I think C is generating spaces without permission

Question

I was writing, in C, a program that gives recommendations from google with the help of another python script. An interesting thing happened when I finally compiled the program and tested it out and as result the output of the python script came out with extra spaces. I searched it on the net but couldn't find anything so here I am.

C code:

#include <stdio.h>
#include <stdlib.h>


void search(){
    system("clear");

    char search[200];
    char a;
    char lsearch[800];
    char esearch[2000];

    printf("\n*Suggestions for your Search*\n\n\n\n");
    printf("> ");
    system ("/bin/stty raw");
    int i = 0;
    while(i > -1) {
        a = getchar();
        if(a == ' ') {
            system("clear");
            search[i] = '!';
            fflush(stdout);
            snprintf(lsearch, 800, "cd python-bash-things && python3 rec.py %s",search);
            fflush(stdout);
            system(lsearch);

            search[i] = ' ';
            printf("\n\n\n> %s",search);
            i++;
            continue;
        }
        else if(a == 3) 
            return;
        else if(a == 13) {
            snprintf(esearch, 2000, "firefox -search '%s'",search);     
            break;
        }
        search[i] = a;
        i++;
    }
    system("clear");
    system ("/bin/stty cooked");
    system(esearch);
}

int main(){
    search();
    return 0;
}

Python code:

#!/usr/bin/env python3

import json
import re
import urllib.parse
import urllib.request
import sys
import os
import datetime
import gzip
import subprocess as sp
import html

SEARCH_ENGINE = 'google'
BROWSER = 'firefox'
TERMINAL = ['gnome-terminal', '--']
CONFIG = {
    'BROWSER_PATH' : {
        'chrome' : ['google-chrome-stable'],
        'firefox' : ['firefox'],
        'chromium' : ['chromium-browser'],
        'brave' : ['brave-browser'],
        'lynx' : TERMINAL + ['lynx']
    },
    'USER_AGENT' : {
        'chrome' : 'Mozilla/5.0 (X11; Fedora; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/77.0.3865.90 Safari/537.36',
        'firefox' : 'Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:69.0) Gecko/20100101 Firefox/69.0',
        'chromium' : 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Ubuntu Chromium/76.0.3809.100 Chrome/76.0.3809.100 Safari/537.36',
        'brave' : 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/77.0.3865.120 Safari/537.36',
        'lynx' : 'Lynx/2.8.9rel.1 libwww-FM/2.14 SSL-MM/1.4.1 OpenSSL/1.1.1d'
    },
    'SEARCH_ENGINE_NAME' : {
        'google' : 'Google',
        'duckduckgo' : 'DuckDuckGo'
    },
    'SEARCH_URL' : {
        'google' : 'https://www.google.com/search?q=',
        'duckduckgo' : 'https://duckduckgo.com/?q='
    },
    'SUGGESTION_URL' : {
        'google' : 'https://www.google.com/complete/search?',
        'duckduckgo' : 'https://duckduckgo.com/ac/?'
    }
}

def cleanhtml(txt):
    return re.sub(r'<.*?>', '', txt)

def fetch_suggestions(search_string):
    if SEARCH_ENGINE == 'google':
        r = {
            'q' : search_string,
            'cp' : '11',
            'client' : 'psy-ab',
            'xssi' : 't',
            'gs_ri' : 'gws-wiz',
            'hl' : 'en-IT',
            'authuser' : '0'
        }
        url = CONFIG['SUGGESTION_URL'][SEARCH_ENGINE] + urllib.parse.urlencode(r)
        headers = {
            'sec-fetch-mode' : 'cors',
            'dnt' : '1',
            'accept-encoding' : 'gzip',
            'accept-language' : 'en-US;q=0.9,en;q=0.8',
            'pragma' : 'no-cache',
            'user-agent' : CONFIG['USER_AGENT'][BROWSER],
            'accept' : '*/*',
            'cache-control' : 'no-cache',
            'authority' : 'www.google.com',
            'referer' : 'https://www.google.com/',
            'sec-fetch-site' : 'same-origin'
        }
        req = urllib.request.Request(url, headers=headers, method='GET')

