- With refactored code, introducing new kinds of actions
is very few lines of code. Allow disabling targets
(services was already possible). Allow masking units,
but as a special case require the complete name.
FIXES#975
- The three steps of modifying services in the target
system do basically the same thing, so factor out
the loops and logging into a systemctl() function.
- Log to warning() instead of just debugging, on failure.
- There is more to failing out of loadModules() than just
emitting modulesFailed, so instead share the failure
code with the code after loading modules -- but don't load any.
- Module dependency-checking is done in two phases:
first, catch any unknown modules that are listed
in *requiredModules* and bail out before loading
anything. Second, check that the modules required
by X occur before X in the sequence.
- The value set in module.desc was never stored for use,
but isn't an attribute of the instance, either. It belongs
with the descriptor, in ModuleManager.
If USE_<foo> is given a value that doesn't match **anything**,
then bail out. Since USE_* is an explicit distro choice for a
specific implementation, it's an error if that implementation
is not there.
When there are multiple modules doing a thing and it really only
makes sense to have one of them in a given Calamares compilation,
the USE_<foo> variables allow you to select one, while ignoring
all the other implementations. If USE_<foo> is not set, all
implementations are included (as usual).
- Use YAML-CPP API for finding out if a node has a value at all.
- Asking for Type() of an undefined or NULL node throws an
exception, so the existing code didn't **actually** catch
cases where a required setting wasn't set at all.
- A valid line (as explained in the comments at the top of
the locale.gen file) is <locale> <encoding> (two fields),
so lines with more than two fields can't be valid locale-
listing lines. For them, pretend they name locale "",
which won't be matched.
- Improved debug-logging
- Fix the actual problem of listing locales more than once,
by listing them all, uniqified, at the end, with an explanitory
comment in the generated file.
- Be more accepting of what constitutes a locale-line; this allows
spaces before and after the `#` comment sign, but because we're
uniquifying, this doesn't cause duplicates.
- Because we write the enabled locales at the end, the full file
comment-header is retained un-mangled (instead of accidentally
enabling a locale mentioned as an example there).
Testing for existence of a file in the live system, and then
copying it in the target system, is not a recipe for success.
- Fix the restore-from-backup part.
- Document that your live and target system must both have
/etc/locale.gen if you want this to work at all.
Also make install for yum and dnf follow the documented syntax: options
(-y) before the command (install), even though yum and dnf also accept
the other order. This also makes it consistent with remove.
Untangle the shortcuts; Create and Cancel had an overlap.
Skip 'r' (Revert all changes) and 'e' (Edit) and settle on
'a' (which might also mean "Add").
FIXES#977
Introduce the notion of emergency modules and emergency jobs.
Initial use will probably center around the preservefiles module,
and possibly umount.
FIXES#928
- After a failure, skip non-emergency jobs.
- After running all emergency jobs, then emit failure message.
- In log, distinguish emergency and non-emergency jobs.
Any job can be an emergency job; emergency modules spawn
emergency jobs (but conversely, a non-emergency module
can spawn an emergency job explicitly).
A potentially emergency module is one that has EMERGENCY
(in CMake) or emergency: true (in module.desc) set.
Any such module must also set emergency: true in the
configuration of the module. This is to allow for
instances of a module that **don't** run as emergency
modules, alongside actual emergency ones.
- In many cases, using QLatin1String is a de-optimization, when
applied to a C string literal. Kevin Kofler pointed out that
those should basically all be QStringLiteral, instead. (Compile
tests with -O3 show that in the optimized object file, the
code size difference is negligible).
- Drop the explicit constructor entirely in cases where we're calling
QProcess::execute(), for consistency.
- Do a little less messing around in the mapping of keyboard locales
to keyboard map names.
As Kevin pointed out, there's an extra conversion involved here --
although with -O3 the difference boils away leaving only a call
to a from-ASCII helper or a from-Latin1 helper.
While here, coding-style.
