"""
analyze_asst_prof_comp.py
==========================
One-off analysis: median total compensation (salary + benefits) for
non-medical Assistant Professors across the whole UC system, 2024.

Data source: CA State Controller's Government Compensation in California
(GCC) dataset, restored from this repo's git history (gcc_data/GCC_*_2024.csv)
because the UC-wages source uc_salary_pipeline.py normally uses has no
benefits field at all.

"Total compensation" = TotalWages (cash pay) + TotalRetirementAndHealthContribution
(employer-paid health/dental/vision + retirement + deferred comp) -- the
standard total-comp figure used for this dataset (e.g. by Transparent
California).

"Non-medical" reuses classifier.py's existing SENATE_FACULTY / is_medical_center
logic, which already flags HCOMP / HS-CLIN-PROF / clinical-workforce titles
as medical regardless of campus.

"Assistant Professor" is a rank-level filter this repo doesn't otherwise
do (classify_title() only buckets at the Senate-faculty-vs-not level) --
_ASST_RANK_RX below mirrors classifier.py's own _SENATE_RX patterns,
restricted to the assistant-rank variants.
"""

import csv
import glob
import re
import statistics

from classifier import classify_title, is_medical_center, normalise

EMPLOYER_TO_CAMPUS = {
    "University of California, Berkeley": "UCB",
    "University of California, Los Angeles": "UCLA",
    "University of California, Davis": "UCD",
    "University of California, Irvine": "UCI",
    "University of California, Santa Barbara": "UCSB",
    "University of California, Riverside": "UCR",
    "University of California, Santa Cruz": "UCSC",
    "University of California, Merced": "UCM",
    "University of California, Office of the President": "UCOP",
    "University of California, San Diego": "UCSD",
    "University of California, San Francisco": "UCSF",
}

# Assistant-rank Senate faculty titles only -- mirrors classifier.py's
# _SENATE_RX but excludes bare/ASSOC/full-professor variants.
# "ASST ACT PROF" (reversed order = non-Senate, per classifier.py's own
# docstring) is NOT matched here and is excluded anyway by the
# SENATE_FACULTY-only filter in main().
_ASST_RANK_RX = [re.compile(p, re.IGNORECASE) for p in (
    r"^ASST\s+PROF-(?:AY|FY|HCOMP|SFT-VM)\b",
    r"^ASST\s+PROF-AY-",
    r"^ASST\s+PROF-FY-",
    r"^ASST\s+PROF OF TEACH\b",
    r"^ASST\s+PROF OF TEACH-",
    r"^ASST\s+PROF OF CLIN\b",
    r"^ASST\s+PROF OF CLIN-",
    r"^ASST\s+PROF IN RES\b",
    r"^ASST\s+PROF IN RES-",
    r"^ACT\s+ASST\s+PROF\b",
    r"^ACT\s+ASST\s+PROF-",
    r"^ACT\s+ASST\s+PROF OF TEACH\b",
    r"^ACT\s+ASST\s+PROF OF TEACH-",
    r"^ACT\s+ASST\s+PROF OF CLIN\b",
    r"^ACT\s+ASST\s+PROF OF CLIN-",
    r"^ACT\s+ASST\s+PROF IN RES\b",
)]

