Better type checking

changedetectionio/__init__.py

@@ -1343,6 +1343,10 @@ def changedetection_app(config=None, datastore_o=None):
     import changedetectionio.blueprint.browser_steps as browser_steps
     app.register_blueprint(browser_steps.construct_blueprint(datastore), url_prefix='/browser-steps')
 
+    import changedetectionio.blueprint.price_data_follower as price_data_follower
+    app.register_blueprint(price_data_follower.construct_blueprint(datastore), url_prefix='/price_data_follower')
+
+
     # @todo handle ctrl break
     ticker_thread = threading.Thread(target=ticker_thread_check_time_launch_checks).start()
     threading.Thread(target=notification_runner).start()

changedetectionio/blueprint/price_data_follower/__init__.py

@@ -0,0 +1,27 @@
+
+from distutils.util import strtobool
+from flask import Blueprint, flash, redirect, url_for
+from flask_login import login_required
+from changedetectionio.store import ChangeDetectionStore
+
+def construct_blueprint(datastore: ChangeDetectionStore):
+
+    price_data_follower_blueprint = Blueprint('price_data_follower', __name__)
+
+    @login_required
+    @price_data_follower_blueprint.route("/<string:uuid>/accept", methods=['GET'])
+    def accept(uuid):
+        datastore.data['watching'][uuid]['track_ldjson_price_data'] = 'accepted'
+        return redirect(url_for("form_watch_checknow", uuid=uuid))
+
+
+    @login_required
+    @price_data_follower_blueprint.route("/<string:uuid>/reject", methods=['GET'])
+    def reject(uuid):
+        datastore.data['watching'][uuid]['track_ldjson_price_data'] = 'rejected'
+        return redirect(url_for("index"))
+
+
+    return price_data_follower_blueprint
+
+

changedetectionio/fetch_site_status.py

@@ -2,7 +2,6 @@ import hashlib
 import logging
 import os
 import re
-import time
 import urllib3
 
 from changedetectionio import content_fetcher, html_tools
@@ -140,7 +139,7 @@ class perform_site_check():
             is_html = False
             is_json = False
 
-        include_filters_rule = watch.get('include_filters', [])
+        include_filters_rule = deepcopy(watch.get('include_filters', []))
         # include_filters_rule = watch['include_filters']
         subtractive_selectors = watch.get(
             "subtractive_selectors", []
@@ -148,6 +147,10 @@ class perform_site_check():
             "global_subtractive_selectors", []
         )
 
+        # Inject a virtual LD+JSON price tracker rule
+        if watch.get('track_ldjson_price_data'):
+            include_filters_rule.append('json:$..price')
+
         has_filter_rule = include_filters_rule and len("".join(include_filters_rule).strip())
         has_subtractive_selectors = subtractive_selectors and len(subtractive_selectors[0].strip())
 
@@ -173,9 +176,13 @@ class perform_site_check():
                 # Don't run get_text or xpath/css filters on plaintext
                 stripped_text_from_html = html_content
             else:
+                # Does it have some ld+json price data? used for easier monitoring
+                update_obj['has_ldjson_price_data'] = html_tools.has_ldjson_product_info(fetcher.content)
+
                 # Then we assume HTML
                 if has_filter_rule:
                     html_content = ""
+
                     for filter_rule in include_filters_rule:
                         # For HTML/XML we offer xpath as an option, just start a regular xPath "/.."
                         if filter_rule[0] == '/' or filter_rule.startswith('xpath:'):

changedetectionio/html_tools.py

@@ -127,8 +127,10 @@ def _get_stripped_text_from_json_match(match):
 
     return stripped_text_from_html
 
-def extract_json_as_string(content, json_filter):
-
+# content - json
+# json_filter - ie json:$..price
+# ensure_is_ldjson_info_type - str "product", optional, "@type == product" (I dont know how to do that as a json selector)
+def extract_json_as_string(content, json_filter, ensure_is_ldjson_info_type=None):
     stripped_text_from_html = False
 
