From 0fd5ec64586ced46b6c91ad07fe33c5886e80ac1 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=C3=89tienne=20Reuh=20Fildadut?= <fildadut@reuh.eu>
Date: Sat, 11 Oct 2025 17:08:22 +0200
Subject: [PATCH] feat: add hash to HTML filename in case of conflict

---
 README.md           |  1 -
 src/feather/data.py | 42 +++++++++++++++++++++++++++---------------
 2 files changed, 27 insertions(+), 16 deletions(-)

diff --git a/README.md b/README.md
index 6ef2c7f..a66d13d 100644
--- a/README.md
+++ b/README.md
@@ -68,7 +68,6 @@ You need Python 3.12 or newer. Then pip it up, as the kids say.
 - [ ] Write documentation
 - [ ] Use inotify for real-time article mark-as-read action
 - [ ] Share the fun somewhere
-- [ ] Actually think about the issues created by the duplicate warning
 - [ ] Get article attachments
 - [ ] Test with FreshRSS
 
diff --git a/src/feather/data.py b/src/feather/data.py
index 572ac9d..7c01449 100644
--- a/src/feather/data.py
+++ b/src/feather/data.py
@@ -87,7 +87,8 @@ class Article(ABC):
     # no default value, computed by compute_fields
     published_formatted: str  # article publication time (text)
     updated_formatted: str  # article publication time (text)
-    html_path: str  # html path, relative to the html_root directory
+    # no default value, computed on save_html
+    html_path: str = None  # html path, relative to the html_root directory (None will force it to be recomputed on next save_html)
     # with default value
     unread: bool = True  # if the article is unread
     title: str = ""  # article title
@@ -104,7 +105,7 @@ class Article(ABC):
     comments_url: str = ""  # article comments URL
     language: str = ""  # article language
     image_url: str = ""  # article main image
-    
+
     def _get_html_path(self):
         config = self.config
 
@@ -123,22 +124,17 @@ class Article(ABC):
             config,
             config.article_filename_template.render(self._get_template_dict()),
         )
+        html_path = category_directory / html_name
+        # Add hash if filename already exists
+        if html_path.exists():
+            id_hash = sha256(str(self.id).encode("utf-8")).hexdigest()[:8]
+            html_path = html_path.parent / sanitize_filename(
+                config, html_path.name, insert_before_suffix=f".{id_hash}"
+            )
 
         return html_path.relative_to(config.html_root)
 
-    def compute_fields(self):
-        config = self.config
-        self.updated_formatted = format_datetime(config, self.updated)
-        self.published_formatted = format_datetime(config, self.published)
-        self.json_path = (
-            config.json_root
-            / f"{sha256(str(self.id).encode('utf-8')).hexdigest()}.json"
-        )
-        self.html_path = str(
-            self.get_html_path().relative_to(config.html_root)
-        )  # TODO: do this dynamically on write, handle overwrite conflict at the same time
-
-    def get_template_dict(self) -> dict:
+    def _get_template_dict(self) -> dict:
         template_fields = (
             "id",
             "unread",
@@ -163,6 +159,16 @@ class Article(ABC):
         d["category"] = self.category.asdict()
         return d
 
+    def compute_fields(self):
+        config = self.config
+        self.updated_formatted = format_datetime(config, self.updated)
+        self.published_formatted = format_datetime(config, self.published)
+        self.json_path = (
+            config.json_root
+            / f"{sha256(str(self.id).encode('utf-8')).hexdigest()}.json"
+        )
+        self.html_path = None
+
     def write_json(self):
         """Write the JSON file associated with this article. Error if it already exists."""
         stored_fields = (
@@ -208,8 +214,14 @@ class Article(ABC):
             return html_path.exists()
 
     def write_html(self):
+        """Write the HTML file associated with this article. Error if it already exists.
+        Note: this may recompute html_path, which is saved into the JSON - so make sure to save the JSON file _after_ the HTML file."""
         # Write HTML file for a JSON object
         config = self.config
+
+        if self.html_path is None:
+            self.html_path = str(self._get_html_path())
+
         html_path = config.html_root / self.html_path
         if html_path.exists():
             raise Exception(