Mercurial > hg > index.fcgi > www > www-1

annotate myrss/myrss_app.py @ 69:ae0f2f438a95

find changesets by author, revision, files, or words in the commit message
myrss: add support for new "purl" RSS type
author paulo
date Thu, 11 Jun 2015 22:03:34 -0700
parents 66a232bae83c
children 3456dd3e8660
rev   line source
paulo@39 1 import os
paulo@40 2 import sys
paulo@39 3 import re
paulo@40 4 import urllib2
paulo@40 5 import threading
paulo@40 6 import Queue
paulo@41 7 import datetime
paulo@41 8 import time
paulo@47 9
paulo@42 10 import logging
paulo@68 11 logging.basicConfig(
paulo@68 12 level=logging.INFO,
paulo@68 13 #filename="_LOG",
paulo@68 14 #format="%(asctime)s %(levelname)-8s %(message)s",
paulo@68 15 )
paulo@39 16
paulo@47 17 import xml.etree.ElementTree
paulo@47 18 import HTMLParser
paulo@47 19
paulo@39 20 import html
paulo@39 21
paulo@39 22
paulo@41 23 FEEDS_FILE = "FEEDS"
paulo@41 24 CACHE_HTML_FILE = "__cache__.html"
paulo@41 25
paulo@44 26 CACHE_LIFE = 1200 # [seconds]
paulo@47 27 MAX_ITEMS = 50
paulo@39 28 MAX_LINK_Z = 4
paulo@40 29 MAX_THREADS = 20
paulo@46 30 URLOPEN_TIMEOUT = 60 # [seconds]
paulo@39 31
paulo@39 32
paulo@39 33 _PARSE_ROOT_TAG_RE = re.compile(r"(\{(.+)\})?(.+)")
paulo@39 34
paulo@39 35 def _parse_root_tag(root_tag):
paulo@39 36 re_match = _PARSE_ROOT_TAG_RE.match(root_tag)
paulo@39 37
paulo@39 38 if re_match is None:
paulo@39 39 return (None, None)
paulo@39 40 else:
paulo@39 41 return re_match.group(2, 3)
paulo@39 42
paulo@39 43
paulo@47 44 def _strip_if_not_none(txt):
paulo@47 45 return txt.strip() if txt is not None else ''
paulo@47 46
paulo@47 47
paulo@39 48 def _go_rss(elementTree):
paulo@47 49 title = _strip_if_not_none(elementTree.find("channel/title").text)
paulo@39 50 link = elementTree.find("channel/link").text
paulo@39 51
paulo@39 52 items = []
paulo@39 53
paulo@39 54 for i in elementTree.findall("channel/item")[:MAX_ITEMS]:
paulo@47 55 it_title = _strip_if_not_none(i.find("title").text)
paulo@39 56 it_link = i.find("link").text
paulo@39 57
paulo@39 58 items.append((it_title, it_link))
paulo@39 59
paulo@39 60 return (title, link, items)
paulo@39 61
paulo@39 62
paulo@39 63 def _go_atom(elementTree):
paulo@39 64 ns = "http://www.w3.org/2005/Atom"
paulo@39 65
paulo@47 66 title = _strip_if_not_none(elementTree.find("{%s}title" % ns).text)
paulo@39 67 link = ''
paulo@39 68
paulo@39 69 for i in elementTree.findall("{%s}link" % ns):
paulo@39 70 if i.get("type") == "text/html" and i.get("rel") == "alternate":
paulo@39 71 link = i.get("href")
paulo@39 72 break
paulo@39 73
paulo@39 74 items = []
paulo@39 75
paulo@39 76 for i in elementTree.findall("{%s}entry" % ns)[:MAX_ITEMS]:
paulo@47 77 it_title = _strip_if_not_none(i.find("{%s}title" % ns).text)
paulo@39 78 it_link = ''
paulo@39 79
paulo@39 80 for j in i.findall("{%s}link" % ns):
paulo@39 81 if j.get("type") == "text/html" and j.get("rel") == "alternate":
paulo@39 82 it_link = j.get("href")
paulo@39 83 break
paulo@39 84
paulo@39 85 items.append((it_title, it_link))
paulo@39 86
paulo@39 87 return (title, link, items)
paulo@39 88
paulo@39 89
paulo@69 90 def _go_purl_rss(elementTree):
