You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@petri.apache.org by wa...@apache.org on 2021/03/25 16:56:39 UTC
[petri-site] branch master updated: Delete sitemap.py
This is an automated email from the ASF dual-hosted git repository.
wave pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/petri-site.git
The following commit(s) were added to refs/heads/master by this push:
new 8cb882b Delete sitemap.py
8cb882b is described below
commit 8cb882be68f11eb96c04dddb0d9c4105c03a93ff
Author: Dave Fisher <da...@comcast.net>
AuthorDate: Thu Mar 25 09:56:31 2021 -0700
Delete sitemap.py
---
plugins/sitemap/sitemap.py | 290 ---------------------------------------------
1 file changed, 290 deletions(-)
diff --git a/plugins/sitemap/sitemap.py b/plugins/sitemap/sitemap.py
deleted file mode 100644
index 45763bb..0000000
--- a/plugins/sitemap/sitemap.py
+++ /dev/null
@@ -1,290 +0,0 @@
-# -*- coding: utf-8 -*-
-"""
-Sitemap
--------
-
-The sitemap plugin generates plain-text or XML sitemaps.
-"""
-
-from __future__ import unicode_literals
-
-from codecs import open
-import collections
-from datetime import datetime
-from logging import info, warning
-import os.path
-import re
-
-from pytz import timezone
-
-from pelican import contents, signals
-from pelican.utils import get_date
-
-TXT_HEADER = ""
-
-XML_HEADER = """<?xml version="1.0" encoding="utf-8"?>
-<urlset xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
-xsi:schemaLocation="http://www.sitemaps.org/schemas/sitemap/0.9 http://www.sitemaps.org/schemas/sitemap/0.9/sitemap.xsd"
-xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
-"""
-
-XML_URL = """
-<url>
-<loc>{0}/{1}</loc>
-<lastmod>{2}</lastmod>
-<changefreq>{3}</changefreq>
-<priority>{4}</priority>
-</url>
-"""
-
-XML_FOOTER = """
-</urlset>
-"""
-
-
-def format_date(date):
- if date.tzinfo:
- tz = date.strftime("%z")
- tz = tz[:-2] + ":" + tz[-2:]
- else:
- tz = "-00:00"
- return date.strftime("%Y-%m-%dT%H:%M:%S") + tz
-
-
-class SitemapGenerator(object):
- def __init__(self, context, settings, path, theme, output_path, *null):
-
- self.output_path = output_path
- self.context = context
- self.now = datetime.now()
- self.siteurl = settings.get("SITEURL")
-
- self.default_timezone = settings.get("TIMEZONE", "UTC")
- self.timezone = getattr(self, "timezone", self.default_timezone)
- self.timezone = timezone(self.timezone)
-
- self.format = "xml"
-
- self.changefreqs = {
- "articles": "monthly",
- "indexes": "daily",
- "pages": "monthly",
- }
-
- self.priorities = {"articles": 0.5, "indexes": 0.5, "pages": 0.5}
-
- self.sitemapExclude = []
-
- config = settings.get("SITEMAP", {})
-
- if not isinstance(config, dict):
- warning("sitemap plugin: the SITEMAP setting must be a dict")
- else:
- fmt = config.get("format")
- pris = config.get("priorities")
- chfreqs = config.get("changefreqs")
- self.sitemapExclude = config.get("exclude", [])
-
- if fmt not in ("xml", "txt"):
- warning("sitemap plugin: SITEMAP['format'] must be 'txt' or 'xml'")
- warning("sitemap plugin: Setting SITEMAP['format'] to 'xml'")
- elif fmt == "txt":
- self.format = fmt
- return
-
- valid_keys = ("articles", "indexes", "pages")
- valid_chfreqs = (
- "always",
- "hourly",
- "daily",
- "weekly",
- "monthly",
- "yearly",
- "never",
- )
-
- if isinstance(pris, dict):
- # We use items for Py3k compat. .iteritems() otherwise
- for k, v in pris.items():
- if k in valid_keys and not isinstance(v, (int, float)):
- default = self.priorities[k]
- warning("sitemap plugin: priorities must be numbers")
- warning(
- "sitemap plugin: setting SITEMAP['priorities']"
- "['{0}'] on {1}".format(k, default)
- )
- pris[k] = default
- self.priorities.update(pris)
- elif pris is not None:
- warning("sitemap plugin: SITEMAP['priorities'] must be a dict")
- warning("sitemap plugin: using the default values")
-
- if isinstance(chfreqs, dict):
- # .items() for py3k compat.
