Merge pull request #149 from yenatch/new-install

reformat install + fix source preprocessing
This commit is contained in:
Bryan Bishop 2013-06-25 20:10:38 -07:00
commit 4b2d4337b4
5 changed files with 198 additions and 154 deletions

View File

@ -1,21 +1,112 @@
# Getting Started
Compiling requires a certain Pokemon Crystal ROM:
Compiling **pokecrystal.gbc** requires a certain **Pokemon Crystal** rom:
```
Pokemon - Crystal Version (UE) (V1.0) [C][!].gbc
md5: 9f2922b235a5eeb78d65594e82ef5dde
```
Save it as `baserom.gbc` in the repository.
Save it as **baserom.gbc** in the repository.
Feel free to ask us on nucleus.kafuka.org #skeetendo if something goes wrong (remember to tell where)!
Feel free to ask us on **[nucleus.kafuka.org #skeetendo](https://kiwiirc.com/client/irc.nolimitzone.com/?#skeetendo)** if something goes wrong!
Don't know how to use IRC? Try [mibbit](http://chat.mibbit.com/?server=nucleus.kafuka.org&channel=#skeetendo).
# Windows
If you are on Windows and can't install Linux, **Cygwin** is a great alternative.
## Installing Cygwin
Cygwin provides a virtual Linux environment on Windows systems. Just get **setup.exe**.
**http://cygwin.com/install.html**
During the install:
* Keep the defaults.
* Most mirrors are molasses. Use **http://mirrors.kernel.org**.
## Linux
## Using Cygwin
Launch the **Cygwin terminal**. Maybe you know your way around the Linux terminal, **bash**. If not, a crash course:
```bash
# list files in current directory
ls
# show current directory
pwd
# change directory
cd /away/we/go
```
## Getting up and running
We need a couple more things to be able to compile.
If you're feeling lazy, just paste these commands into your terminal.
```bash
apt-cyg install make git wget python python-setuptools
easy_install pip
```
**rgbds** will let you compile Game Boy roms.
```bash
cd ~
# download rgbds binaries
wget http://diyhpl.us/~bryan/irc/pokered/rgbds/rgbds.zip
unzip rgbds.zip
rm rgbds.zip
# make rgbds accessible for all time
export PATH=$PATH:`pwd`/rgbds
echo "export PATH=$PATH" >> ~/.bashrc
```
Set up the **pokecrystal** repository:
```bash
git clone https://github.com/kanzure/pokecrystal
cd pokecrystal
# install python requirements
pip install -r requirements.txt
```
## Don't forget baserom.gbc!!
Make sure you downloaded a base rom. Name it **baserom.gbc**.
Now you should be able to build **pokecrystal.gbc** for the first time.
This compiles a new rom from the source code, with any patches filled in from the base rom.
```bash
make
```
This ought to take **between 3 and 15 seconds**, depending on your computer.
If you see `cmp baserom.gbc pokecrystal.gbc` as the last line, the build was successful!
Your first compile processes every source file at once. After that, **only modified source files have to be reprocessed**, so compiling again should be a few seconds faster.
Other **make targets** that may come in handy:
`make clean` deletes any preprocessed source files (.tx), rgbds object files and pokecrystal.gbc, in case something goes wrong.
`make pngs` decompresses any **lz** files in gfx/ and then exports any graphics files to **png**.
`make lzs` does the reverse. This is already part of the build process, so **modified pngs will automatically be converted to 2bpp and lz-compressed** without any additional work.
# Linux
```bash
sudo apt-get install make gcc bison git python python-setuptools
@ -46,36 +137,21 @@ pip install -r requirements.txt
git config diff.hex.textconv hexdump
```
To compile the ROM from ASM source:
```
make clean && make
To compile the rom from asm source:
```bash
make
```
That will take between 3 and 15 seconds, depending on your computer. If you see
`cmp baserom.gbc pokecrystal.gbc` as the last line, the build was successful! Rejoice!
That will take between 3 and 15 seconds, depending on your computer. If you see `cmp baserom.gbc pokecrystal.gbc` as the last line, the build was successful! Rejoice!
