CRAWL ALLTHEM DATA ^^

Crawler.ipynb

@@ -0,0 +1,233 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 220,
+   "id": "83707cda",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import requests\n",
+    "import os\n",
+    "from bs4 import BeautifulSoup\n",
+    "import json"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 229,
+   "id": "72fb7570",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "url = \"https://www.arztsuche-bw.de/index.php?suchen=1&sorting=name&direction=ASC&arztgruppe=psychotherapeut&landkreis=Karlsruhe+-+Stadt\"\n",
+    "offset_str = \"&offset=\"\n",
+    "datapath = \"./data\"\n",
+    "outpath = os.path.join(datapath, \"raw\")\n",
+    "extractedpath = os.path.join(datapath, \"extracted\")\n",
+    "npages = 14 # we can see on the website there are 14 pages\n",
+    "ntherapists = 20 # 20 items per page"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8eff4531",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Crawling page 15\r"
+     ]
+    }
+   ],
+   "source": [
+    "\n",
+    "if not os.path.exists(outpath):\n",
+    "    os.makedirs(outpath)\n",
+    "\n",
+    "for page in range(npages + 1):\n",
+    "    print(f\"Crawling page {page}\", end=\"\\r\")\n",
+    "\n",
+    "    res = requests.get(url + offset_str + str(page * ntherapists)) \n",
+    "\n",
+    "    if res.status_code == 200:\n",
+    "        with open(os.path.join(outpath, f\"page_{page}.html\"), \"w+\") as outfile:\n",
+    "            outfile.write(res.text)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 41,
+   "id": "7d7f9ca1",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "for page in range(npages):\n",
+    "    with open(os.path.join(outpath, f\"page_{page}.html\"), \"r\") as crawled_page:\n",
+    "        parsed = BeautifulSoup(crawled_page.read())\n",
+    "        result = parsed.find_all(\"ul\", {\"class\": \"resultlist\"})[0]\n",
+    "        entries = result.find_all(\"li\", {\"class\": \"resultrow\"})\n",
+    "\n",
+    "        for i, entry in enumerate(entries):\n",
+    "            with open(os.path.join(extractedpath, f\"page_{page}_{i}.html\"), \"w+\") as extractedfile:\n",
+    "                extractedfile.write(str(entry.prettify()))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 114,
+   "id": "0db40ad6",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def parse_appointment_table(tableid, parsed):\n",
+    "    # extract all possible appointments\n",
+    "    termintabelle = parsed.find_all(\"table\", {\"class\": \"termintabelle\"})[tableid]\n",
+    "    rows = termintabelle.find_all(\"tr\")\n",
+    "    appointments = {}\n",
+    "\n",
+    "    for row in rows:\n",
+    "        entries = row.find_all(\"td\")\n",
+    "        day = entries[0].text.strip()\n",
+    "        # Python types are fuckig cursed\n",
+    "        times = list(filter(lambda e: e, map(lambda e: e.text.strip(), entries[1:])))\n",
+    "        appointments[day] = times\n",
+    "    \n",
+    "    return appointments"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 139,
+   "id": "8558c73b",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def extract_qualifications(index, parsed):\n",
+    "    qualis = parsed.find_all(\"dl\", {\"class\": \"bulletlist\"})\n",
+    "    cleaned = list(filter(lambda e: e ,map(lambda e: \"\".join(e.text.splitlines()).strip(), qualis[index].children)))\n",
+    "    return cleaned"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 265,
+   "id": "e5837d3f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def parse_extracted_html(infilepath):\n",
+    "\n",
+    "    doctor = {}\n",
+    "    doctor[\"filepath\"] = infilepath\n",
+    "\n",
+    "    # read in file\n",
+    "    with open(infilepath, \"r\") as infile:\n",
+    "        parsed = BeautifulSoup(infile.read())\n",
+    "\n",
+    "    # extract name\n",
+    "    name = \" \".join(parsed.find_all(\"dd\", {\"class\": \"name\"})[0].findChildren(\"dt\")[0].text.splitlines())\n",
+    "    doctor[\"name\"] = name.strip()\n",
+    "\n",
+    "    ntables = len(parsed.find_all(\"table\", {\"class\": \"termintabelle\"}))\n",
+    "\n",
+    "    # table order changes if no phone appointments are available -_-\n",
+    "    if ntables == 2:\n",
+    "        appointments = parse_appointment_table(1, parsed)\n",
+    "        phone_appointments = parse_appointment_table(0, parsed)\n",
+    "        doctor[\"phone_appointments\"] = phone_appointments\n",
+    "    else:\n",
+    "        appointments = parse_appointment_table(0, parsed)\n",
+    "\n",
+    "    doctor[\"appointments\"] = appointments\n",
+    "\n",
+    "    # extract type of doctor\n",
+    "    resorts = extract_qualifications(0, parsed)\n",
+    "    doctor[\"resorts\"] = resorts[1:]\n",
+    "\n",
+    "    # extract therapy types of doctor\n",
+    "    therapy_types = extract_qualifications(1, parsed)\n",
+    "    doctor[\"therapy_types\"] = therapy_types[1:]\n",
+    "\n",
+    "    # Address\n",
+    "    addresslines = parsed.find_all(\"p\", {\"class\": \"anschrift-arzt\"})[0].text.splitlines()\n",
+    "    doctor[\"address\"] = \" \".join(map(lambda t: t.strip(), addresslines))\n",
+    "\n",
+    "    # contact\n",
+    "    # some have *no* contact info??\n",
+    "    contactscol = parsed.find_all(\"dd\", {\"class\": \"adresse\"})[0].findChildren(\"dd\")\n",
+    "    if len(contactscol) != 0:\n",
+    "        contacts = contactscol[0]\n",
+    "        contact_text = contacts.text.split(\"<br>\")[0].splitlines()\n",
+    "        pseudo_phone_list = list(filter(lambda e: e, map(lambda e: e.strip(), contact_text)))[0].split(\" \")\n",
+    "        \n",
+    "        # appearently some therapists don't have a phone…\n",
+    "        if pseudo_phone_list[0] in [\"Telefon\", \"Mobil\"]:\n",
+    "            doctor[\"phone\"] = pseudo_phone_list[1]\n",
+    "\n",
+    "        # Email (check if it exists)\n",
+    "        mails = contacts.findChildren(\"a\", {\"class\": \"obfuscatedEmail\"})\n",
+    "        mails = list(map(lambda e: e.text.strip(), mails))\n",
+    "\n",
+    "        if len(mails) != 0 :\n",
+    "            doctor[\"mails\"] = mails \n",
+    "\n",
+    "    return doctor"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 269,
+   "id": "8748c681",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "parsing 13|14…\r"
+     ]
+    }
+   ],
+   "source": [
+    "therapists = []\n",
+    "for page in range(npages):\n",
+    "    for therapist in range(ntherapists):\n",
+    "        # only 15 therapists on last page\n",
+    "        if page == npages - 1 and therapist == 15:\n",
+    "            break\n",
+    "        print(f\"parsing {page}|{therapist}…\", end=\"\\r\")\n",
+    "        therapists.append(parse_extracted_html(os.path.join(extractedpath, f\"page_{page}_{therapist}.html\")))\n",
+    "\n",
+    "with open(os.path.join(datapath, \"therapists.json\"), \"w+\") as outfile:\n",
+    "    json.dump(therapists, outfile, indent=2)\n",
+    "\n",
+    "        "
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "python-minimal kernel",
+   "language": "python",
+   "name": "python-minimal"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.12.7"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}