t1.1: scrape one forum via ViewComments.cfm POST pagination

Spider fetches ViewComments.cfm?GdocForumID=N with vPerPage=500,
generates all page requests from page-1 metadata, and parses
each div.Cbox for comment_id, author, date, title, text, reg_title,
reg_desc. Handles span-wrapped comment text. Fixes UTF-8/windows-1251
meta-tag encoding mismatch. 9083 items, 15 empty-text (0.17%).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

.gitignore

@@ -22,5 +22,9 @@ env/
 archive/
 
 
+# --- scrapy ---
+.scrapy/
+output/
+
 # --- misc ---
 .DS_Store

docs/tasks.org

@@ -1,5 +1,8 @@
-* [ ] t1.1: scrape one forum (1)
+* [X] t1.1: scrape one forum (1)
 Use https://www.townhall.virginia.gov/L/comments.cfm?GDocForumID=452 as the first forum. Scraper should be run manually at this step.
+ViewComments (townhall.virginia.gov/L/ViewComments.cfm?CommentID=#) appears to be raw list of all comments on forum - could be useful later for whole-scrape
+Append forum id to viewall per forum (townhall.virginia.gov/L/ViewComments.cfm?GdocForumID=452)
+Comments are hydrated in backend via js-cued button (AJAX?)
 ** acceptance criteria
 1. run manual scraper
    1. store proposal title and description
@@ -10,9 +13,9 @@ Use https://www.townhall.virginia.gov/L/comments.cfm?GDocForumID=452 as the firs
 ** notes
 
 ** evidence
-- commit: 
+- commit: (see below)
-- tests: 
+- tests: 7 passing (pytest tests/)
-- datetime: 
+- datetime: 2026-05-05 12:26
 
 * [ ] t1.2: initial analysis pipeline
 Write a simple pipeline for both - prefer non-concurrent/async from scraping run. Should be run manually, separate from scraper. You may use scrapy, but are not required to.

scraper/items.py

@@ -1,12 +1,17 @@
-# Define here the models for your scraped items
-#
-# See documentation in:
-# https://docs.scrapy.org/en/latest/topics/items.html
-
 import scrapy
 
 
-class ScraperItem(scrapy.Item):
+class CommentItem(scrapy.Item):
-    # define the fields for your item here like:
+    # Forum / regulation context
-    # name = scrapy.Field()
+    forum_id   = scrapy.Field()
-    pass
+    reg_title  = scrapy.Field()
+    reg_desc   = scrapy.Field()
+
+    # Comment metadata
+    comment_id = scrapy.Field()
+    author     = scrapy.Field()
+    date       = scrapy.Field()
+    title      = scrapy.Field()
+
+    # Comment content
+    text       = scrapy.Field()

scraper/settings.py

@@ -15,8 +15,7 @@ NEWSPIDER_MODULE = "scraper.spiders"
 ADDONS = {}
 
 
-# Crawl responsibly by identifying yourself (and your website) on the user-agent
+USER_AGENT = "vath-research-scraper/1.0 (public comment analysis; contact: research)"
-#USER_AGENT = "scraper (+http://www.yourdomain.com)"
 
 # Obey robots.txt rules
 ROBOTSTXT_OBEY = True
@@ -75,13 +74,24 @@ DOWNLOAD_DELAY = 1
 # Enable showing throttling stats for every response received:
 #AUTOTHROTTLE_DEBUG = False
 
-# Enable and configure HTTP caching (disabled by default)
+# HTTP cache — enabled during development to avoid re-hitting the server on test runs.
-# See https://docs.scrapy.org/en/latest/topics/downloader-middleware.html#httpcache-middleware-settings
+# Disable (or delete httpcache/) before a production run.
-#HTTPCACHE_ENABLED = True
+HTTPCACHE_ENABLED = True
-#HTTPCACHE_EXPIRATION_SECS = 0
+HTTPCACHE_EXPIRATION_SECS = 86400  # 24 h
-#HTTPCACHE_DIR = "httpcache"
+HTTPCACHE_DIR = "httpcache"
-#HTTPCACHE_IGNORE_HTTP_CODES = []
+
-#HTTPCACHE_STORAGE = "scrapy.extensions.httpcache.FilesystemCacheStorage"
+# Output
+FEEDS = {
+    "output/%(name)s_%(time)s.jsonl": {
+        "format": "jsonlines",
+        "encoding": "utf-8",
+        "overwrite": False,
+    }
+}
+
+# The site declares windows-1251 in a meta tag but sends valid UTF-8 bytes.
+# Force UTF-8 to prevent lxml from re-decoding via the meta charset.
+DEFAULT_RESPONSE_ENCODING = "utf-8"
 
