Better parser that can do a Set instead of just a list of URLs in a chain.

02-filter-a-log-file/chains_parser.js

@@ -9,13 +9,20 @@ class Parser {
     this.next = "";
   }
 
-  match(reg) {
+  match(reg, consume=true) {
     const n = this.next.match(reg);
 
-    const element = n.length > 1 ? n.slice(1) : n[0];
+    if(n === null) {
+      return undefined;
+    } else {
+      const element = n.length > 1 ? n.slice(1) : n[0];
+
+      if(consume) {
+        this.next = this.next.slice(n[0].length);
+      }
 
-    this.next = this.next.slice(n[0].length);
-    return element;
+      return element;
+    }
   }
 
   start(line) {
@@ -23,122 +30,195 @@ class Parser {
     this.next = line;
   }
 
-  ws() {
-    this.match(/ +/);
+  ws(consume=true) {
+    return this.match(/ +/, consume);
   }
 
-  parse() {
-    try {
-      const [ip, port, conn_id, conn_count] = this.match(/([0-9.]+):([0-9]+):([0-9]+):([0-9]+)/);
+  parse_new_log(ip) {
+    const [port, conn_id, conn_count] = this.match(/^([0-9]+):([0-9]+):([0-9]+)/);
 
-      this.ws();
+    this.ws();
 
-      const [ time ] = this.match(/\[(.*)\]/);
+    const [ time ] = this.match(/\[(.*)\]/);
 
-      this.ws();
+    this.ws();
 
-      const [ url ] = this.match(/"(.+?)"/);
+    const [ full_url ] = this.match(/"(.+?)"/);
 
-      this.ws();
+    const [url, params] = full_url.split("?");
 
-      const code = this.match(/\-|[0-9]+/);
+    this.ws();
 
-      this.ws();
+    const code = this.match(/\-|[0-9]+/);
 
-      const bytes = this.match(/\-|[0-9]+/);
+    this.ws();
 
-      this.ws();
+    const bytes = this.match(/\-|[0-9]+/);
 
-      const [refer] = this.match(/"(.+?)"/);
+    this.ws();
 
-      this.ws();
+    const [refer] = this.match(/"(.+?)"/);
 
-      const [ua] = this.match(/"(.+?)"/);
+    this.ws();
 
-      return {
-        ip,
-        conn_id: parseInt(conn_id),
-        conn_count: parseInt(conn_count),
-        time: new Date(time),
-        url,
-        code: parseInt(code),
-        size: parseInt(bytes),
-        refer: refer === '-' ? undefined : refer,
-        ua: UAParser(ua)
-      };
-    } catch(error) {
-      throw new Error(`Parsing Error: ${ this.next }`);
-    }
+    const [ua] = this.match(/"(.+?)"/);
+
+    return {
+      ip,
+      conn_id: parseInt(conn_id),
+      conn_count: parseInt(conn_count),
+      time: new Date(time),
+      url, params,
+      code: parseInt(code),
+      size: parseInt(bytes),
+      refer: refer === '-' ? undefined : refer,
+      ua: UAParser(ua)
+    };
   }
-}
 
-const read_stream = fs.createReadStream(process.argv[2]);
+  parse_old_log(ip) {
+    this.match(/- -/);
+
+    this.ws();
+
+    // FORMAT: 29/Mar/2022:22:40:52 +0200
+    const [ day, month, year, hour, minute, seconds, tz_offset ] = this.match(/\[([0-9]+)\/([A-Za-z]+)\/([0-9]+):([0-9]+):([0-9]+):([0-9]+) (.+?)\]/);
+
+    this.ws();
 
-const rl = readline.createInterface({
-  input: read_stream,
-  crlfDelay: Infinity
-});
+    const [ method, full_url, http_version ] = this.match(/"([A-Z]+) (.+) HTTP\/([0-9].[0-9])"/);
 
-const parser = new Parser();
-const chains = {};
+    const [ url, params ] = full_url.split("?");
 
-const skip = /(authcheck\/?|.*.svg|.*.webmanifest|.*.js|.*.css|.*.png|.*.txt|.*.woff|.*.jpg|.*.mp4|.*.torrent|\-|.*.ico|\/api\/.*\?.*|.*.html|.*.map)$/
+    this.ws();
 
-for await (let line of rl) {
-  parser.start(line);
+    const code = this.match(/\-|[0-9]+/);
 
-  try {
-    const data = parser.parse();
+    this.ws();
 
-    if(data.ua.os && data.code === 200 && !data.url.match(skip)) {
-      let chain = data.ip in chains ? chains[data.ip] : [];
+    const bytes = this.match(/\-|[0-9]+/);
 
-      chain.push([data.time, data.url, data.refer]);
+    this.ws();
 
-      chains[data.ip] = chain;
+    const [refer] = this.match(/"(.+?)"/);
+
+    this.ws();
+
+    const [ua] = this.match(/"(.+?)"/);
+
+    // this is another IP address sometimes in another log format that I'll ignore
+    const unknown = this.match(/".+?"$/);
+
+    return {
+      ip,
+      method,
+      http_version,
+      time: new Date(`${day} ${month} ${year} ${hour}:${minute}:${seconds} ${tz_offset}`),
+      url, params,
+      code: parseInt(code),
+      size: parseInt(bytes),
+      refer: refer === '-' ? undefined : refer,
+      ua: UAParser(ua)
+    };
+  }
+
+  parse() {
+    const ip = this.match(/^[0-9\.]+/);
+    const test = this.match(/(:| )/);
+
+    // BUG: uhh for some reason it needs == here? === says : doesn't equal :
+    if(test == ":") {
+      return this.parse_new_log(ip);
+    } else if(test == " ") {
+      return this.parse_old_log(ip);
+    } else {
+      // console.log(`PARSE ERROR, expected : or ' ' but got ${test}`);
+      return {};
     }
-  } catch(err) {
-    if(line !== "") console.error(err);
   }
 }
 
