lib init

tests/unit/test_parsers.py

@@ -0,0 +1,128 @@
+import base64
+
+from mail_intake.parsers.subject import normalize_subject
+from mail_intake.parsers.headers import parse_headers, extract_sender
+from mail_intake.parsers.body import extract_body
+
+
+def _b64(text: str) -> str:
+    return base64.b64encode(text.encode("utf-8")).decode("utf-8")
+
+
+# --------------------
+# Subject parsing
+# --------------------
+
+def test_normalize_subject_strips_common_prefixes():
+    assert normalize_subject("Re: Interview Update") == "Interview Update"
+    assert normalize_subject("Fwd: Re: Offer Letter") == "Offer Letter"
+    assert normalize_subject("FW: Re: FW: Status") == "Status"
+
+
+def test_normalize_subject_preserves_content_and_case():
+    subject = "Interview Update – Backend Role"
+    assert normalize_subject(subject) == subject
+
+
+def test_normalize_subject_empty_and_none_safe():
+    assert normalize_subject("") == ""
+
+
+# --------------------
+# Header parsing
+# --------------------
+
+def test_parse_headers_lowercases_keys():
+    raw_headers = [
+        {"name": "From", "value": "Alice <alice@example.com>"},
+        {"name": "Subject", "value": "Hello"},
+    ]
+
+    headers = parse_headers(raw_headers)
+
+    assert headers["from"] == "Alice <alice@example.com>"
+    assert headers["subject"] == "Hello"
+
+
+def test_parse_headers_ignores_invalid_entries():
+    raw_headers = [
+        {"name": "From", "value": "Bob <bob@example.com>"},
+        {"name": None, "value": "X"},
+        {"name": "X-Test", "value": None},
+    ]
+
+    headers = parse_headers(raw_headers)
+    assert "from" in headers
+    assert "x-test" not in headers
+
+
+def test_extract_sender_with_name_and_email():
+    headers = {"from": "Alice Smith <alice@example.com>"}
+    email, name = extract_sender(headers)
+
+    assert email == "alice@example.com"
+    assert name == "Alice Smith"
+
+
+def test_extract_sender_email_only():
+    headers = {"from": "bob@example.com"}
+    email, name = extract_sender(headers)
+
+    assert email == "bob@example.com"
+    assert name is None
+
+
+def test_extract_sender_missing_from():
+    email, name = extract_sender({})
+    assert email == ""
+    assert name is None
+
+
+# --------------------
+# Body parsing
+# --------------------
+
+def test_extract_body_prefers_text_plain():
+    payload = {
+        "parts": [
+            {
+                "mimeType": "text/html",
+                "body": {"data": _b64("<p>Hello <b>HTML</b></p>")},
+            },
+            {
+                "mimeType": "text/plain",
+                "body": {"data": _b64("Hello TEXT")},
+            },
+        ]
+    }
+
+    body = extract_body(payload)
+    assert body == "Hello TEXT"
+
+
+def test_extract_body_falls_back_to_html():
+    payload = {
+        "parts": [
+            {
+                "mimeType": "text/html",
+                "body": {"data": _b64("<p>Hello <b>World</b></p>")},
+            }
+        ]
+    }
+
+    body = extract_body(payload)
+    assert "Hello" in body
+    assert "World" in body
+
+
+def test_extract_body_single_part():
+    payload = {
+        "body": {"data": _b64("Single part body")}
+    }
+
+    body = extract_body(payload)
+    assert body == "Single part body"
+
+
+def test_extract_body_empty_payload():
+    assert extract_body({}) == ""