#!/usr/bin/env python3
"""Tests for the W2 safe NL-query runner (the model-free core).

Boots the REAL schema (server.init_db against a temp DB — exact columns + all migrations),
inserts synthetic fundraising/email/reminder/pipeline data, and exercises every intent plus
the trust-boundary behaviour:
  - each intent returns the right rows over the real schema;
  - SOFT-DELETE is respected on both recency legs (a tombstoned communication and a tombstoned
    email sighting never count), on reminders, and on opportunities; graveyard investors are
    excluded from "live" intents;
  - the validator rejects bad/unknown/unexpected slots WITHOUT crashing (the `?limit=abc` class);
  - LIKE wildcards in a free-text slot are escaped (a city of "%" does NOT return everything);
  - limits clamp to their caps; the audit hook fires with the intent + row count.
Synthetic data only — no real LP substance, no network, no model.

Run: cd backend && python3 nl_query/test_nl_query.py
"""
import os
import sys
import tempfile
from datetime import datetime, timedelta

_DATA = tempfile.mkdtemp()
os.environ["CRM_DATA_DIR"] = _DATA
os.environ["CRM_DB_PATH"] = os.path.join(_DATA, "crm.db")
os.environ["CRM_GMAIL_INTEGRATION_ENABLED"] = "1"

sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))  # backend/
import server  # noqa: E402
import nl_query  # noqa: E402

FAILS = []


def check(cond, msg):
    print(("  PASS " if cond else "  FAIL ") + msg)
    if not cond:
        FAILS.append(msg)


def _ago(days):
    return (datetime.utcnow() - timedelta(days=days)).isoformat() + "Z"


TODAY = datetime.utcnow().date()


def seed(conn):
    c = conn.execute
    # users + mailboxes
    c("INSERT INTO users (id, username, email, password_hash, full_name, role) VALUES "
      "('u_grant','grant','grant@ten31.xyz','x','Grant Smith','admin'),"
      "('u_jon','jonathan','jon@ten31.xyz','x','Jonathan Lee','member')")
    c("INSERT INTO email_accounts (id, user_id, email_address, auth_method) VALUES "
      "('a_grant','u_grant','grant@ten31.xyz','dwd'),"
      "('a_jon','u_jon','jon@ten31.xyz','dwd')")
    # funds
    c("INSERT INTO fundraising_funds (id, column_id, fund_name, display_order) VALUES "
      "('f1','c_f1','Fund I',1),('f2','c_f2','Fund II',2)")

    # investors (graveyard flag is the live/retired axis; no deleted_at on this table)
    def inv(iid, name, lead, total, grave=0):
        c("INSERT INTO fundraising_investors (id, investor_name, lead, graveyard, "
          "source_row_id, total_invested) VALUES (?,?,?,?,?,?)",
          (iid, name, lead, grave, iid, total))
    inv("i_acme", "Acme Capital", "Jonathan Lee", 5_000_000)
    inv("i_beta", "Beta Partners", "Grant Smith", 2_000_000)
    inv("i_cold", "Cold Co", "Grant Smith", 0)                  # never contacted
    inv("i_delta", "Delta LP", "Grant Smith", 1_000_000)        # only a (comms) signal
    inv("i_ghost", "Graveyard Ghost", "Grant Smith", 9_999_999, grave=1)

    # contacts (grid pills) + classic contact rows for the comms leg
    c("INSERT INTO fundraising_contacts (id, investor_id, full_name, email, title, city, "
      "contact_id, sort_order) VALUES "
      "('fc_a','i_acme','Alice Acme','alice@acme.com','GP','Austin','cc_alice',0),"
      "('fc_b','i_beta','Bob Beta','bob@beta.com','LP','Denver',NULL,0),"
      "('fc_d','i_delta','Dana Delta','dana@delta.com','CFO','Miami','cc_dana',0)")
    c("INSERT INTO contacts (id, first_name, last_name, email) VALUES "
      "('cc_alice','Alice','Acme','alice@acme.com'),"
      "('cc_dana','Dana','Delta','dana@delta.com')")

    # commitments — Acme across two funds (3M + 2M = 5M); Beta one fund
    c("INSERT INTO fundraising_commitments (id, investor_id, fund_id, amount) VALUES "
      "('cm1','i_acme','f1',3_000_000),('cm2','i_acme','f2',2_000_000),"
      "('cm3','i_beta','f1',2_000_000)")