paulo@69 91 ns = "http://purl.org/rss/1.0/"
paulo@69 92
paulo@69 93 title = _strip_if_not_none(elementTree.find("{%s}channel/{%s}title" % (ns, ns)).text)
paulo@69 94 link = elementTree.find("{%s}channel/{%s}link" % (ns, ns)).text
paulo@69 95
paulo@69 96 items = []
paulo@69 97
paulo@69 98 for i in elementTree.findall("{%s}item" % ns)[:MAX_ITEMS]:
paulo@69 99 it_title = _strip_if_not_none(i.find("{%s}title" % ns).text)
paulo@69 100 it_link = i.find("{%s}link" % ns).text
paulo@69 101
paulo@69 102 items.append((it_title, it_link))
paulo@69 103
paulo@69 104 return (title, link, items)
paulo@69 105
paulo@69 106
paulo@47 107 _STRIP_HTML_RE = re.compile(r"<.*?>")
paulo@47 108 _htmlParser = HTMLParser.HTMLParser()
paulo@47 109
paulo@47 110 def _strip_html(txt):
paulo@47 111 return _htmlParser.unescape(_STRIP_HTML_RE.sub('', txt))
paulo@47 112
paulo@47 113
paulo@41 114 def _to_html(dtnow, docstruct):
paulo@41 115 datetime_str = dtnow.strftime("%Y-%m-%d %H:%M %Z")
paulo@41 116 page_title = "myrss -- %s" % datetime_str
paulo@41 117
paulo@42 118 root = html.HTML("html")
paulo@39 119
paulo@39 120 header = root.header
paulo@41 121 header.title(page_title)
paulo@39 122 header.link(rel="stylesheet", type="text/css", href="index.css")
paulo@39 123
paulo@41 124 body = root.body
paulo@41 125 body.h1(page_title)
paulo@41 126
paulo@39 127 link_z = 0
paulo@39 128
paulo@39 129 for feed in docstruct:
paulo@40 130 if feed is None:
paulo@40 131 continue
paulo@40 132
paulo@39 133 (title, link, items) = feed
paulo@39 134
paulo@47 135 body.h2.a(_strip_html(title), href=link, klass="z%d" % (link_z % MAX_LINK_Z))
paulo@39 136 link_z += 1
paulo@41 137 p = body.p
paulo@39 138
paulo@39 139 for (i, (it_title, it_link)) in enumerate(items):
paulo@39 140 if i > 0:
paulo@39 141 p += " - "
paulo@39 142
paulo@47 143 p.a(_strip_html(it_title), href=it_link, klass="z%d" % (link_z % MAX_LINK_Z))
paulo@39 144 link_z += 1
paulo@39 145
paulo@46 146 dtdelta = datetime.datetime.now() - dtnow
paulo@46 147 root.div("%.3f" % (dtdelta.days*86400 + dtdelta.seconds + dtdelta.microseconds/1e6), klass="debug")
paulo@46 148
paulo@39 149 return unicode(root).encode("utf-8")
paulo@39 150
paulo@39 151
paulo@47 152 def _fetch_url(url):
paulo@40 153 try:
paulo@42 154 logging.info("processing %s" % url)
paulo@46 155 feed = urllib2.urlopen(urllib2.Request(url, headers={"User-Agent": ''}), timeout=URLOPEN_TIMEOUT)
paulo@40 156 except urllib2.HTTPError as e:
paulo@42 157 logging.info("(%s) %s" % (url, e))
paulo@47 158 return None
paulo@47 159
paulo@47 160 return feed
paulo@47 161
paulo@47 162
paulo@47 163 def _process_feed(feed):
paulo@47 164 ret = None
paulo@40 165
paulo@40 166 elementTree = xml.etree.ElementTree.parse(feed)
paulo@40 167 root = elementTree.getroot()
paulo@40 168
paulo@40 169 parsed_root_tag = _parse_root_tag(root.tag)
paulo@40 170
paulo@40 171 if parsed_root_tag == (None, "rss"):
paulo@40 172 version = float(root.get("version", 0.0))
paulo@40 173 if version >= 2.0:
paulo@40 174 ret = _go_rss(elementTree)
paulo@40 175 else:
paulo@40 176 raise NotImplementedError("Unsupported rss version")
