match

src/pkg/utils/match.test.ts

@@ -1,10 +1,157 @@
 import { describe, expect, it } from "vitest";
-import { dealMatches } from "./match";
+import { dealPatternMatches, parsePatternMatchesURL, UrlMatch } from "./match";
 
 // https://developer.chrome.com/docs/extensions/mv3/match_patterns/
-describe("dealMatches", () => {
-  it("*://link.17173.com*", () => {
-    const matches = dealMatches(["*://link.17173.com*"]);
-    expect(matches).toEqual(["*://link.17173.com/*"]);
+describe("UrlMatch-google", () => {
+  const url = new UrlMatch<string>();
+  url.add("https://*/*", "ok1");
+  url.add("https://*/foo*", "ok2");
+  url.add("https://*.google.com/foo*bar", "ok3");
+  url.add("https://example.org/foo/bar.html", "ok4");
+  url.add("http://127.0.0.1/*", "ok5");
+  url.add("*://mail.google.com/*", "ok6");
+  it("match1", () => {
+    expect(url.match("https://www.google.com/")).toEqual(["ok1"]);
+    expect(url.match("https://example.org/foo/bar.html")).toEqual(["ok1", "ok2", "ok4"]);
+  });
+  it("match2", () => {
+    expect(url.match("https://example.com/foo/bar.html")).toEqual(["ok1", "ok2"]);
+    expect(url.match("https://www.google.com/foo")).toEqual(["ok1", "ok2"]);
+    expect(url.match("https://www.google.com/foo2")).toEqual(["ok1", "ok2"]);
+  });
+  it("match3", () => {
+    expect(url.match("https://www.google.com/foo/baz/bar")).toEqual(["ok1", "ok2", "ok3"]);
+    expect(url.match("https://docs.google.com/foobar")).toEqual(["ok1", "ok2", "ok3"]);
+  });
+  it("match4", () => {
+    expect(url.match("https://example.org/foo/bar.html")).toEqual(["ok1", "ok2", "ok4"]);
+  });
+  it("match5", () => {
+    expect(url.match("http://127.0.0.1/")).toEqual(["ok5"]);
+    expect(url.match("http://127.0.0.1/foo/bar.html")).toEqual(["ok5"]);
+  });
+  it("match6", () => {
+    expect(url.match("http://mail.google.com/foo/baz/bar")).toEqual(["ok6"]);
+    expect(url.match("https://mail.google.com/foobar")).toEqual(["ok1", "ok2", "ok3", "ok6"]);
+  });
+});
+
+describe("UrlMatch-google-error", () => {
+  const url = new UrlMatch<string>();
+  it("error-1", () => {
+    expect(() => {
+      url.add("https://*foo/bar", "ok1");
+    }).toThrow(Error);
+  });
+  it("error-2", () => {
+    expect(() => {
+      url.add("https://foo.*.bar/baz", "ok1");
+    }).toThrow(Error);
+  });
+  it("error-3", () => {
+    expect(() => {
+      url.add("http:/bar", "ok1");
+    }).toThrow(Error);
+  });
+});
+
+// 从tm找的一些特殊的匹配规则
+describe("UrlMatch-search", () => {
+  const url = new UrlMatch<string>();
+  url.add("https://www.google.com/search?q=*", "ok1");
+  it("match1", () => {
+    expect(url.match("https://www.google.com/search?q=foo")).toEqual(["ok1"]);
+    expect(url.match("https://www.google.com/search?q1=foo")).toEqual([]);
+  });
+
+  url.add("https://bbs.tampermonkey.net.cn", "ok2");
+  it("match2", () => {
+    expect(url.match("https://bbs.tampermonkey.net.cn")).toEqual(["ok2"]);
+    expect(url.match("https://bbs.tampermonkey.net.cn/")).toEqual(["ok2"]);
+    expect(url.match("https://bbs.tampermonkey.net.cn/foo/bar.html")).toEqual([]);