The plymouthcfg Calamares module is optional. Distributions which
write filesystems with a full plymouth configuration won't even
want to use it (see plymouthcfg docs).
However, now grubcfg depends on plymouthcfg to run because
the globalstorage value to trigger setting 'splash' in grub,
is set in the plymouthcfg module.
Just check for plymouth existence separately in the grub module.
Fixes ea1c8a0e5d
Since this is a new language, it is currently 0% translated.
That is why it goes into _tx_bad. It will move to one of the
other categories once some translation has happened. Add the
(still empty) Transifex files already.
- Add a TODO for allowing modules to come from somewhere other
than the module loader (this would allow "internal" modules
that are always present)
- Warnings are warnings
- Add -v (verbose) and -b (load via bytearray)
- Verbose prints the keys read from the file,
- Bytes reads via an indirection through QByteArray, like Settings does
- Collect the failed modules, instead of bailing out on the first one
(this also prevents crashes caused by quit() called from a timer).
- Introduce a slot to report on failed module loading (no UI yet).
- Fudge the numbers in the North, to improve location
of the pins and lines of latitude.
- Inuvik, Yellowknife, Cambridge Bay, Resolute look ok
- Thule, Scoresbysund look ok; Danmarkshavn a pixel or so too far North
- Reykjavik is a bit too far North
- Longyearbyen is a bit too far North
Since these places are off by one or two pixels, this becomes
invisible when a large pin + text label is placed on it.
The scaling on the map was a little off; the degrees of latitude
are a little wider there than around the equator and Europe.
- Johannesburg is in the right spot
- Hobart is no longer a suburb of Melbourne
- Punta Arenas is in Chile
Replace pin and text label with just a dot (to pinpoint where
locations are) and draw latitude lines on the globe when
DEbUG_TIMEZONE is set at compile time. Since there's probably
still timezone-related bugs (in particular in the images that
map points on the globe to timezones), leave this in the codebase.
This is orthogonal to the SKIP_* mechanism already documented
for avoiding modules. If the module is enabled, but its dependencies
are not present, don't bother building the module. This follows
e.g. plasmalnf as an "avoidably heavy dependency".
Related to a misplaced comment in ISSUE #956
This test-application should load a single module and execute it --
that can be used to quickly test configurations, loading, etc. This
is preparation for loading all sorts of Python modules by PythonQt.
The loader does some internals initialization and gets the module,
but doesn't actually run it yet.
- Add a *userShell* key, which can be left out (default, backwards-
compatible) to retain the old /bin/bash behavior, or explicitly
set to empty to defer to useradd-configuration, or explicitly
set to something non-empty to use that shell.
- This is prep-work for #964, which was caused by #955
- Original assumption was that distro's would have a working
useradd configuration; @abucodonosor already pointed out that
this was probably not the case, but I ignored that.
- more flexible way to keep (all kinds of) files from the host
system, into the target system.
- WIP: substitutions like in shellprocess (@@ROOT@@, @@HOME@@ probably)
- WIP: creating a JSON file from global settings
Extensive go-over on the partitioning code. #622 is maybe "possibly fixed",
but there's no real indication of what constitutes an invalid combination
of flags.
FIXES#884FIXES#951FIXES#953FIXES#622
The substitution of @@ROOT@@ should happen when running in the
host, not in the target, system. Also only complain about it
if @@ROOT@@ is actually used in the commands.
FIXES#954
- If there is a partition already (newly) created, then pass that
to the dialog so that it can use the setings previously applied
(e.g. mount point and flags).
- This avoids the case where you create or format a partition,
then click on it again to edit it and the previous settings are lost.
- Setup the lsit of flags consistently, by providing the available
and to-be-checked flags.
- In CreatePartitionDialog, assume that ~0 is all the flags.
This file is full of helper functions for the partition-editing
dialogs. At first it was just mount-point helper functions,
but there is other functionality that can be refactored.
- If we're changing the flags to enable EFI boot, then that's
enough to satisfy the (future) EFI bootability check.
