sdlc-hub/tests/test_wiki_convert.py

"""Перевод HTML Eva в Markdown (wiki-module, spec wiki «Markdown-копия вики», design 10)."""
import re
from pathlib import Path

from sdlc_hub.modules.wiki import convert

FIX = Path(__file__).parent / "fixtures" / "wiki"
FILE_URL = "https://eva.example.test/files/obj/CmfDocument/CmfDocument%3Aa08/CmfDocument%3A3f2a9c1e-0b7d-4e8a-9c21-5d6e7f8a9b0c/model.xlsx?version=1"
IMG_REF = "/files/obj/CmfDocument/CmfDocument%3Aa08/CmfDocument%3A3f2a9c1e-0b7d-4e8a-9c21-5d6e7f8a9b0c/image2026-7-27_14-20-6.png?version=1"
THUMB_REF = "/files/obj/CmfDocument/CmfDocument%3Aa08/CmfDocument%3A3f2a9c1e-0b7d-4e8a-9c21-5d6e7f8a9b0c/model.xlsx.meta/thumbnail.jpg?version=1"
FILE_REF = "/files/obj/CmfDocument/CmfDocument%3Aa08/CmfDocument%3A3f2a9c1e-0b7d-4e8a-9c21-5d6e7f8a9b0c/model.xlsx?version=1"


def _html(name):
    return (FIX / name).read_text(encoding="utf-8")


def test_image_becomes_attachment():
    md, atts = convert.to_markdown(
        _html("page_image.html"),
        [{"ref": IMG_REF, "url": "https://eva.example.test" + IMG_REF, "name": "image2026-7-27_14-20-6.png"}],
        [],
    )
    assert "[📎 image2026-7-27_14-20-6.png](attachment:@0)" in md.splitlines()
    assert atts == [{"url": "https://eva.example.test" + IMG_REF, "name": "image2026-7-27_14-20-6.png"}]


def test_thumbnail_and_file_one_link():
    md, atts = convert.to_markdown(
        _html("page_file.html"),
        [
            {"ref": THUMB_REF, "url": FILE_URL, "name": "model.xlsx"},
            {"ref": FILE_REF, "url": FILE_URL, "name": "model.xlsx"},
        ],
        [],
    )
    assert [a["name"] for a in atts] == ["model.xlsx"] and len(atts) == 1
    assert md.count("](attachment:@0)") == 1
    assert "[📎 model.xlsx](attachment:@0)" in md
    assert "attach-file" not in md


def test_page_link():
    md, _ = convert.to_markdown(
        _html("page_links.html"),
        [],
        [{"ref": "project/Document/DOC-019834", "code": "DOC-019834"}],
    )
    assert "[Модель данных](page:DOC-019834)" in md


def test_external_link_kept():
    md, _ = convert.to_markdown(_html("page_links.html"), [], [])
    assert "(https://example.org)" in md


def test_no_raw_html():
    for name in ("page_image.html", "page_file.html", "page_links.html"):
        md, _ = convert.to_markdown(
            _html(name),
            [{"ref": IMG_REF, "url": "https://eva.example.test" + IMG_REF, "name": "image.png"},
             {"ref": THUMB_REF, "url": FILE_URL, "name": "model.xlsx"},
             {"ref": FILE_REF, "url": FILE_URL, "name": "model.xlsx"}],
            [{"ref": "project/Document/DOC-019834", "code": "DOC-019834"}],
        )
        assert re.search(r"<[a-zA-Z/]", md) is None, name


def test_table_kept():
    md, _ = convert.to_markdown(_html("page_links.html"), [], [])
    table_lines = [line for line in md.splitlines() if "|" in line]
    assert table_lines and any("---" in line for line in table_lines)


def test_headings_atx():
    md, _ = convert.to_markdown(_html("page_links.html"), [], [])
    assert "# Требования" in md.splitlines()


def test_static_links_removed():
    html = '<p>Текст <img src="/static/x.png" alt=""> и <a href="static/y.html">ссылка</a></p>'
    md, atts = convert.to_markdown(html, [], [])
    assert "static" not in md and atts == []


def test_fill_ids():
    assert convert.fill_ids("[x](attachment:@1)", [7, 9]) == "[x](attachment:9)"


# Колонка управления строками таблицы редактора Eva (TinyMCE, `table-row-controls`) попадает в HTML страницы и
# раньше давала абзацы «№», «1», «2»… перед таблицей (#95)
CARD_REF = "/files/obj/CmfDocument/CmfDocument%3Aa08/CmfDocument%3A3f2a9c1e-0b7d-4e8a-9c21-5d6e7f8a9b0c/model.dbml?version=1"


def test_table_row_controls_removed():
    md, _ = convert.to_markdown(_html("page_table_controls.html"), [], [])
    lines = [line.strip() for line in md.splitlines() if line.strip()]
    assert not [line for line in lines if re.fullmatch(r"№|\d{1,3}", line)], lines
    assert "| № | Замечание | Решение |" in lines
    assert "| 2-бис | Дополнение | Принято |" in lines
    assert "Текст после таблицы." in lines


def test_card_preview_kept_with_controls():
    md, atts = convert.to_markdown(
        _html("page_table_controls.html"),
        [{"ref": CARD_REF, "url": "https://eva.example.test" + CARD_REF, "name": "model.dbml"}],
        [],
    )
    assert "[📎 model.dbml](attachment:@0)" in md and len(atts) == 1


