mirror of
https://github.com/skelsec/aardwolf
synced 2026-06-08 17:28:11 +00:00
skelsec
256 lines
9.2 KiB
Python
256 lines
9.2 KiB
Python
import os
|
|
import json
|
|
import base64
|
|
import traceback
|
|
import xml.etree.ElementTree as ET
|
|
|
|
from tqdm import tqdm
|
|
|
|
from aardwolf.keyboard import KeyboardLayout, VK_MODIFIERS
|
|
|
|
"""
|
|
This script will fetch all keyboard layouts from https://kbdlayout.info/ and parse the results.
|
|
Results will be stored in a JSON format to be consumed by the 'KeyboardLayout' class.
|
|
As I do not control the website, I can't guarantee that this script will work in the future.
|
|
|
|
"""
|
|
|
|
def full_refresh():
|
|
loc_info = {}
|
|
import re
|
|
import urllib.request, urllib.error, urllib.parse
|
|
|
|
klids_to_filename = {}
|
|
names_to_filename = {}
|
|
shortnames_to_filename = {}
|
|
|
|
|
|
dirname = os.path.dirname(__file__)
|
|
layout_dir_path = os.path.join(dirname,'layouts/')
|
|
layout_manager_py_path = os.path.join(dirname,'layoutmanager.py')
|
|
with open(layout_dir_path + '__init__.py' , 'w') as f:
|
|
f.write('')
|
|
|
|
|
|
|
|
dlocations = {}
|
|
url = 'https://kbdlayout.info/'
|
|
response = urllib.request.urlopen(url)
|
|
webContent = response.read().decode('UTF-8')
|
|
|
|
#print(webContent)
|
|
hrefs = re.findall(r'href=[\'"]?([^\'" >]+)', webContent)
|
|
for href in hrefs:
|
|
if href.startswith('/00') is True or href.startswith('/k') is True or href.startswith('/K') is True:
|
|
dlocations[href[1:-1]] = 1
|
|
|
|
pbar = tqdm(dlocations)
|
|
i = 0
|
|
for loc in pbar:
|
|
#i += 1
|
|
#if i == 10:
|
|
# break
|
|
try:
|
|
iurl = url + '/' + loc +'/'
|
|
#print('Fetching url: %s' % kurl)
|
|
response = urllib.request.urlopen(iurl)
|
|
webContent = response.read().decode('UTF-8')
|
|
|
|
match = re.findall(r'(?<=<th>KLID</th><td>)(.*)(?=</td>|$)', webContent)
|
|
if match != '':
|
|
pbar.set_description('Processing %s' % ','.join(match)[:50].ljust(50))
|
|
loc_info[loc] = match
|
|
else:
|
|
loc_info[loc] = []
|
|
|
|
murl = url + '/' + loc +'/download/cldr'
|
|
response = urllib.request.urlopen(murl)
|
|
webContent = response.read().decode('UTF-8')
|
|
root = ET.fromstring(webContent)
|
|
locale = ''
|
|
names = []
|
|
if 'locale' in root.attrib:
|
|
locale = root.attrib['locale']
|
|
for entry in root:
|
|
if entry.tag == 'names':
|
|
for name in entry:
|
|
if name.tag == 'name':
|
|
if 'value' in name.attrib:
|
|
names.append(name.attrib['value'])
|
|
|
|
kurl = url + '/' + loc +'/download/xml'
|
|
#print('Fetching url: %s' % kurl)
|
|
response = urllib.request.urlopen(kurl)
|
|
webContent = response.read().decode('UTF-8')
|
|
#print('Parsing XML data...')
|
|
layout = kbdlayout_xml_to_kayboardlayout(loc, loc_info[loc], locale, names, webContent)
|
|
fname = 'layout_%s.py' % loc
|
|
layouts_file_path = os.path.join(layout_dir_path,fname)
|
|
|
|
with open(layouts_file_path, 'w', newline = '') as f:
|
|
layoutdata = 'layoutdata = "%s"\r\n' % base64.b64encode(layout.to_json().encode()).decode()
|
|
f.write(layoutdata)
|
|
|
|
|
|
for klid in loc_info[loc]:
|
|
shortname = klid.split(' ',1)[1][1:-1].replace('-','').lower()
|
|
klids_to_filename[klid] = fname
|
|
shortnames_to_filename[shortname] = fname
|
|
|
|
for name in names:
|
|
names_to_filename[name] = fname
|
|
|
|
|
|
except Exception as e:
|
|
print('Error! Parsing failed: %s' % loc)
|
|
traceback.print_exc()
|
|
continue
|
|
|
|
|
|
with open(layout_manager_py_path, 'w') as f:
|
|
f.write("""
|
|
import json
|
|
import base64
|
|
import importlib
|
|
|
|
from aardwolf.keyboard import KeyboardLayout
|
|
|
|
### Importlib is used here, as there are many keyboard layouts and you probably don't wish to use all of them
|
|
### Importing all of them would take a considerable time however, hence dynamic loading
|
|
|
|
class KeyboardLayoutManager:
|
|
def __init__(self):
|
|
self.klids_to_filename = json.loads(base64.b64decode("%s").decode())
|
|
self.names_to_filename = json.loads(base64.b64decode("%s").decode())
|
|
self.shortnames_to_filename = json.loads(base64.b64decode("%s").decode())
|
|
