chore: polish for public release

- Real contact in SEC User-Agent (required by EDGAR) - Fix Dockerfile cache dir (was missing, build would fail) - Pin flask>=3.0,<4 - Harden clean_ereader to strip single-quoted attributes - Remove unused templates (archive.html, more.html) - Add pytest suite (10 tests) and CI workflow
2026-08-20 21:11:29 +00:00 · 2026-08-20 21:11:29 +00:00 · fa7e3eb545
commit fa7e3eb545
parent 873a6b6717
9 changed files with 149 additions and 66 deletions
--- a/.github/workflows/test.yml
+++ b/.github/workflows/test.yml
@ -0,0 +1,25 @@
+name: Test
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+
+      - name: Install dependencies
+        run: |
+          pip install -r requirements.txt
+          pip install pytest
+
+      - name: Run tests
+        run: python -m pytest tests/ -v
--- a/.gitignore
+++ b/.gitignore
@ -1,5 +1,6 @@
 __pycache__/
 *.pyc
-cache/
 .env
 *.log
+cache/*
+!cache/.gitkeep
--- a/README.md
+++ b/README.md
@ -40,6 +40,13 @@ docker run -p 5001:5001 10k-viewer

 Enter a stock ticker (e.g. `AAPL`, `MSFT`, `GOOGL`) to fetch and browse the latest 10-K filing data. Supports any company with a valid CIK identifier.

+## Tests
+
+```bash
+pip install -r requirements.txt pytest
+python -m pytest tests/
+```
+
 ## License

 MIT
--- a/app.py
+++ b/app.py
@ -9,8 +9,9 @@ app = Flask(__name__)
 logging.basicConfig(level=logging.INFO)
 log = logging.getLogger("10k")

-UA = "10k-viewer/1.0 (test@test.com)"
+UA = "10k-viewer/1.0 (Jarian Cottingham jarianc@proton.me)"
 CACHE_DIR = Path(__file__).parent / "cache"
+CACHE_DIR.mkdir(parents=True, exist_ok=True)
 MAX_RETRIES = 1
 PAGE_SIZE = 10

@ -155,9 +156,9 @@ def clean_ereader(html):
    t = html.decode("utf-8", errors="replace")
    t = re.sub(r'<script[^>]*>.*?</script>', '', t, flags=re.DOTALL|re.I)
    t = re.sub(r'<style[^>]*>.*?</style>', '', t, flags=re.DOTALL|re.I)
-    t = re.sub(r'\s+style="[^"]*"', '', t, flags=re.I)
-    t = re.sub(r'\s+on\w+="[^"]*"', '', t, flags=re.I)
-    t = re.sub(r'\s+(class|id)="[^"]*"', '', t, flags=re.I)
+    t = re.sub(r"""\s+style=["'][^"']*["']""", '', t, flags=re.I)
+    t = re.sub(r"""\s+on\w+=["'][^"']*["']""", '', t, flags=re.I)
+    t = re.sub(r"""\s+(class|id)=["'][^"']*["']""", '', t, flags=re.I)
    t = re.sub(r'<!--.*?-->', '', t, flags=re.DOTALL)
    s = '<style>body{font-family:Georgia,serif!important;font-size:18px!important;line-height:1.6!important;max-width:700px;margin:0 auto;padding:20px;color:#000;background:#fff}table{border-collapse:collapse;width:100%;margin:10px 0;font-size:14px}th,td{border:1px solid #000;padding:4px 8px;text-align:left}th{font-weight:bold;background:#eee}a{color:#000}h1,h2,h3,h4,h5,h6{font-family:Georgia,serif!important;margin:16px 0 8px}p{margin:8px 0}</style>'
    if '<head>' in t.lower():
@ -229,5 +230,4 @@ def view(ticker, filename):