+  });
+});
+
+describe("UrlMatch-port1", () => {
+  const url = new UrlMatch<string>();
+  url.add("http://test.list.ggnb.top/search", "ok1");
+  it("match1", () => {
+    expect(url.match("http://test.list.ggnb.top/search")).toEqual(["ok1"]);
+    expect(url.match("http://test.list.ggnb.top/search?")).toEqual([]);
+    expect(url.match("http://test.list.ggnb.top/search?foo=bar")).toEqual([]);
+  });
+
+  it("port", () => {
+    expect(url.match("http://test.list.ggnb.top:80/search")).toEqual(["ok1"]);
+  });
+});
+
+describe("UrlMatch-port2", () => {
+  const url = new UrlMatch<string>();
+  url.add("http://test.list.ggnb.top:80/search", "ok1");
+  url.add("http://test.list.ggnb.top*/search", "ok2");
+  url.add("http://test.list.ggnb.top:*/search", "ok3");
+  url.add("http://localhost:3000/", "ok4");
+  it("match1", () => {
+    expect(url.match("http://test.list.ggnb.top:80/search")).toEqual(["ok1", "ok2", "ok3"]);
+    expect(url.match("http://test.list.ggnb.top:81/search")).toEqual(["ok2", "ok3"]);
+    expect(url.match("http://test.list.ggnb.top/search")).toEqual(["ok1", "ok2", "ok3"]);
+  });
+  it("case2", () => {
+    expect(url.match("http://localhost:3000/")).toEqual(["ok4"]);
+    expect(url.match("http://localhost:8000/")).toEqual([]);
+  });
+});
+
+// https://developer.chrome.com/docs/extensions/mv3/match_patterns/
+describe("dealPatternMatches", () => {
+  it("https://developer.chrome.com/docs/extensions/develop/concepts/match-patterns?hl=zh-cn#examples", () => {
+    const matches = dealPatternMatches(["https://*/*", "http://127.0.0.1/*", "http://127.0.0.1/"]);
+    expect(matches.patternResult).toEqual(["https://*/*", "http://127.0.0.1/*", "http://127.0.0.1/"]);
+  });
+  // 处理一些特殊情况
+  it("*://link.17173.com*", () => {
+    const matches = dealPatternMatches(["*://link.17173.com*"]);
+    expect(matches.patternResult).toEqual(["*://link.17173.com/*"]);
+  });
+});
+
+describe("parsePatternMatchesURL", () => {
+  it("https://developer.chrome.com/docs/extensions/develop/concepts/match-patterns?hl=zh-cn#examples", () => {
+    const matches = parsePatternMatchesURL("https://*/*");
+    expect(matches).toEqual({
+      scheme: "https",
+      host: "*",
+      path: "*",
+    });
+    const matches2 = parsePatternMatchesURL("https://*/foo*");
+    expect(matches2).toEqual({
+      scheme: "https",
+      host: "*",
+      path: "foo*",
+    });
+    const matches3 = parsePatternMatchesURL("http://127.0.0.1/");
+    expect(matches3).toEqual({
+      scheme: "http",
+      host: "127.0.0.1",
+      path: "",
+    });
+  });
+  it("search", () => {
+    // 会忽略掉search部分
+    const matches = parsePatternMatchesURL("https://*/*?search");
+    expect(matches).toEqual({
+      scheme: "https",
+      host: "*",
+      path: "*",
+    });
+  });
+  it("*://link.17173.com*", () => {
+    const matches = parsePatternMatchesURL("*://link.17173.com*");
+    expect(matches).toEqual({
+      scheme: "*",
+      host: "link.17173.com",
+      path: "*",
+    });
   });
 });

