Actually, many *globs*. It additionally prints the glob if no files
matched it, since that's clearer.
Also move the ComputerTestDelegate's filesystem to be disk-based. This
is what actual computers use, and the MemoryMount is a little broken.
- Languages are converted to JSON
- Rename most *(_advanced) blocks to *_{advanced,normal}. It's more
verbose, but means they're sorted together.
- A couple of changes to the ROM to work with some Java changes.
- Update recipes and advancements to not use damage values.
Are most of these changes small and petty? Yes. However, IMO they do
make the code more readable. Anyway, a summary of some of the more
interesting changes:
- Expose Abstract*Upgrade classes in the API
- Fix the spelling of Jonathan in the API docs (*shakes fist*)
- Fix bug with printout not working in the offhand.
- Rename any argments/variables accidentally named "m_*", and add an
inspection to prevent it happening again.
- Remove most of the Block*.Properties classes - just inline them in
the parent class.
- Return super.writeToNBT instead of reassigning at the top.
This makes use of the "sent" variable, which would otherwise go unused. It also makes rednet.send compliant to the behaviour specified in the Wiki: http://www.computercraft.info/wiki/Rednet.send
This is largely invisible (it's marked as a child of the main
"computercraft" mod), but allows other mods (such as Plethora) to add
hard/soft dependencies on CC:T in a user-friendly manner.
- Some performance improvements to JEI recipe resolver
- Use a shared map for upgrade items, meaning we only need one map
lookup.
- Cache the basic upgrade recipes.
- Use the MC version within project rather than version name.
Before IPocketAccess.getEntity would return the entity which last held
fthis computer, even if not holding it any more. As
ba823bae13 describes, this caused
pocket.equip/pocket.unequip to dupe items.
We move the validation from the PocketAPI into the main IPocketAccess
implementation, to ensure this issue does not occur elsewhere. Note, we
require a separate method, as this is no longer thread-safe.
We also now return ok, err instead of throwing an exception, in order to
be consistent with the turtle functions. See dan200/ComputerCraft#328.
Oh goodness, this is going to painful to update to 1.13.
We now translate:
- Computer/Disk ID tooltips
- /computercraft descriptions, synopsises and usages. The last of these
may not always be translated when in SMP, as it is sometimes done on
the server, but the alternative would be more complex than I'm happy
with.
- Tracking field names. Might be worth adding descriptions too in the
future.
Also cleanup a couple of other translation keys, so they're more
consistent with Minecraft.
Closes#141
Previously we would register the recipes within our code, but the
advancements were written manually. This now generates JSON files for
both the advancement and recipe.
While this does mean we're shipping even more JSON, we'll need to do
this for 1.13 anyway, and means our advancements are guaranteed to be
consistent.
On a side note, a couple of other changes:
- Turtle upgrades are now mounted on the right in the creative
menu/fake recipes. This means the upgrade is now clearly visible in
the inventory.
- We no longer generate legacy turtle items at all: we'll always
construct turtle_expanded.
- Several peripheral items are no longer registered as having sub-types
(namely advanced and full-block modems).
- We only have one disk advancement now, which unlocks all 16 recipes.
- We have removed the disk conversion recipes - these can be
exposed through JEI if needed.
- Share the ILuaContext across all method calls, as well as shifting it
into an anonymous class.
- Move the load/loadstring prefixing into bios.lua
- Be less militant in prefixing chunk names:
- load will no longer do any auto-prefixing.
- loadstring will not prefix when there no chunk name is supplied.
Before we would do `"=" .. supplied_program`, which made no sense.
For instance, `pastebin run https://pastebin.com/LYAxmSby` will now
extract the code and download appropriately. Also add an error message
when we received something which is not a valid pastebin code.
See #134.
Before it was not actually selected until the task had yielded for the
first time. If a computer did not yield (or took a while to do so),
nothing would actually show up.
We now generate a table and concatinate the elements together. This has
several benefits:
- We no longer emit emit trailing spaces, which caused issues on 1.13's
command system.
