#!/usr/bin/env python3
"""Pārbauda visas data/*.xml datnes:
  1. atbilst XSD schemas/mk-noteikumi-0.1.xsd (arī identifikatoru unikalitāte);
  2. datnes nosaukums = noteikumu identifikators, un identifikators = gads un numurs no Act;
  3. Source/SHA256 = sources/<id>.txt sha256;
  4. katrs teksts (Text, Note, Title, LegalBasis, Lead, Item, Line, Heading) ir avota teksta datnē tādā pašā secībā;
  5. katra elementa identifikators sākas ar noteikumu identifikatoru, punkta identifikators atbilst drukātajam numuram.

  python3 tools/validate.py            → kopsavilkums; izejas kods 1, ja kāda pārbaude neizdodas
"""
import glob
import hashlib
import os
import re
import sys

from lxml import etree

ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
NS = {"n": "urn:pppa:vpk:mk-noteikumi:0.1"}
SUPC = "⁰¹²³⁴⁵⁶⁷⁸⁹"
schema = etree.XMLSchema(etree.parse(f"{ROOT}/schemas/mk-noteikumi-0.1.xsd"))


def pid(num):
    return "p" + re.sub("[" + SUPC + "]+", lambda m: "s" + m.group(0).translate(str.maketrans(SUPC, "0123456789")), num)


def check(path):
    errs = []
    doc = etree.parse(path)
    if not schema.validate(doc):
        errs += [f"XSD: {e.message}" for e in list(schema.error_log)[:5]]
    r = doc.getroot()
    did = r.get("id")
    if os.path.basename(path) != f"{did}.xml":
        errs.append(f"datnes nosaukums {os.path.basename(path)} ≠ {did}.xml")
    num, adopted = r.findtext("n:Act/n:Number", namespaces=NS), r.findtext("n:Act/n:Adopted", namespaces=NS) or ""
    if did != f"lv-mkn-{adopted[:4]}-{num}":
        errs.append(f"identifikators {did} ≠ lv-mkn-{adopted[:4]}-{num}")
    src = f"{ROOT}/{r.findtext('n:Source/n:File', namespaces=NS)}"
    if not os.path.exists(src):
        return errs + [f"nav avota datnes {src}"]
    body = open(src, encoding="utf-8").read()
    if hashlib.sha256(body.encode("utf-8")).hexdigest() != r.findtext("n:Source/n:SHA256", namespaces=NS):
        errs.append("SHA256 nesakrīt ar avota datni")
    pos, missing = 0, []
    for e in r.iter():
        if not isinstance(e.tag, str):
            continue
        tag = etree.QName(e).localname
        if tag in ("Text", "Note", "LegalBasis", "Lead", "Line", "Heading") or (tag == "Title" and etree.QName(e.getparent()).localname != "Act"):
            for part in (e.text or "").split("\n"):
                i = body.find(part, pos)
                if i < 0:
                    missing.append(f"{tag}: {part[:70]}")
                else:
                    pos = i
        if e is not r and e.get("id") and not e.get("id").startswith(did + "."):
            errs.append(f"{e.get('id')} nesākas ar {did}")
        if tag == "Point" and not e.get("n").endswith(")") and not e.get("id").endswith("." + pid(e.get("n"))):
            errs.append(f"punkts n={e.get('n')} id={e.get('id')}")
    errs += [f"nav avotā secībā — {m}" for m in missing[:5]]
    return errs


if __name__ == "__main__":
    bad = 0
    for p in sorted(glob.glob(f"{ROOT}/data/*.xml")):
        e = check(p)
        r = etree.parse(p).getroot()
        pts = len(r.findall(".//n:Point", NS))
        print(f"{'OK  ' if not e else 'FAIL'} {os.path.basename(p)}  punkti {pts}")
        for x in e:
            print("     ", x)
        bad += bool(e)
    sys.exit(1 if bad else 0)