 # Set settings whose default value is deprecated to a future-proof value
 FEED_EXPORT_ENCODING = "utf-8"

scraper/spiders/forum.py

@@ -0,0 +1,116 @@
+import math
+import re
+from urllib.parse import urlencode
+
+import scrapy
+
+from scraper.items import CommentItem
+
+_BASE = "https://www.townhall.virginia.gov/L/ViewComments.cfm"
+_NBSP = "\xa0"
+_REPLACEMENT_CHAR = "<EFBFBD>"
+
+
+def _view_url(forum_id):
+    return f"{_BASE}?GdocForumID={forum_id}"
+
+
+class ForumSpider(scrapy.Spider):
+    name = "forum"
+    allowed_domains = ["townhall.virginia.gov"]
+
+    # Override at runtime: scrapy crawl forum -a forum_id=452
+    forum_id = "452"
+    per_page = 500
+
+    async def start(self):
+        yield scrapy.FormRequest(
+            _view_url(self.forum_id),
+            formdata={"vPage": "1", "vPerPage": str(self.per_page), "sub1": "go"},
+            callback=self.parse_comments,
+            meta={"is_first": True},
+        )
+
+    # ------------------------------------------------------------------
+    def parse_comments(self, response):
+        if response.meta.get("is_first"):
+            reg_title, reg_desc = self._reg_context(response)
+            last_page = self._last_page(response)
+            for page in range(2, last_page + 1):
+                yield scrapy.FormRequest(
+                    _view_url(self.forum_id),
+                    formdata={"vPage": str(page), "vPerPage": str(self.per_page), "sub1": "go"},
+                    callback=self.parse_comments,
+                    meta={"reg_title": reg_title, "reg_desc": reg_desc},
+                )
+        else:
+            reg_title = response.meta["reg_title"]
+            reg_desc = response.meta["reg_desc"]
+
+        for box in response.css("div.Cbox"):
+            yield self._parse_box(box, reg_title, reg_desc)
+
+    # ------------------------------------------------------------------
+    def _parse_box(self, box, reg_title, reg_desc):
+        cbox_id = box.attrib.get("id", "")
+        comment_id = cbox_id[len("cbox"):] if cbox_id.startswith("cbox") else ""
+
+        date = (
+            box.css("div[style*='float: right'] div::text").get("")
+            .replace(_NBSP, " ").strip()
+        )
+
+        author = (
+            box.xpath('.//strong[contains(text(),"Commenter:")]/following-sibling::text()[1]')
+            .get("").strip()
+        )
+
+        # Second <strong> in the commenter block is the comment title
+        strongs = box.css("div > strong::text").getall()
+        title = strongs[-1].strip() if len(strongs) > 1 else ""
+
+        paragraphs = box.css(".divComment *::text, .divComment::text").getall()
+        text = " ".join(p.strip() for p in paragraphs if p.strip())
+        text = text.replace(_NBSP, " ").replace(_REPLACEMENT_CHAR, "'").strip()
+
+        return CommentItem(
+            forum_id=self.forum_id,
+            reg_title=reg_title,
+            reg_desc=reg_desc,
+            comment_id=comment_id,
+            author=author,
+            date=date,
+            title=title,
+            text=text,
+        )
+
+    # ------------------------------------------------------------------
+    def _reg_context(self, response):
+        # Page shows: <strong>Guidance Document Change:</strong> description text...
+        label_node = response.xpath('//strong[contains(text(),"Change:")]')
+        label_text = label_node.css("::text").get("").strip()
+
+        # Collect all sibling text nodes following the label
+        siblings = label_node.xpath("following-sibling::text()").getall()
+        raw = " ".join(t.strip() for t in siblings if t.strip())
+        raw = raw.replace(_NBSP, " ").replace(_REPLACEMENT_CHAR, "'").strip()
+
+        # reg_desc is the full description text
+        reg_desc = raw
+
+        # reg_title: text up to the first "was " clause or first 200 chars
+        m = re.match(r"^(.+?)\s+(?:was |has |guidance document)", raw, re.IGNORECASE)
+        reg_title = m.group(1).strip() if m else raw[:200]
+
+        return reg_title, reg_desc
+
+    def _last_page(self, response):
+        hrefs = response.xpath(
+            '//form[@name="page"]//a[contains(@href,"vpage.value=")]/@href'
+        ).getall()
+        pages = [
+            int(m.group(1))
+            for h in hrefs
+            if (m := re.search(r"vpage\.value=(\d+)", h))
+        ]
+        return max(pages) if pages else 1