- We no longer need the error level variable, nor have the weird
recursion system - it's just easier to understand.
- Run optipng on all our images. This has very little effect on most of
them (as they're all so small anyway), but has resulted in a 50%
reduction in some cases.
- Run Proguard on our shadowed dependencies (Cobalt).
- Minify our JSON files, stripping all whitespace. This is mostly
useful for FML's annotation cache, as that's a massive file, but
still a semi-useful optimisation to make.
This has helped reduce the jar by about 110kb, which isn't much but
still feels somewhat worth it.
- All "named" entries (blocks, items, recipes, TEs and pocket/turtle
upgrades) are registeredin one place.
- Most client side models/textures are registered in ClientRegistry -
we can't do item colours or TEs for now, as these aren't event based.
- A little cleanup to how we handle ItemPocketComputer models.
- Move all HTTP tasks to a unified "MonitoredResource" model. This
provides a uniform way of tracking object's lifetimes and disposing
of them when complete.
- Rewrite HTTP requests to use Netty instead of standard Java. This
offers several advantages:
- We have access to more HTTP verbs (mostly PATCH).
- We can now do http -> https redirects.
- We no longer need to spawn in a new thread for each HTTP request.
While we do need to run some tasks off-thread in order to resolve
IPs, it's generally a much shorter task, and so is less likely to
inflate the thread pool.
- Introduce several limits for the http API:
- There's a limit on how many HTTP requests and websockets may exist
at the same time. If the limit is reached, additional ones will be
queued up until pending requests have finished.
- HTTP requests may upload a maximum of 4Mib and download a maximum
of 16Mib (configurable).
- .getResponseCode now returns the status text, as well as the status
code.
- Move configuration loading into a separate file, just so it doesn't
clutter up ComputerCraft.java.
- Normalise property names, so they're all snake_case.
- Split properties into separate categories (http, turtle, peripheral),
so the main one is less cluttered.
- Define an explicit ordering of each category.
- Replace BlockCableCableVariant with a boolean, removing any custom
logic around it.
- Replace our previous cable Forge-blockstate with a vanilla one, which
handles cable connections in the model logic.
This ends up being slightly cleaner as we can rely on Minecraft's own
model dependency system. Also reduces reliance on Forge's APIs, which
_potentially_ makes porting a little easier.
This is implemented in a rather ugly way: we register a client command
(/computercraft_copy) which updates the clipboard, and run that via a
click handler on the chat message.
This hopefully makes wired modems a little easier to use. We'll see.
This replaces the existing IMount openFor* method with openChannelFor*
ones, which return an appropriate byte channel instead.
As channels are not correctly closed when GCed, we introduce a
FileSystemWrapper. We store a weak reference to this, and when it is
GCed or the file closed, we will remove it from our "open file" set and
ensure any underlying buffers are closed.
While this change may seem a little odd, it does introduce some
benefits:
- We can replace JarMount with a more general FileSystemMount. This
does assume a read-only file system, but could technically be used
for other sources.
- Add support for seekable (binary) handles. We can now look for
instances of SeekableByteChannel and dynamically add it. This works
for all binary filesystem and HTTP streams.
- Rewrite the io library to more accurately emulate PUC Lua's
implementation. We do not correctly implement some elements (most
noticably "*n", but it's a definite improvement.
This crash can be triggered remotely by specially constructed rednet messages, making this a bit of a problem, as any repeaters can be remotely crashed.
- Make window.reposition's argument validation a little more strict.
Previously it would accept `window.reposition(x, y, width)` (no
height argument), just not act upon it.
- Use select instead of table.unpack within `pastebin run`.
- Use `parallel.waitForAny` instead of `waitForAll` within the dance
program.
- Pipe the entire help file into `textutils.pagedPrint`, rather than
doing it line by line.
- Remove bytecode loading disabling from bios.lua. This never worked
correctly, and serves little purpose as LuaJ is not vulnerable to
such exploits.
- Slight airbrush effect for normal turtles. While this is only really
visible when upping the contrast, it's probably nice to fix.
