[pull] master from sherlock-project:master

.github/CODEOWNERS

@@ -1,5 +1,5 @@
 ### REPOSITORY
-/.github/CODEOWNERS @sdushantha
+/.github/CODEOWNERS @sdushantha @ppfeister
 /.github/FUNDING.yml @sdushantha
 /LICENSE @sdushantha
 

.github/workflows/regression.yml

@@ -11,6 +11,7 @@ on:
       - '**/*.py'
       - '**/*.ini'
       - '**/*.toml'
+      - 'Dockerfile'
   push:
     branches:
       - master
@@ -21,11 +22,13 @@ on:
       - '**/*.py'
       - '**/*.ini'
       - '**/*.toml'
+      - 'Dockerfile'
 
 jobs:
   tox-lint:
-    # Linting is ran through tox to ensure that the same linter is used by local runners
     runs-on: ubuntu-latest
+    # Linting is ran through tox to ensure that the same linter
+    # is used by local runners
     steps:
       - uses: actions/checkout@v4
       - name: Set up linting environment
@@ -41,7 +44,8 @@ jobs:
   tox-matrix:
     runs-on: ${{ matrix.os }}
     strategy:
-      fail-fast: false # We want to know what specicic versions it fails on
+      # We want to know what specicic versions it fails on
+      fail-fast: false
       matrix:
         os: [
           ubuntu-latest,
@@ -67,3 +71,22 @@ jobs:
           pip install tox-gh-actions
       - name: Run tox
         run: tox
+  docker-build-test:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+      - name: Get version from pyproject.toml
+        id: get-version
+        run: |
+          VERSION=$(grep -m1 'version = ' pyproject.toml | cut -d'"' -f2)
+          echo "version=$VERSION" >> $GITHUB_OUTPUT
+      - name: Build Docker image
+        run: |
+          docker build \
+            --build-arg VERSION_TAG=${{ steps.get-version.outputs.version }} \
+            -t sherlock-test:latest .
+      - name: Test Docker image runs
+        run: docker run --rm sherlock-test:latest --version

Dockerfile

@@ -4,7 +4,7 @@
   # 3. Build image with BOTH latest and version tags
     # i.e. `docker build -t sherlock/sherlock:0.16.0 -t sherlock/sherlock:latest .`
 
-FROM python:3.12-slim-bullseye as build
+FROM python:3.12-slim-bullseye AS build
 WORKDIR /sherlock
 
 RUN pip3 install --no-cache-dir --upgrade pip

pyproject.toml

@@ -46,14 +46,10 @@ PySocks = "^1.7.0"
 requests = "^2.22.0"
 requests-futures = "^1.0.0"
 stem = "^1.8.0"
-torrequest = "^0.1.0"
 pandas = "^2.2.1"
 openpyxl = "^3.0.10"
 tomli = "^2.2.1"
 
-[tool.poetry.extras]
-tor = ["torrequest"]
-
 [tool.poetry.group.dev.dependencies]
 jsonschema = "^4.0.0"
 rstr = "^3.2.2"

