Split code into another file

This commit is contained in:
Edward Betts 2022-08-17 13:35:26 +01:00
parent 8432632aae
commit 78de5cc139
2 changed files with 222 additions and 209 deletions

220
dab_mechanic/wikipedia.py Normal file
View file

@ -0,0 +1,220 @@
from collections import defaultdict
from typing import Any, Iterator, Optional, TypedDict
import flask
import lxml.html
import requests
disambig_templates = [
"Template:Disambiguation",
"Template:Airport disambiguation",
"Template:Biology disambiguation",
"Template:Call sign disambiguation",
"Template:Caselaw disambiguation",
"Template:Chinese title disambiguation",
"Template:Disambiguation cleanup",
"Template:Genus disambiguation",
"Template:Hospital disambiguation",
"Template:Human name disambiguation",
"Template:Human name disambiguation cleanup",
"Template:Letter-number combination disambiguation",
"Template:Mathematical disambiguation",
"Template:Military unit disambiguation",
"Template:Music disambiguation",
"Template:Number disambiguation",
"Template:Opus number disambiguation",
"Template:Phonetics disambiguation",
"Template:Place name disambiguation",
"Template:Portal disambiguation",
"Template:Road disambiguation",
"Template:School disambiguation",
"Template:Species Latin name abbreviation disambiguation",
"Template:Species Latin name disambiguation",
"Template:Station disambiguation",
"Template:Synagogue disambiguation",
"Template:Taxonomic authority disambiguation",
"Template:Taxonomy disambiguation",
"Template:Template disambiguation",
"Template:WoO number disambiguation",
]
def link_params(enwiki: str) -> dict[str, str | int]:
"""Parameters for finding article links from the API."""
params: dict[str, str | int] = {
"action": "query",
"format": "json",
"formatversion": 2,
"titles": enwiki,
"generator": "links",
"gpllimit": "max",
"gplnamespace": 0,
"tllimit": "max",
"redirects": 1,
"tlnamespace": 10,
"tltemplates": "|".join(disambig_templates),
"prop": "templates",
}
return params
def needs_disambig(link: dict[str, Any]) -> bool:
"""Is this a disambiguation link."""
return bool(
not link["title"].endswith(" (disambiguation)") and link.get("templates")
)
def get_article_links(enwiki: str) -> list[str]:
"""Get links that appear in this article."""
url = "https://en.wikipedia.org/w/api.php"
params: dict[str, str | int] = link_params(enwiki)
links: set[str] = set()
redirects = defaultdict(set)
while True:
data = requests.get(url, params=params).json()
pages = data["query"].pop("pages")
for r in data["query"].pop("redirects"):
redirects[r["to"]].add(r["from"])
links.update(page["title"] for page in pages if needs_disambig(page))
if "continue" not in data:
break
params["gplcontinue"] = data["continue"]["gplcontinue"]
for link in set(links):
if link in redirects:
links.update(redirects[link])
return list(links)
# return {link["title"] for link in r.json()["query"]["pages"][0]["links"]}
def call_parse_api(enwiki: str) -> dict[str, Any]:
"""Call mediawiki parse API for given article."""
url = "https://en.wikipedia.org/w/api.php"
params: dict[str, str | int] = {
"action": "parse",
"format": "json",
"formatversion": 2,
"disableeditsection": 1,
"page": enwiki,
"prop": "text|links|headhtml",
"disabletoc": 1,
}
r = requests.get(url, params=params)
parse: dict[str, Any] = r.json()["parse"]
return parse
def get_article_html(enwiki: str) -> str:
"""Parse article wikitext and return HTML."""
text: str = call_parse_api(enwiki)["text"]
return text
class DabItem(TypedDict):
"""Represent a disabiguation page."""
num: int
title: str
html: str
def delete_toc(root: lxml.html.HtmlElement) -> None:
"""Delete table of contents from article HTML."""
for toc in root.findall(".//div[@class='toc']"):
toc.getparent().remove(toc)
def get_dab_html(dab_num: int, title: str) -> str:
"""Parse dab page and rewrite links."""
dab_html = get_article_html(title)
root = lxml.html.fromstring(dab_html)
delete_toc(root)
element_id_map = {e.get("id"): e for e in root.findall(".//*[@id]")}
for a in root.findall(".//a[@href]"):
href: str | None = a.get("href")
if not href:
continue
if not href.startswith("#"):
a.set("href", "#")
a.set("onclick", f"return select_dab(this, {dab_num})")
continue
destination_element = element_id_map[href[1:]]
assert destination_element is not None
destination_element.set("id", f"{dab_num}{href[1:]}")
a.set("href", f"#{dab_num}{href[1:]}")
html: str = lxml.html.tostring(root, encoding=str)
return html
class Article:
"""Current article we're working on."""
def __init__(self, enwiki: str) -> None:
"""Make a new Article object."""
self.enwiki = enwiki
self.links = get_article_links(enwiki)
self.dab_list: list[DabItem] = []
self.dab_lookup: dict[int, str] = {}
self.dab_order: list[str] = []
self.parse: Optional[dict[str, Any]] = None
def save_endpoint(self) -> str:
"""Endpoint for saving changes."""
href: str = flask.url_for("save", enwiki=self.enwiki.replace(" ", "_"))
return href
def load(self) -> None:
"""Load parsed article HTML."""
self.parse = call_parse_api(self.enwiki)
self.root = lxml.html.fromstring(self.parse.pop("text"))
def iter_links(self) -> Iterator[tuple[lxml.html.Element, str]]:
"""Disambiguation links that need fixing."""
seen = set()
for a in self.root.findall(".//a[@href]"):
title = a.get("title")
if title is None or title not in self.links:
continue
a.set("class", "disambig")
if title in seen:
continue
seen.add(title)
yield a, title
def process_links(self) -> None:
"""Process links in parsed wikitext."""
for dab_num, (a, title) in enumerate(self.iter_links()):
a.set("id", f"dab-{dab_num}")
dab: DabItem = {
"num": dab_num,
"title": title,
"html": get_dab_html(dab_num, title),
}
self.dab_list.append(dab)
self.dab_order.append(title)
self.dab_lookup[dab_num] = title
def get_html(self) -> str:
"""Return the processed article HTML."""
html: str = lxml.html.tostring(self.root, encoding=str)
return html

