You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
openmw/components/config/gamesettings.cpp

619 lines
21 KiB
C++

#include "gamesettings.hpp"
#include <QDir>
#include <QRegularExpression>
#include <components/files/configurationmanager.hpp>
#include <components/files/qtconversion.hpp>
#include <components/misc/utf8qtextstream.hpp>
const char Config::GameSettings::sArchiveKey[] = "fallback-archive";
const char Config::GameSettings::sContentKey[] = "content";
const char Config::GameSettings::sDirectoryKey[] = "data";
namespace
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
template <typename T>
QList<T> reverse(QList<T> values)
{
std::reverse(values.begin(), values.end());
return values;
}
}
Config::GameSettings::GameSettings(const Files::ConfigurationManager& cfg)
: mCfgMgr(cfg)
{
#if QT_VERSION < QT_VERSION_CHECK(6, 0, 0)
// this needs calling once so Qt can see its stream operators, which it needs when dragging and dropping
// it's automatic with Qt 6
qRegisterMetaTypeStreamOperators<SettingValue>("Config::SettingValue");
#endif
}
void Config::GameSettings::validatePaths()
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QList<SettingValue> paths = mSettings.values(QString("data"));
mDataDirs.clear();
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
for (const auto& dataDir : paths)
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
if (QDir(dataDir.value).exists())
{
SettingValue copy = dataDir;
copy.value = QDir(dataDir.value).canonicalPath();
mDataDirs.append(copy);
}
}
// Do the same for data-local
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
const QString& local = mSettings.value(QString("data-local")).value;
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
if (!local.isEmpty() && QDir(local).exists())
{
mDataLocal = QDir(local).canonicalPath();
}
}
QString Config::GameSettings::getResourcesVfs() const
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QString resources = mSettings.value(QString("resources"), { "./resources", "", "" }).value;
resources += "/vfs";
return QFileInfo(resources).canonicalFilePath();
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QList<Config::SettingValue> Config::GameSettings::values(
const QString& key, const QList<SettingValue>& defaultValues) const
{
if (!mSettings.values(key).isEmpty())
return mSettings.values(key);
return defaultValues;
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
bool Config::GameSettings::containsValue(const QString& key, const QString& value) const
{
auto [itr, end] = mSettings.equal_range(key);
while (itr != end)
{
if (itr->value == value)
return true;
++itr;
}
return false;
}
bool Config::GameSettings::readFile(QTextStream& stream, const QString& context, bool ignoreContent)
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
if (readFile(stream, mSettings, context, ignoreContent))
{
mContexts.push_back(context);
return true;
}
return false;
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
bool Config::GameSettings::readUserFile(QTextStream& stream, const QString& context, bool ignoreContent)
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
return readFile(stream, mUserSettings, context, ignoreContent);
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
bool Config::GameSettings::readFile(
QTextStream& stream, QMultiMap<QString, SettingValue>& settings, const QString& context, bool ignoreContent)
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QMultiMap<QString, SettingValue> cache;
QRegularExpression replaceRe("^\\s*replace\\s*=\\s*(.+)$");
QRegularExpression keyRe("^([^=]+)\\s*=\\s*(.+)$");
auto initialPos = stream.pos();
while (!stream.atEnd())
{
QString line = stream.readLine();
if (line.isEmpty() || line.startsWith("#"))
continue;
10 months ago
QRegularExpressionMatch match = replaceRe.match(line);
if (match.hasMatch())
{
QString key = match.captured(1).trimmed();
// Replace composing entries with a replace= line
if (key == QLatin1String("config") || key == QLatin1String("replace") || key == QLatin1String("data")
|| key == QLatin1String("fallback-archive") || key == QLatin1String("content")
|| key == QLatin1String("groundcover") || key == QLatin1String("fallback"))
settings.remove(key);
}
}
stream.seek(initialPos);
while (!stream.atEnd())
{
QString line = stream.readLine();
if (line.isEmpty() || line.startsWith("#"))
continue;
QRegularExpressionMatch match = keyRe.match(line);
if (match.hasMatch())
{
QString key = match.captured(1).trimmed();
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
SettingValue value{ match.captured(2).trimmed(), value.value, context };
// Don't remove composing entries
if (key != QLatin1String("config") && key != QLatin1String("replace") && key != QLatin1String("data")
&& key != QLatin1String("fallback-archive") && key != QLatin1String("content")
&& key != QLatin1String("groundcover") && key != QLatin1String("fallback"))
settings.remove(key);
if (key == QLatin1String("config") || key == QLatin1String("user-data") || key == QLatin1String("resources")
|| key == QLatin1String("data") || key == QLatin1String("data-local")
|| key == QLatin1String("load-savegame"))
{
// Path line (e.g. 'data=...'), so needs processing to deal with ampersands and quotes
// The following is based on boost::io::detail::quoted_manip.hpp, but we don't actually use
// boost::filesystem::path anymore, and use a custom class MaybeQuotedPath which uses Boost-like quoting
// rules but internally stores as a std::filesystem::path.
