Commit Graph

1226 Commits

Author SHA1 Message Date
Matt Holt
81e26970a3 proxy: Move handling of headers around to prevent memory usage spikes (#784)
* Move handling of headers around to prevent memory use spikes

While debugging #782, I noticed that using http2 and max_fails=0,
X-Forwarded-For grew infinitely when an upstream request failed after
refreshing the test page. This change ensures that headers are only
set once per request rather than appending in a time-terminated loop.

* Refactor some code into its own function
2016-04-27 12:50:18 -06:00
W. Mark Kubacki
f561dc0bc1 browse: Serve a datetime format the IE11 and Firefox understand
Another new safeguard is that we check whether the datetime has been read
correctly. If not then the listing will not be localized.

closes #793
2016-04-27 19:24:42 +02:00
Matthew Holt
21382702d2
Update readme version, 0.8.3 2016-04-26 09:28:05 -06:00
Matthew Holt
e97649493b
Update change log; version 0.8.3 2016-04-26 08:36:59 -06:00
Matthew Holt
19d6f666aa
Fix build script with default git repo path (cwd) 2016-04-25 22:10:16 -06:00
W. Mark Kubacki
6f5cff5393 tls: Prevent Go stdlib from overwriting our very first tls ticket key (#785)
[1] 57e459e02b/src/crypto/tls/common.go (L424)
[2] 57e459e02b/src/crypto/tls/common.go (L392-L407)

[2] has overwritten the first tls ticket key on round N=0, that has previously
been written using [1].

Go's stdlib does not use c.sessionTicketKeys≥1 as indicator if those values had
already been set; initializing that lone SessionTicketKey does the job for for
now.
    If c.serverInit() were called in round N+1 all existing tls ticket keys
would be overwritten (in round N<4 except the very first one, of course).
As member variables of tls.Config are read-only by then, we cannot keep
updating SessionTicketKey as well.

This has been escalated to Go's authors with golang/go#15421 here:
https://github.com/golang/go/issues/15421

Thanks to Matthew Holt for the initial report!
2016-04-23 17:49:48 +02:00
Matthew Holt
5c96ee1d9c
Fix bug in renewing default certificate 2016-04-22 10:14:56 -06:00
W. Mark Kubacki
3c578dfbc1 Catch whitespace code style violations locally (#774)
Those settings enforce convergence on common coding style with respect to whitespace.

Do not use tabs to indent with shell scripts because those tabs most often
serve the function of triggering command completion. Which could end a
command before it is pasted completely.
  Traditionally indentation is two spaces here, not four.

Other rules will catch stray whitespace at the end of lines or files, which,
once committed, would annoy the next developer because his editor would strip
them from lines he did not intended to modify in the first place.
2016-04-20 16:56:57 +02:00
Abiola Ibrahim
a093aea797 Merge pull request #773 from eliasp/patch-1
Typ (creatation → creation)
2016-04-20 14:52:14 +01:00
Elias Probst
9f1762873a Typ (creatation → creation) 2016-04-20 13:08:31 +02:00
Matt Holt
c3417a0757 Merge pull request #772 from mholt/fix-browse
Make Browse Great Again ★★★
2016-04-19 10:16:08 -06:00
W-Mark Kubacki
72bc6932b0 browse: Jail the root directory using http.Dir() 2016-04-19 03:48:27 +02:00
W-Mark Kubacki
a41e3d2515 browse: Use JavaScript to localize the shown datetime
The datetime format is whatever the user sets in his/her browser or OS.

This converts timezones.

Tested with Chrome 50 and Internet Explorer 11.
2016-04-19 03:48:27 +02:00
W-Mark Kubacki
7f35600b28 browse: Emit datetime in UTC instead of the server's timezone
Makes sure the view is the same no matter where a site is hosted.
2016-04-19 00:56:47 +02:00
W-Mark Kubacki
cc6aa6b54b browse: Remove whitespace from template's output, annotate output
Fixes a surplus — next to "go up".

Identifies the preamble as the table's summary.

Emits filesizes in bytes, which can be consumed by any browser-side scripts
or utilized in sorting when the table is copy-and-pasted into a spreadsheet
software.

