edward/eurotunnel-scrape
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity

Add more detail to web page

Browse source
This commit is contained in:
Edward Betts 2023-08-08 18:57:33 +01:00
parent 036c0cfbd8
commit 348162169f
2 changed files with 20 additions and 9 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
templates/index.html
View file

@ -2,7 +2,7 @@
<html lang="en"> <html lang="en">
<head> <head>
<meta charset="utf-8"> <meta charset="utf-8">
<title></title> <title>Eurotunnel prices</title>
<link href="https://cdn.jsdelivr.net/npm/bootstrap@5.3.1/dist/css/bootstrap.min.css" rel="stylesheet" integrity="sha384-4bw+/aepP/YC94hEpVNVgiZdgIC5+VKNBQNGCHeKRQN+PtmoHDEXuppvnDJzQIu9" crossorigin="anonymous"> <link href="https://cdn.jsdelivr.net/npm/bootstrap@5.3.1/dist/css/bootstrap.min.css" rel="stylesheet" integrity="sha384-4bw+/aepP/YC94hEpVNVgiZdgIC5+VKNBQNGCHeKRQN+PtmoHDEXuppvnDJzQIu9" crossorigin="anonymous">
</head> </head>
@ -11,12 +11,14 @@
<body> <body>
<div class="container mt-3"> <div class="container mt-3">
<p>All Flexi Long Stay tickets are £269</p> <h1>Eurotunnel prices</h1>
<p>Pay an extra £40 to make a price return refundable.</p>
<p>All Flexi Long Stay tickets are £269. Pay an extra £40 to make a return ticket refundable.</p>
<div class="row"> <div class="row">
<div class="col"> <div class="col">
<h3>Out: {{ out_date.strftime("%a, %-d %b %Y") }}</h3>
<p>Updated: {{ out_ts.strftime("%a, %-d %b %Y at %H:%M") }}</p> <p>Updated: {{ out_ts.strftime("%a, %-d %b %Y at %H:%M") }}</p>
<table class="table table-sm w-auto"> <table class="table table-sm w-auto">
<tr> <tr>
@ -36,6 +38,7 @@
<div class="col"> <div class="col">
<h3>Return: {{ back_date.strftime("%a, %-d %b %Y") }}</h3>
<p>Updated: {{ back_ts.strftime("%a, %-d %b %Y at %H:%M") }}</p> <p>Updated: {{ back_ts.strftime("%a, %-d %b %Y at %H:%M") }}</p>
<table class="table table-sm w-auto"> <table class="table table-sm w-auto">
<tr> <tr>

20
web_view.py
View file

@ -2,7 +2,7 @@
import os import os
from dataclasses import dataclass from dataclasses import dataclass
from datetime import UTC, datetime from datetime import UTC, datetime, date
from decimal import Decimal from decimal import Decimal
import flask import flask
@ -12,7 +12,7 @@ import pytz
app = flask.Flask(__name__) app = flask.Flask(__name__)
app.debug = True app.debug = True
data_loc = os.path.expanduser("~/lib/data/eurotunnel") data_loc = os.path.expanduser("~edward/lib/data/eurotunnel")
def get_filename(direction: str) -> tuple[datetime, str]: def get_filename(direction: str) -> tuple[datetime, str]:
@ -35,7 +35,7 @@ class Train:
price: Decimal | None = None price: Decimal | None = None
def get_tickets(filename: str) -> list[Train]: def get_tickets(filename: str) -> tuple[date, list[Train]]:
"""Get trains and prices.""" """Get trains and prices."""
tree = lxml.html.parse(filename) tree = lxml.html.parse(filename)
root = tree.getroot() root = tree.getroot()
@ -43,6 +43,12 @@ def get_tickets(filename: str) -> list[Train]:
trains = [] trains = []
by_time = {} by_time = {}
day_div = root.find(".//div[@class='col-md-1 als-item selected']")
assert day_div
day_id = day_div.get("data-id")
assert day_id
d = date.fromisoformat(day_id[1:])
for mission in root.findall(".//div[@data-mission]"): for mission in root.findall(".//div[@data-mission]"):
dep_text = mission.findtext(".//b") dep_text = mission.findtext(".//b")
assert dep_text assert dep_text
@ -69,25 +75,27 @@ def get_tickets(filename: str) -> list[Train]:
by_time[mission_time].price = price by_time[mission_time].price = price
return trains return (d, trains)
@app.route("/") @app.route("/")
def index() -> str: def index() -> str:
"""Index.""" """Index."""
out_ts, out_filename = get_filename("outbound") out_ts, out_filename = get_filename("outbound")
out = get_tickets(out_filename) out_date, out = get_tickets(out_filename)
out = [t for t in out if t.dep > "0800" and "0700" < t.arr < "2200"] out = [t for t in out if t.dep > "0800" and "0700" < t.arr < "2200"]
back_ts, back_filename = get_filename("return") back_ts, back_filename = get_filename("return")
back = get_tickets(back_filename) back_date, back = get_tickets(back_filename)
back = [t for t in back if t.dep > "1100" and "0700" < t.arr < "2200"] back = [t for t in back if t.dep > "1100" and "0700" < t.arr < "2200"]
return flask.render_template( return flask.render_template(
"index.html", "index.html",
out_ts=out_ts, out_ts=out_ts,
out_date=out_date,
out=out, out=out,
back_ts=back_ts, back_ts=back_ts,
back_date=back_date,
back=back, back=back,
) )