Sometimes people's texture packs contain weird things, and it'd
be good if instead of throwing an unhandled IOError, we raise
a TextureException with the filename in it.
Also, make verbose mode actually run find_file in verbose mode,
which is useful for finding where Overviewer loaded that texture
file from.
This reverts commit 9895fe875b.
The original code was correct, but confusing. Overviewer did regenerate
the texture in each worker process because Python pickling is incredibly
fragile fucking garbage designed by idiots who have no clue of what
a good language looks like.
We used to generate textures multiple times for each worker process,
which didn't seem to be intended as it was done gated behind a reversed
if statement. It just so happened to fix an issue introduced by the
pickling code which was deleting vital attributes like the actual
generated textures.
I'm honestly not sure why this was ever done, so let's just throw this
out and call it a day.
Fixes#1728
People could always set this through the config file, but somehow
not the command line. This commit fixes this oddity by adding the
option to the genPOI script's argument parser.
Don't use .append(), we already know the final length. Instead,
have python initialise the list of a certain size and set the elements
in it.
Makes the function like 4% faster.
Turns out my previous grasp of how Minecraft does this was wrong.
This seems to be the correct way. One side effect is that biome data
now has less resolution. One only really notices this when looking at
water, for which Minecraft does not even use the water colours for
in-game, otherwise I can't really tell a big difference.
Fixes#1698.
Okay so you know how you have chunks and they map to 4 levels of biomes?
Logically you'd go hmmmm, let's just map 4 Y levels to one biome level.
Except there aren't 16 chunk heights, there are about 17 if you count
Y=16 (which can contain data), but there's also -1 technically but I've
never seen this have data in an NBT structure I don't think.
Anyway, let's just hope we're doing the right thing by giving Y=16 the
same biome as Y=15. If someone wants to check whether that is actually
correct please feel free to, but for now it's better than crashing at
the very least.
Would be nice if we had a spec for these kinds of things.
Fixes#1685.
Apparently rotated worlds would try to access the biome for Y-level
-1. I don't know why they do that (probably some dumb reason that may
be a bug in of itself) but returning the first level should be harmless.
Fixes#1683.
Fixes#1650.
Mojang changed the biomes code so that it now can have different
biomes for different Y levels. We need to adjust our logic accordingly,
which is done through some small BiomeDispensary class where we shove
a numpy'd Mojang array in and can then read out the biomes for each level.
Biome data is now stored per-section, which needed some changes on the C
side of things. I didn't change anything in the biome overlay code so
I wouldn't be surprised if it's broken now, but for the time being I'd
rather have 1.15 fixed than some obscure overlay.
Tested to work with 1.14 and 1.15 data. No new biomes have been added
to the code yet.
Observers get initialised at config parse time, before Overviewer
creates the outputdir. So we can just tell them to create the
outputdir if it doesn't already exist to avoid an awkward error
on the initial render.
Fixes#1657.
Apparently the gzip stream can be malformed in a way where it
does read it, but then throws an EOFError when it abruptly ends.
Catch this so we don't error out.
Apparently this is a status sometimes seen with worlds converted from
1.12 to 1.14. From the sample I have, it looks like they're properly
populated with light and terrain data, so hopefully this doesn't break
anything.
Possibly fixes#1630.
python3-imaging doesn't exist but python3-pil does, which
python3-pillow is aliased to. It also includes the necessary headers.
Add a note about pip missing the headers and warn of ABI mismatches if
people do weird things.
While dicking around with the tests I noticed that you could make
them fail if you ran them twice. Many hours were spent investigating,
and it turns out that Overviewer's config stuff has global state that
gets modified by having parsed configurations actually modify the
default config values. Not good!
We can fix this by having settingsDefinition return a dict of the
defaults, instead of assigning it to module-level names.
We can also get rid of test_all.py, because what it would do with
pytest is run all tests *twice*. Instead, do the module path stuff
in __init__.py. Also, instead of throwing a non-specific Exception
if exmaple isn't checked out, just skip the test thank you very much.