- A few off-colour pixels for advanced turtles
This implements an argument format similar to LuaReqeust, as described
in dan200/ComputerCraft#515. The Lua argument checking code is a little
verbose and repetitive, but I'm not sure how to avoid that - we should
look into improving it in the future.
Closes#21
- fixed circumflex typo: you can now cîrcûmflêx on all your friends
- added comment making it clearer that the following lines are intended for backwards compatibility.
The two recipes are pretty similar, so this allows us to substantially
simplify the code. This now introduces the additional requirement that
computers must be used to create turtles, rather than pocket computers
or another turtle.
This adds IComputerItem.withFamily(ItemStack, ComputerFamily) as well as
a ComputerFamilyRecipe class. Each type of computer (normal, turtle,
pocket) defines a recipe using this class, as they require a different
number of gold ingots to upgrade.
These act similarly to conventional wired modems, but with the advantage
that they are a full block. This means they can be attached to
peripherals which are not solid (such as chests). Further more, as they
do not have a direction, they allow wrapping peripherals on all 6 sides.
It's worth noting that wired modems do not require a cable - they will
automatically form connections to adjacent network elements when placed.
Mostly intended for those people who don't like .inspect() or
.getItemDetail(), but could allow modpacks to block equipping upgrades,
placing blocks, etc...
Whilst I'm pretty sure this is safe for general use, I'm disabling this
by default for now. I may consider enabling it in the future if no
issues are found.
This uses Netty's websocket functionality, meaning we do not have to
depend on another library.
As websockets do not fit neatly into the standard polling socket model,
the API is significantly more event based than CCTweaks's. One uses
http.websocket to connect, which will wait until a connection is
established and then returns the connection object (an async variant is
available).
Once you have a websocket object, you can use .send(msg) to transmit a
message. Incoming messages will fire a "websocket_message" event, with
the URL and content as arguments. A convenience method (.receive())
exists to aid waiting for valid messages.
if you call write(nil), you will get the error "bios.lua:229: bad argument: string expected, got nil", so nil is not a valid argument for write() and should be removed.
While adding checks to the io API in #424, I had forgot that the io API has his own type() function who overwrite the default one. This PR fix this. Sorry for that.
- Path containing '/' or '\' are resolved relative to the current
directory, rather than using the path. Paths starting with '/' still
resolve relative to the root directory.
- Shell completion will also include sub-directories of the current
directory.
Closes#219
If fs.open() failed, it will return nil and a a error message like "No such file" or "Out of space". This PR make, that io.open() returned this error too.
- Convert most recipes to JSON
- Add JSON factories for impostor and turtle recipes.
- Several mappings changes
- Migrate to Forge's new registry system
While the check (most likely) wouldn't break anything, it's better to fix it and not worry then to not fix it and find out it breaks image loading. Also fixes how files are loaded.
- textutils.serializeJSON now takes a table/string/number/boolean instead of just a string as the t argument, and checks for nil in the tNBTStyle argument, so the parameter becomes optional again, as described in the wiki.
- textutils.slowWrite and textutils.slowPrint now checks if the rate
parameter is a number, if not nil.
- Adds a 1px margin around every glyph. This is generally empty,
with the exception of teletext characters where it continues their
pattern.
- Uses GL_CLAMP with the font texture.
Closes#300
edit currently deletes the previous 4 spaces when pressing backspace,
wherever you are on the line. This can be frustrating when you are
trying to align text in comments or strings.
This changes edit to only delete 4 spaces if all preceding characters
are whitespace, otherwise it only deletes a single character.
Attempting to use pastebin with plain HTTP results in a 301 redirect to the HTTPS version. This PR (and associated commit) makes the pastebin command work again.
Reduces amount of string operations and write calls inside writeHighlighted function by switching it to blit and using the 3rd argument in string.match.
Remove drawPixelInternal / drawLineHorizontal. At runtime they're both just excess calls.
Localise high-use term functions.
Arg-related function errors reflect that "colour" parameter is optional.
Add saveImage(tImage, sPath).
Use term.blit for images.