# --- fix round 1 (review Task 6: I1, I2, I3, M3, M4) ---

def test_unsafe_schemes_dropped_from_links():
    html = ('<p><a href="javascript:alert(1)">x1</a> <a href="JaVaScript:alert(2)">x2</a> '
            '<a href="data:text/html,<b>x</b>">x3</a> <a href=" java\tscript:alert(3)">x4</a> '
            '<a href="vbscript:msgbox">x5</a></p>')
    md, _ = convert.to_markdown(html, [], [])
    for text in ("x1", "x2", "x3", "x4", "x5"):
        assert text in md
    assert "(javascript" not in md.lower() and "(data:" not in md and "(vbscript" not in md.lower()
    assert "java" not in md.lower().replace("x4", "")  # схема не осталась ни в каком виде


def test_unsafe_scheme_img_dropped():
    md, atts = convert.to_markdown('<p>a<img src="javascript:x" alt="z">b</p>', [], [])
    assert md.strip() == "ab" and atts == []


def test_attachment_name_and_link_text_escaped():
    md, atts = convert.to_markdown(
        '<p><a href="/files/a">x</a></p>',
        [{"ref": "/files/a", "url": "https://eva.example.test/files/a", "name": "a]b[c).png"}],
        [],
    )
    assert "[📎 a\\]b\\[c).png](attachment:@0)" in md.splitlines()
    assert atts[0]["name"] == "a]b[c).png"

    md, _ = convert.to_markdown(
        '<p><a href="project/Document/DOC-1">a ] b [x](http://evil)</a></p>',
        [],
        [{"ref": "project/Document/DOC-1", "code": "DOC-1"}],
    )
    assert "[a \\] b \\[x\\](http://evil)](page:DOC-1)" in md.splitlines()


def test_newlines_in_link_text_collapsed():
    md, _ = convert.to_markdown(
        '<p><a href="project/Document/DOC-1">первая\nвторая</a></p>', [],
        [{"ref": "project/Document/DOC-1", "code": "DOC-1"}],
    )
    assert "[первая вторая](page:DOC-1)" in md.splitlines()


def test_literal_attachment_marker_in_text_is_safe():
    md, atts = convert.to_markdown("<p>[x](attachment:@5) и attachment:@0</p>", [], [])
    assert atts == []
    assert convert.fill_ids(md, []) == md
    assert "(attachment:@5)" not in md.replace("\\(attachment:\\@5)", "")


def test_fill_ids_by_index_not_prefix():
    ids = list(range(100, 111))
    assert convert.fill_ids("[a](attachment:@1) [b](attachment:@10)", ids) == \
        "[a](attachment:101) [b](attachment:110)"


def test_fill_ids_out_of_range_raises():
    try:
        convert.fill_ids("[a](attachment:@3)", [1, 2])
    except ValueError as exc:
        assert "@3" in str(exc)
    else:
        raise AssertionError("ожидалась ValueError")


def test_none_lists_are_empty():
    md, atts = convert.to_markdown("<p>текст</p>", None, None)
    assert md == "текст\n" and atts == []


def test_literal_tag_text_has_no_raw_html():
    md, _ = convert.to_markdown("<p>&lt;tag&gt; и &lt;script&gt;alert(1)&lt;/script&gt;</p>", [], [])
    assert re.search(r"<[a-zA-Z/]", md) is None
    assert "&lt;tag&gt;" in md or "&lt;tag" in md


# --- fix round 2 (re-review 1: B1, B2, B3) ---

def _no_forged_marker(md, ids=(7,)):
    """Авторский текст не порождает ссылку вложения: fill_ids ничего не меняет и шаблона нет."""
    assert "](attachment:@" not in md
    assert convert.fill_ids(md, list(ids)) == md


def test_page_link_text_escaped_once():
    md, _ = convert.to_markdown(
        '<p><a href="project/Document/DOC-1">a_b *x* c</a></p>', [],
        [{"ref": "project/Document/DOC-1", "code": "DOC-1"}],
    )
    assert "[a\\_b \\*x\\* c](page:DOC-1)" in md.splitlines()
    assert "\\\\" not in md


def test_page_link_at_sign_not_double_escaped():
    md, _ = convert.to_markdown(
        '<p><a href="project/Document/DOC-1">user@example.com</a></p>', [],
        [{"ref": "project/Document/DOC-1", "code": "DOC-1"}],
    )
    assert "[user@example.com](page:DOC-1)" in md.splitlines()


