user, so missing is just supplied as a
        summary.)'''
        success = []
        missing = []
        ui = self.ui

        at = 0
        available = self.exists(set(hash for (_filename, hash) in files))
        for filename, hash in files:
            ui.progress(_('getting largefiles'), at, unit=_('files'),
                total=len(files))
            at += 1
            ui.note(_('getting %s:%s\n') % (filename, hash))

            if not available.get(hash):
                ui.warn(_('%s: largefile %s not available from %s\n')
                        % (filename, hash, util.hidepassword(self.url)))
                missing.append(filename)
                continue

            if self._gethash(filename, hash):
                success.append((filename, hash))
            else:
                missing.append(filename)

        ui.progress(_('getting largefiles'), None)
        return (success, missing)

    def _gethash(self, filename, hash):
        """Get file with the provided hash and store it in the local repo's
        store and in the usercache.
        filename is for informational messages only.
        """
        util.makedirs(lfutil.storepath(self.repo, ''))
        storefilename = lfutil.storepath(self.repo, hash)

        tmpname = storefilename + '.tmp'
        with util.atomictempfile(tmpname,
                createmode=self.repo.store.createmode) as tmpfile:
            try:
                gothash = self._getfile(tmpfile, filename, hash)
            except StoreError as err:
                self.ui.warn(err.longmessage())
                gothash = ""

        if gothash != hash:
            if gothash != "":
                self.ui.warn(_('%s: data corruption (expected %s, got %s)\n')
                             % (filename, hash, gothash))
            util.unlink(tmpname)
            return False

        util.rename(tmpname, storefilename)
        lfutil.linktousercache(self.repo, hash)
        return True

    def verify(self, revs, contents=False):
        '''Verify the existence (and, optionally, contents) of every big
        file revision referenced by every changeset in revs.
        Return 0 if all is well, non-zero on any errors.'''

        self.ui.status(_('searching %d changesets for largefiles\n') %
                       len(revs))
        verified = set()                # set of (filename, filenode) tuples
        filestocheck = []               # list of (cset, filename, expectedhash)
        for rev in revs:
            cctx = self.repo[rev]
            cset = "%d:%s" % (cctx.rev(), node.short(cctx.node()))

            for standin in cctx:
                filename = lfutil.splitstandin(standin)
                if filename:
                    fctx = cctx[standin]
                    key = (filename, fctx.filenode())
                    if key not in verified:
                        verified.add(key)
                        expectedhash = lfutil.readasstandin(fctx)
                        filestocheck.append((cset, filename, expectedhash))

        failed = self._verifyfiles(contents, filestocheck)

        numrevs = len(verified)
        numlfiles = len(set([fname for (fname, fnode) in verified]))
        if contents:
            self.ui.status(
                _('verified contents of %d revisions of %d largefiles\n')
                % (numrevs, numlfiles))
        else:
            self.ui.status(
                _('verified existence of %d revisions of %d largefiles\n')
                % (numrevs, numlfiles))
        return int(failed)

    def _getfile(self, tmpfile, filename, hash):
        '''Fetch one revision of one file from the store and write it
        to tmpfile.  Compute the hash of the file on-the-fly as it
        downloads and return the hash.  Close tmpfile.  Raise
        StoreError if unable to download the file (e.g. it does not
        exist in the store).'''
        raise NotImplementedError('abstract method')

    def _verifyfiles(self, contents, filestocheck):
        '''Perform the actual verification of files in the store.
        'contents' controls verification of content hash.
        'filestocheck' is list of files to check.
        Returns _true_ if any problems are found!
        '''
        raise NotImplementedError('abstract method')