Reify Chunk

[paperdoorknob] / glowfic.py

diff --git a/glowfic.py b/glowfic.py
index 901246b10533f390ff565263fc9d5b1465b0af3c..41697867b2453e0af35175de866f02fe71409c54 100644 (file)
--- a/glowfic.py
+++ b/glowfic.py
@@ -5,6 +5,7 @@
 # Free Software Foundation, version 3.
 
 
+from dataclasses import dataclass
 import itertools
 
 from typing import Iterable
@@ -12,15 +13,26 @@ from typing import Iterable
 from bs4 import BeautifulSoup
 from bs4.element import Tag
 
+from images import ImageStore
+
+
+@dataclass(frozen=True)
+class Chunk:
+    icon: str | None
+    character: str | None
+    screen_name: str | None
+    author: str | None
+    content: Tag
+
 # We avoid the name "post" because the Glowfic community uses the term
 # inconsistently:
 #  * The Glowfic software sometimes uses "post" to refer to a whole thread
-#    (eg: in the URL), but more often uses "post" to refer to just the first
-#    chunk in a thread.  The non-first chunks are "replies".
+#    (in the URL), sometimes uses "post" to refer to chunks (in the CSS),
+#    but mostly uses "post" to refer to just the first chunk in a thread
+#    (in the HTML and UI).  The non-first chunks are "replies".
 #  * Readers and this software don't need to distinguish first-chunks and
 #    non-first-chunks.
-#  * Humans in the community tend to use "posts" to mean "chunks" ("replies"
-#    in the Glowfic software's lexicon).
+#  * Humans in the community tend to use "posts" to mean chunks.
 
 
 def chunkDOMs(html: BeautifulSoup) -> Iterable[Tag]:
@@ -37,3 +49,31 @@ def chunkDOMs(html: BeautifulSoup) -> Iterable[Tag]:
         return rs
 
     return itertools.chain([text()], the_replies())
+
+
+def makeChunk(chunk_dom: Tag, image_store: ImageStore) -> Chunk:
+
+    def getIcon() -> str | None:
+        icon_div = chunk_dom.find_next('div', class_='post-icon')
+        if icon_div is None:
+            return None
+        icon_img = icon_div.find_next('img')
+        if icon_img is None:
+            return None
+        assert isinstance(icon_img, Tag)
+        return image_store.get_image(icon_img.attrs['src'])
+
+    def getTextByClass(css_class: str) -> str | None:
+        div = chunk_dom.find_next('div', class_=css_class)
+        if div is None:
+            return None
+        return div.text.strip()
+
+    content = chunk_dom.find_next('div', class_='post-content')
+    assert isinstance(content, Tag)
+
+    return Chunk(getIcon(),
+                 getTextByClass('post-character'),
+                 getTextByClass('post-screenname'),
+                 getTextByClass('post-author'),
+                 content)