Implement new ubus rpcd backend calls for later use in the frontend:
- netdevs: dump information about Linux network devices present
- boardjson: dump /etc/board.json if present
- offload_support: query whether netfilter offloading is supported
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Properly handle ubus POST requests exceeding the default chunk size
and fix a possible nil dereference when rejecting incoming requests
due to bad JSON message framing.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Introduce a new tools.prng JS library which implements the musl libc
srand() and rand() calls to produce identical pseudo random number
sequences in the browser which is needed for the string to color
conversion later on.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Refactor L.require() to use L.raise() instead of L.error() to signal
class loading failures. This allows callers to handle class loading
errors in a graceful manner.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Factor out an L.raise() function out of L.error() which constructs and
throws an exception object.
Rework the remaining L.error() function to internally use L.raise() to
construct exceptionts to render.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
- Do not artificially cutoff dropdown items, use all available space
- Close open dropdown when clicking into the preview area
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Now that LuCI.Request is transparently coalescing requests, there
is no need for the rather cumbersome batch()/flush() api in rpc.js.
Also refactor the code to get rid of the rpcRequestRegistry indirection
and remove superfluous promises which simplifies the code a lot.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Extend LuCI.Request to automatically coalesce subsequent requests
to ubus resources into single batch requests.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Implement setters and getters for the base URL prefix used and also
implement a session ID getter while we're at it.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Introduce a new LuCI.Poll class which is able to repeat any
promise based function and not strictly tied to HTTP request
semantics.
Also rework LuCI.Request.Poll and XHR.Poll as well as
LuCI.start(), LuCI.stop(), LuCI.halt() etc. to redirect to
the new api.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
- Fix discovering base url if cache buster is appended to luci.js href
- Fix extracting require tokens in minified sources
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Add a new client side form.js library which is a more or less direct
reimplementation of the Lua side cbi.lua in JavaScript.
Due to its client side nature, it supports a number of features which
would be hard to realize on the server side, such as drag&drop sorting,
modal sub-map dialogs, reload-free editing etc.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Do not render standard widgets like checkboxes, select boxes,
text input fields etc. on the server side anymore but utilize
the ui.js primitives instead.
This avoids logic duplication between server side cbi templates
and JS widgets in the future.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Implement further widget primitives like text inputs or checkboxes and
support custom validation callback functions when instantiating CBI
validators.
Also add support initializing ui.js widgets from the "data-ui-widget"
HTML attribute.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Introduce a new view() target for CBI dispatch nodes, as long with the
required template and plumbing work in luci.js to allow requiring view
classes.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
If the ui class is loaded, use its modalDialog facility to display runtime
errors, alternatively render them into the maincontent area.
Also prevent duplication of stack trace information and throw a low level
error on session expiration to allow higher layers to properly handle it.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Add a new rpc.js class which provides low level facilities to exchanges
messages with the ubus rpc endpoint.
Also introduce a new uci.js class which provides client side uci
manipulation routines.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Add an admin/ubus route mimicking the native uhttpd-mod-ubus protocol.
The main difference to the native protocol is that this gateway requires
no additional per-object/procedure ACL setup on the router side and that
it is located under the same prefix as LuCI itself, allowing the reuse
of the session login cookie.
This route is meant to be a transitional mechanism until client side
RPC calls are eventually migrated to uhttpd-mod-ubus completely.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
In order to prepare the move of the dynlist and dropdown widgets into
the LuCI ui class, remove the direct calls to cbi_d_update() and replace
them with custom events instead.
Extend cbi_init() to handle these custom events and to invoke
cbi_d_update() when receiving them.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Add a new function L.error() which creates and throws a custom
error object with stack information and given type.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Use the new class loader infrastructure to move gui specific
functionality out of the luci.js core and dispatch a new event
'luci-loaded' which is fired once all external classes have
been fetched.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Add a fetch() inspired HTTP request utility class to luci.js and
replace the old xhr.js class with a stub using the new request api.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
The former name "Flashmemory write access" is wrong. The triggers also
indicate read/erase access to the flash memories.
Signed-off-by: Martin Schiller <ms@dev.tdt.de>
* fix escaping of magic lua chars (#2800)
* fix redundant second gsub line in shellstartsqescape function
* fix return value of shellstartsqescape function
Signed-off-by: Dirk Brenken <dev@brenken.org>
Modify the process status filter to allow also the processes
with a nice value to be visible on the Status/Processes page.
Filter out the top process itself.
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
commit 8a34a54b6aa6 ("base-files: use OPENWRT prefix for os-release
variables") changes the prefix of the os-release variables from LEDE_
to OPENWRT_. Use the new name.
Signed-off-by: Bjørn Mork <bjorn@mork.no>
When a value identical to the stored one is submitted, the CBI framework
will not emit an option write event and therfore not store the value in
the form data dictionary passed to SimpleForm.handle().
This usage pattern usally works be accident for file editor views such
as admin_system/crontab because \r\n windows style line endings are
substituted with unix \n ones before writing the data, defeating the
equality check in CBI.
When a single line without trailing newline is submitted however, the
CBI will not see a difference to the data stored in the file and clear
out the value on subsequent saves.