// Caution: This is intentional behaviour to duplicate how Boost and what we replaced it with worked,
// and we rely on that.
QChar delim = '\"';
QChar escape = '&';
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
if (value.value.at(0) == delim)
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QString valueOriginal = value.value;
value.value = "";
for (QString::const_iterator it = valueOriginal.begin() + 1; it != valueOriginal.end(); ++it)
{
if (*it == escape)
++it;
else if (*it == delim)
break;
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
value.value += *it;
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
value.originalRepresentation = value.value;
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
2 months ago
value = processPathSettingValue(value);
}
if (ignoreContent && (key == QLatin1String("content") || key == QLatin1String("data")))
continue;
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QList<SettingValue> values = cache.values(key);
values.append(settings.values(key));
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
bool exists = false;
for (const auto& existingValue : values)
{
if (existingValue.value == value.value)
{
exists = true;
break;
}
}
if (!exists)
{
cache.insert(key, value);
}
}
}
if (settings.isEmpty())
{
settings = cache; // This is the first time we read a file
validatePaths();
return true;
}
// Merge the changed keys with those which didn't
settings.unite(cache);
validatePaths();
return true;
}
bool Config::GameSettings::writeFile(QTextStream& stream)
{
// Iterate in reverse order to preserve insertion order
auto i = mUserSettings.end();
while (i != mUserSettings.begin())
{
i--;
// path lines (e.g. 'data=...') need quotes and ampersands escaping to match how boost::filesystem::path uses
// boost::io::quoted
// We don't actually use boost::filesystem::path anymore, but use a custom class MaybeQuotedPath which uses
// Boost-like quoting rules but internally stores as a std::filesystem::path.
// Caution: This is intentional behaviour to duplicate how Boost and what we replaced it with worked, and we
// rely on that.
if (i.key() == QLatin1String("config") || i.key() == QLatin1String("user-data")
|| i.key() == QLatin1String("resources") || i.key() == QLatin1String("data")
|| i.key() == QLatin1String("data-local") || i.key() == QLatin1String("load-savegame"))
{
stream << i.key() << "=";
// Equivalent to stream << std::quoted(i.value(), '"', '&'), which won't work on QStrings.
QChar delim = '\"';
QChar escape = '&';
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QString string = i.value().originalRepresentation;
stream << delim;
for (auto& it : string)
{
if (it == delim || it == escape)
stream << escape;
stream << it;
}
stream << delim;
stream << '\n';
continue;
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
stream << i.key() << "=" << i.value().originalRepresentation << "\n";
}
return true;
}
bool Config::GameSettings::isOrderedLine(const QString& line)
{
return line.contains(QRegularExpression("^\\s*fallback-archive\\s*="))
|| line.contains(QRegularExpression("^\\s*fallback\\s*="))
|| line.contains(QRegularExpression("^\\s*data\\s*="))
|| line.contains(QRegularExpression("^\\s*data-local\\s*="))
|| line.contains(QRegularExpression("^\\s*resources\\s*="))
|| line.contains(QRegularExpression("^\\s*groundcover\\s*="))
|| line.contains(QRegularExpression("^\\s*content\\s*="));
}
// Policy:
//
// - Always ignore a line beginning with '#' or empty lines; added above a config
// entry.
//
// - If a line in file exists with matching key and first part of value (before ',',
// '\n', etc) also matches, then replace the line with that of mUserSettings.