     # Try to parse/filter out the JSON, if we get some parser error, then maybe it's embedded <script type=ldjson>
@@ -139,7 +141,12 @@ def extract_json_as_string(content, json_filter):
         # Foreach <script json></script> blob.. just return the first that matches json_filter
         s = []
         soup = BeautifulSoup(content, 'html.parser')
-        bs_result = soup.findAll('script')
+
+        if ensure_is_ldjson_info_type:
+            bs_result = soup.findAll('script', {"type": "application/ld+json"})
+        else:
+            bs_result = soup.findAll('script')
+
 
         if not bs_result:
             raise JSONNotFound("No parsable JSON found in this document")
@@ -156,7 +163,12 @@ def extract_json_as_string(content, json_filter):
                 continue
             else:
                 stripped_text_from_html = _parse_json(json_data, json_filter)
-                if stripped_text_from_html:
+                if ensure_is_ldjson_info_type:
+                    # Could sometimes be list, string or something else random
+                    if isinstance(json_data, dict):
+                        if json_data.get('@type', False) and json_data.get('@type','').lower() == ensure_is_ldjson_info_type.lower():
+                            break
+                elif stripped_text_from_html:
                     break
 
     if not stripped_text_from_html:
@@ -243,6 +255,18 @@ def html_to_text(html_content: str, render_anchor_tag_content=False) -> str:
 
     return text_content
 
+
+# Does LD+JSON exist with a @type=='product' and a .price set anywhere?
+def has_ldjson_product_info(content):
+    try:
+        pricing_data = extract_json_as_string(content=content, json_filter='json:$..price', ensure_is_ldjson_info_type="product")
+    except JSONNotFound as e:
+        # Totally fine
+        return False
+    x=bool(pricing_data)
+    return x
+
+
 def workarounds_for_obfuscations(content):
     """
     Some sites are using sneaky tactics to make prices and other information un-renderable by Inscriptis

changedetectionio/model/Watch.py

@@ -26,6 +26,8 @@ class model(dict):
             'extract_title_as_title': False,
             'fetch_backend': None,
             'filter_failure_notification_send': strtobool(os.getenv('FILTER_FAILURE_NOTIFICATION_SEND_DEFAULT', 'True')),
+            'has_ldjson_price_data': None,
+            'track_ldjson_price_data': None,
             'headers': {},  # Extra headers to send
             'ignore_text': [],  # List of text to ignore when calculating the comparison checksum
             'include_filters': [],

changedetectionio/static/styles/scss/styles.scss

@@ -1009,3 +1009,20 @@ ul {
   border-radius: 5px;
   color: var(--color-warning);
 }
+
+/* automatic price following helpers */
+.tracking-ldjson-price-data {
+  background-color: var(--color-background-button-green);
+  color: #000;
+  padding: 3px;
+  border-radius: 3px;
+  white-space: nowrap;
+}
+
+.ldjson-price-track-offer {
+  a.pure-button {
+    border-radius: 3px;
+    padding: 3px;
+    background-color: var(--color-background-button-green);
+  }
+}

changedetectionio/static/styles/styles.css

@@ -945,3 +945,16 @@ ul {
     display: inline;
     height: 26px;
     vertical-align: middle; }
+
+/* automatic price following helpers */
+.tracking-ldjson-price-data {
+  background-color: var(--color-background-button-green);
+  color: #000;
+  padding: 3px;
+  border-radius: 3px;
+  white-space: nowrap; }
+
+.ldjson-price-track-offer a.pure-button {
+  border-radius: 3px;
+  padding: 3px;
+  background-color: var(--color-background-button-green); }

changedetectionio/store.py

@@ -250,12 +250,15 @@ class ChangeDetectionStore:
     def clear_watch_history(self, uuid):
         import pathlib
 
-        self.__data['watching'][uuid].update(
-            {'last_checked': 0,
-             'last_viewed': 0,
-             'previous_md5': False,
-             'last_notification_error': False,
-             'last_error': False})
+        self.__data['watching'][uuid].update({
+                'last_checked': 0,
+                'has_ldjson_price_data': None,
+                'last_error': False,
+                'last_notification_error': False,
+                'last_viewed': 0,
+                'previous_md5': False,
+                'track_ldjson_price_data': None,
+            })
 