Uses <time> along with proper datetime representation, which a browser could
utilize to display the datetime rendered according to the requestor's locale.
2016-04-19 00:20:44 +02:00
W-Mark Kubacki
239f6825f7 browse: When sorting by size, offset directories
Assigns negative sizes to directories in order to have them listed reliably
before any zero-sized files. That order is what most users expect when
sorting by size.

As side effect directories will appear before files on all filesystem
implementations. To give an example: before this change directories had a size
of 4 KiB when using Linux with ext4 or tmpfs, and with ZFS a size resembling
an estimation of the number of leaves within said directory.
2016-04-19 00:20:44 +02:00
W-Mark Kubacki
1d38d113f8 browse: Move predicate 'limit' to ServeListing
This keeps the interface of all available formatters honest,
and allows for truncated listings all formats.
2016-04-19 00:20:44 +02:00
W-Mark Kubacki
6908136092 browse: Split ServeHTTP into small specialized functions 2016-04-19 00:20:44 +02:00
Matt Holt
da016f8d5a Merge pull request #760 from Burmudar/hostname-placeholder
Add hostname placeholder. Headers use replacer
2016-04-18 15:54:31 -06:00
W. Mark Kubacki
2f2d357fb6 browse: Fix known bugs (#770)
* browse: Catch the case of a directory disappearing before having been read

* browse: Revert to old pass-through behaviour

PROPFIND is a request for an alternate view on a directory's contents, which
response is indeed not implemented but ideally allowed to ask for.
OPTIONS would ideally return (at least) what methods the requestor could use,
which is an allowed request method, too.

This addresses #767.
2016-04-18 11:42:36 -06:00
Matthew Holt
924b53eb3c Minor changes 2016-04-18 09:43:28 -06:00
Tobias Weingartner
2b51be7fd7 Better error message. (#768)
* Fix PrivateKeyBytes to error out and fail tests on error.

* Better error message.
2016-04-17 15:15:14 -06:00
Tobias Weingartner
376e1090a3 Fix PrivateKeyBytes to error out and fail tests on error. 2016-04-17 14:17:42 -06:00
elcore
a682100c5e Merge pull request #765 from mholt/elcore-error-ecdsa
Error if we are unable to marshal the ECDSA private key
2016-04-17 18:40:02 +02:00
elcore
aba3d37c88 Error if we are unable to marshal the ECDSA private key 2016-04-17 18:19:41 +02:00
W. Mark Kubacki
0890e330e2 Merge pull request #759 from mholt/paths-cleanup
Tell usage of 'path' from 'filepath' and fix *path checking
2016-04-17 14:16:57 +02:00
W-Mark Kubacki
c05c5163e2 browse: Don't leak Cookies to sessions in HTTP from HTTPS 2016-04-17 01:16:17 +02:00
W-Mark Kubacki
3513b6f2f7 fileserver: When out of filedescriptors, spread retry attempts 2016-04-17 01:16:17 +02:00
W-Mark Kubacki
4a6121f989 Tell usage of 'path' from 'filepath' and fix *path checking 2016-04-17 01:16:16 +02:00
William Bezuidenhout
a762bec06d Add hostname placeholder. Header uses replacer
On matched header rules, replacer is used to replace any placeholders
defined in header rules iex. X-Backend {hostname} where {hostname} will
be replaced by the hostname key present in the replacer

hostname key added to replacer. The value is determined by the output of
`os.Hostname()`
2016-04-16 19:06:39 +02:00
Matthew Holt
b75016e646 Fix lint warning 2016-04-15 15:13:44 -06:00
W. Mark Kubacki
ddf4b1fd3b Merge pull request #757 from mholt/extend-tls-client-auth
Extend tls client auth
2016-04-15 22:57:45 +02:00
W-Mark Kubacki
69c2d78f69 Support configuring less restrictive TLS client auth requirements
Caddyfile parameter "clients" of "tls" henceforth accepts a special
first modifier. It is one of, and effects:

 * request         = tls.RequestClientCert
 * require         = tls.RequireAnyClientCert
 * verify_if_given = tls.VerifyClientCertIfGiven
 * (none)          = tls.RequireAndVerifyClientCert

The use-case for this is as follows: A middleware would serve items to the
public, but if a certificate were given the middleware would permit file
manipulation.