// - else remove line
//
// - If there is no corresponding line in file, add at the end
//
// - Removed content items are saved as comments if the item had any comments.
// Content items prepended with '##' are considered previously removed.
//
bool Config::GameSettings::writeFileWithComments(QFile& file)
{
QTextStream stream(&file);
Misc::ensureUtf8Encoding(stream);
// slurp
std::vector<QString> fileCopy;
QString line = stream.readLine();
while (!line.isNull())
{
fileCopy.push_back(line);
line = stream.readLine();
}
stream.seek(0);
// empty file, no comments to keep
if (fileCopy.empty())
return writeFile(stream);
// start
// |
// | +----------------------------------------------------------+
// | | |
// v v |
// skip non-"ordered" lines (remove "ordered" lines) |
// | ^ |
// | | |
// | non-"ordered" line, write saved comments |
// | ^ |
// v | |
// blank or comment line, save in temp buffer <--------+ |
// | | | |
// | +------- comment line ------+ |
// v (special processing '##') |
// "ordered" line |
// | |
// v |
// save in a separate map of comments keyed by "ordered" line |
// | |
// +----------------------------------------------------------+
//
//
QRegularExpression settingRegex("^([^=]+)\\s*=\\s*([^,]+)(.*)$");
std::vector<QString> comments;
auto commentStart = fileCopy.end();
std::map<QString, std::vector<QString>> commentsMap;
for (auto iter = fileCopy.begin(); iter != fileCopy.end(); ++iter)
{
if (isOrderedLine(*iter))
{
// save in a separate map of comments keyed by "ordered" line
if (!comments.empty())
{
QRegularExpressionMatch match = settingRegex.match(*iter);
if (match.hasMatch())
{
commentsMap[match.captured(1) + "=" + match.captured(2)] = comments;
comments.clear();
commentStart = fileCopy.end();
}
// else do nothing, malformed line
}
*iter = QString(); // "ordered" lines to be removed later
}
else if ((*iter).isEmpty() || (*iter).contains(QRegularExpression("^\\s*#")))
{
// comment line, save in temp buffer
if (comments.empty())
commentStart = iter;
// special removed content processing
if ((*iter).contains(QRegularExpression("^##content\\s*=")))
{
if (!comments.empty())
{
commentsMap[*iter] = comments;
comments.clear();
commentStart = fileCopy.end();
}
}
else
comments.push_back(*iter);
*iter = QString(); // assume to be deleted later
}
else
{
QRegularExpressionMatch match = settingRegex.match(*iter);
// blank or non-"ordered" line, write saved comments
if (!comments.empty() && match.hasMatch() && settingRegex.captureCount() >= 2
&& mUserSettings.find(match.captured(1)) != mUserSettings.end())
{
if (commentStart == fileCopy.end())
throw std::runtime_error(
"Config::GameSettings: failed to parse settings - iterator is past of end of settings file");
for (const auto& comment : comments)
{
*commentStart = comment;
++commentStart;
}
comments.clear();
commentStart = fileCopy.end();
}
// keep blank lines and non-"ordered" lines other than comments
// look for a key in the line
if (!match.hasMatch() || settingRegex.captureCount() < 2)
{
// no key or first part of value found in line, replace with a null string which
// will be removed later
*iter = QString();
comments.clear();
commentStart = fileCopy.end();
continue;
}
// look for a matching key in user settings
*iter = QString(); // assume no match
QString key = match.captured(1);
QString keyVal = match.captured(1) + "=" + match.captured(2);
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QMultiMap<QString, SettingValue>::const_iterator i = mUserSettings.find(key);
while (i != mUserSettings.end() && i.key() == key)
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
// todo: does this need to handle paths?
