Use the MediaWiki API to get Wiki information

Good morning: relaxed: There are times when you want to get information on wikipedia. Let's use the MediaWiki API that can be used in such cases.

What is MediaWiki API?

You can use wiki functions such as adding, updating, and searching login pages to the wiki.

Official: MediaWiki API

Basic form of API

API: https://ja.wikipedia.org/w/api.php

By giving the above URL a processing type as the parameter "action", you can use various functions such as search and update. This time, we will learn how to search and obtain the basic "acquisition (action = query)" Wiki information.

Set titles = XXX to specify the page title and get the information for that page

getWikiData


import requests
import json


def getWikiData(url, params):
  res = requests.get( url,
                      params = params)
  return res.json()


url      ="https://ja.wikipedia.org/w/api.php"
params = { "action"  : "query",
           "titles"  : "Python",
           "format"  : "json"
         }

print(getWikiData(url,params))

【result】

{
 'batchcomplete': '',
  'query': {
    'pages': {
      '993': {
        'pageid': 993, ##Unique ID
        'ns': 0,## nameSpace 
        'title': 'Python' ##page title
      }
    }
  }
}

Set prop = XXX when adding the acquired contents

For example ... If you want basic information on a page called "Python"

Official: prop = info (in)


params = { "action"  : "query",
           "titles"  : "Python",
           "prop"    : "info",
           "format"  : "json"
         }
[Result] * Folds because it is a little long 

{
  'batchcomplete': '',
  'query': {
    'pages': {
      '993': {
        'pageid': 993, 
        'ns': 0,
        'title': 'Python',
        'contentmodel': 'wikitext',
        'pagelanguage': 'ja',
        'pagelanguagehtmlcode': 'ja',
        'pagelanguagedir': 'ltr',
        'touched': '2020-09-21T07:44:48Z',
        'lastrevid': 79623671,
        'length': 50355
      }
    }
  }
}

More detailed properties can be specified for each prop

For example ... If you want to see basic information + the number of people who visited the page

(Most of the time, [shortened major item prop + prop] is the item name)


params = { "action"  : "query",
           "titles"  : "Python",
           "prop"    : "info",##Major item standing position
           "inprop"  : "watchers",##Sub-item standing position
           "format"  : "json"
         }
Result * Folds because it is a little long 

{
  'batchcomplete': '',
  'query': {
    'pages': {
      '993': {
        'pageid': 993,
        'ns': 0,
        'title': 'Python',
        'contentmodel': 'wikitext',
        'pagelanguage': 'ja',
        'pagelanguagehtmlcode': 'ja',
        'pagelanguagedir': 'ltr',
        'touched': '2020-09-21T07:44:48Z',
        'lastrevid': 79623671,
        'length': 50355,
        'watchers': 157 ##here
      }
    }
  }
}

For example ... If you want to get the "category" and "key item for sorting" included in the page "Python"

Official: prop = categories (cl)

  • If you want to get other items, refer to the above page.

params = { "action"  : "query",
           "titles"  : "Python",
           "prop"    : "categories",
           "clprop"  : "sortkey",
           "format"  : "json"
         }
Result * Folds because it is a little long 

{
  'batchcomplete': '',
  'query': {
    'pages': {
      '993': {
        'pageid': 993,
        'ns': 0,
        'title': 'Python',
        'categories': [
          {
            'ns': 14,
            'title': 'Category:Python',
            'sortkey': '2a0a505954484f4e',
            'sortkeyprefix': '*'
          },
          {
            'ns': 14,
            'title': 'Category:Object-oriented language',
            'sortkey': '505954484f4e0a505954484f4e',
            'sortkeyprefix': 'PYTHON'
          },
          {
            'ns': 14,
            'title': 'Category:Open source software',
            'sortkey': '505954484f4e0a505954484f4e',
            'sortkeyprefix': 'PYTHON'
          },
          {
            'ns': 14,
            'title': 'Category:Scripting language',
            'sortkey': '505954484f4e0a505954484f4e',
            'sortkeyprefix': 'PYTHON'
          },
          {
            'ns': 14,
            'title': 'Category:Basic Information Technology Engineer Examination',
            'sortkey': 'e381afe38184e3819de382930a505954484f4e',
            'sortkeyprefix': 'Yes'
          },
          {
            'ns': 14,
            'title': 'Category:Articles containing invalid sources/2018',
            'sortkey': '420a505954484f4e',
            'sortkeyprefix': 'B'
          }
        ]
      }
    }
  }
}

Set list = XXX to get search results in list

For example ... To get all the categories on the Wiki

list=allcategories (ac)

  • If you want to get other items, refer to the above page.

params = { "action"  : "query",
           "titles"  : "Python",
           "list"    : "allcategories", 
           "acprop"  : "size", ##Different properties for each list
           "aclimit" :  5, ##Maximum number of acquisitions Item name changes for each list
           "format"  : "json"
         }
Result * Folds because it is a little long 


{
  'batchcomplete': '',
  'continue': {
    'accontinue': '.22_LR_firearms',
    'continue': '-||'
  },
  'query': {
    'pages': {
      '993': {
        'pageid': 993,
        'ns': 0,
        'title': 'Python'
      }
    },
    'allcategories': [
      {
        'size': 1,
        'pages': 1,
        'files': 0,
        'subcats': 0,
        '*': '" + afterCat + "'
      },
      {
        'size': 1,
        'pages': 1,
        'files': 0,
        'subcats': 0,
        '*': '" + afterCat + "$2'
      },
      {
        'size': 2,
        'pages': 2,
        'files': 0,
        'subcats': 0,
        '*': '$1'
      },
      {
        'size': 3,
        'pages': 3,
        'files': 0,
        'subcats': 0,
        '*': '((documentation))Pages with unusual use of'
      },
      {
        'size': 9,
        'pages': 9,
        'files': 0,
        'subcats': 0,
        '*': '+Ultra'
      }
    ]
  }
}

For example ... To get a list of pages included in a specific category

params = { "action"  : "query",
           "titles"  : "Python",
           "list"    : "categorymembers",
           "cmtitle" : "Category:Object-oriented language", ##If you use categorymembers, you must set this here.
           "cmlimit" :  5, ##Maximum number of acquisitions
           "cmprop"  : "ids|title|sortkey",##Items to get.[|]You can specify multiple delimiters.
           "format"  : "json"
         }
Result * Folds because it is a little long 
{
  'batchcomplete': '',
  'continue': {
    'cmcontinue': 'page|4345594c4f4e|2496222',
    'continue': '-||'
  },
  'query': {
    'pages': {
      '993': {
        'pageid': 993,
        'ns': 0,
        'title': 'Python'
      }
    },
    'categorymembers': [
      {
        'pageid': 821212,
        'ns': 0,
        'title': 'Comparison of object-oriented languages',
        'sortkey': '2ae381b2e3818be3818f0ae382aae38396e382b8e382a7e382afe38388e68c87e59091e8a880e8aa9ee381aee6af94e8bc83'
      },
      {
        'pageid': 181337,
        'ns': 0,
        'title': 'ActiveBasic',
        'sortkey': '41435449564542415349430a4143544956454241534943'
      },
      {
        'pageid': 3785500,
        'ns': 0,
        'title': 'Ballerina',
        'sortkey': '42414c4c4552494e41'
      },
      {
        'pageid': 2066745,
        'ns': 0,
        'title': 'Boo (Programming language)',
        'sortkey': '424f4f2028e38397e383ade382b0e383a9e3839fe383b3e382b0e8a880e8aa9e29'
      },
      {
        'pageid': 1503,
        'ns': 0,
        'title': 'C Sharp',
        'sortkey': '43230a43205348415250'
      }
    ]
  }
}

Parameter supplement

Multiple selection of elements
info and images with prop(Get image information)If you want to get multiple items such as[|]It can be separated by a pipe.
"prop" : "info"|"images"It will be the shape you want to see.

Return value supplement

ns(namespace)
The type of page retrieved is displayed.
Content page for ns: 0, category for ns: 14
continue
This item is created when there is a continuation of the result acquired this time.
If you set the item of "short form of prop + continue" included in the return value as an argument, you can get the data from the continuation.

Last concrete example

When creating a user dictionary for mecab, I created the following code because I wanted to get all the content titles included in the corresponding category.

makeOwaraiList



import requests
import json
import csv
import re

def getWikiData(url, params):
  res = requests.get( url,
                      params = params)
  return res.json()


fileName = "wikiList"
url      ="https://ja.wikipedia.org/w/api.php"
params   = { "action" : "query",
             "list"   : "categorymembers",
             "cmlimit": "50",
             "format" : "json"
           }

categories = ['Japanese comedy combination','Japanese comedy combination','Japanese comedy duo','Japanese comedy trio','Japanese comedy group','Japan>Couple comedy combination']


with open( fileName + ".csv",'a', encoding="utf-8") as f:
  writer = csv.writer(f)

  for category in categories :
    params['cmtitle'] = 'Category:' + category;
    params['cmcontinue'] = '';
    wikiData = getWikiData(url,params)
    while True:
      for page in wikiData['query']['categorymembers']:
        if page['ns']  == 0:
          title = (re.sub("\(comedy\)|\(comedyコンビ\)|\(comedytrio\)|\(talent\)|\(comedyグループ\)|\(trio\)|\(unit\)|\(Entertainer\)", "" ,page['title'])).strip()
          writer.writerow([title])
      if ('continue' in wikiData and wikiData['continue']['cmcontinue']):
        params['cmcontinue'] = wikiData['continue']['cmcontinue']
        wikiData = getWikiData(url,params)
      else :
        break

Again, I fully trusted the formula and studied and used the API: relaxed: There are many situations where you can use the information on the wiki, so if you can master this API, it will be more fun to implement.

Recommended Posts