# Other Senate-faculty patterns (non-assistant) so we can flag anything
# that matched SENATE_FACULTY but neither this nor the assistant patterns
# recognise -- a diagnostic for titles the rank regex list above missed.
_NON_ASST_SENATE_RX = [re.compile(p, re.IGNORECASE) for p in (
    r"^(?:ASSOC\s+)?PROF-(?:AY|FY|HCOMP|SFT-VM)\b",
    r"^(?:ASSOC\s+)?PROF-AY-",
    r"^(?:ASSOC\s+)?PROF-FY-",
    r"^(?:ASSOC\s+)?PROF OF TEACH\b",
    r"^(?:ASSOC\s+)?PROF OF TEACH-",
    r"^(?:ASSOC\s+)?PROF OF CLIN\b",
    r"^(?:ASSOC\s+)?PROF OF CLIN-",
    r"^(?:ASSOC\s+)?PROF IN RES\b",
    r"^(?:ASSOC\s+)?PROF IN RES-",
    r"^ACT\s+PROF\b", r"^ACT\s+PROF-",
    r"^ACT\s+ASSOC\s+PROF\b", r"^ACT\s+ASSOC\s+PROF-",
    r"^ACT\s+PROF OF TEACH\b", r"^ACT\s+PROF OF TEACH-",
    r"^ACT\s+ASSOC\s+PROF OF TEACH\b", r"^ACT\s+ASSOC\s+PROF OF TEACH-",
    r"^ACT\s+PROF OF CLIN\b", r"^ACT\s+PROF OF CLIN-",
    r"^ACT\s+ASSOC\s+PROF OF CLIN\b", r"^ACT\s+ASSOC\s+PROF OF CLIN-",
    r"^ACT\s+PROF IN RES\b", r"^ACT\s+ASSOC\s+PROF IN RES\b",
    r"^INSTR\b", r"^INSTR-", r"^INSTRUCTOR\b",
    r"^(?:ASSOC\s+)?INSTR IN RES\b",
    r"^REGENTS'? PROF\b",
    r"\bPROF(?:ESSOR)?[-\s]EMERITUS\b", r"\bPROF(?:ESSOR)?\s+EMERITUS\b",
    r"^RECALL FACULTY\b",
    r"^(?:ASSOC\s+)?AGRON AES(?:-AY|-FY|-B/E/E|-SFT-VM)?$",
    r"^(?:ASSOC\s+)?ASTRONOMER(?:-AY|-FY)?$",
)]

def is_assistant_rank(raw_title: str) -> bool:
    t = normalise(raw_title)
    return any(rx.search(t) for rx in _ASST_RANK_RX)


def is_recognised_non_assistant_senate(raw_title: str) -> bool:
    t = normalise(raw_title)
    return any(rx.search(t) for rx in _NON_ASST_SENATE_RX)


def parse_money(s: str) -> float:
    s = (s or "").replace(",", "").replace("$", "").strip()
    return float(s) if s else 0.0


def main():
    rows = []
    unrecognised_senate_titles = set()

    for path in sorted(glob.glob("gcc_data/GCC_*_2024.csv")):
        if path.endswith("GCC_ALL_2024.csv"):
            continue
        with open(path, newline="", encoding="utf-8-sig") as f:
            for row in csv.DictReader(f):
                employer = row["EmployerName"]
                campus = EMPLOYER_TO_CAMPUS.get(employer)
                if campus is None:
                    raise ValueError(f"Unmapped EmployerName: {employer!r} in {path}")

                title = row["Position"]
                if classify_title(title) != "SENATE_FACULTY":
                    continue
                if is_medical_center(title, campus):
                    continue
                if not is_assistant_rank(title):
                    if not is_recognised_non_assistant_senate(title):
                        unrecognised_senate_titles.add(title)
                    continue

                total_comp = (parse_money(row["TotalWages"])
                             + parse_money(row["TotalRetirementAndHealthContribution"]))
                rows.append({"campus": campus, "title": title, "total_comp": total_comp})

    if unrecognised_senate_titles:
        print("NOTE: Senate-faculty titles not matched by any rank pattern "
              "(not counted as assistant, but check these aren't mis-missed asst titles):")
        for t in sorted(unrecognised_senate_titles):
            print(f"  - {t}")
        print()

    all_comp = [r["total_comp"] for r in rows]
    print(f"Non-medical Assistant Professors, UC system-wide, 2024: n={len(rows)}")
    print(f"Median total compensation (salary + benefits): ${statistics.median(all_comp):,.2f}")
    print(f"Mean total compensation:                        ${statistics.mean(all_comp):,.2f}")
    print()
    print(f"{'Campus':<8}{'N':>6}{'Median':>16}")
    for campus in sorted(EMPLOYER_TO_CAMPUS.values()):
        campus_comp = [r["total_comp"] for r in rows if r["campus"] == campus]
        if campus_comp:
            print(f"{campus:<8}{len(campus_comp):>6}{statistics.median(campus_comp):>16,.2f}")
        else:
            print(f"{campus:<8}{0:>6}{'--':>16}")


if __name__ == "__main__":
    main()