#!/usr/bin/env python3
"""Scan cloned RISC-V repos and print a compact completeness report."""
import os, subprocess, io, sys

sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8")
REPO_DIR = "/data/repos"
REPOS = ["serv", "neorv32", "picorv32", "ibex", "VexRiscv",
         "e203_hbirdv2", "pulpino", "pulpissimo", "opentitan"]

RTL_EXT = {".v", ".sv", ".svh", ".vh", ".vhd", ".vhdl", ".fhdl", ".scala"}
TB_DIRS = {"tb", "tb_core", "testbench", "verif", "dv", "dv_uvm", "tests", "test", "tbench"}
DOC_EXT = {".md", ".rst", ".tex", ".pdf"}

def head_date(repo):
    try:
        out = subprocess.run(["git", "log", "-1", "--format=%ci %h"],
                             cwd=os.path.join(REPO_DIR, repo), capture_output=True, text=True).stdout
        return out.strip()[:16]
    except Exception:
        return "?"

def scan(repo):
    root = os.path.join(REPO_DIR, repo)
    rtl_files = rtl_lines = doc_files = tb_rtl = 0
    core_files = makefiles = pyscripts = tcl = bazel = cmake = 0
    workflows, licenses = [], []
    gitmods = []
    for dirpath, dirnames, filenames in os.walk(root):
        dirnames[:] = [d for d in dirnames if d != ".git"]
        rel = os.path.relpath(dirpath, root)
        parts = set(rel.replace("\\", "/").split("/"))
        for f in filenames:
            p = os.path.join(dirpath, f)
            ext = os.path.splitext(f)[1].lower()
            lf = f.lower()
            if ext in RTL_EXT:
                try:
                    with open(p, "rb") as fh:
                        n = sum(1 for _ in fh)
                except Exception:
                    n = 0
                rtl_files += 1
                rtl_lines += n
                if parts & TB_DIRS:
                    tb_rtl += 1
            elif ext in DOC_EXT:
                doc_files += 1
            if lf == "license" or lf.startswith("license."):
                licenses.append(os.path.relpath(p, root))
            if lf.endswith(".core"):
                core_files += 1
            if lf in ("makefile",) or lf.startswith("makefile.") or lf.endswith(".mk"):
                makefiles += 1
            if ext == ".py":
                pyscripts += 1
            if ext == ".tcl":
                tcl += 1
            if lf.startswith("bazelrc") or lf == "build" or lf.endswith(".bzl") or lf == "module_bazel":
                bazel += 1
            if lf in ("cmakelists.txt",):
                cmake += 1
        if rel == os.path.join(".github", "workflows"):
            workflows = [f for f in filenames if f.endswith((".yml", ".yaml"))]
    gm = os.path.join(root, ".gitmodules")
    if os.path.exists(gm):
        with open(gm, encoding="utf-8", errors="ignore") as fh:
            gitmods = [l.split("=", 1)[1].strip() for l in fh if l.strip().startswith("path =")]
    tops = sorted(d for d in os.listdir(root)
                  if os.path.isdir(os.path.join(root, d)) and d != ".git")
    lic = licenses[0] if licenses else "MISSING"
    readme = os.path.exists(os.path.join(root, "README.md")) or os.path.exists(os.path.join(root, "README.rst"))
    print(f"=== {repo} ===")
    print(f"HEAD: {head_date(repo)} | LICENSE: {lic} | README: {'Y' if readme else 'N'}")
    print(f"RTL files: {rtl_files} | RTL lines: {rtl_lines} | TB rtl files: {tb_rtl} | doc files: {doc_files}")
    print(f"FuseSoC .core: {core_files} | Makefile: {makefiles} | .py: {pyscripts} | .tcl: {tcl} | bazel-ish: {bazel} | cmake: {cmake}")
    print(f"CI workflows: {len(workflows)} {workflows[:4]}")
    print(f"submodules({len(gitmods)}): {gitmods[:12]}")
    print(f"top dirs: {tops}")
    print()

for r in REPOS:
    try:
        scan(r)
    except Exception as e:
        print(f"=== {r} === ERROR {e}\n")