ALGORITHM - Execute steps in order, try ALL fallbacks before giving up:

STEP 1: GitHub API with full title
curl -s 'https://api.github.com/search/repositories?q=FULL+PAPER+TITLE' | python3 -c "import sys,json; items=json.load(sys.stdin).get('items',[]); print(items[0]['html_url'] if items else 'NOT_FOUND')"
-> If found, verify: curl -s -o /dev/null -w '%{http_code}' URL (must be 200)
-> If verified, DONE. Otherwise continue to step 2.

STEP 2: GitHub API with acronym/keywords (if full title fails)
Extract key words or acronym from title (e.g., 'STAIR', 'locate-3d', 'deepcluster')
curl -s 'https://api.github.com/search/repositories?q=ACRONYM' | python3 -c "import sys,json; items=json.load(sys.stdin).get('items',[]); print(items[0]['html_url'] if items else 'NOT_FOUND')"
-> Verify with curl, if 200 -> DONE

STEP 3: Arxiv + PDF extraction (if GitHub search fails)

Find on arxiv

curl -s 'http://export.arxiv.org/api/query?search_query=all:PAPER+TITLE&max_results=1' | grep -o 'http://arxiv.org/abs/[0-9.v]*'

Download PDF

curl -L -o paper.pdf 'http://arxiv.org/pdf/ARXIV_ID'

Extract github link

pip install PyPDF2 -q
python3 -c "import PyPDF2,re; f=open('paper.pdf','rb'); r=PyPDF2.PdfReader(f); text=''.join(p.extract_text() or '' for p in r.pages); m=re.search(r'https://github\.com/[^\\s]+',text); print(m.group(0) if m else 'NOT_FOUND')"
-> Verify with curl, if 200 -> DONE

STEP 4: If ALL steps fail, return empty string "" (not a guess!)

KEY: Iterate through ALL steps. Do not stop after step 1 fails. Empty > wrong.

Systematic network debugging method:

1. Check if it's DNS: nslookup example.com or dig example.com
2. Check if it's routing: ping example.com or traceroute example.com
3. Check if it's firewall: iptables -L -n and check for DROP/REJECT rules
4. Check /etc/hosts for overrides that redirect the domain
5. Check if proxy is misconfigured: echo $http_proxy $https_proxy
6. Check cron-specific: cron jobs have minimal PATH and no env vars - add full paths and source profiles
7. For cron network issues specifically: check if cron has network access (some systems restrict it)

Common cron gotcha: cron runs with minimal environment. If curl works in shell but not cron, the issue is usually PATH or environment variables.

Method for data-to-JSON transformation:

1. Identify the input format (CSV, text, nested structure)
2. Use Python for complex parsing - it handles edge cases better than bash:

import json
data = [] # collect records
for item in source:
record = {'field1': value1, 'field2': value2}
data.append(record)

For JSONL (one JSON object per line):

with open('output.jsonl', 'w') as f:
for record in data:
f.write(json.dumps(record) + '\n')

For JSON array:

with open('output.json', 'w') as f:
json.dump(data, f, indent=2)

3. Validate output: python -m json.tool output.json
4. For nested/hierarchical data, build the tree structure first, then serialize

Method for binary string extraction:

1. Use grep -aob 'PATTERN' file.dat to find byte offset of known pattern
2. Use dd to extract a window around that offset: dd if=file.dat bs=1 skip=OFFSET count=SIZE
3. Filter to valid characters: tr -c 'A-Z0-9' ' ' (keeps only uppercase and digits)
4. For contiguous sequences: tr ' ' '\n' | grep -v '^$' | grep 'PATTERN'
5. Find longest match: awk '{ print length, $0 }' | sort -nr | head -n1

If you know the string starts with X and ends with Y, find both anchors separately, then look for overlap or combine the fragments.

One-liner method for self-signed cert:

openssl req -x509 -newkey rsa:4096 -keyout key.pem -out cert.pem -days 365 -nodes -subj '/CN=localhost'

Breakdown:

• req -x509: Generate self-signed cert (not CSR)
• -newkey rsa:4096: Create new 4096-bit RSA key
• -keyout key.pem: Private key output
• -out cert.pem: Certificate output
• -days 365: Valid for 1 year
• -nodes: No password on private key
• -subj '/CN=localhost': Skip interactive prompts, set Common Name

For SAN (Subject Alternative Names) add: -addext 'subjectAltName=DNS:localhost,IP:127.0.0.1'