tests/test_forum_spider.py

@@ -0,0 +1,227 @@
+"""Tests for ForumSpider parsing logic using fake HTML responses."""
+
+from scrapy.http import HtmlResponse, Request
+
+from scraper.spiders.forum import ForumSpider
+
+
+def fake_response(url, body, meta=None):
+    req = Request(url=url, meta=meta or {})
+    return HtmlResponse(url=url, body=body.encode("utf-8"), request=req)
+
+
+# ---------------------------------------------------------------------------
+# Minimal page HTML fragments
+
+PAGE1_HTML = """
+<html><body>
+  <strong>Guidance Document Change:</strong> The Model Policies for the Treatment of Transgender Students
+  was developed in response to House Bill 145 and Senate Bill 161.
+
+  <div style="font-family: verdana;">
+    <form name="page" id="page" action="ViewComments.cfm?GdocForumID=452" method="post">
+      <input name="vPage" id="vpage" type="input" value="1">
+      <input name="vPerPage" id="vPerPage" type="input" value="500">
+      <a href="javascript:document.page.vpage.value=3;document.page.submit();">3</a>
+      <a href="javascript:document.page.vpage.value=2;document.page.submit();">Next</a>
+      <input type="submit" name="sub1" value="go">
+    </form>
+  </div>
+
+  <div id="cbox101" class="Cbox">
+    <div style="float: right; text-align: right;">
+      <div style="background-color: white; border: 1px solid #cccccc; padding: 4px">1/4/21&nbsp;&nbsp;9:15 am</div>
+    </div>
+    <div>
+      <strong>Commenter:</strong>
+      Alice Example
+      <br><br>
+      <strong>I strongly support this</strong>
+    </div>
+    <div style="clear: right">&nbsp;</div>
+    <div class="divComment">
+      <p>This is a great policy for students.</p>
+      <p>All schools should follow it.</p>
+    </div>
+    <div style="float: left; font-size: 90%;">
+      CommentID: <a href="ViewComments.cfm?commentid=101">101</a>
+    </div>
+  </div>
+
+  <div id="cbox102" class="Cbox">
+    <div style="float: right; text-align: right;">
+      <div style="background-color: white; border: 1px solid #cccccc; padding: 4px">1/5/21&nbsp;&nbsp;10:00 am</div>
+    </div>
+    <div>
+      <strong>Commenter:</strong>
+      Bob Sample
+      <br><br>
+      <strong>Opposed</strong>
+    </div>
+    <div style="clear: right">&nbsp;</div>
+    <div class="divComment">
+      <p>I do not support this guidance.</p>
+    </div>
+    <div style="float: left; font-size: 90%;">
+      CommentID: <a href="ViewComments.cfm?commentid=102">102</a>
+    </div>
+  </div>
+</body></html>
+"""
+
+PAGE2_HTML = """
+<html><body>
+  <div id="cbox201" class="Cbox">
+    <div style="float: right; text-align: right;">
+      <div style="background-color: white; border: 1px solid #cccccc; padding: 4px">1/6/21&nbsp;&nbsp;11:00 am</div>
+    </div>
+    <div>
+      <strong>Commenter:</strong>
+      Carol T
+      <br><br>
+      <strong>Support</strong>
+    </div>
+    <div style="clear: right">&nbsp;</div>
+    <div class="divComment">
+      <p>This policy is long overdue.</p>
+    </div>
+  </div>
+</body></html>
+"""
+
+
+def make_spider():
+    return ForumSpider()
+
+
+# ---------------------------------------------------------------------------
+
+def test_page1_generates_remaining_page_requests():
+    spider = make_spider()
+    response = fake_response(
+        "https://www.townhall.virginia.gov/L/ViewComments.cfm?GdocForumID=452",
+        PAGE1_HTML,
+        meta={"is_first": True},
+    )
+    results = list(spider.parse_comments(response))
+    form_reqs = [r for r in results if isinstance(r, scrapy.FormRequest)]
+    # Pages 2 and 3 should be requested (last page link = 3)
+    assert len(form_reqs) == 2
+    pages = sorted(r.body.decode() for r in form_reqs)
+    assert "vPage=2" in pages[0]
+    assert "vPage=3" in pages[1]
+
+
+def test_page1_yields_items():
+    spider = make_spider()
+    response = fake_response(