And, in a different plugin such as a forum or blog, not verifying a client
cert would be nice for registration: said blog would subsequently only
compare the SPKI of a client certificate.
2016-04-15 22:21:55 +02:00
W-Mark Kubacki
f31875dfde Move sanitization of URL.Path to Server
No need to have this in every plugin. And, even in flat filesystems
filenames with dots and slashes are best avoided.
2016-04-15 21:02:36 +02:00
W-Mark Kubacki
4e98cc3005 browse: Return HTTP errors on unhandled HTTP methods
For example, a HTTP POST should not be answered with StatusOK,
and a response to HTTP OPTIONS should not carry any contents.
2016-04-15 18:22:51 +02:00
Matt Holt
d3a77ce3c3 Use binExt 2016-04-13 15:21:18 -06:00
W. Mark Kubacki
b149a86bc2 server: Rotate TLS ticket "keys" (#742) 2016-04-12 10:09:45 -06:00
Matt Holt
ac80f6edc3 Merge pull request #744 from tboerger/feature/breadcrumb
browse: Dropped LinkedPath and updated browse template
2016-04-12 10:01:01 -06:00
Thomas Boerger
ef95173827
Dropped LinkedPath and updated browse template
As discussed with @mholt I have dropped the old LinkedPath function and
replaced it within the browse template with the new BreadcrumbMap
function. Visually it looks exactly the same as before, now the template
functionality is just more powerful.

Signed-off-by: Thomas Boerger <tboerger@suse.de>
2016-04-12 17:18:21 +02:00
Matt Holt
36a3e204b6 Merge pull request #740 from tboerger/feature/browse-breadcrumb
Added breadcrumb map function to browse
2016-04-11 14:08:18 -06:00
Thomas Boerger
e0b63d92f4
Added breadcrumb map function to browse
In order to being able to really build a custom template for the browse
directive I have added another function to build even custom breadcrumb
paths. The other function `LinkedPath` is not that easy styleable as
this map function. That way we are able to build the breadcrumb path
matching different CSS frameworks like Bootstrap.

Signed-off-by: Thomas Boerger <thomas@webhippie.de>
2016-04-11 21:47:44 +02:00
Thomas Boerger
004a7f84ef Make test case less dependent on exact error string (#741) 2016-04-11 13:08:30 -06:00
Matt Holt
ed8a48e7f1 Merge pull request #739 from tboerger/feature/browse-go-up
Added "go up" link to browse template
2016-04-11 12:28:13 -06:00
Thomas Boerger
c86c26a056 Added "go up" link to browse template
In order to have directly a link within the browse listing I have added
a link to the top of the table to get one level up in the tree. Added
that after a chat with @mholt.

Signed-off-by: Thomas Boerger <thomas@webhippie.de>
2016-04-11 20:10:41 +02:00
Matt Holt
0a7ca64f53 gzipResponseWriter should also be a Flusher
To be consistent with 3faad41b43 and c64cf218b0
2016-04-11 00:24:26 -06:00
Matt Holt
c4e2cf96e7 Merge pull request #737 from tw4452852/closeNotifier
http.CloseNotifier implementation for http.ResponseWriter wrapper
2016-04-11 00:18:06 -06:00
Tw
c64cf218b0 http.CloseNotifier implementation for http.ResponseWriter wrapper
Signed-off-by: Tw <tw19881113@gmail.com>
2016-04-11 14:10:33 +08:00
Matt Holt
6a7b777f14 panic if not a Flusher
Caddy recovers panics that occur in the middleware stack so this is not a risk to process termination. This way is also preferable to hiding the error. See 3faad41b43 (commitcomment-17035158)
2016-04-09 22:11:36 -06:00
Matthew Holt
67b137175e
Replaced automate.sh with Go program 2016-04-09 10:02:16 -06:00
Matthew Holt
dfa3b8645d
Who uses 32-bit Mac anyway. :P 2016-04-09 00:40:37 -06:00