src/pkg/utils/match.ts

@@ -1,3 +1,5 @@
+import Logger from "@App/app/logger/logger";
+
 export interface Url {
   scheme: string;
   host: string;
@@ -5,50 +7,269 @@ export interface Url {
   search: string;
 }
 
-// 根据https://developer.chrome.com/docs/extensions/develop/concepts/match-patterns?hl=zh-cn进行匹配
-export class Match {}
+export default class Match<T> {
+  protected cache = new Map<string, T[]>();
 
-export function parseURL(url: string): Url | undefined {
-  const match = /^(.+?):\/\/(.*?)((\/.*?)(\?.*?|)|)$/.exec(url);
-  if (match) {
-    return {
-      scheme: match[1],
-      host: match[2],
-      path: match[4] || (url[url.length - 1] === "*" ? "*" : "/"),
-      search: match[5],
-    };
-  }
-  // 处理一些特殊情况
-  switch (url) {
-    case "*":
+  protected rule = new Map<string, T[]>();
+
+  protected parseURL(url: string): Url | undefined {
+    if (url.indexOf("*http") === 0) {
+      url = url.substring(1);
+    }
+    const match = /^(.+?):\/\/(.*?)((\/.*?)(\?.*?|)|)$/.exec(url);
+    if (match) {
       return {
-        scheme: "*",
-        host: "*",
-        path: "*",
-        search: "*",
+        scheme: match[1],
+        host: match[2],
+        path: match[4] || (url[url.length - 1] === "*" ? "*" : "/"),
+        search: match[5],
       };
-    default:
+    }
+    // 处理一些特殊情况
+    switch (url) {
+      case "*":
+        return {
+          scheme: "*",
+          host: "*",
+          path: "*",
+          search: "*",
+        };
+      default:
+    }
+    return undefined;
+  }
+
+  protected compileRe(url: string): string {
+    const u = this.parseURL(url);
+    if (!u) {
+      return "";
+    }
+    switch (u.scheme) {
+      case "*":
+        u.scheme = ".+?";
+        break;
+      case "http*":
+        u.scheme = "http[s]?";
+        break;
+      default:
+    }
+    let pos = u.host.indexOf("*");
+    if (u.host === "*" || u.host === "**") {
+      pos = -1;
+    } else if (u.host.endsWith("*")) {
+      // 处理*结尾
+      if (!u.host.endsWith(":*")) {
+        u.host = u.host.substring(0, u.host.length - 1);
+      }
+    } else if (pos !== -1 && pos !== 0) {
+      return "";
+    }
+    u.host = u.host.replace(/\*/g, "[^/]*?");
+    // 处理 *.开头
+    if (u.host.startsWith("[^/]*?.")) {
+      u.host = `([^/]*?\\.?)${u.host.substring(7)}`;
+    } else if (pos !== -1) {
+      if (u.host.indexOf(".") === -1) {
+        return "";
+      }
+    }
+    // 处理顶域
+    if (u.host.endsWith("tld")) {
+      u.host = `${u.host.substr(0, u.host.length - 3)}.*?`;
+    }
+    // 处理端口
+    const pos2 = u.host.indexOf(":");
+    if (pos2 === -1) {
+      u.host = `${u.host}(:\\d+)?`;
+    } else {
+      const port = u.host.substring(pos2 + 1);
+      if (port === "*") {
+        u.host = `${u.host.substring(0, pos2)}(:\\d+)?`;
+      } else {
+        u.host = `${u.host.substring(0, pos2)}(:${port})?`;
+      }
+    }
+    let re = `^${u.scheme}://${u.host}`;
+    if (u.path === "/") {
+      re += "[/]?";
+    } else {
+      re += u.path.replace(/\*/g, ".*?");
+    }
+    if (u.search) {
+      re += u.search.replace(/([\\?])/g, "\\$1").replace(/\*/g, ".*?");
+    }
+    return `${re.replace(/\//g, "/")}$`;
+  }
+
+  public add(url: string, val: T) {
+    const re = this.compileRe(url);
+    if (!re) {
+      throw new Error(`invalid url: ${url}`);
+    }
+    let rule = this.rule.get(re);
+    if (!rule) {
+      rule = [];
+      this.rule.set(re, rule);
+    }
+    rule.push(val);
+    this.delCache();
+  }
+
+  public match(url: string): T[] {
+    let ret = this.cache.get(url);
+    if (ret) {
+      return ret;
+    }
+    ret = [];
+    try {
+      this.rule.forEach((val, key) => {
+        const re = new RegExp(key);
+        if (re.test(url) && ret) {
+          ret.push(...val);
+        }
+      });
+    } catch (e) {
+      // eslint-disable-next-line no-console
+      console.warn("bad match rule", Logger.E(e));
+      // LoggerCore.getLogger({ component: "match" }).warn(
+      //   "bad match rule",
+      //   Logger.E(e)
+      // );
+    }
+    this.cache.set(url, ret);
+    return ret;
+  }
+
+  protected static getId(val: any): string {
+    if (typeof val === "object") {
+      return (<{ uuid: string }>(<unknown>val)).uuid;
+    }
+    return <string>(<unknown>val);
+  }
+
+  public del(val: T) {
+    const id = Match.getId(val);
+    this.rule.forEach((rules, key) => {
+      const tmp: T[] = [];
+      rules.forEach((rule) => {
+        if (Match.getId(rule) !== id) {
+          tmp.push(rule);
+        }
+      });
+      if (tmp) {
+        this.rule.set(key, tmp);
+      } else {
+        this.rule.delete(key);
+      }
+    });
+    this.delCache();
+  }
+
+  protected delCache() {
+    this.cache.clear();
   }
-  return undefined;
 }
 
