From e678081a779ac6ea55f731b2ae756614ffc57c6d Mon Sep 17 00:00:00 2001
From: lub <git@lubiland.de>
Date: Wed, 5 Jun 2024 16:09:37 +0200
Subject: [PATCH] init

---
 .gitignore       |  1 +
 requirements.txt |  2 ++
 scrape.py        | 46 ++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 49 insertions(+)
 create mode 100644 .gitignore
 create mode 100644 requirements.txt
 create mode 100644 scrape.py

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..5ceb386
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+venv
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..9d981c3
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,2 @@
+bs4
+requests
diff --git a/scrape.py b/scrape.py
new file mode 100644
index 0000000..815b06b
--- /dev/null
+++ b/scrape.py
@@ -0,0 +1,46 @@
+import requests
+from bs4 import BeautifulSoup
+
+overview_url = 'https://gruene-hohenlohe.de/kalender'
+overview_html = requests.get(overview_url, timeout=60).text
+overview_soup = BeautifulSoup(overview_html, 'html.parser')
+
+
+for href in overview_soup.select('.media-body h2 a'):
+    event_url = 'https://gruene-hohenlohe.de/' + href.attrs['href']
+    event_html = requests.get(event_url, timeout=60).text
+    event_soup = BeautifulSoup(event_html, 'html.parser')
+
+    data = event_soup.select('.calendarize dl dd')
+
+    output = ''
+
+    # date
+    output += data[0].text.strip()
+    output += ' '
+
+    # time
+    timespan = data[1].text.strip()
+    time = timespan.split(' ')[0]
+
+    output += time
+    output += ' Uhr'
+    output += ', '
+
+    # place
+    output += data[2].text.strip()
+    output += ', '
+
+    # title
+    #output += '<b>'
+    #output += event_soup.select('.calendarize h1')[0].text.strip()
+    #output += '</b>'
+
+    #try:
+    description = event_soup.select('.calendarize .text p')[0].text.strip()
+    #output += ' '
+    output += description
+    #except IndexError:
+    #    pass
+
+    print(output)