         # JSON Data, Screenshots, Textfiles (history index and snapshots), HTML in the future etc
         for item in pathlib.Path(os.path.join(self.datastore_path, uuid)).rglob("*.*"):

changedetectionio/templates/watch-overview.html

@@ -98,6 +98,12 @@
                     {% if watch.last_notification_error is defined and watch.last_notification_error != False %}
                     <div class="fetch-error notification-error"><a href="{{url_for('notification_logs')}}">{{ watch.last_notification_error }}</a></div>
                     {% endif %}
+                    {% if watch['has_ldjson_price_data'] and not watch['track_ldjson_price_data']  %}
+                    <div class="ldjson-price-track-offer">Embedded price data detected, follow only price data? <a href="{{url_for('price_data_follower.accept', uuid=watch.uuid)}}" class="pure-button button-xsmall">Yes</a> <a href="{{url_for('price_data_follower.reject', uuid=watch.uuid)}}" class="">No</a></div>
+                    {% endif %}
+                    {% if watch['track_ldjson_price_data'] %}
+                    <span class="tracking-ldjson-price-data">Price</span>
+                    {% endif %}
                     {% if not active_tag %}
                     <span class="watch-tag-list">{{ watch.tag}}</span>
                     {% endif %}

changedetectionio/tests/test_automatic_follow_ldjson_price.py

@@ -0,0 +1,112 @@
+#!/usr/bin/python3
+
+import time
+from flask import url_for
+from .util import live_server_setup, extract_UUID_from_client, extract_api_key_from_UI
+
+from ..html_tools import *
+
+
+def set_response_with_ldjson():
+    test_return_data = """<html>
+       <body>
+     Some initial text</br>
+     <p>Which is across multiple lines</p>
+     </br>
+     So let's see what happens.  </br>
+     <div class="sametext">Some text thats the same</div>
+     <div class="changetext">Some text that will change</div>
+     <script type="application/ld+json">
+        {
+           "@context":"https://schema.org/",
+           "@type":"Product",
+           "@id":"https://www.some-virtual-phone-shop.com/celular-iphone-14/p",
+           "name":"Celular Iphone 14 Pro Max 256Gb E Sim A16 Bionic",
+           "brand":{
+              "@type":"Brand",
+              "name":"APPLE"
+           },
+           "image":"https://www.some-virtual-phone-shop.com/15509426/image.jpg",
+           "description":"You dont need it",
+           "mpn":"111111",
+           "sku":"22222",
+           "offers":{
+              "@type":"AggregateOffer",
+              "lowPrice":8097000,
+              "highPrice":8099900,
+              "priceCurrency":"COP",
+              "offers":[
+                 {
+                    "@type":"Offer",
+                    "price":8097000,
+                    "priceCurrency":"COP",
+                    "availability":"http://schema.org/InStock",
+                    "sku":"102375961",
+                    "itemCondition":"http://schema.org/NewCondition",
+                    "seller":{
+                       "@type":"Organization",
+                       "name":"ajax"
+                    }
+                 }
+              ],
+              "offerCount":1
+           }
+        }
+       </script>
+     </body>
+     </html>
+"""
+
+    with open("test-datastore/endpoint-content.txt", "w") as f:
+        f.write(test_return_data)
+    return None
+
+
+# actually only really used by the distll.io importer, but could be handy too
+def test_check_ldjson_price_autodetect(client, live_server):
+    live_server_setup(live_server)
+
+    # Give the endpoint time to spin up
+    time.sleep(1)
+
+    set_response_with_ldjson()
+
+    # Add our URL to the import page
+    test_url = url_for('test_endpoint', _external=True)
+    res = client.post(
+        url_for("import_page"),
+        data={"urls": test_url},
+        follow_redirects=True
+    )
+    assert b"1 Imported" in res.data
+    time.sleep(3)
+
+    # Should get a notice that it's available
+    res = client.get(url_for("index"))
+    assert b'ldjson-price-track-offer' in res.data
+
+    # Accept it
+    uuid = extract_UUID_from_client(client)
+
+    client.get(url_for('price_data_follower.accept', uuid=uuid, follow_redirects=True))
+    time.sleep(2)
+
+    # Trigger a check
+    client.get(url_for("form_watch_checknow"), follow_redirects=True)
+    time.sleep(2)
+    # Offer should be gone
+    res = client.get(url_for("index"))
+    assert b'Embedded price data' not in res.data
+    assert b'tracking-ldjson-price-data' in res.data
+
+    # and last snapshop (via API) should be just the price
+    api_key = extract_api_key_from_UI(client)
+    res = client.get(
+        url_for("watchsinglehistory", uuid=uuid, timestamp='latest'),
+        headers={'x-api-key': api_key},
+    )
+
+    # Should just see the price in the API reply
+    assert res.data == b'8097000'
+
+    client.get(url_for("form_delete", uuid="all"), follow_redirects=True)