|
|
def __layout_loader(self, filename):
|
|
layoutdata = importlib.import_module('aardwolf.keyboard.layouts.%%s' %% filename[:-3])
|
|
return KeyboardLayout.from_layoutdata(layoutdata.layoutdata)
|
|
|
|
|
|
def get_layout_by_klid(self, klid):
|
|
if klid not in self.klids_to_filename:
|
|
return None
|
|
return self.__layout_loader(self.klids_to_filename[klid])
|
|
|
|
def get_layout_by_name(self, name):
|
|
if name not in self.names_to_filename:
|
|
return None
|
|
|
|
return self.__layout_loader(self.names_to_filename[name])
|
|
|
|
def get_layout_by_shortname(self, name):
|
|
if name not in self.shortnames_to_filename:
|
|
return None
|
|
return self.__layout_loader(self.shortnames_to_filename[name])
|
|
|
|
def get_klids(self):
|
|
for klid in self.klids_to_filename:
|
|
yield klid
|
|
|
|
def get_names(self):
|
|
for name in self.names_to_filename:
|
|
yield name
|
|
|
|
def get_shortnames(self):
|
|
for name in self.shortnames_to_filename:
|
|
yield name
|
|
|
|
if __name__ == '__main__':
|
|
kl = KeyboardLayoutManager()
|
|
kl.get_layout_by_shorname('sq')
|
|
""" % ( base64.b64encode(json.dumps(klids_to_filename).encode()).decode(), base64.b64encode(json.dumps(names_to_filename).encode()).decode(), base64.b64encode(json.dumps(shortnames_to_filename).encode()).decode() ))
|
|
|
|
|
|
|
|
def kbdlayout_xml_to_kayboardlayout(loc, klids, locale, layout_names, xml_data):
|
|
layout = KeyboardLayout()
|
|
layout.klid = klids
|
|
layout.name = layout_names
|
|
root = ET.fromstring(xml_data)
|
|
#root = tree.getroot()
|
|
#print(root)
|
|
#print(root.tag)
|
|
#print(root.attrib)
|
|
for k in root.attrib:
|
|
if k == 'RightAltIsAltGr':
|
|
layout.RightAltIsAltGr = True if root.attrib[k].lower() == 'true' else False
|
|
elif k == 'ShiftCancelsCapsLock':
|
|
layout.ShiftCancelsCapsLock = True if root.attrib[k].lower() == 'true' else False
|
|
elif k == 'ChangesDirectionality':
|
|
layout.ChangesDirectionality = True if root.attrib[k].lower() == 'true' else False
|
|
else:
|
|
print('Not porcessed attrib key: %s' % k)
|
|
input()
|
|
for child in root:
|
|
for entry in child:
|
|
if entry.tag == 'PK':
|
|
sc = int.from_bytes(bytes.fromhex((entry.attrib['SC'])), byteorder='big', signed = False)
|
|
layout.sc_to_vk[sc] = entry.attrib['VK']
|
|
|
|
for result in entry:
|
|
if result.tag == 'Result':
|
|
extra_with = []
|
|
if 'With' in result.attrib:
|
|
extra_with = result.attrib['With'].split(' ')
|
|
children = list(result.getchildren())
|
|
for rc in children:
|
|
if rc.tag == 'DeadKeyTable':
|
|
process_tag(layout, sc, rc, extra_with)
|
|
|
|
process_tag(layout, sc, result, [])
|
|
#if 'Text' in result.attrib:
|
|
# data = result.attrib['Text']
|
|
#if 'TextCodepoints' in result.attrib:
|
|
# data = result.attrib['TextCodepoints']
|
|
#
|
|
#modifiers = 0
|
|
#if 'With' in result.attrib:
|
|
# for wtag in result.attrib['With'].split(' '):
|
|
# modifiers |= VK_MODIFIERS[wtag]
|
|
#
|
|
#if sc not in layout.sc_to_char:
|
|
# layout.sc_to_char[sc] = {}
|
|
#
|
|
#if data is not None:
|
|
# layout.sc_to_char[sc][modifiers] = data
|
|
# layout.char_to_sc[data] = (sc, '')
|
|
# if modifiers != 0:
|
|
# layout.char_to_sc[data] = (sc, modifiers)
|
|
|
|
|
|
else:
|
|
print('Unknown tag %s' % entry.tag)
|
|
input()
|
|
|
|
return layout
|
|
|
|
def process_tag(layout, sc, result, extra_with):
|
|
data = None
|
|
if 'Text' in result.attrib:
|
|
data = result.attrib['Text']
|
|
if 'Accent' in result.attrib:
|
|
data = result.attrib['Accent']
|
|
if 'TextCodepoints' in result.attrib:
|
|
data = result.attrib['TextCodepoints']
|
|
|
|
modifiers = 0
|
|
for wtag in extra_with:
|
|
modifiers |= VK_MODIFIERS[wtag]
|
|
if 'With' in result.attrib:
|
|
for wtag in result.attrib['With'].split(' '):
|
|
modifiers |= VK_MODIFIERS[wtag]
|
|
|
|
if sc not in layout.sc_to_char:
|
|
layout.sc_to_char[sc] = {}
|
|
|
|
if data is not None:
|
|
layout.sc_to_char[sc][modifiers] = data
|
|
layout.char_to_sc[data] = (sc, '')
|
|
if modifiers != 0:
|
|
layout.char_to_sc[data] = (sc, modifiers)
|
|
|
|
if __name__ == '__main__':
|
|
full_refresh()
|
|
#pass |