From 71672da7afc68a7e63a52753a872e0c833f460e6 Mon Sep 17 00:00:00 2001
From: emdee <emdee>
Date: Wed, 16 Nov 2022 13:49:17 +0000
Subject: [PATCH] add lookupdns.py

---
 README.md    | 13 +++++++--
 lookupdns.py | 79 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 89 insertions(+), 3 deletions(-)
 create mode 100644 lookupdns.py

diff --git a/README.md b/README.md
index e92564d..ce62184 100644
--- a/README.md
+++ b/README.md
@@ -26,11 +26,16 @@ way of knowing when that script has finished doing its work. For this
 reason, the external script should execute at the end
 ```console.log("__PHANTOM_PY_DONE__");``` when done. This will trigger
 the PDF generation or the file saving, after which phantompy will exit.
+If you do not want to run any javascipt file, this trigger is provided
+in the code by default.
 
 It is important to remember that since you're just running WebKit, you can
 use everything that WebKit supports, including the usual JS client
 libraries, CSS, CSS @media types, etc.
 
+Qt picks up proxies from the environment, so this will respect
+```https_proxy``` or ```http_proxy``` if set.
+
 ## Dependencies
 
 * Python3
@@ -73,9 +78,11 @@ The standalone program is ```quash_phantompy.py```
 ### Arguments
 
 ```
-<url> Can be a http(s) URL or a path to a local file
-<pdf-file> Path and name of PDF file to generate
-[<javascript-file>] (optional) Path and name of a JavaScript file to execute
+--js_input (optional) Path and name of a JavaScript file to execute on the HTML
+--html_output <html-file> (optional)  Path a HTML output file to generate after JS is applied
+--pdf_output <pdf-file> (optional)  Path and name of PDF file to generate after JS is applied
+--log_level 10=debug 20=info 30=warn 40=error
+html_or_url - required argument, a http(s) URL or a path to a local file.
 ```
 Setting ```DEBUG=1``` in the environment will give debugging messages
 on ```stderr```.
diff --git a/lookupdns.py b/lookupdns.py
new file mode 100644
index 0000000..e792aec
--- /dev/null
+++ b/lookupdns.py
@@ -0,0 +1,79 @@
+#!/usr/local/bin/python3.sh
+# -*-mode: python; indent-tabs-mode: nil; py-indent-offset: 4; coding: utf-8 -*
+
+# Looks for urls https://dns.google/resolve?
+# and parses them to extract a magic field.
+# https://dns.google/resolve?name=domain.name&type=TXT&cd=true&do=true
+
+import sys
+import os
+
+from phantompy import Render
+
+global LOG
+import logging
+import warnings
+warnings.filterwarnings('ignore')
+LOG = logging.getLogger()
+
+class LookFor(Render):
+    
+  def __init__(self, app, do_print=True, do_save=False):
+    app.lfps = []
+    self._app = app
+    self.do_print = do_print
+    self.do_save = do_save
+    self.progress = 0
+    self.we_run_this_tor_relay = None
+    Render.__init__(self, app, do_print, do_save)
+
+  def _exit(self, val):
+    Render._exit(self, val)
+    self.percent = 100
+    LOG.debug(f"phantom.py: Exiting with val {val}")
+    i = self.uri.find('name=')
+    fp = self.uri[i+5:]
+    i = fp.find('.')
+    fp = fp[:i]
+    # threadsafe?
+    self._app.lfps.append(fp)
+    
+  def _html_callback(self, *args):
+    """print(self, QPrinter, Callable[[bool], None])"""
+    if type(args[0]) is str:
+        self._save(args[0])
+        i = self.ilookfor(args[0])
+        self._onConsoleMessage(i, "__PHANTOM_PY_SAVED__", 0 , '')
+
+  def ilookfor(self, html):
+      import json
+      marker = '<pre style="word-wrap: break-word; white-space: pre-wrap;">'
+      if marker not in html: return -1
+      i = html.find(marker) + len(marker)
+      html = html[i:]
+      assert html[0] == '{', html
+      i = html.find('</pre')
+      html = html[:i]
+      assert html[-1] == '}', html
+      LOG.debug(f"Found {len(html)} json")
+      o = json.loads(html)
+      if "Answer" not in o.keys() or type(o["Answer"]) != list:
+          LOG.warn(f"FAIL {self.uri}")
+          return 1
+      for elt in o["Answer"]:
+          assert type(elt) == dict, elt
+          assert 'type' in elt, elt
+          if elt['type'] != 16: continue
+          assert 'data' in elt, elt
+          if elt['data'] == 'we-run-this-tor-relay':
+              LOG.info(f"OK {self.uri}")
+              self.we_run_this_tor_relay = True
+              return 0
+      self.we_run_this_tor_relay = False
+      LOG.warn(f"BAD {self.uri}")
+      return 2
+  
+  def _loadFinished(self, result):
+      LOG.debug(f"phantom.py: Loading finished {self.uri}")
+      self.toHtml(self._html_callback)
+