-const uniques = {};
+const parse_logs = async (file_name) => {
+  const read_stream = fs.createReadStream(file_name);
 
-for(let key in chains) {
-  const chain = chains[key];
+  const rl = readline.createInterface({
+    input: read_stream,
+    crlfDelay: Infinity
+  });
 
-  const first = chain[0][2];
+  const parser = new Parser();
 
-  const urls = chain.map(s => {
-    const ref = s[2];
-    if(ref && ref !== first && !ref.includes("learnjsthehardway.com") ) {
-      return `${s[1]}(${s[2]})`;
-    } else {
-      return s[1];
+  const stats = {
+    lines: 0,
+    chains: 0,
+    excluded: 0,
+    errors: 0,
+    roots: 0,
+    firsts: 0
+  };
+
+  const skip = /(authcheck|.*\.svg|.*\.webmanifest|.*\.js|.*\.css|.*php|socket\.io|\.env|.*\.png|.*\.txt|.*\.woff|.*\.jpg|.*\.mp4|.*\.torrent|\-|.*\.ico|\/api\/.*\?.*|.*\.html|.*\.map|.*.php)/
+
+  const by_ip = {};
+
+  for await (let line of rl) {
+    try {
+      stats.lines += 1;
+      parser.start(line);
+
+      const data = parser.parse();
+
+      // skip lines that have content we don't care about
+      if(data.url.match(skip)) continue;
+
+      // store or update the chain in the by_ip chain
+      const ip_chain = by_ip[data.ip] || [];
+
+      ip_chain.push(data);
+
+      by_ip[data.ip] = ip_chain;
+    } catch(error) {
+      stats.errors += 1;
     }
-  });
+  }
 
+  return [by_ip, stats];
+}
+
+const chain_to_set = (requests) => {
+  const path = new Set();
+
+  for(let r of requests) {
+    path.add(r.url);
+  }
+
+  return path.values();
+}
+
+const sort_request_chains = (by_ip) => {
+  let ip_chains = {};
   let seen;
 
-  let full = urls.filter((s, index, self) => {
-    if(s === seen) {
-      return false;
-    } else {
-      seen = s;
-      return true;
-    }
-  }).join(" ");
+  for(let [ip, requests] of Object.entries(by_ip)) {
+    const chain = chain_to_set(requests);
+
+    const ref = requests[0].refer ? `[${requests[0].refer}]` : "";
+    const url_set = [ref, ...chain].join(" ");
 
-  if(first) {
-    full = `[${first}] ${full}`;
+    ip_chains[url_set] = url_set in ip_chains ? ip_chains[url_set] + 1 : 1;
   }
 
-  uniques[full] = full in uniques ? uniques[full] + 1 : 1;
+  const chains_sorted = Object.entries(ip_chains);
+  chains_sorted.sort((a, b) => b[1] - a[1]);
+
+  return chains_sorted;
 }
 
-const sorted = Object.entries(uniques);
-sorted.sort((a, b) => b[1] - a[1]);
+const [by_ip, stats] = await parse_logs(process.argv[2]);
+const chains_sorted = sort_request_chains(by_ip);
 
-for(let [url, count] of sorted) {
+for(let [url, count] of chains_sorted) {
   console.log(count, url);
 }
+
+console.log(stats);

02-filter-a-log-file/package-lock.json

@@ -10,6 +10,7 @@
       "license": "BSD",
       "dependencies": {
         "ava": "^4.3.1",
+        "date-fns": "^2.29.1",
         "ua-parser-js": "^1.0.2"
       }
     },
@@ -500,6 +501,18 @@
         "node": ">=0.10.0"
       }
     },
+    "node_modules/date-fns": {
+      "version": "2.29.1",
+      "resolved": "https://registry.npmjs.org/date-fns/-/date-fns-2.29.1.tgz",
+      "integrity": "sha512-dlLD5rKaKxpFdnjrs+5azHDFOPEu4ANy/LTh04A1DTzMM7qoajmKCBc8pkKRFT41CNzw+4gQh79X5C+Jq27HAw==",
+      "engines": {
+        "node": ">=0.11"
+      },
+      "funding": {
+        "type": "opencollective",
+        "url": "https://opencollective.com/date-fns"
+      }
+    },
     "node_modules/date-time": {
       "version": "3.1.0",
       "resolved": "https://registry.npmjs.org/date-time/-/date-time-3.1.0.tgz",
@@ -2164,6 +2177,11 @@
         "array-find-index": "^1.0.1"
       }
     },
+    "date-fns": {
+      "version": "2.29.1",
+      "resolved": "https://registry.npmjs.org/date-fns/-/date-fns-2.29.1.tgz",
+      "integrity": "sha512-dlLD5rKaKxpFdnjrs+5azHDFOPEu4ANy/LTh04A1DTzMM7qoajmKCBc8pkKRFT41CNzw+4gQh79X5C+Jq27HAw=="
+    },
     "date-time": {
       "version": "3.1.0",
       "resolved": "https://registry.npmjs.org/date-time/-/date-time-3.1.0.tgz",

02-filter-a-log-file/package.json

@@ -11,6 +11,7 @@
   "license": "BSD",
   "dependencies": {
     "ava": "^4.3.1",
+    "date-fns": "^2.29.1",
     "ua-parser-js": "^1.0.2"
   }
 }