## Windows
# Now what?
Set up [GitHub for Windows](http://windows.github.com/) and clone this repository.
**main.asm** is a good starting point. The structure of the source is laid out here.
If you haven't already, get [Python 2.7](http://www.python.org/ftp/python/2.7.3/python-2.7.3.msi) ([64-bit](http://www.python.org/ftp/python/2.7.3/python-2.7.3.amd64.msi)).
* **Can't find something?** Anyone can add to the source. There's lots to be uncovered.
Extract the following files from the [RGBDS](https://github.com/downloads/bentley/rgbds/rgbds-0.0.1.zip) package into the repository:
`rgbasm.exe`
`rgbds.exe`
`rgbfix.exe`
`rgblink.exe`
* **Do your own thing!** The asm source is hack-friendly, and the supplementary scripts in extras/ can be used for other projects.
Install [make](http://gnuwin32.sourceforge.net/downlinks/make.php) for Windows.
To compile the ROM from ASM source, run `pokecrystal.bat`.
That will take between 3 and 15 seconds, depending on your computer. If you see
`FC: no differences encountered`, the build was successful! Rejoice!
Now you may try messing around with `main.asm`, or just do whatever you wanted to.
# Contributions are welcome!
* We'll be happy to answer any **questions** on **[nucleus.kafuka.org #skeetendo](https://kiwiirc.com/client/irc.nolimitzone.com/?#skeetendo)**.

View File

@ -1,29 +1,24 @@
.SUFFIXES: .asm .tx .o .gbc .png .2bpp .lz
TEXTFILES = \
text/sweethoney.tx \
text/phone/bill.tx \
text/phone/elm.tx \
text/phone/mom.tx \
text/phone/trainers1.tx \
text/common.tx \
text/common_2.tx \
text/common_3.tx \
main.tx
TEXTFILES := $(shell find ./ -type f -name '*.asm' | grep -v pokecrystal.asm | grep -v constants.asm | grep -v gbhw.asm | grep -v hram.asm | grep -v constants | grep -v wram.asm)
TEXTQUEUE :=
PNG_GFX = $(shell find gfx/ -type f -name '*.png')
LZ_GFX = $(shell find gfx/ -type f -name '*.lz')
TWOBPP_GFX = $(shell find gfx/ -type f -name '*.2bpp')
PNG_GFX := $(shell find gfx/ -type f -name '*.png')
LZ_GFX := $(shell find gfx/ -type f -name '*.lz')
TWOBPP_GFX := $(shell find gfx/ -type f -name '*.2bpp')
all: pokecrystal.gbc
cmp baserom.gbc $<
clean:
rm -f main.tx pokecrystal.o pokecrystal.gbc ${TEXTFILES}
pokecrystal.o: pokecrystal.asm constants.asm wram.asm ${TEXTFILES} lzs
rm -f pokecrystal.o pokecrystal.gbc
@echo 'rm -f $(TEXTFILES:.asm=.tx)'
@rm -f $(TEXTFILES:.asm=.tx)
pokecrystal.o: $(TEXTFILES:.asm=.tx) $(LZ_GFX) $(TWOBPP_GFX)
python prequeue.py $(TEXTQUEUE)
rgbasm -o pokecrystal.o pokecrystal.asm
.asm.tx:
python preprocessor.py < $< > $@
$(eval TEXTQUEUE := $(TEXTQUEUE) $<)
@rm $@
pokecrystal.gbc: pokecrystal.o
rgblink -o $@ $<
@ -33,6 +28,7 @@ pngs:
cd extras && python gfx.py mass-decompress && python gfx.py dump-pngs
lzs: $(LZ_GFX) $(TWOBPP_GFX)
@:
gfx/pics/%/front.lz: gfx/pics/%/front.png gfx/pics/%/tiles.2bpp
python extras/gfx.py png-to-lz --front $^

View File

@ -7,7 +7,7 @@ The source code in this project successfully converts back into a ROM image. All
## Base ROM
The following ROM is required for compiling:
The following rom is required for compiling:
```
Pokemon - Crystal Version (UE) (V1.0) [C][!].gbc
@ -17,16 +17,18 @@ md5: 9f2922b235a5eeb78d65594e82ef5dde
Eventually this will not be necessary.
## See also
* disassembly of [Pokémon Red](http://bitbucket.org/iimarckus/pokered).
## Contributing
* Hang out with us on IRC: `nucleus.kafuka.org #skeetendo` (for example, by
using [mibbit](http://chat.mibbit.com/)).
## What can I do?
* Are we missing something? Make a pull request! Contributions are welcome.
* Tackle some [issues](https://github.com/kanzure/pokecrystal/issues)!
* Take a look at some of the disasm tools in **extras/**. Most of the scripts are generalized enough to take apart other Game Boy games.
* Tackle some **[issues](https://github.com/kanzure/pokecrystal/issues)**!
## See also
* Hang out with us on irc: **[nucleus.kafuka.org #skeetendo](https://kiwiirc.com/client/irc.nolimitzone.com/?#skeetendo)**
* Disassembly of **[Pokémon Red](http://bitbucket.org/iimarckus/pokered)**.