This is the weirdest rabbit hole I've ever gone down and I haven't
slept in about 30 hours. I'm going to push this commit, and if it
breaks anything, I'll be blissfully asleep as the unwashed masses
begin to riot over exception this traceback that.
CamelCase does not bode well in Python land, so it's best we rename
these ill-named files before the Guidoists get us and throw us into
a damp dungeon.
This adds:
- mossy stone brick stairs
- mossy cobblestone stairs
- mossy stone brick walls
We also add another block class for walls, and while we're at it,
clean up the stairs texture function to not have a huge sprawling
elif mess for loading textures, but instead to a cheeky dictionary
lookup.
In case you're wondering, yes I am just as disgusted by this code as
you are.
genPOI needs this for groupby, and since we can now throw cropped
regionsets into genPOI we definitely need this too.
Probably also fixes rotated regionsets with multiple genPOI filters.
Use CroppedRegionSets to generate markers, which should only affect
handleEntities. I have not yet checked whether this generates
dupicate players and manualpois for maps with multiple crop zones,
but that's a fire to put out for another day.
Fixes#1574.
If the first texture loaded was not in the pack, i.e. we fell back
on system textures, then the pack was ignored altogether for the
remainder of the render.
Here we fix this by using an OrderedDict instead of the split jars
and jarpath nonsense, and then always checking the custom textures
first and opening them if they are not already open, and putting
the jar file handle at the start of the OrderedDict.
We also switch around some logic to avoid unnecessary isdir calls
when we're using a resource pack.
Sometimes a texture pack is not complete, but system textures are
available to make up the difference. In that case, we should load
those.
Current caveat is that we might not notice texture names changing
with newer minecraft clients because we have old ones installed.
I'll have to think some more about that one.
If loading the grass and foliage biome colour fails, Overviewer doesn't
necessarily crash, it just renders everything gray, apart from one small
issue: in trying to call base_finish, it'd crash trying to decrease the
reference of a null pointer.
This option allows you to specify your own initial center for a
tileset, which is useful if your map is extremely asymmetric or
you don't really care about what's around the spawn.
Future work needs to be done on the JS side in order to fix the
fromWorldToLatLng and friends, as they're currently off by -24 in X
and +24 in Z direction.
Closes#1350.
find_file contained a lot of code for searching textures in ways that
haven't been relevant for years. For example, it had workarounds for
MCPatcher, a tool that hasn't been updated since 2015.
Instead of searching in multiple paths that will never succeed in a
resource pack and Minecraft version that actually work, just look
for the filename that the block functions ask for.
Also, if we already have a manually specified resource pack open,
we shouldn't repeatedly re-open it. Furthermore, it's best if
the very first thing find_file does is try to find the file in
the resource pack or client jar we've already opened previously.
We can now also remove chest.png, ender.png, foliage.png, grass.png
and normal_double.png because Minecraft ships them, so we don't
need to ship them ourselves too. Us shipping them actually hid the
whole find_file garbagefire.
Concerns issue #1604.
Previously, textures that couldn't be found the first time around
would not have this fact cached, so we'd run through the whole
slow path of find_file with each blockid and data entry.
This is what caused the chests texture generation function to take
almost 12 seconds on my system, as it is looking for "normal.png"
that hasn't existed in Minecraft textures for ages.
Shoving the exception, if one is encountered, into the cache, and
then re-raising it if we load it from the texture cache brings this
number down to about 2 seconds. That's six times as fast!
Concerns issue #1604.
We can try each jar in descending order of versions until we actually
find one that works. This fixes#1328.
Future work might move the "if self.jar is not None" higher up in the
function so that we don't keep re-checking the overviewer dir and texture
path for files that we will then not use at all for some reason.
Somebody reported that genPOI was throwing an exception in two
different places related to trying to read bytes as a string.
I personally could not reproduce (possibly needs some special player
names or something?) but I think this change should fix it either way.
So turns out genPOI wants to use sorted() so it can use itertools.groupby.
Python 2 somehow just... sorted objects so that groupby (which is basically
the UNIX uniq) could group them properly. Python 3 no longer does this,
so as comparison method let's just compare regiondir strings since that
gives us the desired result.
Fixes#1597.