QString settingLine = i.key() + "=" + i.value().originalRepresentation;
QRegularExpressionMatch keyMatch = settingRegex.match(settingLine);
if (keyMatch.hasMatch())
{
if ((keyMatch.captured(1) + "=" + keyMatch.captured(2)) == keyVal)
{
*iter = std::move(settingLine);
break;
}
}
++i;
}
}
}
// comments at top of file
for (auto& iter : fileCopy)
{
if (iter.isNull())
continue;
// Below is based on readFile() code, if that changes corresponding change may be
// required (for example duplicates may be inserted if the rules don't match)
if (/*(*iter).isEmpty() ||*/ iter.contains(QRegularExpression("^\\s*#")))
{
stream << iter << "\n";
continue;
}
}
// Iterate in reverse order to preserve insertion order
QString settingLine;
auto it = mUserSettings.end();
while (it != mUserSettings.begin())
{
it--;
// path lines (e.g. 'data=...') need quotes and ampersands escaping to match how boost::filesystem::path uses
// boost::io::quoted
// We don't actually use boost::filesystem::path anymore, but use a custom class MaybeQuotedPath which uses
// Boost-like quoting rules but internally stores as a std::filesystem::path.
// Caution: This is intentional behaviour to duplicate how Boost and what we replaced it with worked, and we
// rely on that.
if (it.key() == QLatin1String("config") || it.key() == QLatin1String("user-data")
|| it.key() == QLatin1String("resources") || it.key() == QLatin1String("data")
|| it.key() == QLatin1String("data-local") || it.key() == QLatin1String("load-savegame"))
{
settingLine = it.key() + "=";
// Equivalent to settingLine += std::quoted(it.value(), '"', '&'), which won't work on QStrings.
QChar delim = '\"';
QChar escape = '&';
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QString string = it.value().originalRepresentation;
settingLine += delim;
for (auto& iter : string)
{
if (iter == delim || iter == escape)
settingLine += escape;
settingLine += iter;
}
settingLine += delim;
}
else
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
settingLine = it.key() + "=" + it.value().originalRepresentation;
QRegularExpressionMatch match = settingRegex.match(settingLine);
if (match.hasMatch())
{
auto i = commentsMap.find(match.captured(1) + "=" + match.captured(2));
// check if previous removed content item with comments
if (i == commentsMap.end())
i = commentsMap.find("##" + match.captured(1) + "=" + match.captured(2));
if (i != commentsMap.end())
{
std::vector<QString> cLines = i->second;
for (const auto& cLine : cLines)
stream << cLine << "\n";
commentsMap.erase(i);
}
}
stream << settingLine << "\n";
}
// flush any removed settings
if (!commentsMap.empty())
{
auto i = commentsMap.begin();
for (; i != commentsMap.end(); ++i)
{
if (i->first.contains(QRegularExpression("^\\s*content\\s*=")))
{
std::vector<QString> cLines = i->second;
for (const auto& cLine : cLines)
stream << cLine << "\n";
// mark the content line entry for future preocessing
stream << "##" << i->first << "\n";
// commentsMap.erase(i);
}
}
}
// flush any end comments
if (!comments.empty())
{
for (const auto& comment : comments)
stream << comment << "\n";
}
file.resize(file.pos());
return true;
}
bool Config::GameSettings::hasMaster()
{
bool result = false;
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QList<SettingValue> content = mSettings.values(QString(Config::GameSettings::sContentKey));
for (int i = 0; i < content.count(); ++i)
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
if (content.at(i).value.endsWith(QLatin1String(".omwgame"), Qt::CaseInsensitive)
|| content.at(i).value.endsWith(QLatin1String(".esm"), Qt::CaseInsensitive))
{
result = true;
break;
}
}
return result;
}
void Config::GameSettings::setContentList(
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
const QList<SettingValue>& dirNames, const QList<SettingValue>& archiveNames, const QStringList& fileNames)
{
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
remove(sDirectoryKey);
for (auto const& item : dirNames)
setMultiValue(sDirectoryKey, item);
remove(sArchiveKey);
for (auto const& item : archiveNames)
setMultiValue(sArchiveKey, item);
remove(sContentKey);
for (auto const& item : fileNames)
setMultiValue(sContentKey, { item });
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QList<Config::SettingValue> Config::GameSettings::getDataDirs() const
{
return reverse(mDataDirs);
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QList<Config::SettingValue> Config::GameSettings::getArchiveList() const
{
// QMap returns multiple rows in LIFO order, so need to reverse
return reverse(values(sArchiveKey));
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
QList<Config::SettingValue> Config::GameSettings::getContentList() const
{
// QMap returns multiple rows in LIFO order, so need to reverse
return reverse(values(sContentKey));
}
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
bool Config::GameSettings::isUserSetting(const SettingValue& settingValue) const
{
return settingValue.context.isEmpty() || settingValue.context == getUserContext();
}
2 months ago