In no particular order:
bios.lua - added missing test for ensure turtle folder exists.
paintutils.lua - added drawLineHorizontal() to reduce most (not all) cases of running multiple writes on same horizontal line that can be solved with one write.
textutils.lua - Added exception to complete function - will not complete if provided with LuaKeyword - solves do->dofile problem and removes other LuaKeyword related ones in advance.
edit.lua - Changed logic in handling the paste event - if paste event is received when menu is open it automatically closed said menu and lets it paste - resolves ctrl+ctrl+v annoyance. Added Jump to Menu functions - allows for fast jump to needed line - must for bigger files and a nice feature to have.
set.lua - Switched set to use pagedPrint - this will ensure that even if there are more settings than lines on screen the set command will show you all of them.
startup.lua - Added autocompletition for turtle programs go, turn, equip, unequip and command program exec.
lua.lua - Changed return function to print returned stuff correctly -
will print all returned variables even if there are any nils in there.
- Adds support for blacklisting domains
- Adds support for blacklisting & whitelisting IP addresses and
IP ranges.
- Reuse threads for HTTP requests
AddressPredicate will parse a series of patterns and convert them into
regexes or CIDR ranges. When checking whether an address is accessible,
we first ensure the domain is whitelisted and isn't blacklisted.
If everything is OK, then we start create a new thread for the HTTP
request and resolve the IP, ensuring that is whitelisted & not
blacklisted. Then the normal HTTP request is continued.
However, http.checkURL also needs to resolve the IP address. In order to
avoid blocking the Lua thread, this method will return instantly and
create a new thread which will queue an event.
As both http.request and http.checkURL are now creating threads and
queuing events, some logic is abstracted into a separate HTTPTask class
- this allows us to share the thread creation, finishing and cancelling
logic.
I fixed 2 Bugs:
1. If you call this function, without 2 numbers, you get a error in the window API and not in your Program
2. If you call, this function with 2 numbers lower then 1 (e.g. term.setCursorPos(0,0) ), CraftOS will hang forever and need to press Ctrl+R or rejoin the world.
"shell" now runs each program in a new lua environment, instead of
sharing that lua environment between all programs launched under a one
shell. Said environment now includes an implemenation of "require" and
the "package" API, so that programs can require modules into that
environment.
This means that programs can require in libraries without polluting the
global namespace, and without breaking the virtual computer model, as
each program has it's own set of requires, which are discarded when the
program ends.
There was a very small gap between the turtle frame and the top of the
turtle body. This increases the height of the body by one pixel, adding
a little bit of overlap and ensuring the frame renders on top fixes
this issue.
- Move the encoding/decoding from the Filesystem implementation to the
individual handles.
- Move each handle into an core.apis.handles package from the main fs
API.
- Move the HTTP response to inherit from these handles.
- Allow binary handles' read function to accept a number, specifying
how many characters to read - these will be returned as a Lua string.
- Add readAll to binary handles
- Allow binary handles' write function to accept a string which is
decoded into the individual bytes.
- Add "binary" argument to http.request and friends in order to return
a binary handle.
- Ensure file handles are open when reading from/writing to them.
- Return the error message when opening a file fails.
This makes block/model names a little mode consistent and should help
with porting to 1.11.2.
- Prefix all tile entities with "computercraft:".
- Change all "pascalCase" and "CC-*" items to use underscore_case
- Listen to the missing mappings event and gracefully convert
blocks/items.
- Lower case all model and texture names
- Move model registration code into preInit - this ensures we don't
get texture/model errors in the preInit stage.
This uses Minecraft's colour tinting system in order to change the
colour of turtle models. This removes the need to have 16 models and
textures for each colour, reducing texture atlas space and hopefully
memory consumption.
See #145
table.unpack will often stop at the first nil value, meaning some event
arguments may be discarded. By storing the number of arguments through
table.pack, and then using that count when unpacking, we can ensure all
values are returned/resumed with.
Updated the source code to the version shipped as the 1.80pr0 alpha
release. Also removed some unnecessary files from the LuaJ subfolder
which were bulking up the repository.