- for k, v in chfreqs.items():
- if k in valid_keys and v not in valid_chfreqs:
- default = self.changefreqs[k]
- warning("sitemap plugin: invalid changefreq '{0}'".format(v))
- warning(
- "sitemap plugin: setting SITEMAP['changefreqs']"
- "['{0}'] on '{1}'".format(k, default)
- )
- chfreqs[k] = default
- self.changefreqs.update(chfreqs)
- elif chfreqs is not None:
- warning("sitemap plugin: SITEMAP['changefreqs'] must be a dict")
- warning("sitemap plugin: using the default values")
-
- def write_url(self, page, fd): # NOQA C901
-
- if getattr(page, "status", "published") != "published":
- return
-
- if getattr(page, "private", "False") == "True":
- return
-
- # We can disable categories/authors/etc by using False instead of ''
- if not page.save_as:
- return
-
- page_path = os.path.join(self.output_path, page.save_as)
- if not os.path.exists(page_path):
- return
-
- lastdate = getattr(page, "date", self.now)
- try:
- lastdate = self.get_date_modified(page, lastdate)
- except ValueError:
- warning(
- "sitemap plugin: " + page.save_as + " has invalid modification date,"
- )
- warning("sitemap plugin: using date value as lastmod.")
- lastmod = format_date(lastdate)
-
- if isinstance(page, contents.Article):
- pri = self.priorities["articles"]
- chfreq = self.changefreqs["articles"]
- elif isinstance(page, contents.Page):
- pri = self.priorities["pages"]
- chfreq = self.changefreqs["pages"]
- else:
- pri = self.priorities["indexes"]
- chfreq = self.changefreqs["indexes"]
-
- pageurl = "" if page.url == "index.html" else page.url
-
- # Exclude URLs from the sitemap:
- if self.format == "xml":
- flag = False
- for regstr in self.sitemapExclude:
- if re.match(regstr, pageurl):
- flag = True
- break
- if not flag:
- fd.write(XML_URL.format(self.siteurl, pageurl, lastmod, chfreq, pri))
- else:
- fd.write(self.siteurl + "/" + pageurl + "\n")
-
- def get_date_modified(self, page, default):
- if hasattr(page, "modified"):
- if isinstance(page.modified, datetime):
- return page.modified
- return get_date(page.modified)
- else:
- return default
-
- def set_url_wrappers_modification_date(self, wrappers):
- for (wrapper, articles) in wrappers:
- lastmod = datetime.min.replace(tzinfo=self.timezone)
- for article in articles:
- lastmod = max(lastmod, article.date.replace(tzinfo=self.timezone))
- try:
- modified = self.get_date_modified(article, datetime.min).replace(
- tzinfo=self.timezone
- )
- lastmod = max(lastmod, modified)
- except ValueError:
- # Supressed: user will be notified.
- pass
- setattr(wrapper, "modified", str(lastmod))
-
- def generate_output(self, writer):
- path = os.path.join(self.output_path, "sitemap.{0}".format(self.format))
-
- pages = (
- self.context["pages"]
- + self.context["articles"]
- + [c for (c, a) in self.context["categories"]]
- + [t for (t, a) in self.context["tags"]]
- + [a for (a, b) in self.context["authors"]]
- )
-
- self.set_url_wrappers_modification_date(self.context["categories"])
- self.set_url_wrappers_modification_date(self.context["tags"])
- self.set_url_wrappers_modification_date(self.context["authors"])
-
- for article in self.context["articles"]:
- pages += article.translations
-
- info("writing {0}".format(path))
-
- with open(path, "w", encoding="utf-8") as fd:
-
- if self.format == "xml":
- fd.write(XML_HEADER)
- else:
- fd.write(TXT_HEADER.format(self.siteurl))
-
- FakePage = collections.namedtuple(
- "FakePage", ["status", "date", "url", "save_as"]
- )
-
- for standard_page in self.context["DIRECT_TEMPLATES"]:
- standard_page_url = self.context.get(
- "{}_URL".format(standard_page.upper())
- )
- standard_page_save_as = self.context.get(
- "{}_SAVE_AS".format(standard_page.upper())
- )
-
- # No save _SAVE_AS field means no output file. Skip.
- if not standard_page_save_as:
- continue
-
- fake = FakePage(
- status="published",
- date=self.now,
- url=standard_page_url or "{}.html".format(standard_page),
- save_as=standard_page_save_as,
- )
- self.write_url(fake, fd)
-
- # add template pages
- # We use items for Py3k compat. .iteritems() otherwise
- for path, template_page_url in self.context["TEMPLATE_PAGES"].items():
-
- # don't add duplicate entry for index page
- if template_page_url == "index.html":
- continue
-
- fake = FakePage(
- status="published",
- date=self.now,
- url=template_page_url,
- save_as=template_page_url,
- )
- self.write_url(fake, fd)
-
- for page in pages:
- self.write_url(page, fd)
-
- if self.format == "xml":
- fd.write(XML_FOOTER)
-
-
-def get_generators(generators):
- return SitemapGenerator
-
-
-def register():
- signals.get_generators.connect(get_generators)