 if __name__ == "__main__":
-    CACHE_DIR.mkdir(parents=True, exist_ok=True)
    app.run(host="0.0.0.0", port=5001)
--- a/cache/.gitkeep
+++ b/cache/.gitkeep
--- a/requirements.txt
+++ b/requirements.txt
@ -1 +1 @@
-flask
+flask>=3.0,<4
--- a/templates/archive.html
+++ b/templates/archive.html
@ -1,26 +0,0 @@
-<!DOCTYPE html>
-<html>
-<head>
-<meta charset="utf-8">
-<title>{{ ticker }} - Archive Status</title>
-<style>
-body { font-family: Georgia, serif; font-size: 18px; max-width: 600px; margin: 0 auto; padding: 20px; }
-h1 { font-size: 1.4em; }
-.back { font-size: 0.85em; }
-.stats { margin: 20px 0; padding: 15px; background: #f5f5f5; }
-.ok { color: #080; }
-.err { color: #c00; }
-</style>
-</head>
-<body>
-<p class="back"><a href="/{{ ticker }}">&larr; Back to {{ ticker }}</a> | <a href="/">All companies</a></p>
-<h1>Archive Status</h1>
-<div class="stats">
-<p class="ok">{{ archived }} of {{ total }} filings archived</p>
-{% if failed > 0 %}
-<p class="err">{{ failed }} failed (SEC may have rate-limited)</p>
-{% endif %}
-</div>
-<p>Files saved to local cache for offline reading.</p>
-</body>
-</html>
--- a/templates/more.html
+++ b/templates/more.html
@ -1,33 +0,0 @@
-<!DOCTYPE html>
-<html>
-<head>
-<meta charset="utf-8">
-<title>{{ ticker }} - More Filings</title>
-<style>
-body { font-family: Georgia, serif; font-size: 18px; max-width: 600px; margin: 0 auto; padding: 20px; }
-h1 { font-size: 1.4em; }
-ul { list-style: none; padding: 0; }
-li { padding: 8px 0; border-bottom: 1px solid #ccc; }
-a { color: #000; }
-small { color: #555; }
-.back { font-size: 0.85em; }
-.form-tag { display: inline-block; padding: 1px 5px; font-size: 0.75em; margin-left: 5px; }
-.form-10k { background: #000; color: #fff; }
-.form-10q { background: #666; color: #fff; }
-</style>
-</head>
-<body>
-<p class="back"><a href="/{{ ticker }}?form={{ form_filter }}">&larr; Back to filings</a> | <a href="/">All companies</a></p>
-<h1>{{ ticker }} - More Filings</h1>
-<ul>
-{% for f in filings %}
-<li>
-<a href="{{ f.sec_url }}" target="_blank">{{ f.filing_date }}</a>
-<span class="form-tag {% if f.form == '10-K' %}form-10k{% else %}form-10q{% endif %}">{{ f.form }}</span>
-</li>
-{% else %}
-<li>No more filings.</li>
-{% endfor %}
-</ul>
-</body>
-</html>
--- a/tests/test_app.py
+++ b/tests/test_app.py
@ -0,0 +1,109 @@
+"""Tests for the 10-K viewer Flask app."""
+
+import sys
+from pathlib import Path
+
+import pytest
+
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
+
+import app as viewer  # noqa: E402
+
+
+SAMPLE_SUBS = {
+    "filings": {
+        "recent": {
+            "form": ["10-K", "8-K", "10-Q", "10-K"],
+            "filingDate": ["2024-11-01", "2024-10-15", "2024-08-01", "2023-11-02"],
+            "primaryDocument": [
+                "aapl-20240928.htm",
+                "aapl-20241015.htm",
+                "aapl-20240628.htm",
+                "aapl-20230930.htm",
+            ],
+            "accessionNumber": [
+                "0000320193-24-000123",
+                "0000320193-24-000100",
+                "0000320193-24-000090",
+                "0000320193-23-000050",
+            ],
+        }
+    }
+}
+
+
+@pytest.fixture
+def client():
+    viewer._filings_cache.clear()
+    viewer._tickers = None
+    viewer.app.config["TESTING"] = True
+    with viewer.app.test_client() as c:
+        yield c
+
+
+def test_parse_filings_filters_forms():
+    r = viewer.parse_filings(SAMPLE_SUBS, "0000320193", {"10-K"})
+    assert len(r) == 2
+    assert all(f["form"] == "10-K" for f in r)
+
+
+def test_parse_filings_builds_url_and_file():
+    r = viewer.parse_filings(SAMPLE_SUBS, "0000320193", {"10-K", "10-Q"})
+    first = r[0]
+    assert first["file"] == "2024-11-01_000032019324000123.html"
+    assert (
+        first["url"]
+        == "https://www.sec.gov/Archives/edgar/data/0000320193/000032019324000123/aapl-20240928.htm"
+    )
+
+
+def test_parse_filings_no_matching_forms():
+    assert viewer.parse_filings(SAMPLE_SUBS, "0000320193", {"20-F"}) == []
+
+
+def test_clean_ereader_strips_scripts_and_handlers():
+    html = b"<html><head><script>evil()</script></head><body><div onclick='x()' class='c' id='i'>hi</div></body></html>"
+    out = viewer.clean_ereader(html).decode("utf-8")
+    assert "<script>" not in out
+    assert "onclick" not in out
+    assert "class=" not in out
+    assert "id=" not in out
+    assert "Georgia" in out
+
+
+def test_get_tickers_fallback():
+    viewer._tickers = None
+    t = viewer.get_tickers()
+    assert t["AAPL"]["cik"] == "0000320193"
+
+
+def test_index_default_lists_companies(client):
+    resp = client.get("/")
+    assert resp.status_code == 200
+    assert "AAPL" in resp.get_data(as_text=True)
+
+
+def test_index_search_filters(client):
+    resp = client.get("/?q=apple")
+    assert resp.status_code == 200
+    body = resp.get_data(as_text=True)
+    assert "AAPL" in body
+    assert "MSFT" not in body
+
+
+def test_company_renders_filings(client, monkeypatch):
+    monkeypatch.setattr(viewer, "get_filings", lambda cik, forms: [
+        {"form": "10-K", "date": "2024-11-01", "file": "f1.html", "url": "https://x/f1.html"},
+    ])
+    resp = client.get("/AAPL")
+    assert resp.status_code == 200
+    assert "Apple Inc." in resp.get_data(as_text=True)
+    assert "f1.html" in resp.get_data(as_text=True)
+
+
+def test_company_unknown_ticker_404(client):
+    assert client.get("/NOPE").status_code == 404
+
+
+def test_view_unknown_ticker_404(client):
+    assert client.get("/NOPE/view/f1.html").status_code == 404