analise-artigo/analise_artigo/bib.py

"""Parser simples de arquivos .bib (BibTeX).

Usa apenas a biblioteca padrão; suporta campos com valores em chaves
aninhadas e valores multi-linha.
"""

from __future__ import annotations

import re
from dataclasses import dataclass, field
from pathlib import Path


@dataclass
class Reference:
    key: str
    type: str  # article, inproceedings, misc, ...
    fields: dict[str, str] = field(default_factory=dict)

    @property
    def authors(self) -> str:
        value = self.fields.get("author", "").strip()
        # "X, A. and B, C. and others" -> "X, A., B, C. et al." (apenas campos author)
        value = re.sub(r"\s+and\s+others\b", " et al.", value, flags=re.I)
        return re.sub(r"\s+and\s+", ", ", value, flags=re.I)

    @property
    def title(self) -> str:
        return self.fields.get("title", "").strip()

    @property
    def year(self) -> str:
        return self.fields.get("year", "").strip()


    @property
    def journal(self) -> str:
        return self.fields.get(
            "journal", self.fields.get("booktitle", "")
        ).strip()

    def label(self) -> str:
        """Rótulo curto autor (ano)."""
        first = re.split(r",|and", self.authors)[0].strip()
        if "," in first:
            surname = first.split(",")[0].strip()
        else:
            surname = last_token(first)
        return f"{surname} ({self.year})" if self.year else surname


def last_token(name: str) -> str:
    parts = name.replace("-", " ").split()
    return parts[-1] if parts else name


def _split_entries(text: str) -> list[tuple[str, str, str]]:
    """Retorna [(type, key, body), ...] respeitando aninhamento de chaves."""
    entries: list[tuple[str, str, str]] = []
    i = 0
    while True:
        m = re.compile(r"@(\w+)\s*\{").search(text, i)
        if not m:
            break
        etype, key = m.group(1), ""
        start = m.end()
        depth = 1
        j = start
        while j < len(text) and depth:
            if text[j] == "{":
                depth += 1
            elif text[j] == "}":
                depth -= 1
            j += 1
        body = text[start : j - 1]

        # chave = conteúdo até a primeira vírgula fora de chaves
        depth = 0
        k = 0
        while k < len(body):
            c = body[k]
            if c == "{":
                depth += 1
            elif c == "}":
                depth -= 1
            elif c == "," and depth == 0:
                break
            if c == "\n" and depth == 0:
                break
            k += 1
        key = body[:k].strip()
        entries.append((etype, key, body[k + 1 :]))
        i = j
    return entries


def _parse_fields(body: str) -> dict[str, str]:
    fields: dict[str, str] = {}
    i = 0
    n = len(body)
    while i < n:
        while i < n and body[i] in " \t\n,":
            i += 1
        if i >= n:
            break
        m = re.compile(r"(\w+)\s*=").match(body, i)
        if not m:
            break
        fname = m.group(1).lower()
        i = m.end()
        while i < n and body[i] in " \t\n":
            i += 1
        if i >= n:
            break
        if body[i] == "{":
            depth = 1
            j = i + 1
            while j < n and depth:
                if body[j] == "{":
                    depth += 1
                elif body[j] == "}":
                    depth -= 1
                j += 1
            value = body[i + 1 : j - 1]
            i = j
        else:
            m2 = re.compile(r"([^,]+)").match(body, i)
            if not m2:
                i += 1
                continue
            value = m2.group(1)
            i = m2.end()
        value = value.strip()
        value = re.sub(r"\s*\n\s*", " ", value)
        value = value.replace("{", "").replace("}", "").strip()
        value = _clean_bib_value(value)
        fields[fname] = value
    return fields


def _clean_bib_value(value: str) -> str:
    """Remove escapes LaTeX correntes em .bib (diacríticos, \\& etc.).

    Ex.: ``Ki\\vs\\vs`` -> ``Kiss``, ``L'opez`` -> ``Lopez``,
    ``Tirke\\cs`` -> ``Tirkes``.
    """
    value = re.sub(
        r"\\[\'\"~vc](?:\{([A-Za-z])\}|([A-Za-z]))",
        lambda m: m.group(1) or m.group(2),
        value,
        flags=re.I,
    )
    for esc, ch in (("\\&", "&"), ("\\$", "$"), ("\\%", "%"),
                    ("\\_", "_"), ("\\#", "#")):
        value = value.replace(esc, ch)
    value = value.replace("---", " — ").replace("--", " – ")
    return value


def parse_bib(bib_path: Path | str) -> dict[str, Reference]:
    text = Path(bib_path).read_text(encoding="utf-8")
    refs: dict[str, Reference] = {}
    for etype, key, body in _split_entries(text):
        if not key:
            continue
        refs[key] = Reference(key=key, type=etype, fields=_parse_fields(body))
    return refs