View File

@ -310,24 +310,14 @@ def separate_comment(l):
"""
Separates asm and comments on a single line.
"""
asm = ""
comment = None
in_quotes = False
# token either belongs to the line or to the comment
for token in l:
if comment:
comment += token
else:
if not in_quotes:
if token == ";":
comment = ";"
continue
if token == "\"":
in_quotes = not in_quotes
asm += token
return asm, comment
for i in xrange(len(l)):
if not in_quotes:
if l[i] == ";":
break
if l[i] == "\"":
in_quotes = not in_quotes
return l[:i], l[i:] or None
def quote_translator(asm):
"""
@ -335,63 +325,50 @@ def quote_translator(asm):
"""
# split by quotes
asms = asm.split("\"")
asms = asm.split('"')
# skip asm that actually does use ASCII in quotes
lowasm = asms[0].lower()
if "section" in lowasm \
or "incbin" in lowasm:
sys.stdout.write(asm)
return
if "SECTION" in asms[0]\
or "INCBIN" in asms[0]\
or "INCLUDE" in asms[0]:
return asm
print_macro = False
if asms[0].strip() == 'print':
asms[0] = asms[0].replace('print','db 0,')
print_macro = True
output = ""
output = ''
even = False
i = 0
for token in asms:
i = i + 1
if even:
characters = []
# token is a string to convert to byte values
while len(token):
# read a single UTF-8 codepoint
char = token[0]
if ord(char) >= 0xFC:
char = char + token[1:6]
token = token[6:]
elif ord(char) >= 0xF8:
char = char + token[1:5]
token = token[5:]
elif ord(char) >= 0xF0:
char = char + token[1:4]
token = token[4:]
elif ord(char) >= 0xE0:
char = char + token[1:3]
token = token[3:]
elif ord(char) >= 0xC0:
if ord(char) < 0xc0:
token = token[1:]
# certain apostrophe-letter pairs are considered a single character
if char == "'" and token:
if token[0] in 'dlmrstv':
char += token[0]
token = token[1:]
elif ord(char) < 0xe0:
char = char + token[1:2]
token = token[2:]
elif ord(char) < 0xf0:
char = char + token[1:3]
token = token[3:]
elif ord(char) < 0xf8:
char = char + token[1:4]
token = token[4:]
elif ord(char) < 0xfc:
char = char + token[1:5]
token = token[5:]
else:
token = token[1:]
# certain apostrophe-letter pairs are only a single byte
if char == "'" and len(token) > 0 and \
(token[0] == "d" or \
token[0] == "l" or \
token[0] == "m" or \
token[0] == "r" or \
token[0] == "s" or \
token[0] == "t" or \
token[0] == "v"):
char = char + token[0]
token = token[1:]
char = char + token[1:6]
token = token[6:]
characters += [char]
if print_macro:
@ -421,32 +398,26 @@ def quote_translator(asm):
output += ", ".join(["${0:02X}".format(chars[char]) for char in characters])
# if not even
else:
output += (token)
output += token
even = not even
sys.stdout.write(output)
return
return output
def extract_token(asm):
token = asm.split(" ")[0].replace("\t", "").replace("\n", "")
return token
return asm.split(" ")[0].strip()
