rvt123

#!/usr/bin/env python3
# task_complete.py
"""Task completion sound using Windows from WSL2 - no pip install needed"""
import subprocess

def task_complete():
    try:
        # Call Windows PowerShell to play a beep
        # Format: powershell.exe -Command "[console]::beep(frequency, duration)"
        subprocess.run(

rvt123

HANDOFF.md

Maintain a persistent handoff file for continuity.

Update after each task:

• current status
• next steps
• test results
• environment details

rvt123

## Completion notification

At the end of every task:

- Run: ~/.codex/notify.py

Purpose:
- Notify user when task is complete
- Avoid manual polling
- Enable async workflows

rvt123

name	caveman-compress
description	Compress prose in a file using caveman-style brevity while preserving full technical meaning and structure. Creates a backup and rewrites the original file.
user-invocable	true
argument-hint	[file path]

Purpose

Reduce token usage in large markdown/docs while keeping meaning intact.

rvt123

Default response style

Use caveman-style brevity by default.

Default = Full. Auto-fallback = Lite when readability would suffer.

Rules:

• Keep technical meaning exact
• Remove filler, hedging, repetition
• Answer first, then reasoning, then next step

rvt123

tfidf = TfidfVectorizer(tokenizer=lambda x: x,preprocessor=lambda x: x,ngram_range=(1,1),min_df=10)
tfidf_vectorizer_vectors_train = tfidf.fit_transform(df_train.loc[:,'clean_text'].values)
features = tfidf_vectorizer_vectors_train.toarray()
vocab = dict(zip(tfidf.get_feature_names(),features[:,:len(tfidf.get_feature_names())].sum(axis=0)))

embeddings_dict = {}
with open("glove.6B.100d.txt", 'r', encoding='utf-8') as f:
    for line in f:
        values = line.split()
        token = values[0]

rvt123

def clean_data(data,col,re_emoji,re_inc_boostr,re_dec_boostr,re_pos,re_neg,re_bad):
    dataframe = data.copy()
    link_regex = re.compile(r'(?:ftp|https?|www|file)\.?:?[//|\\\\]?[\w\d:#@%/;$()~_?\+-=\\\&]+\.[\w\d:#@%/;$~_?\+-=\\\&]+')
    dataframe = dataframe.assign(**dict(zip([col,'num_link'], zip(*dataframe['text'].apply(lambda x: re.subn(link_regex,'LINK',x) ) ))))
    dataframe = dataframe.assign(**dict(zip([col,'num_usermention'], zip(*dataframe[col].apply(lambda x: re.subn(r'@[\w]*','USERMENTION',x)) ))))
    dataframe = dataframe.assign(**dict(zip([col,'num_hashtag'], zip(*dataframe[col].apply(lambda x: re.subn(r'#[\w]*','HASHTAG',x) ) ))))
    dataframe = dataframe.assign(**dict(zip([col,'num_emoji'], zip(*dataframe[col].apply(lambda x: re.subn(re_emoji, lambda m: EMOJIS.get(m.group(), 'EMOJI') , x) )))))
    dataframe[col] = dataframe[col].apply(lambda x : re.sub(r'(.)\1{2,}', r'\1',x)) # make looong as long
    dataframe[col] = dataframe[col].apply(lambda x : expandContractions(x) ) # expand cont

rvt123

DATA_LIST = [] ## list to which all data is appended to 
SMALL_DICT = {} # Dictionary to which countries are added if there is some error
for index,country_name_id in enumerate(COUNTRIES_DICT.items()): # Looping in the all the countries 
    try: # try and except block to catch error 
        country_name,country_id = country_name_id
        print(index,country_name)
        ser = Service('./chromedriver.exe')
        chrome_options = webdriver.ChromeOptions()
        chrome_options.add_argument("--incognito")
        driver = webdriver.Chrome(service=ser,options=chrome_options)

rvt123

DATA_LIST = []
country_name = 'United Kingdom'
ser = Service('./chromedriver.exe')
chrome_options = webdriver.ChromeOptions()
chrome_options.add_argument("--incognito")
driver = webdriver.Chrome(service=ser,options=chrome_options)
driver.get(BASE_URL + COUNTRIES_DICT.get(country_name))
element = driver.find_element(By.XPATH,'//*[@id="disclaimer-content"]/button')
driver.execute_script("arguments[0].click();", element)

rvt123

def get_data(page_source,DATA,category,country,page_num):
    content = BeautifulSoup(page_source,'html5lib')
    soups = content.findAll('div',{'class':'thumb-block'})
    for so in soups:
        if 'thumb-ad' not in so.get('class'):
            this_data = {}
            this_data['TITLE'] = so.find('p',{'class':'title'}).find('a').get('title')
            meta_data = so.find('p',{'class':'metadata'})
            quality = meta_data.find('span',{'class':'video-hd-mark'})
            if quality: