From 3feb4640970094682203f5a39e0c99c752d1db2a Mon Sep 17 00:00:00 2001
From: Corban-Lee <corbz.jones@gmail.com>
Date: Thu, 15 May 2025 10:32:40 +0100
Subject: [PATCH] test(bot): write tests for ANY filter

---
 src/bot/__tests__/filters.test.ts | 67 +++++++++++++++++++++++++++++--
 src/bot/filter.ts                 | 35 ++++++----------
 2 files changed, 77 insertions(+), 25 deletions(-)

diff --git a/src/bot/__tests__/filters.test.ts b/src/bot/__tests__/filters.test.ts
index e1664a8..7beb54a 100644
--- a/src/bot/__tests__/filters.test.ts
+++ b/src/bot/__tests__/filters.test.ts
@@ -14,13 +14,72 @@ const templateFilter: prisma.Filter = {
 
 };
 
-describe("Regex Matching Test", () => {
+describe("Match: ALL", () => {  });
+
+describe("Match: ANY", () => { 
+    const filter: prisma.Filter = {
+        ...templateFilter,
+        name: "Block American Politics",
+        value: String.raw`trump biden democrat republican gop dnc kamala harris`,
+        is_insensitive: true
+    };
+
+    const testCases: { input: string, expected: boolean }[] = [
+{
+            input: "Republicans float new tax breaks for tips, local taxes in Trump budget package",
+            expected: true  // Contains 'republican' and 'trump'.
+        },
+        {
+            input: "Biden’s Approval Rating Dips Amid Economic Concerns",
+            expected: true  // Contains 'biden'.
+        },
+        {
+            input: "GOP Governors Call for Tougher Border Security Measures",
+            expected: true  // Contains 'gop'.
+        },
+        {
+            input: "Biden Announces New Initiative to Tackle Climate Change",
+            expected: true  // Contains 'biden'.
+        },
+        {
+            input: "Biden Announces New Initiative to Tackle Climate Change",
+            expected: true  // Contains 'biden'.
+        },
+        {
+            input: "Joe Biden gives thoughts on Donald Trump, Vladimir Putin, Ukraine, and Kamala Harris in first interview since leaving office",
+            expected: true  // Contains 'biden', 'trump', 'kamala' and 'harris'.
+        },
+        {
+            input: "UK Prime Minister Keir Starmer hails limited US-UK trade deal, but 10% duties remain",
+            expected: false
+        },
+        {
+            input: "Scientists discover new species of fish in Pacific Ocean",
+            expected: false
+        },
+        {
+            input: "Federal Reserve signals cautious approach to interest rate cuts in 2025",
+            expected: false
+        }
+    ];
+
+    for (const { input, expected } of testCases) {
+        test(`Test input: ${input}`, () => {
+            const result = filters.any(filter, input);
+            expect(result).toBe(expected);
+        });
+    }
+});
+
+describe("Match: LITERAL", () => {  });
+
+describe("Match: REGEX", () => {
     const filter: prisma.Filter = {
         ...templateFilter,
         name: "Block American Politics",
         value: String.raw`\b(trump|biden|democrat|republican|gop|dnc|kamala|harris)\b`,
         is_insensitive: true
-    }
+    };
 
     const testCases: { input: string, expected: boolean }[] = [
         {
@@ -67,4 +126,6 @@ describe("Regex Matching Test", () => {
             expect(result).toBe(expected);
         });
     }
-});
\ No newline at end of file
+});
+
+describe("Match: FUZZY", () => {  });
diff --git a/src/bot/filter.ts b/src/bot/filter.ts
index 5b78461..f425690 100644
--- a/src/bot/filter.ts
+++ b/src/bot/filter.ts
@@ -1,11 +1,13 @@
 import fuzz from "fuzzball";  // todo: implement for fuzzy match
 import { Filter, MatchingAlgorithms } from "../../generated/prisma";
 
-function splitMatch(filterValue: string): string[] {
+function splitWords(filterValue: string): string[] {
     const findTerms = [...filterValue.matchAll(/"([^"]+)"|(\S+)/g)];
     return findTerms.map(value => {
         const term = value[1] || value[2];
-        return term.trim().replace(/\s+/g, "\\s+").replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
+        return term.trim()
+            .replace(/\s+/g, "\\s+")                 // Replace whitespace with equivelant regex characters
+            .replace(/[.*+?^${}()|[\]\\]/g, '\\$&')  // Escape common regex characters
     });
 }
 
@@ -14,26 +16,15 @@ export const all = (filter: Filter, input: string) => {
 };
 
 export const any = (filter: Filter, input: string) => {
-    // const words = splitMatch(filter.value);
-    // console.log(JSON.stringify(words));
-    // return words.some(word => new RegExp(`\\b${word}\\b`).test(input));
-
-    // TODO:
-    //      The below code works, but I'd like to use regex as was done in pyrss.
-    //      The above commented code does not work, figure it out.
-
-
-    const inputWords = input.split(" ").map(word => filter.is_insensitive ? word.toLowerCase() : word);
-    const filterWords = filter.value.split(" ").map(word => filter.is_insensitive ? word.toLowerCase() : word);
-
-    for (const filterWord of filterWords) {
-        if (inputWords.includes(filterWord)) {
-            console.log(input);
-            return true;
-        }
+    try {
+        const flags = filter.is_insensitive ? "i" : "";
+        const filterWords = splitWords(filter.value).toString().replace(/,/g, "|");
+        const filterValue = String.raw`\b(${filterWords})\b`
+        return new RegExp(filterValue, flags).test(input);
+    } catch (error) {
+        console.error(`ANY: Invalid regex pattern: ${filter.value}`, error);
+        return false;
     }
-
-    return false;
 };
 
 export const literal = (filter: Filter, input: string) => {
@@ -45,7 +36,7 @@ export const regex = (filter: Filter, input: string) => {
         const flags = filter.is_insensitive ? "i" : "";
         return new RegExp(filter.value, flags).test(input);
     } catch (error) {
-        console.error(`Invalid regex pattern: ${filter.value}`, error);
+        console.error(`REGEX: Invalid regex pattern: ${filter.value}`, error);
         return false;
     }
 };