merge_feeds.py

import os
import xml.etree.ElementTree as ET

# Merge multiple RSS feeds into a single feed, ensuring unique items by guid

feeds_dir = "feeds"
output_file = "webservices.xml"

# Store unique items by guid
items_by_guid = {}

# Store channel metadata from the first file
channel_metadata = None

for filename in sorted(os.listdir(feeds_dir)):
    if not filename.endswith(".xml"):
        continue
    path = os.path.join(feeds_dir, filename)
    try:
        tree = ET.parse(path)
        root = tree.getroot()
        channel = root.find("channel")
        if channel is None:
            continue
        # Save channel metadata from the first file
        if channel_metadata is None:
            channel_metadata = [elem for elem in channel if elem.tag != "item"]
        # Process items
        for item in channel.findall("item"):
            guid_elem = item.find("guid")
            if guid_elem is not None:
                guid = guid_elem.text
                if guid not in items_by_guid:
                    items_by_guid[guid] = item
    except Exception as e:
        print(f"Error processing {filename}: {e}")

# Build the combined RSS feed
rss = ET.Element("rss", version="2.0")
channel = ET.SubElement(rss, "channel")
if channel_metadata:
    for elem in channel_metadata:
        channel.append(elem)
for item in items_by_guid.values():
    channel.append(item)

# Write to output file
tree = ET.ElementTree(rss)
tree.write(output_file, encoding="utf-8", xml_declaration=True)
print(f"Combined RSS written to {output_file}")