    # emails: matched + a per-mailbox sighting. is_sent + from_email decide direction.
    def email(eid, frm, frm_name, days, inv_id, account, is_sent, deleted=False):
        c("INSERT INTO emails (id, rfc_message_id, from_email, from_name, sent_at, subject, "
          "is_matched, match_status) VALUES (?,?,?,?,?,?,1,'matched')",
          (eid, "rfc_" + eid, frm, frm_name, _ago(days), "Re: " + eid))
        c("INSERT INTO email_account_messages (id, email_id, account_id, gmail_message_id, "
          "gmail_thread_id, is_sent, deleted_at) VALUES (?,?,?,?,?,?,?)",
          ("eam_" + eid, eid, account, "g_" + eid, "t_" + eid, is_sent,
           _ago(days) if deleted else None))
        c("INSERT INTO email_investor_links (id, email_id, fundraising_investor_id, "
          "matched_address, match_kind) VALUES (?,?,?,?, 'exact_email')",
          ("eil_" + eid, eid, inv_id, frm))
    email("ea_recent", "grant@ten31.xyz", "Grant Smith", 0, "i_acme", "a_grant", 1)   # Acme: today
    email("eb_old", "grant@ten31.xyz", "Grant Smith", 40, "i_beta", "a_grant", 1)     # Beta: 40d
    email("edel", "grant@ten31.xyz", "Grant Smith", 0, "i_beta", "a_grant", 1, deleted=True)  # tombstoned
    email("ej", "jon@ten31.xyz", "Jonathan Lee", 0, "i_acme", "a_jon", 1)             # jonathan today
    email("ein", "alice@acme.com", "Alice Acme", 3, "i_acme", "a_grant", 0)           # inbound 3d
    # an UNMATCHED sent email by Grant (NO email_investor_links row) — captured, but not to a
    # known investor. The investor-email intents are matched-only, so it must be EXCLUDED from
    # comms_by_user / email_counts_by_user; without the matched-only filter it would inflate both.
    c("INSERT INTO emails (id, rfc_message_id, from_email, from_name, sent_at, subject, "
      "is_matched, match_status) VALUES ('eunm','rfc_eunm','grant@ten31.xyz','Grant Smith',?,"
      "'Internal: team lunch',0,'unmatched')", (_ago(0),))
    c("INSERT INTO email_account_messages (id, email_id, account_id, gmail_message_id, "
      "gmail_thread_id, is_sent, deleted_at) VALUES "
      "('eam_eunm','eunm','a_grant','g_eunm','t_eunm',1,NULL)")

    # communications (the other recency leg) — Delta has ONLY comms: one live (5d), one tombstoned
    # (today). If the soft-delete filter broke, Delta would read as contacted today.
    c("INSERT INTO communications (id, contact_id, type, communication_date, created_by) VALUES "
      "('cmm_live','cc_dana','email',?,'u_grant')", (_ago(5),))
    c("INSERT INTO communications (id, contact_id, type, communication_date, created_by, deleted_at) "
      "VALUES ('cmm_del','cc_dana','email',?,'u_grant',?)", (_ago(0), _ago(0)))

    # reminders — open(overdue) / open(future) / done / deleted / standalone
    def rem(rid, inv_id, title, due, status="open", deleted=False):
        c("INSERT INTO reminders (id, investor_id, investor_name, title, due_date, status, "
          "deleted_at) VALUES (?,?,?,?,?,?,?)",
          (rid, inv_id, title, title, due, status, _ago(0) if deleted else None))
    rem("r_over", "i_beta", "Send deck", (TODAY - timedelta(days=1)).isoformat())       # overdue
    rem("r_future", "i_acme", "Quarterly check-in", (TODAY + timedelta(days=10)).isoformat())
    rem("r_done", "i_acme", "Old task", (TODAY - timedelta(days=2)).isoformat(), status="done")
    rem("r_del", "i_acme", "Tombstoned", (TODAY - timedelta(days=2)).isoformat(), deleted=True)
    rem("r_standalone", None, "Team chore", (TODAY - timedelta(days=1)).isoformat())

    # opportunities — committed / meeting (live) / lost (terminal) / deleted
    def opp(oid, inv_id, contact, stage, expected, owner, deleted=False):
        c("INSERT INTO opportunities (id, name, contact_id, stage, expected_amount, owner_id, "
          "fundraising_investor_id, deleted_at) VALUES (?,?,?,?,?,?,?,?)",
          (oid, oid, contact, stage, expected, owner, inv_id, _ago(0) if deleted else None))
    # opp contact_id must reference a real contacts row (FK on); reuse the two we made
    opp("o_acme", "i_acme", "cc_alice", "committed", 4_000_000, "u_jon")
    opp("o_beta", "i_beta", "cc_dana", "meeting", 1_000_000, "u_grant")
    opp("o_lost", "i_acme", "cc_alice", "lost", 9_000_000, "u_jon")
    opp("o_del", "i_beta", "cc_dana", "due_diligence", 7_000_000, "u_grant", deleted=True)
    conn.commit()


def names(res):
    return [r["investor_name"] for r in res["rows"]]


def main():
    server.init_db()
    conn = server.get_db()
    seed(conn)
    run = lambda *a, **k: nl_query.run_query(conn, *a, **k)