View file

@ -3,7 +3,6 @@
import inspect
import json
import re
from typing import Any, Iterator, Optional, TypedDict
import flask
import lxml.html
@ -13,7 +12,7 @@ from requests_oauthlib import OAuth1Session
from werkzeug.debug.tbtools import get_current_traceback
from werkzeug.wrappers import Response
from dab_mechanic import wikidata_oauth
from dab_mechanic import wikidata_oauth, wikipedia
app = flask.Flask(__name__)
app.config.from_object("config.default")
@ -89,145 +88,6 @@ def index():
return flask.render_template("index.html", articles=articles)
def call_parse_api(enwiki: str) -> dict[str, Any]:
"""Call mediawiki parse API for given article."""
url = "https://en.wikipedia.org/w/api.php"
params: dict[str, str | int] = {
"action": "parse",
"format": "json",
"formatversion": 2,
"disableeditsection": 1,
"page": enwiki,
"prop": "text|links|headhtml",
"disabletoc": 1,
}
r = requests.get(url, params=params)
parse: dict[str, Any] = r.json()["parse"]
return parse
def get_article_html(enwiki: str) -> str:
"""Parse article wikitext and return HTML."""
text: str = call_parse_api(enwiki)["text"]
return text
disambig_templates = [
"Template:Disambiguation",
"Template:Airport disambiguation",
"Template:Biology disambiguation",
"Template:Call sign disambiguation",
"Template:Caselaw disambiguation",
"Template:Chinese title disambiguation",
"Template:Disambiguation cleanup",
"Template:Genus disambiguation",
"Template:Hospital disambiguation",
"Template:Human name disambiguation",
"Template:Human name disambiguation cleanup",
"Template:Letter-number combination disambiguation",
"Template:Mathematical disambiguation",
"Template:Military unit disambiguation",
"Template:Music disambiguation",
"Template:Number disambiguation",
"Template:Opus number disambiguation",
"Template:Phonetics disambiguation",
"Template:Place name disambiguation",
"Template:Portal disambiguation",
"Template:Road disambiguation",
"Template:School disambiguation",
"Template:Species Latin name abbreviation disambiguation",
"Template:Species Latin name disambiguation",
"Template:Station disambiguation",
"Template:Synagogue disambiguation",
"Template:Taxonomic authority disambiguation",
"Template:Taxonomy disambiguation",
"Template:Template disambiguation",
"Template:WoO number disambiguation",
]
def link_params(enwiki: str) -> dict[str, str | int]:
"""Parameters for finding article links from the API."""
params: dict[str, str | int] = {
"action": "query",
"format": "json",
"formatversion": 2,
"titles": enwiki,
"generator": "links",
"gpllimit": "max",
"gplnamespace": 0,
"tllimit": "max",
"tlnamespace": 10,
"tltemplates": "|".join(disambig_templates),
"prop": "templates",
}
return params
def needs_disambig(link: dict[str, Any]) -> bool:
"""Is this a disambiguation link."""
return bool(
not link["title"].endswith(" (disambiguation)") and link.get("templates")
)
def get_article_links(enwiki: str) -> list[str]:
"""Get links that appear in this article."""
url = "https://en.wikipedia.org/w/api.php"
params: dict[str, str | int] = link_params(enwiki)
links: set[str] = set()
while True:
data = requests.get(url, params=params).json()
links.update(