This commit alignes the logic used by various SimpleForm views to
behave identically and predictable:
- File data is handled in the SimpleForm.handle() callback
- The forcewrite property is used to disable equality checks
- Submission of an empty string empties the backing file
Fixes: #2737
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Currently only the dnsmasq lease file is read for hostname hint. If someone use only odhcpd as dhcp daemon the hostname resolution on wifi_assoc list is broken.
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Move duid_to_mac function from status.lua to sys.lua.
Implement translation from DUID-LL without header to MAC
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
The following changes are included:
* properly handle "Präfix" as neuter
* add missing hyphens to many appearances of "Präfix"
* fix spelling
* add few missing translations
Signed-off-by: Lars Kruse <devel@sumpfralle.de>
Currently the lookup loop is very inefficient. This improve the lookup loop by moving the entities to a separate array and adds them only one time. Every entities is checked 3 times, after that it will be moved to dns cache as a raw addres (not resolved). This also introduce the Type column to identify common packet using the source port.
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Since openwrt-18.06 the auto option is set/unset if we press
connect/disconnect on the interface page. So I think we should move this
to the general tab so that we see at once if this is started on boot or
not.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Currently on the system status page the the memory usage is shown in kB
only. If you have a system with 1 GB of Ram its hard to read the large
numbers. This automatically formats the byte numbers more human readable
into KB / MB / GB.
Submitted-by: Claudio Marelli <camarelli@gmx.net>
[squash commits, rewrap commit message]
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Switch the LuaSrcDiet download to direct cached SCM clones and
fix the tarball name while we're at it.
Supersedes: #2522
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
In some situation it is usefull to inform the use that this section
could not get delete, because this config is referenced or is in use.
Example pseudo code:
function s.remove(self, sid)
<check if secation could get deleted>
if output == nil then
self.error_msg = translatef("Unable to remove this section")
return nil
end
end
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Especially for systems without RTC this change has two advantages
* manual time sync with time server during configuration
* test possibility of the time sync over the configured servers
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
This checks if the installed hostapd supports SAE and allows to select
WPA3 Personal SAE in that case.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Even Freifunk was one of the major factory to create the LuCI-system, it's
now only a very small part of LuCI. LuCI has become a much bigger thing
and it seems that it's time to move the packages only relating to Freifunk
into it's own feed.
On the mailinglist it was discussed [1] and a repo below the general
Freifunk team on github was created.
This commit removes all packages that will be hosted in the new repo [2]
1 - http://lists.freifunk.net/pipermail/wlannews-freifunk.net/2019-February/004818.html
2 - https://github.com/freifunk/openwrt-packages
Signed-off-by: Sven Roederer <freifunk@it-solutions.geroedel.de>
Since the section labels are already HTML-escaped implicitely by the
striptags() function, we must not escape them again in attr() or
ifattr().
Fixes: #2524
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Extend the attr() and ifattr() template functions to take an optional
further parameter indicating that the passed value should not be escaped.
This is needed for cases where the input already is escaped through
other means, e.g. when the value was previously filtered through the
striptags() template helper.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Rework the regexp pattern to allow both IP addresses without mask
and IPv6 addresses with netmask instead of CIDR notation.
Fixes: #2495, #2530
Fixes: f6bfac211 ("luci-mod-status: rework iptables status page")
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Drop the "secret" value from RPC session objects in order to make them
compatible with ordinary web sessions used by the LuCI web interface.
That secret value was never used for anything and is the only difference
compared to normal LuCI login sessions.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
The get_all() function has changed and how it was used before no
longer works properly. These changes not only change to individual
get() calls, but actually adds to the readability of the code and reduces
file size.
Signed-off-by: pmelange <isprotejesvalkata@gmail.com>
In cbi.js there is an error which causes ipaddrport validation to always fail.
JS match() return the entire match as ret[0] and individual matches (for multiple ()) as the
subsequent list members. So we fix it by just fixing the index in the calls that want the
individual parts.
Signed-off-by: Daniel F. Dickinson <cshored@thecshore.com>
Update timezone data to 2018i
http://mm.icann.org/pipermail/tz-announce/2018-December/000053.htmlhttp://mm.icann.org/pipermail/tz-announce/2018-December/000054.html
2018h:
* Qyzylorda, Kazakhstan moved from +06 to +05 on 2018-12-21.
* New zone Asia/Qostanay because Qostanay, Kazakhstan didn't move.
* Metlakatla, Alaska observes PST this winter only.
* Guess Morocco will continue to adjust clocks around Ramadan.
* Add predictions for Iran from 2038 through 2090.
2018i:
* São Tomé and Príncipe switches from +01 to +00 on 2019-01-01.
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
Do not call cbi_init() from the browser field template but lazily initialize
the field, like it is being done for all other widgets as well.
Fixes: #2398
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Replace all XHR poll time number with -1 so they will use the default
poll interval time value from "/etc/config/luci". If this is not set then
5 seconds as default is used.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Make poll interval configurable. Use option from
----
config core 'main'
option pollinterval '5'
---
This will only be used if the interval option is less or equal zero.
If the poll interval is not configured then a default value of 5 seconds
is used.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