Config::SettingValue Config::GameSettings::processPathSettingValue(const SettingValue& value)
Store original representation of paths in content lists Also compare against existing content lists in a more forgiving way. The first improvement makes it possible to use relative paths in openmw.cfg without the launcher canonicalising them. This was really annoying if you used a relative path on purpose. It also stops the launcher converting all paths to Qt's convention, where forward slashes are used on Windows even though they're not native. The engine doesn't care, so you could always put either in the config file, but the launcher wouldn't stand for that, and would make them match. To make this work, we need to store a path's originalRepresentation in the content list, compare paths loaded from openmw.cfg based on their originalRepresentation, and convert paths from originalRepresentation to absolute value when loading them from a content list. The second improvement means that paths that are equivalent, but expressed differently (e.g. mismatched case on Windows, mismatched separators on Windows, or mild differences like unnecessary `./`es and doubled separators) don't trigger the creation of a new effectively-identical content list. To make this work, we had to switch the comparison to lexicaly normalise the path first. It could only be lexical normalisation as originalRepresentation might be absolute, relative, or absolute-but-based-on-a-path-slug, and we didn't want slugs to break things or relative paths to count as equivalent to absolute ones that refer to the same file. The comparison is case-insensitive on Windows, and case-sensitive elsewhere. This isn't strictly right, as you can have case-sensitive things mounted on Windows or tell a Linux directory to be case-insensitive, but we can't tell when that might happen based on a lexical path as it depends on real directory properties (and might differ for different parts of the path, which is too much hassle to support).
2 months ago
{
std::filesystem::path path = Files::pathFromQString(value.value);
std::filesystem::path basePath = Files::pathFromQString(value.context.isEmpty() ? getUserContext() : value.context);
mCfgMgr.processPath(path, basePath);
return SettingValue{ Files::pathToQString(path), value.originalRepresentation, value.context };
}
void Config::GameSettings::clear()
{
mSettings.clear();
Track source of settings This one's a biggie. The basic idea's that GameSettings should know: * what the interpreted value of a setting is, so it can actually be used. * what the original value the user put in their config was, so it can be put back when the config's saved. * which path it's processing the openmw.cfg from so relative paths can be resolved correctly. * whether a setting's a user setting that can be modified, or from one of the other openmw.cfg files that can't necessarily be modified. This had fairly wide-reaching implications. The first is that paths are resolved properly in cases where they previously wouldn't have been. Without this commit, if the launcher saw a relative path in an openmw.cfg, it'd be resolved relative to the process' working directory (which we always set to the binary directory for reasons I won't get into). That's not what the engine does, so is bad. It's also not something a user's likely to suspect. This mess is no longer a problem as paths are resolved correctly when they're loaded instead of on demand when they're used by whatever uses them. Another problem was that if paths used slugs like ?userconfig? would be written back to openmw.cfg with the slugs replaced, which defeats the object of using the slugs. This is also fixed. Tracking which settings are user settings and which are in a non-editable openmw.cfg allows the launcher to grey out rows so they can't be edited (which is sensible as they can't be edited on-disk) while still being aware of content files that are provided by non-user data directories etc. This is done in a pretty straightforward way for the data directories and fallback-archives, as those bits of UI are basic, but it's more complicated for content files as that uses a nmodel/view approach and has a lot more moving parts. Thankfully, I'd already implemented that when dealing with builtin.omwscripts, so it just needed wiring up. One more thing of note is that I made the SettingValue struct storable as a QVariant so it could be attached to the UI widgets as userdata, and then I could just grab the original representation and use it instead of needing any complicated mapping from display value to on-disk value.
10 months ago
mContexts.clear();
mUserSettings.clear();
mDataDirs.clear();
mDataLocal.clear();
}
QDataStream& Config::operator<<(QDataStream& out, const SettingValue& settingValue)
{
out << settingValue.value;
out << settingValue.originalRepresentation;
out << settingValue.context;
return out;
}
QDataStream& Config::operator>>(QDataStream& in, SettingValue& settingValue)
{
in >> settingValue.value;
in >> settingValue.originalRepresentation;
in >> settingValue.context;
return in;
}