This is for issue #622 as well. Fixes#884.
- Use the desired (future) flags, if set, to initialize the
flags checkboxes. If there are no future flags set, this
returns active flags as before.
- This fixes the situation where editing a partition, changing
flags, then editing it *again* re-starts with the original
flags instead of the modified flags.
- PartitionInfo maintains information on "what is desired" for
a given Partition. Now we can set desired flags, alongside
the flags already supported by Partition (where activeFlags()
gives you the flags currently set on that partition).
- Avoids case where you edit a partition with a mountpoint
set; previously, calling setText() would update the text
but leave the selected index unchanged (usually 0), so that
later calling selectedMountPoint() would return empty.
- Move to one place which handles the standard mount points
- While here, introduce explicit "(no mount point)" string
into the combo box. This is prep-work for issue #951.
- Some locales have no nativeLanguageName(), so instead display
the locale id (e.g. "eo") and the resulting language in English
(which, if it is really unsupported, will be "C").
- Include all headers for types that need to be fully-defined
(e.g. return types). This guards against uses in contexts where
those headers have not been implicitly or previously included.
FIXES#948
QLocale::name() doesn't include script information, and if it did
it would probably use sr_RS@Latin; when searching for translation
files it won't consider dropping just the country.
- The QLocale constructor which takes a string (locale name) doesn't
understand sr@latin, and returns the Cyrillic locale. Fix that
by creating locales ourselves for @latin locales.
- sr and sr@latin now display correctly in the right script in the
native language dropdown.
- The (RTL) text "Arabiy (Misr)" should be entirely RTL, so
make the parenthetical insert -- which would otherwise be LTR
and so mess up the placing of those parenthesis around the country --
explicitly RTL.
- Since there are no RTL languages in Calamares right now with
country-local translations, this isn't visible.
- A locale suggests it is country-specific by having the form <lang>_<country>
- This mostly fixes locale "ar" being presented as "Arabiy (Misr)" when
there is no need to (and the RTL is messed up then, too).
- Introduce intermediate data class for building up the list
of languages to present.
- Sort on the English names, with en_US at the top (ugh).
- Show the native names.
- Much like std::find_if, but slightly muddled because there's
no iterator that we can sensibly use.
- Scan the ComboBox for a locale that matches a predicate.
- Log more as the search for a good locale progresses.
- Don't mix matching the locale with filling the ComboBox
(even though that's slightly more efficient).
- Since these tests use network resources, they are not enabled by default.
Set the environment variable TEST_HTTP_GET to actually do them.
- Do one request for each provider and check that they are all consistent.
(This works for me, yielding Europe/Amsterdam for all).
- Some providers don't provide a single flat JSON object
(e.g. "{time_zone: foo}") but a nested structure
(e.g. "{location: {time_zone: foo}}"), so allow dots
in the selector to do multi-level selection.
- Some providers return weirdly escaped data; strip out useless
escaping before splitting (there are no characters in correct
time zone names that need escaping)
- Add some tests for TZ splitting
- In GeoIP handler constructors that take a string (to configure the
selector to use), interpret the empty string (which generally isn't
a meaningful selector) as meaning "use the default".
- Drop the no-argument constructors in favor of a default-argument
which is empty.
- GeoIP gets a string selector; the interpretation is up to derived classes.
- GeoIPXML and GeoIPJSON use the selector to select an element by tag
or an attribute, respectively.
- The handler for JSON data should be called that, not named
specially after the original provider it was implemented for.
- Make filename and classname consistent, GeoIPJSON.
- The screenshot stays one size, but different ThemeWidgets may overlap
partially when you shrink the screen or have more than three / four
themes listed.
- Probably needs work in the surrounding container and overall better
page-scrollbar support.
- Unchanged config files will continue to use the weird addition
of /json, and interpret JSON data.
- Allow to specify full URL with data format through one of
geoipStyle: json
geoipStyle: xml
- XML support is optional
- Don't insert a space before the output of a process
- To do this, suppress space and quoting on the output, and to do
that move the labeling-output for warnings and errors into
the constructor (so that an idiomatic .nospace() does the right thing).