-// 处理油猴的match和include为chrome的matches
-export function dealMatches(matches: string[]) {
-  const result: string[] = [];
-  for (let i = 0; i < matches.length; i++) {
-    const url = parseURL(matches[i]);
-    if (url) {
-      // *开头但是不是*.的情况
-      if (url.host.startsWith("*")) {
-        if (!url.host.startsWith("*.")) {
-          // 删除开头的*号
-          url.host = url.host.slice(1);
-        }
-      } else if (url.host.endsWith("*")) {
-        url.host = url.host.slice(0, -1);
+export class UrlMatch<T> extends Match<T> {
+  protected excludeMatch = new Match<T>();
+
+  public exclude(url: string, val: T) {
+    this.excludeMatch.add(url, val);
+  }
+
+  public del(val: T): void {
+    super.del(val);
+    this.excludeMatch.del(val);
+    this.cache.clear();
+  }
+
+  public match(url: string): T[] {
+    const cache = this.cache.get(url);
+    if (cache) {
+      return cache;
+    }
+    let ret = super.match(url);
+    // 排除
+    const includeMap = new Map();
+    ret.forEach((val) => {
+      includeMap.set(Match.getId(val), val);
+    });
+    const exclude = this.excludeMatch.match(url);
+    const excludeMap = new Map();
+    exclude.forEach((val) => {
+      excludeMap.set(Match.getId(val), 1);
+    });
+    ret = [];
+    includeMap.forEach((val: T, key) => {
+      if (!excludeMap.has(key)) {
+        ret.push(val);
+      }
+    });
+    this.cache.set(url, ret);
+    return ret;
+  }
+}
+
+export interface PatternMatchesUrl {
+  scheme: string;
+  host: string;
+  path: string;
+}
+
+// 解析URL, 根据https://developer.chrome.com/docs/extensions/develop/concepts/match-patterns?hl=zh-cn进行处理
+export function parsePatternMatchesURL(url: string): PatternMatchesUrl | undefined {
+  let result: PatternMatchesUrl | undefined;
+  const match = /^(.+?):\/\/(.*?)(\/(.*?)(\?.*?|)|)$/.exec(url);
+  if (match) {
+    result = {
+      scheme: match[1],
+      host: match[2],
+      path: match[4] || (url[url.length - 1] === "*" ? "*" : ""),
+    };
+  } else {
+    // 处理一些特殊情况
+    switch (url) {
+      case "*":
+        result = {
+          scheme: "*",
+          host: "*",
+          path: "*",
+        };
+        break;
+      default:
+    }
+  }
+  if (result) {
+    if (result.host !== "*") {
+      // *开头但是不是*.的情况
+      if (result.host.startsWith("*")) {
+        if (!result.host.startsWith("*.")) {
+          // 删除开头的*号
+          result.host = result.host.slice(1);
+        }
+      }
+      // 结尾是*的情况
+      if (result.host.endsWith("*")) {
+        result.host = result.host.slice(0, -1);
       }
-      result.push(`${url.scheme}://${url.host}/${url.path}` + (url.search ? "?" + url.search : ""));
     }
   }
   return result;
 }
+
+// 处理油猴的match和include为chrome的pattern-matche
+export function dealPatternMatches(matches: string[]) {
+  const patternResult: string[] = [];
+  const result: string[] = [];
+  for (let i = 0; i < matches.length; i++) {
+    const url = parsePatternMatchesURL(matches[i]);
+    if (url) {
+      patternResult.push(`${url.scheme}://${url.host}/${url.path}`);
+      result.push(matches[i]);
+    }
+  }
+  return {
+    patternResult,
+    result,
+  };
+}