        reply_data = gzip.decompress(urllib.request.urlopen(req).read()).split(b'\n')[1]
        reply_data = json.loads(reply_data)
        return [ cleanhtml(res[0]).strip() for res in reply_data[0] ]
    else:  
        if search_string.startswith('!'):
            bang_search = True
            search_string = search_string.lstrip('!')
        else:
            bang_search = False
        r = {
            'q' : search_string,
            'callback' : 'autocompleteCallback',
            'kl' : 'wt-wt',
            '_' : str(int((datetime.datetime.now().timestamp())*1000))
        }
        url = CONFIG['SUGGESTION_URL'][SEARCH_ENGINE] + urllib.parse.urlencode(r)
        if bang_search:
            url = url.replace('?q=', '?q=!')
        headers = {
            'pragma' : 'no-cache',
            'dnt' : '1',
            'accept-encoding' : 'gzip',
            'accept-language' : 'en-US;q=0.9,en;q=0.8',
            'user-agent' : CONFIG['USER_AGENT'][BROWSER],
            'sec-fetch-mode' : 'no-cors',
            'accept' : '*/*',
            'cache-control' : 'no-cache',
            'authority' : 'duckduckgo.com',
            'referer' : 'https://duckduckgo.com/',
            'sec-fetch-site' : 'same-origin',
        }
        req = urllib.request.Request(url, headers=headers, method='GET')
        reply_data = gzip.decompress(urllib.request.urlopen(req).read()).decode('utf8')
        reply_data = json.loads(re.match(r'autocompleteCallback\((.*)\);', reply_data).group(1))
        return [ cleanhtml(res['phrase']).strip() for res in reply_data ]

def main():
    search_string = html.unescape((' '.join(sys.argv[1:])).strip())

    if search_string.endswith('!'):
        
        search_string = search_string.rstrip('!').strip()
        results = fetch_suggestions(search_string)
        for r in results:
            print(html.unescape(r))
    else:
        url = CONFIG['SEARCH_URL'][SEARCH_ENGINE] + urllib.parse.quote_plus(search_string)
        sp.Popen(CONFIG['BROWSER_PATH'][BROWSER] + [url], stdout=sp.DEVNULL, stderr=sp.DEVNULL, shell=False)

def validate_config(c):
    if type(c) != dict:
        print('Configuration file must be a JSON object', file=sys.stderr)
        sys.exit(1)
    for k in ('SEARCH_ENGINE', 'BROWSER', 'TERMINAL'):
        if k not in c:
            print('Configuration file is missing %s' % k, file=sys.stderr)
            sys.exit(1)
    for k in ('SEARCH_ENGINE', 'BROWSER'):
        if type(c[k]) != str:
            print('Configuration Error: The value of %s must be a string' % k, file=sys.stderr)
    if type(c['TERMINAL']) != list:
        print('Configuration Error: The value of TERMINAL must be a list of strings', file=sys.stderr)
        sys.exit(1)
    for x in c['TERMINAL']:
        if type(x) != str:
            print('Configuration Error: The value of TERMINAL must be a list of strings', file=sys.stderr)
            sys.exit(1)

if __name__ == "__main__":
    try:
        fname = os.path.expanduser('~/.config/rofi-web-search/config.json')
        if os.path.exists(fname):
            try:
                config = json.loads(open(fname, 'r').read())
            except json.JSONDecodeError:
                print('Configuration file %s is not a valid JSON' % fname, file=sys.stderr)
                sys.exit(1)
            validate_config(config)
            SEARCH_ENGINE = config['SEARCH_ENGINE']
            BROWSER = config['BROWSER']
            TERMINAL = config['TERMINAL']
        else:
            config = {
                    'SEARCH_ENGINE' : SEARCH_ENGINE,
                    'BROWSER' : BROWSER,
                    'TERMINAL' : TERMINAL
                }
            os.makedirs(os.path.dirname(fname))
            f = open(fname, 'w')
            f.write(json.dumps(config, indent=4))
            f.write('\n')
            f.close()
        main()
    except:
        sys.exit(1)

output after typing mario:

marion cotillard
                mario
                     mario balotelli
                                    mario badescu
                                                 mario götze
                                                            mario gomez
                                                                       mario oyna
                                                                                 mario mandzukic
          mariobet
                  mario lemina



                              > mario 

All answers are appreciated

Edit: An interesting thing is that the terminal demonstrates the same behavior after running the program.

Answer 1

The problem code is:

char search[200];
...
search[i] = '!';
... and other assignments to search
snprintf(esearch, 2000, "firefox -search '%s'",search);//search has never been `\0` terminated

By calling a string function with a string that is not null terminated is undefined behavior.

And, from another post:

if you have a string that is not null-terminated, you cannot use the C string manipulation routines on it. You can't use strlen, strcpy, [snprintf] or strcat . Basically, any function that takes a char* but no separate length is not usable. (emphasis mine)

To address this one issue, simply initialize the buffer at the time of its declaration:

char search[200] = {0};

Then limit the content written to search to 199 characters, (search[198]) so as not to overwrite the null terminator which should be preserved at position search[199].