- Settings is just a settings class, no UI involved, so
move to libcalamares where it can be used also from
system helpers.
- YAML utilities are useful at a lower level of the stack, too.
- Put the (constant) 'Calamares will now quit' on its own debug line.
- Tell the user what the search paths are if a module is not found
(prompted by a mis-configuration in a Neon live image).
- Make the BCP47 value explicitly lower-case.
- Add some constness and encapsulation.
- Fix up documentation in the packages module explaining the
format of the ${LOCALE} replacement (now forced to lower-case,
but it is also only the language part, not e.g. en-UK).
FIXES#922
- Count only the packages that will be changed, given the current
locale settings.
- Preserve global storage unchanged (don't remove any locale-packages).
- Use DebugRow for one-row-at-a-time output with continuations.
- Use DebugList for one-item-per-line with continuations.
- Use DebugMap for one-row-at-a-time output of a QVariantMap.
- This enables working in three modes:
- No themes listed; all are shown without screenshots,
- Themes listed, showAll false; only those are shown,
- Themes listed, showAll true; the installed-but-not-listed
themes are shown after the listed ones, and have limited info.
- Although it's not necessarily accurate for an extensively-modified
Plasma configuration, we can read the Look-and-Feel from the
configuration files. Allows auto-detection.
- For OEM modes where there is already a theme, add a preselect:
key to pick a specific theme and have that one come up as already-
selected in the list.
- Don't re-run the lnftool if an already-selected theme is clicked
again. Use toggled() instead of clicked().
- Don't bother with the address of the main thread
- Do put a marker on restart into the log file
- Do put the Calamares version into the log file (previously, the version
was printed through cDebug() before the log file was opened, so it was
lost to the on-disk log).
Documentation lives both in Calamares and in the calamares-branding
repo. Whether a slideshow has navigation arrows or not is up to the
slideshow / branding component author, but Calamares now ships some
support QML to make that easy.
FIXES#841FIXES#904
It is the distro's responsibility to produce screenshots that look
good; previously I chose to preserve the aspect ratio on the grounds
that this would keep the look of the screenshot even if the distro
had done one in a weird size. This makes the screenshot part
of the LNF selection look weird, though, since then you get
blank parts.
Switch to ignoring the aspect ration; distro's should produce
screenshots in a 12x8 (i.e. 3:2) aspect ratio, preferrably at
least 120x80 pixels -- but keep in mind hiDPI and the default
font sizes of the distro, which may make other sizes look better.
(this follows discussion with BlueStar Linux)
- Auto-advance the default presentation
- Add more example slides to the fancy presentation
- Expand README.md explaining what the default classes can do
Removed these features that make sense in a presentation slideshow
(e.g. during a talk) but that are potentially confusing during
a passive slideshow like Calamares has:
- Using 'c' blanks the slideshow.
- Entering a slide number + enter changes slides.
- Add a NavButton, which shows a directional arrow, and fades in on hover.
It can be used left- or right- by setting an image source
and click handler.
- Specialize NavButton to Forward and BackButton.
- Add a SlideCounter navigation aid.
the original code does not distinguish the document comments inside the locale.gen file from the real locale list. The language was then enabled from the header comments of the file instead of the correct value in the list.
The new code verify tha the complete locale string is just after the first character of the string, enablig only the correct value of the locale list.
An example:
# en_US.UTF-8 UTF-8 --> document header, should not be enabled
#en_US.UTF-8 UTF-8 --> correct section to enable
Related to this request:
https://code.chakralinux.org/tools/calamares-chakra/issues/2
The install-bits branch commit 83639b182b
dropped .so-versioning for libcalamares and the creation of the Python-
support symlink. This broke KDE Neon dev-unstable because the embedded
Python can no longer find libcalamares.
Installing unversioned .so's straight to LIBDIR is also not a good thing
(according to Debian), so revert to the original scheme with versioned
.so and a Python-support symlink.
Medium-term fix is to install unversioned straight into LIBDIR/calamares
and fix up the RPATH for the executable.
- Applies to libcalamares and libcalamaresui.so, install with no
version, just the bare .so. Since Calamares doesn't do versioning
anyway, and its plugins should be re-compiled for any change,
putting them in lib as unversioned .so's should make Calamares
happy and silence lintian.
- The Python testmodule script can end up calling in to System
methods (via System::instance()). This is unusual, and the
System instance has not been created at that point.
Now, create an instance and warn about it.
- ValueCheck shouldn't own the pointer, since it's just a QPair
and there are temporary copies made (e.g. in
ContextualProcessBinding::append() ) and we get double-deletes.
- Do deletion by hand; going full unique_ptr would be a bit overkill.
- Re-build the structures for doing value-checks, is now more tree-like.
- Document pointer ownership.
- Introduce wildcard matches ("*")
- Don't drop empty command-lists, since they can be used to avoid
wildcard matches. (E.g. "in this case, do nothing, but don't
fall through to wildcard").
- Since the image size isn't known a priori (due to sizing based on fonts),
load the image and then resize in all code paths.
- Use the right resizing flags.
- .. and actually use the resulting scaled pixmap.
Thanks to Jeff Hodd.
- If the next step will be an install-step (e.g. hit the optional
confirmation step) then change the text on the 'next' button to
'install'.
- Do a little refactoring to make that more pleasant.
FIXES#905
- Move logging-levels to an enum
- (re-)Order logging-levels so that the normal debug statement is
not the most-important (lowest level).
- Drop using namespace std;
These additional pointers were introduced for translations,
and needed their own tricks to get lupdate to recognize the
strings. Using QCoreApplication::translate() removes the
need to a QObject to provide context. Drop the now-unneeded
parameters.
Instead of using tr and some macro hacks to get lupdate to
recognize the translation, instead use QCoreApplication::translate()
which takes its own context for translation.
- For both shellprocess and contextualprocess, add a top-level key
"timeout" that defaults to 10 seconds (which it already did).
- Allows setting "global" timeout for command-lists, while still
allowing individual timeouts per-command.
- Setting timeout per global variable in contextualprocess is not
supported; that would restrict the possible space of comparisions,
while not supporting a global setting timeout seems reasonable enough.
Use instances if you need wildly variable timeouts and don't want to
set them individually.
- Replace plain StringList with a list of <String, timeout> pairs,
and run that instead. All code paths still use the default 10sec
timeout and there's no way to change that.
- Copy stdout from timed-out process into the output variable,
instead of just dumping it into the log file. This will
improve the user experience, too, because they will get some
feedback / explanation of what the process has done.
- add license file from libpwquality for provenance
- translate pwquality_strerror() into the PWSettingsHolder convenience class
- use Qt translations, since we'd otherwise also have to wire up, and
wire in, libpwquality gettext translations.
- Use shared_ptr and a helper class to hide away raw pointer use
from libpwquality. Provide a convenience C++ API.
- Simplify configuration through helper class.
Use the samegame example from the Qt Quick demos as a branding "slideshow".
Instead of watching slides go by, you can play samegame! Click on
"new game" to start, and then click on groups of same-colored balls to make
them go away -- at least two same-colored balls must be touching.
Once the exec step is done, the game vanishes automatically.
This is an additional example for #841
- Remove some superfluous intermediate defines
- baseFactory was not used (always Calamares::PluginFactory)
- Move DECLARATION and DEFINITIONS apart
- CALAMARES_PLUGIN_FACTORY_DEFINITION was redefined (identically)
- CALAMARES_PLUGIN_FACTORY_DECLARATION was redefined (identically)
- __VA_ARGS__ was constant
- Issue asks to make the setting more visible, which seems sensible to
me. It **is** kind of hidden away for those distro's that make
the setting visible (not everyone does).