    print("investors_cold")
    r = run("investors_cold", {"days": 30})
    check(names(r) == ["Cold Co", "Beta Partners"], f"cold(30) never-first then stale: {names(r)}")
    check(run("investors_cold", {"days": 90})["row_count"] == 1, "cold(90): only never-contacted")
    check("Graveyard Ghost" not in names(run("investors_cold", {"days": 3650})),
          "cold excludes graveyard investors")
    check("Delta LP" in names(run("investors_cold", {"days": 3})), "cold(3) sees Delta (comms 5d)")
    check("Delta LP" not in names(run("investors_cold", {"days": 7})),
          "cold(7): Delta's tombstoned comm (today) did NOT count")

    print("investor_lookup")
    r = run("investor_lookup", {"name": "acme"})
    check(r["row_count"] == 1 and r["rows"][0]["total_invested"] == 5_000_000, "lookup total committed")
    check({c["fund_name"] for c in r["rows"][0]["commitments"]} == {"Fund I", "Fund II"},
          "lookup per-fund breakdown")
    check(r["rows"][0]["contacts"][0]["email"] == "alice@acme.com", "lookup surfaces contact email")

    print("investors_by_city / by_lead / top / follow_up")
    check(names(run("investors_by_city", {"city": "Austin"})) == ["Acme Capital"], "by_city")
    check(set(names(run("investors_by_lead", {"lead": "Grant"}))) == {"Beta Partners", "Cold Co", "Delta LP"},
          "by_lead excludes graveyard + other leads")
    check(names(run("top_investors_committed", {"limit": 2})) == ["Acme Capital", "Beta Partners"],
          "top by committed (graveyard + zero excluded)")
    r = run("investors_follow_up")
    check(names(r) == ["Beta Partners", "Acme Capital"], f"follow_up overdue-first, open-only: {names(r)}")
    check(r["rows"][0]["overdue"] == 1 and r["rows"][1]["overdue"] == 0, "follow_up overdue flag")

    print("pipeline")
    r = run("pipeline_totals")
    stages = {row["stage"]: row for row in r["rows"]}
    check(set(stages) == {"committed", "meeting"}, f"pipeline_totals excludes lost+deleted: {set(stages)}")
    check(stages["committed"]["expected_total"] == 4_000_000, "pipeline_totals stage sum")
    r = run("pipeline_top", {"limit": 10})
    check(names(r) == ["Acme Capital", "Beta Partners"], "pipeline_top furthest-stage first")
    check(r["rows"][0]["last_activity_at"] is not None, "pipeline_top enriches last activity")

    print("emails")
    check(run("recent_emails", {"direction": "outbound"})["row_count"] == 3,
          "recent_emails(outbound): 3 live (tombstoned sighting excluded)")
    check(run("recent_emails", {"direction": "inbound"})["row_count"] == 1, "recent_emails(inbound)")
    check(run("recent_emails")["row_count"] == 4, "recent_emails(any): 4 live")
    r = run("investor_last_contact", {"name": "beta"})
    check(r["rows"][0]["days_since"] >= 39, "investor_last_contact days_since")
    check(run("comms_by_user", {"user": "Grant"})["row_count"] == 2,
          "comms_by_user: grant's 2 live MATCHED outbound (tombstoned + unmatched excluded)")
    r = run("email_counts_by_user", {"user": "grant"})
    check(r["rows"][0]["this_week"] == 1,
          "email_counts this_week = 1 live matched (tombstoned + unmatched excluded)")
    check(r["rows"][0]["ytd"] >= 1, "email_counts ytd")

    print("trust boundary")
    check(run("investors_cold", {"days": "abc"})["error"] == "bad_slot", "bad int slot -> bad_slot, no crash")
    check(run("nope")["error"] == "unknown_intent", "unknown intent rejected")
    check(run("pipeline_totals", {"foo": 1})["error"] == "bad_slot", "unexpected slot rejected")
    check(run("investor_lookup", {})["error"] == "bad_slot", "missing required slot rejected")
    check(run("investors_by_city", {"city": "%"})["row_count"] == 0,
          "LIKE wildcard escaped — '%' does not match every row")
    check(run("investors_cold", {"days": 0})["slots"]["days"] == 1, "int slot clamps to min")
    check(run("top_investors_committed", {"limit": 99999})["slots"]["limit"] == nl_query.INTENTS
          ["top_investors_committed"]["slots"]["limit"]["max"], "int slot clamps to max")

    print("audit hook + catalog")
    seen = []
    run("pipeline_totals", audit_fn=seen.append, actor="tester", source="test")
    check(len(seen) == 1 and seen[0]["intent"] == "pipeline_totals" and seen[0]["error"] is None
          and seen[0]["actor"] == "tester", "audit hook fires with intent/actor/no-error")
    run("nope", audit_fn=seen.append)
    check(seen[-1]["error"] == "unknown_intent", "audit hook fires on rejection too")
    check(len(nl_query.catalog()) == len(nl_query.INTENTS), "catalog covers every intent")

    conn.close()
    print()
    if FAILS:
        print(f"{len(FAILS)} FAILED")
        for f in FAILS:
            print("  - " + f)
        sys.exit(1)
    print("ALL PASS")


if __name__ == "__main__":
    main()