Subj : The ARRL Letter
To   : Mortar M.
From : Sean Dennis
Date : Mon Aug 17 2026 10:09 pm

Hello Mortar,

17 Aug 26 09:40, you wrote to me:

 MM> Is this OK with ARRL?  Site owners can get rather touchy about
 MM> scraping.

From the bottom of each ARRL Letter:

"Copyright (C) 2026 American Radio Relay League, Incorporated. Use and 
distribution of this publication, or any portion thereof, is permitted for 
non-commercial or educational purposes, with attribution. All other purposes 
require written permission."

I make sure that's left in every "letter" I scrape.

 MM> Can you post the script?  I'm interested in how that's done.

=== Cut ===
#!/usr/bin/env python3

### ARRL Letter Posting Script
###
### By Sean Dennis with assistance
### from Microsoft Copilot
###
### (C) Sean Dennis KS4TD
###
### Released under the MIT License.

import requests
import re
import textwrap
import unicodedata
import os
import sys
import time

BASE = "https://www.arrl.org"
LETTER_LIST = "https://www.arrl.org/arrlletter"
CACHE_FILE = os.path.expanduser("~/.arrlletter_cache")

# ------------------------------------------------------------
# Retry-capable fetcher with browser headers
# ------------------------------------------------------------
def fetch_with_retry(url, retries=5, delay=3):
    headers = {
        "User-Agent": (
            "Mozilla/5.0 (X11; Linux x86_64) "
            "AppleWebKit/537.36 (KHTML, like Gecko) "
            "Chrome/124.0 Safari/537.36"
        )
    }

    for attempt in range(1, retries + 1):
        try:
            print("Fetching {} (attempt {}/{})".format(url, attempt, retries))
            r = requests.get(url, headers=headers, timeout=20)
            r.raise_for_status()
            return r.text
        except Exception as e:
            print("Attempt {} failed: {}".format(attempt, e))
            time.sleep(delay)

    print("ERROR: All attempts failed.")
    return None

# ------------------------------------------------------------
# Utility functions
# ------------------------------------------------------------
def log(msg):
    print(msg)

def to_ascii(s):
    return unicodedata.normalize("NFKD", s).encode("ascii", 
"ignore").decode("ascii")

def load_cached_issue():
    if not os.path.exists(CACHE_FILE):
        return None
    try:
        with open(CACHE_FILE, "r") as f:
            return f.read().strip()
    except:
        return None

def save_cached_issue(issue):
    try:
        with open(CACHE_FILE, "w") as f:
            f.write(issue)
    except Exception as e:
        log("Warning: could not save cache: {}".format(e))

# ------------------------------------------------------------
# 1. Fetch ARRL Letter index page (with retry)
# ------------------------------------------------------------
index_html = fetch_with_retry(LETTER_LIST)
if index_html is None:
    sys.exit(1)

# ------------------------------------------------------------
# 2. Find newest issue link (single or double quotes)
# ------------------------------------------------------------
issues = 
re.findall(r"href=['\"](/arrlletterissue\?issue=\d{4}-\d{2}-\d{2})['\"]", 
index_html)
if not issues:
    log("ERROR: No ARRL Letter issues found.")
    sys.exit(1)

latest_issue_path = issues[0]
latest_issue_url = BASE + latest_issue_path
latest_issue_id = latest_issue_path.split("=")[-1]

log("Latest ARRL Letter issue: {}".format(latest_issue_id))

# ------------------------------------------------------------
# 3. Check cache
# ------------------------------------------------------------
cached = load_cached_issue()
if cached == latest_issue_id:
    log("Cached issue matches latest. Nothing new to download.")
    sys.exit(0)

# ------------------------------------------------------------
# 4. Download the issue HTML (with retry)
# ------------------------------------------------------------
page_html = fetch_with_retry(latest_issue_url)
if page_html is None:
    sys.exit(1)

# ------------------------------------------------------------
# 5. Improved HTML cleanup
# ------------------------------------------------------------
import html

# Remove scripts and styles
clean = re.sub(r"<script.*?>.*?</script>", "", page_html, flags=re.DOTALL)
clean = re.sub(r"<style.*?>.*?</style>", "", clean, flags=re.DOTALL)

# Remove all HTML tags
clean = re.sub(r"<[^>]+>", "", clean)

# Decode HTML entities (&nbsp;, &#39;, etc.)
clean = html.unescape(clean)

# Remove "undefined" junk lines
clean = clean.replace("undefined", "")

# Collapse multiple spaces
clean = re.sub(r"[ \t]+", " ", clean)

# Remove repeated blank lines
clean = re.sub(r"\n\s*\n\s*\n+", "\n\n", clean)

# Remove repeated photo captions (ARRL duplicates them)
clean = re.sub(r" \[Photo.*?\] ", "", clean)

# Remove ARRL navigation/header lines
clean = clean.replace("ARRL Home Page", "")
clean = clean.replace("ARRL Audio News", "")
clean = clean.replace("ARRL Letter Archive", "")

# Remove unsubscribe footer
clean = clean.replace("Unsubscribe from this list.", "")

# Clean up leftover blank lines
clean = re.sub(r"\n\s*\n+", "\n\n", clean)

# Normalize line endings
clean = clean.strip()

# Normalize line endings
clean = clean.strip()

# ------------------------------------------------------------
# 6. Convert to ASCII
# ------------------------------------------------------------
ascii_text = to_ascii(clean)

1# ------------------------------------------------------------
# 7. Normalize whitespace and wrap at 72 columns
# ------------------------------------------------------------
out_lines = []
for line in ascii_text.splitlines():
    line = line.strip()
    if not line:
        out_lines.append("")
        continue
    wrapped = textwrap.wrap(line, width=72)
    out_lines.extend(wrapped)

# ------------------------------------------------------------
# 8. Write final file
# ------------------------------------------------------------
output_file = "/tmp/arrlletter-latest.asc"
try:
    with open(output_file, "w") as f:
        f.write("\n".join(out_lines))
    log("Created {}".format(output_file))
    save_cached_issue(latest_issue_id)
except Exception as e:
    log("ERROR: Could not write output file: {}".format(e))
    sys.exit(1)
# ------------------------------------------------------------
=== Cut ===

Here's the actual script called by MBSE (note that the "mbmsg" lines are 
wrapped):

=== Cut ===
#!/bin/bash

# Environment
# PATH needed since cron doesn't inherit bash's environment
export PATH=/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin
# Required by MBSE
export MBSE_ROOT=/opt/mbse

$MBSE_ROOT/bin/arrl_letter.py

# Post in MIN_HAM
mbmsg post "ARS KS4TD" "All" 61 "The ARRL Letter" "/tmp/arrlletter-latest.txt" 

# Post in Fidonet's HAM
mbmsg post "ARS KS4TD" "All" 117 "The ARRL Letter" 
"/tmp/arrlletter-latest.txt" -
=== Cut ===

I believe the Python script will run under Windows with minor modifications.

-- Sean

.... "I have a love interest in every one of my films: a gun." - Schwarzenegger
--- GoldED+/LNX 1.1.5-b20260304
 * Origin: Outpost BBS * Johnson City, TN (1:18/200)

.