- While here, add a tooltip explaining what it does.
FIXES#893
Allow running one or more commands based on the value of
a global configuration variable. This could, of course,
be done in a Python module with some custom code,
but for simple cases this is more straightforward
to configure through module instances.
Uses the CommandList developed for the ShellProcess
module to do the actual work.
FIXES#874
- Move CommandList so it can be used from more modules than
just ShellProcess
- Allow a CommandList to run itself. This centralizes
code for executing one or more commands and simplifies
the ShellProcess module.
Various small cleanups:
- mention instance id in log message
- code formatting / style
- This turns off the space-available check in the welcome module;
without libparted, always fail that check.
- Allows running the welcome module on OS without libparted.
- Also allow a single string instead of a list
- Add count() method to CommandList
- Drop over-engineering, add more logging
- Expand tests with some more examples
This is basically dummyprocess, except with an expanded configuration
interface so you can run 1 or more shell commands in the live
or target system with a suitable configuration file and instance
of shellprocess in settings.conf.
It can replace downstream modules that implement their own
process modules with a command, by an instance of shellprocess.
Back targetEnvCommand() with a more general runCommand()
that takes an argument selecting the location to run
the command in. This allows us also to use the same
API for running processes in the host during install,
as we do for running them in the target system.
One reason for this change is wanting to run (user-specified)
commands and independently from the global dontChroot setting,
run those commands in the live system or the target.
This changes the ABI of the DLL, since targetEnvCommand()
is no longer exported. Plugins will need to be recompiled.
- refactor targetEnvCommand() into more general runCommand().
- While here, allow host system commands to run even if
there is no global storage.
- provide convenience accessors for ProcessResult members
- Move explanation of process errors out of ProcessJob
- Move from ProcessJob to ProcessResult, so it can be
reused outside of ProcessJob (e.g. from ShellProcessJob).
- Add some convenience functions, too.
Make a function out of explaining-skipped-modules, and call it
not only after collecting all the modules, but also after
the feature summary, so that it's quite clear which modules
are skipped.
Scenario is this: you have no suitable installation devices on
your system (everything is mounted, or HDD has died), click through
to partition page, where you have all the buttons available, but no
devices in the list. The following actions then cause a crash:
- clicking "back"
- clicking any button
Prevent that:
- you can click "back", but if there is no device selected
nothing happens to the device state (no nullptr deref,
and no crash)
- button code is now more resilient to this scenario
- buttons are hidden until a device is available, so you
can't even click on them to trigger the code.
- remove hide-close-button hack
- refactor code in viewmanager for confirming quit
- hook up confirm-and-quit to WM close button
- also works for alt-F4 and other quit methods
- while here, update copyright year
FIXES#870
Modules may be skipped for different reasons: SKIP_MODULES
is the traditional approach to suppress some, but other modules
may have unmet build requirements (e.g. Plasma Look-and-Feel,
or the Partitioning module) and should be able to opt-out
of being built. For all those skipped, log it explicitly after
all the modules have been examined.
Only CMake-based (e.g. C++) modules support opting-out in this way.
This is meant to run one or more jobs based on specific global
configuration values; if could also be done by a Python
module with just some if's, but this one can be used with
just the config file and covers a bunch of use-cases.
- calculate a hash of the filename, and use that
- makes it possible to distinguish different screenshots
even when the image file is missing / badly configured
- most colors will be dreadful
Patch by Gabriel C. (@abucodonosor).
- use libcalamares functions
- no need to copy /etc/adjtime over we can run hwclock in chroot
since we have /proc , /sys , /dev , /run/* already bind mounted.
- added RTC and ISA probing methode ( see issue #873)
- we probe default from /dev/rtc* ,
- fall back to ISA
- still doesn't work we just print a BIOS/Kernel BUG message and continue
- NOTE: issue #873 is about broken ArchLinux kernel config but there
are HP boxes with real RTC problems no matter what kernel config
is used so let us be nice and don't error out..
FIXES#873