sherlock_project/resources/data.json

@@ -79,13 +79,13 @@
     "username_claimed": "pink"
   },
   "AllMyLinks": {
-    "errorMsg": "Not Found",
-    "errorType": "message",
-    "regexCheck": "^[a-z0-9][a-z0-9-]{2,32}$",
-    "url": "https://allmylinks.com/{}",
-    "urlMain": "https://allmylinks.com/",
-    "username_claimed": "blue"
-  },
+  "errorMsg": "Page not found",
+  "errorType": "message",
+  "regexCheck": "^[a-z0-9][a-z0-9-]{2,32}$",
+  "url": "https://allmylinks.com/{}",
+  "urlMain": "https://allmylinks.com/",
+  "username_claimed": "blue"
+},
   "AniWorld": {
     "errorMsg": "Dieses Profil ist nicht verf\u00fcgbar",
     "errorType": "message",
@@ -115,7 +115,7 @@
     "username_claimed": "lio24d"
   },
   "Apple Discussions": {
-    "errorMsg": "The page you tried was not found. You may have used an outdated link or may have typed the address (URL) incorrectly.",
+    "errorMsg": "Looking for something in Apple Support Communities?",
     "errorType": "message",
     "url": "https://discussions.apple.com/profile/{}",
     "urlMain": "https://discussions.apple.com",
@@ -572,8 +572,7 @@
     "username_claimed": "brown"
   },
   "CyberDefenders": {
-    "errorMsg": "<title>Blue Team Training for SOC analysts and DFIR - CyberDefenders</title>",
-    "errorType": "message",
+    "errorType": "status_code",
     "regexCheck": "^[^\\/:*?\"<>|@]{3,50}$",
     "request_method": "GET",
     "url": "https://cyberdefenders.org/p/{}",
@@ -600,6 +599,12 @@
     "urlMain": "https://www.dailymotion.com/",
     "username_claimed": "blue"
   },
+  "dcinside": {
+    "errorType": "status_code",
+    "url": "https://gallog.dcinside.com/{}",
+    "urlMain": "https://www.dcinside.com/",
+    "username_claimed": "anrbrb"
+  },
   "Dealabs": {
     "errorMsg": "La page que vous essayez",
     "errorType": "message",
@@ -608,13 +613,14 @@
     "urlMain": "https://www.dealabs.com/",
     "username_claimed": "blue"
   },
-  "DeviantART": {
-    "errorType": "status_code",
-    "regexCheck": "^[a-zA-Z][a-zA-Z0-9_-]*$",
-    "url": "https://{}.deviantart.com",
-    "urlMain": "https://deviantart.com",
-    "username_claimed": "blue"
-  },
+ "DeviantArt": {
+  "errorType": "message",
+  "errorMsg": "Llama Not Found",
+  "regexCheck": "^[a-zA-Z][a-zA-Z0-9_-]*$",
+  "url": "https://www.deviantart.com/{}",
+  "urlMain": "https://www.deviantart.com/",
+  "username_claimed": "blue"
+},
   "DigitalSpy": {
       "errorMsg": "The page you were looking for could not be found.",
       "errorType": "message",
@@ -1440,12 +1446,12 @@
     "username_claimed": "blue"
   },
   "Mydramalist": {
-    "errorMsg": "Sign in - MyDramaList",
-    "errorType": "message",
-    "url": "https://www.mydramalist.com/profile/{}",
-    "urlMain": "https://mydramalist.com",
-    "username_claimed": "elhadidy12398"
-  },
+  "errorMsg": "The requested page was not found",
+  "errorType": "message",
+  "url": "https://www.mydramalist.com/profile/{}",
+  "urlMain": "https://mydramalist.com",
+  "username_claimed": "elhadidy12398"
+},
   "Myspace": {
     "errorType": "status_code",
     "url": "https://myspace.com/{}",
@@ -1459,6 +1465,13 @@
     "urlMain": "https://www.native-instruments.com/forum/",
     "username_claimed": "jambert"
   },
+  "namuwiki": {
+    "__comment__": "This is a Korean site and it's expected to return false negatives in certain other regions.",
+    "errorType": "status_code",
+    "url": "https://namu.wiki/w/%EC%82%AC%EC%9A%A9%EC%9E%90:{}",
+    "urlMain": "https://namu.wiki/",
+    "username_claimed": "namu"
+  },
   "NationStates Nation": {
     "errorMsg": "Was this your nation? It may have ceased to exist due to inactivity, but can rise again!",
     "errorType": "message",
@@ -1809,8 +1822,7 @@
     "username_claimed": "blue"
   },
   "Roblox": {
-    "errorMsg": "Page cannot be found or no longer exists",
-    "errorType": "message",
+    "errorType": "status_code",
     "url": "https://www.roblox.com/user.aspx?username={}",
     "urlMain": "https://www.roblox.com/",
     "username_claimed": "bluewolfekiller"
@@ -1918,7 +1930,7 @@
   },
   "SlideShare": {
     "errorType": "message",
-    "errorMsg": "<title>Username available</title>",
+    "errorMsg": "<title>Page no longer exists</title>",
     "url": "https://slideshare.net/{}",
     "urlMain": "https://slideshare.net/",
     "username_claimed": "blue"
@@ -1952,6 +1964,13 @@
     "urlMain": "https://www.snapchat.com",
     "username_claimed": "teamsnapchat"
   },
+  "SOOP": {
+    "errorType": "status_code",
+    "url": "https://www.sooplive.co.kr/station/{}",
+    "urlMain": "https://www.sooplive.co.kr/",
+    "urlProbe": "https://api-channel.sooplive.co.kr/v1.1/channel/{}/station",
+    "username_claimed": "udkn"
+  },
   "SoundCloud": {
     "errorType": "status_code",
     "url": "https://soundcloud.com/{}",
@@ -2119,6 +2138,12 @@
     "urlMain": "https://themeforest.net/",
     "username_claimed": "user"
   },
+  "tistory": {
+    "errorType": "status_code",
+    "url": "https://{}.tistory.com/",
+    "urlMain": "https://www.tistory.com/",
+    "username_claimed": "notice"
+  },
   "TnAFlix": {
     "errorType": "status_code",
     "isNSFW": true,
@@ -2793,7 +2818,7 @@
     "username_claimed": "green"
   },
   "threads": {
-    "errorMsg": "<title>Threads</title>",
+    "errorMsg": "<title>Threads • Log in</title>",
     "errorType": "message",
     "headers": {
       "Sec-Fetch-Mode": "navigate"