Simple scraper for greek parliamentary election data from the official web site (https://ekloges.ypes.gr/)
As gioiliop7 suggested there is an API for scraping some of the data.
The older python scipt using headless selenium is also included.
The code consists of a python script for downloading the election data, and saving them to a JSON file as a dictionary
A second script -exportcsv.py- reads the JSON files and exports data to csv
File Kwdikoi.csv contains the correspondance between codes and districts/municipalites names
- ekloges_parser_json.py is run, probably more than one time, to download all data into two JSON files
- exportscv.py reads the two JSON files and outputs csv
Two csv files with election data for the 2023 May and 2023 Jube elections are included
Each file has 7 columns
- Datestring
- String identification of whether the data is MUNICIPALITY of DISTRICT
- name of MUNICIPALITY or DISTRICT (in greek)
- String DISTRICT (could be ommited)
- String whith the district name in which "name" belongs
- party/code name (codes are LEFKA/AKYRA/GRAMMENOI/EGYRA correspondig to WHITE/INVALID/REGISTERED/VALID)
- number of votes
The scaper with minimal changes should be able to also read the 2019 elections correctly.
A csv file with the 2019 elections is also included
There are two problems
- division into municipalities and districts is not constant
- parties come and go
The first problem is easier to solve.The are in the main site stat data (names etc), it is relatively easy to parse them and reconstruct votes for previous elections as well.
The second problem is more difficult, as an online source has not yet identified
The 2019 file could be incomplete.
The file size is considerably smaller - this could simply indicate missing names and/or less parties.
- impelement sanity cheks (districts = sum of municipalities, country data, party data)
- use online data to navigate througout changing districts/municipalities
- parse web site to automatically correspond party codes to code names
- include older data
- use json data for date extraction / currently hardcoded in script
- parse higher granularity data