page["title"] for page in data["query"]["pages"] if needs_disambig(page)
)
if "continue" not in data:
break
params["gplcontinue"] = data["continue"]["gplcontinue"]
return list(links)
# return {link["title"] for link in r.json()["query"]["pages"][0]["links"]}
def delete_toc(root: lxml.html.HtmlElement) -> None:
"""Delete table of contents from article HTML."""
for toc in root.findall(".//div[@class='toc']"):
toc.getparent().remove(toc)
def get_dab_html(dab_num: int, title: str) -> str:
"""Parse dab page and rewrite links."""
dab_html = get_article_html(title)
root = lxml.html.fromstring(dab_html)
delete_toc(root)
element_id_map = {e.get("id"): e for e in root.findall(".//*[@id]")}
for a in root.findall(".//a[@href]"):
href: str | None = a.get("href")
if not href:
continue
if not href.startswith("#"):
a.set("href", "#")
a.set("onclick", f"return select_dab(this, {dab_num})")
continue
destination_element = element_id_map[href[1:]]
assert destination_element is not None
destination_element.set("id", f"{dab_num}{href[1:]}")
a.set("href", f"#{dab_num}{href[1:]}")
html: str = lxml.html.tostring(root, encoding=str)
return html
def make_disamb_link(edit: tuple[str, str]) -> str:
"""Given an edit return the appropriate link."""
return f"[[{edit[1]}|{edit[0]}]]"
@ -278,73 +138,6 @@ def save(enwiki: str) -> Response | str:
)
class DabItem(TypedDict):
"""Represent a disabiguation page."""
num: int
title: str
html: str
class Article:
"""Current article we're working on."""
def __init__(self, enwiki: str) -> None:
"""Make a new Article object."""
self.enwiki = enwiki
self.links = get_article_links(enwiki)
self.dab_list: list[DabItem] = []
self.dab_lookup: dict[int, str] = {}
self.dab_order: list[str] = []
self.parse: Optional[dict[str, Any]] = None
def save_endpoint(self) -> str:
"""Endpoint for saving changes."""
href: str = flask.url_for("save", enwiki=self.enwiki.replace(" ", "_"))
return href
def load(self) -> None:
"""Load parsed article HTML."""
self.parse = call_parse_api(self.enwiki)
self.root = lxml.html.fromstring(self.parse.pop("text"))
def iter_links(self) -> Iterator[tuple[lxml.html.Element, str]]:
"""Disambiguation links that need fixing."""
seen = set()
for a in self.root.findall(".//a[@href]"):
title = a.get("title")
if title is None or title not in self.links:
continue
a.set("class", "disambig")
if title in seen:
continue
seen.add(title)
yield a, title
def process_links(self) -> None:
"""Process links in parsed wikitext."""
for dab_num, (a, title) in enumerate(self.iter_links()):
a.set("id", f"dab-{dab_num}")
dab: DabItem = {
"num": dab_num,
"title": title,
"html": get_dab_html(dab_num, title),
}
self.dab_list.append(dab)
self.dab_order.append(title)
self.dab_lookup[dab_num] = title
def get_html(self) -> str:
"""Return the processed article HTML."""
html: str = lxml.html.tostring(self.root, encoding=str)
return html
@app.route("/enwiki/<path:enwiki>")
def article_page(enwiki: str) -> Response:
"""Article Page."""
@ -356,7 +149,7 @@ def article_page(enwiki: str) -> Response:
flask.url_for(flask.request.endpoint, enwiki=enwiki_underscore)
)
article = Article(enwiki)
article = wikipedia.Article(enwiki)
article.load()
article.process_links()