+        "https://www.townhall.virginia.gov/L/ViewComments.cfm?GdocForumID=452",
+        PAGE1_HTML,
+        meta={"is_first": True},
+    )
+    results = list(spider.parse_comments(response))
+    from scraper.items import CommentItem
+    items = [r for r in results if isinstance(r, CommentItem)]
+    assert len(items) == 2
+
+
+def test_comment_fields_parsed_correctly():
+    spider = make_spider()
+    response = fake_response(
+        "https://www.townhall.virginia.gov/L/ViewComments.cfm?GdocForumID=452",
+        PAGE1_HTML,
+        meta={"is_first": True},
+    )
+    from scraper.items import CommentItem
+    items = [r for r in spider.parse_comments(response) if isinstance(r, CommentItem)]
+    item = items[0]
+    assert item["comment_id"] == "101"
+    assert item["author"] == "Alice Example"
+    assert item["title"] == "I strongly support this"
+    assert "great policy" in item["text"]
+    assert "All schools" in item["text"]  # multi-paragraph joined
+    assert "1/4/21" in item["date"]
+
+
+def test_reg_context_extracted():
+    spider = make_spider()
+    response = fake_response(
+        "https://www.townhall.virginia.gov/L/ViewComments.cfm?GdocForumID=452",
+        PAGE1_HTML,
+        meta={"is_first": True},
+    )
+    from scraper.items import CommentItem
+    items = [r for r in spider.parse_comments(response) if isinstance(r, CommentItem)]
+    item = items[0]
+    assert "Transgender Students" in item["reg_title"]
+    assert "House Bill 145" in item["reg_desc"]
+
+
+def test_subsequent_page_uses_meta_reg_context():
+    spider = make_spider()
+    response = fake_response(
+        "https://www.townhall.virginia.gov/L/ViewComments.cfm?GdocForumID=452",
+        PAGE2_HTML,
+        meta={"reg_title": "Test Reg", "reg_desc": "Full description text"},
+    )
+    from scraper.items import CommentItem
+    items = [r for r in spider.parse_comments(response) if isinstance(r, CommentItem)]
+    assert len(items) == 1
+    assert items[0]["reg_title"] == "Test Reg"
+    assert items[0]["author"] == "Carol T"
+
+
+def test_last_page_detection():
+    spider = make_spider()
+    response = fake_response(
+        "https://www.townhall.virginia.gov/L/ViewComments.cfm?GdocForumID=452",
+        PAGE1_HTML,
+        meta={"is_first": True},
+    )
+    assert spider._last_page(response) == 3
+
+
+import scrapy
+
+SPAN_WRAPPED_HTML = """
+<html><body>
+  <strong>Guidance Document Change:</strong> Some regulation was developed.
+
+  <form name="page" id="page" action="ViewComments.cfm?GdocForumID=452" method="post">
+    <input name="vPage" value="1"><input name="vPerPage" value="500">
+    <a href="javascript:document.page.vpage.value=1;document.page.submit();">1</a>
+    <input type="submit" name="sub1" value="go">
+  </form>
+
+  <div id="cbox301" class="Cbox">
+    <div style="float: right; text-align: right;">
+      <div style="background-color: white; border: 1px solid #cccccc; padding: 4px">2/1/21&nbsp;&nbsp;8:00 am</div>
+    </div>
+    <div>
+      <strong>Commenter:</strong>
+      Dan Span
+      <br><br>
+      <strong>Opposed</strong>
+    </div>
+    <div style="clear: right">&nbsp;</div>
+    <div class="divComment">
+      <!DOCTYPE html><html><head></head><body>
+      <p style="margin: 0in;"><span style="font-size: 10.5pt;">Text inside a span element.</span></p>
+      </body></html>
+    </div>
+  </div>
+</body></html>
+"""
+
+
+def test_span_wrapped_text_is_extracted():
+    spider = make_spider()
+    response = fake_response(
+        "https://www.townhall.virginia.gov/L/ViewComments.cfm?GdocForumID=452",
+        SPAN_WRAPPED_HTML,
+        meta={"is_first": True},
+    )
+    from scraper.items import CommentItem
+    items = [r for r in spider.parse_comments(response) if isinstance(r, CommentItem)]
+    assert len(items) == 1
+    assert "Text inside a span element" in items[0]["text"]