def make_macro_table():
return dict([(macro.macro_name, macro) for macro in macros])
return dict(((macro.macro_name, macro) for macro in macros))
macro_table = make_macro_table()
def macro_test(asm):
"""
Returns a matching macro, or None/False.
"""
# macros are determined by the first symbol on the line
token = extract_token(asm)
# check against all names
if token in macro_table:
return (macro_table[token], token)
@ -600,64 +571,46 @@ def macro_translator(macro, token, line):
sys.stdout.write(output)
def include_file(asm):
"""This is more reliable than rgbasm/rgbds including files on its own."""
prefix = asm.split("INCLUDE \"")[0] + '\n'
filename = asm.split("\"")[1]
suffix = asm.split("\"")[2]
read_line(prefix)
lines = open(filename, "r").readlines()
for line in lines:
read_line(line)
read_line(suffix)
def read_line(l):
"""Preprocesses a given line of asm."""
# strip and store any comment on this line
if ";" in l:
asm, comment = separate_comment(l)
else:
asm = l
comment = None
# strip comments from asm
asm, comment = separate_comment(l)
# handle INCLUDE as a special case
if "INCLUDE \"" in l:
include_file(asm)
# export all labels
if ':' in asm[:asm.find('"')]:
sys.stdout.write('GLOBAL ' + asm.split(':')[0] + '\n')
# expect preprocessed .asm files
if "INCLUDE" in asm:
asm = asm.replace('.asm','.tx')
sys.stdout.write(asm)
# ascii string macro preserves the bytes as ascii (skip the translator)
elif len(asm) > 6 and "\tascii " in [asm[:7], "\t" + asm[:6]]:
elif len(asm) > 6 and "ascii " == asm[:6] or "\tascii " == asm[:7]:
asm = asm.replace("ascii", "db", 1)
sys.stdout.write(asm)
# convert text to bytes when a quote appears (not in a comment)
elif "\"" in asm:
quote_translator(asm)
sys.stdout.write(quote_translator(asm))
# check against other preprocessor features
else:
macro, token = macro_test(asm)
if macro:
macro_translator(macro, token, asm)
else:
sys.stdout.write(asm)
# show line comment
if comment != None:
sys.stdout.write(comment)
if comment: sys.stdout.write(comment)
def preprocess(lines=None):
"""Main entry point for the preprocessor."""
if not lines:
# read each line from stdin
lines = sys.stdin
lines = (sys.stdin.readlines())
elif not isinstance(lines, list):
# split up the input into individual lines
lines = lines.split("\n")

17
prequeue.py Normal file
View File

@ -0,0 +1,17 @@
# coding: utf-8
# Starting a new python process to preprocess each source file
# creates too much overhead. Instead, a list of files to preprocess
# is fed into a script run from a single process.
import os
import sys
import preprocessor
if __name__ == '__main__':
for source in sys.argv[1:]:
dest = os.path.splitext(source)[0] + '.tx'
sys.stdin = open(source, 'r')
sys.stdout = open(dest, 'w')
preprocessor.preprocess()