paulo@40 177 elif parsed_root_tag == ("http://www.w3.org/2005/Atom", "feed"):
paulo@40 178 ret = _go_atom(elementTree)
paulo@69 179 elif parsed_root_tag == ("http://www.w3.org/1999/02/22-rdf-syntax-ns#", "RDF"):
paulo@69 180 ret = _go_purl_rss(elementTree)
paulo@40 181 else:
paulo@40 182 raise NotImplementedError("Unknown root tag")
paulo@40 183
paulo@40 184 return ret
paulo@40 185
paulo@40 186
paulo@40 187 class WorkerThread(threading.Thread):
paulo@40 188 def __init__(self, *args, **kwargs):
paulo@40 189 self._input_queue = kwargs.pop("input_queue")
paulo@40 190 self._output_queue = kwargs.pop("output_queue")
paulo@40 191 threading.Thread.__init__(self, *args, **kwargs)
paulo@40 192 self.daemon = True
paulo@40 193
paulo@40 194 def run(self):
paulo@40 195 while True:
paulo@40 196 (idx, url) = self._input_queue.get()
paulo@40 197 docfeed = None
paulo@40 198 try:
paulo@47 199 feed = _fetch_url(url)
paulo@47 200 if feed is not None:
paulo@47 201 docfeed = _process_feed(feed)
paulo@40 202 except Exception as e:
paulo@42 203 logging.info("(%s) exception: %s" % (url, e))
paulo@40 204 self._output_queue.put((idx, docfeed))
paulo@40 205
paulo@40 206
paulo@44 207 def main(input_queue, output_queue, lock):
paulo@41 208 ret = ''
paulo@41 209
paulo@44 210 with lock:
paulo@44 211 epoch_now = time.time()
paulo@44 212 dtnow = datetime.datetime.fromtimestamp(epoch_now)
paulo@41 213
paulo@44 214 if os.path.exists(CACHE_HTML_FILE) and (epoch_now - os.stat(CACHE_HTML_FILE).st_mtime) < float(CACHE_LIFE):
paulo@44 215 with open(CACHE_HTML_FILE) as cache_html_file:
paulo@44 216 ret = cache_html_file.read()
paulo@41 217
paulo@44 218 else:
paulo@44 219 with open(FEEDS_FILE) as feeds_file:
paulo@44 220 feedlines = feeds_file.readlines()
paulo@41 221
paulo@44 222 docstruct = [None]*len(feedlines)
paulo@44 223 num_input = 0
paulo@44 224 for (i, l) in enumerate(feedlines):
paulo@44 225 if l[0] != '#':
paulo@44 226 l = l.strip()
paulo@44 227 input_queue.put((i, l))
paulo@44 228 num_input += 1
paulo@41 229
paulo@44 230 for _ in range(num_input):
paulo@44 231 (idx, docfeed) = output_queue.get()
paulo@44 232 docstruct[idx] = docfeed
paulo@41 233
paulo@44 234 ret = _to_html(dtnow, docstruct)
paulo@41 235
paulo@44 236 with open(CACHE_HTML_FILE, 'w') as cache_html_file:
paulo@44 237 cache_html_file.write(ret)
paulo@41 238
paulo@41 239 return ret
paulo@41 240
paulo@41 241
paulo@42 242 class MyRssApp:
paulo@42 243 def __init__(self):
paulo@42 244 self._iq = Queue.Queue(MAX_THREADS)
paulo@42 245 self._oq = Queue.Queue(MAX_THREADS)
paulo@44 246 self._main_lock = threading.Lock()
paulo@39 247
paulo@42 248 for _ in range(MAX_THREADS):
paulo@42 249 WorkerThread(input_queue=self._iq, output_queue=self._oq).start()
paulo@42 250
paulo@42 251 def __call__(self, environ, start_response):
paulo@44 252 response_body = main(self._iq, self._oq, self._main_lock)
paulo@42 253 response_headers = [
paulo@47 254 ("Content-Type", "text/html; charset=UTF-8"),
paulo@42 255 ("Content-Length", str(len(response_body))),
paulo@42 256 ]
paulo@42 257 start_response("200 OK", response_headers)
paulo@42 258
paulo@42 259 return [response_body]