From 321ab19ffbf492ed34a567604ba93109dea3e826 Mon Sep 17 00:00:00 2001
From: dgtlmoon <dgtlmoon@gmail.com>
Date: Tue, 30 Jul 2024 17:10:21 +0200
Subject: [PATCH] Ability to use our own plugins to scrape extra data

---
 .../processors/restock_diff/hookspecs.py      | 23 +++++++++++++++++++
 .../processors/restock_diff/plugin_manager.py | 17 ++++++++++++++
 .../processors/restock_diff/processor.py      | 15 ++++++++++++
 .../templates/watch-overview.html             |  2 +-
 requirements.txt                              |  3 +++
 5 files changed, 59 insertions(+), 1 deletion(-)
 create mode 100644 changedetectionio/processors/restock_diff/hookspecs.py
 create mode 100644 changedetectionio/processors/restock_diff/plugin_manager.py

diff --git a/changedetectionio/processors/restock_diff/hookspecs.py b/changedetectionio/processors/restock_diff/hookspecs.py
new file mode 100644
index 00000000..e14cb5a9
--- /dev/null
+++ b/changedetectionio/processors/restock_diff/hookspecs.py
@@ -0,0 +1,23 @@
+import pluggy
+from typing import Dict
+from changedetectionio.model import Watch as Watch
+
+plugin_namespace = "changedetectionio.restock_price_scraper"
+hookspec = pluggy.HookspecMarker(plugin_namespace)
+
+class HookSpec:
+    @hookspec
+    def scrape_price_restock(self, watch: Watch.model, html_content: str, screenshot: bytes, update_obj: Dict) -> Dict:
+        """
+         Scrape price and restock data from html_content and/or screenshot and return via update_obj
+
+         Args:
+             watch (Watch.model): The watch object containing watch configuration.
+             html_content (str): The HTML content to scrape.
+             screenshot (bytes): The screenshot data.
+             update_obj (Dict): The dictionary to update with scraped data.
+
+         Returns:
+             Optional[Dict]: The updated dictionary with the scraped price data, or None if no update is made.
+         """
+
diff --git a/changedetectionio/processors/restock_diff/plugin_manager.py b/changedetectionio/processors/restock_diff/plugin_manager.py
new file mode 100644
index 00000000..c3702f65
--- /dev/null
+++ b/changedetectionio/processors/restock_diff/plugin_manager.py
@@ -0,0 +1,17 @@
+import pluggy
+from .hookspecs import HookSpec
+import importlib.metadata
+
+# Define the plugin namespace
+plugin_namespace = "changedetectionio.restock_price_scraper"
+
+# Create a pluggy.PluginManager instance
+pm = pluggy.PluginManager(plugin_namespace)
+
+# Register the hook specifications
+pm.add_hookspecs(HookSpec)
+
+# Automatically discover and register plugins using entry points
+for entry_point in importlib.metadata.entry_points().get(plugin_namespace, []):
+    plugin = entry_point.load()
+    pm.register(plugin())
diff --git a/changedetectionio/processors/restock_diff/processor.py b/changedetectionio/processors/restock_diff/processor.py
index b2184e35..2dc728b7 100644
--- a/changedetectionio/processors/restock_diff/processor.py
+++ b/changedetectionio/processors/restock_diff/processor.py
@@ -119,6 +119,8 @@ class perform_site_check(difference_detection_processor):
     xpath_data = None
 
     def run_changedetection(self, watch, skip_when_checksum_same=True):
+        from .plugin_manager import pm
+
         if not watch:
             raise Exception("Watch no longer exists.")
 
@@ -198,6 +200,19 @@ class perform_site_check(difference_detection_processor):
             update_obj['restock']["in_stock"] = True if self.fetcher.instock_data == 'Possibly in stock' else False
             logger.debug(f"Watch UUID {watch.get('uuid')} restock check returned '{self.fetcher.instock_data}' from JS scraper.")
 
+        # Ask any "changedetectionio.restock_price_scraper" namespace plugins if they can add something
+        # (Should return an updated 'update_obj')
+        plugin_price_scraping = pm.hook.scrape_price_restock(watch=watch,
+                                                             html_content=self.fetcher.content,
+                                                             screenshot=self.fetcher.screenshot,
+                                                             update_obj=update_obj)
+        if plugin_price_scraping:
+            for plugin_result in plugin_price_scraping:
+                update_obj.update(plugin_result)
+                if plugin_result.get('restock'):
+                    update_obj['restock'].update(plugin_result.get('restock'))
+
+
         # What we store in the snapshot
         price = update_obj.get('restock').get('price') if update_obj.get('restock').get('price') else ""
         snapshot_content = f"In Stock: {update_obj.get('restock').get('in_stock')} - Price: {price}"
diff --git a/changedetectionio/templates/watch-overview.html b/changedetectionio/templates/watch-overview.html
index 736e19da..bdb48406 100644
--- a/changedetectionio/templates/watch-overview.html
+++ b/changedetectionio/templates/watch-overview.html
@@ -168,7 +168,7 @@
                         {% if watch.get('restock') and watch['restock']['price'] != None %}
                             {% if watch['restock']['price'] != None %}
                                 <span class="restock-label price" title="Price">
-                                {{ watch['restock']['price']|format_number_locale }} {{ watch['restock']['currency'] }}
+                                {{ watch['restock']['price']|format_number_locale }} {% if watch['restock']['currency'] %} {{ watch['restock']['currency'] }}{% endif %}
                                 </span>
                             {% endif %}
                         {% elif not watch.has_restock_info %}
diff --git a/requirements.txt b/requirements.txt
index 2e085cf6..cbaca1c9 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -92,3 +92,6 @@ babel
 
 # Needed for > 3.10, https://github.com/microsoft/playwright-python/issues/2096
 greenlet >= 3.0.3
+
+# Our own plugins
+changedetection.io-amazon-price-scraper>=0.21
\ No newline at end of file