def test_attachment_name_star_and_underscore_at_word_edge_escaped():
    md, _ = convert.to_markdown(
        '<p><a href="/files/a">x</a></p>',
        [{"ref": "/files/a", "url": "https://eva.example.test/files/a", "name": "_x_*.png"}],
        [],
    )
    assert "[📎 \\_x\\_\\*.png](attachment:@0)" in md.splitlines()


def test_attachment_name_intraword_underscore_kept():
    md, _ = convert.to_markdown(
        '<p><a href="/files/a">x</a></p>',
        [{"ref": "/files/a", "url": "https://eva.example.test/files/a", "name": "my_file.png"}],
        [],
    )
    assert "[📎 my_file.png](attachment:@0)" in md.splitlines()


def test_author_href_attachment_scheme_is_text():
    md, atts = convert.to_markdown('<p><a href="attachment:@5">x</a></p>', [], [])
    assert md.strip() == "x" and atts == []
    _no_forged_marker(md)


def test_author_img_src_attachment_scheme_dropped():
    md, atts = convert.to_markdown('<p>a<img src="attachment:@5" alt="z">b</p>', [], [])
    assert md.strip() == "ab" and atts == []
    _no_forged_marker(md)


def test_author_alt_cannot_forge_marker():
    md, _ = convert.to_markdown('<p><img src="/x.png" alt="a](attachment:@0)"></p>', [], [])
    _no_forged_marker(md)


def test_author_title_cannot_forge_marker():
    md, _ = convert.to_markdown('<p><a href="https://e.org" title="](attachment:@0)">t</a></p>', [], [])
    _no_forged_marker(md)


def test_author_code_cannot_forge_marker():
    md, _ = convert.to_markdown("<p><code>](attachment:@0)</code></p>", [], [])
    _no_forged_marker(md)


def test_author_text_cannot_forge_marker():
    md, _ = convert.to_markdown("<p>](attachment:@0) и &#93;(attachment:&#64;0)</p>", [], [])
    _no_forged_marker(md)


def test_author_private_chars_cannot_forge_marker():
    md, _ = convert.to_markdown("<p>&#57344;0&#57345;</p>", [], [])
    _no_forged_marker(md)


def test_plain_at_sign_kept():
    md, _ = convert.to_markdown("<p>user@example.com и @ivan</p>", [], [])
    assert "user@example.com и @ivan" in md


def test_code_keeps_literal_text():
    md, _ = convert.to_markdown("<p><code>&lt;script&gt; a@b</code></p>", [], [])
    assert "`<script> a@b`" in md


# пакет D (#135): T6.1–T6.5
IMG_URL = "https://eva.example.test" + IMG_REF
FILE = {"ref": FILE_REF, "url": FILE_URL, "name": "model.xlsx"}
THUMB = {"ref": THUMB_REF, "url": FILE_URL, "name": "model.xlsx"}
IMG = {"ref": IMG_REF, "url": IMG_URL, "name": "scheme.png"}
LINK = {"ref": "project/Document/DOC-1", "code": "DOC-1"}


def test_thumbnail_in_wrapper_keeps_link():
    md, atts = convert.to_markdown(f'<p><a href="{FILE_REF}"><span><img src="{THUMB_REF}"></span></a> текст</p>', [THUMB, FILE], [])
    assert md == "[📎 model.xlsx](attachment:@0) текст\n" and len(atts) == 1


def test_other_attachment_inside_attachment_link():
    md, atts = convert.to_markdown(f'<p><a href="{FILE_REF}"><img src="{IMG_REF}"></a></p>', [FILE, IMG], [])
    assert md == "[📎 scheme.png](attachment:@0)[📎 model.xlsx](attachment:@1)\n"
    assert [a["name"] for a in atts] == ["scheme.png", "model.xlsx"]


def test_attachment_inside_page_link():
    md, _ = convert.to_markdown(f'<p><a href="project/Document/DOC-1"><img src="{IMG_REF}"></a></p>', [IMG], [LINK])
    assert md == "[📎 scheme.png](attachment:@0) [DOC-1](page:DOC-1)\n"


def test_attachment_and_text_inside_page_link():
    md, _ = convert.to_markdown(f'<p><a href="project/Document/DOC-1"><img src="{IMG_REF}">Схема</a></p>', [IMG], [LINK])
    assert md == "[📎 scheme.png](attachment:@0) [Схема](page:DOC-1)\n"


def test_attachment_inside_external_link():
    md, _ = convert.to_markdown(f'<p><a href="https://ex.com/x"><img src="{IMG_REF}"></a></p>', [IMG], [])
    assert md == "[📎 scheme.png](attachment:@0) [https://ex.com/x](https://ex.com/x)\n"


def test_duplicate_ref_first_wins():
    md, atts = convert.to_markdown(f'<p><img src="{IMG_REF}"></p>', [dict(IMG, name="first"), dict(IMG, url="u2", name="second")], [])
    assert atts == [{"url": IMG_URL, "name": "first"}]
    md, _ = convert.to_markdown('<p><a href="project/Document/DOC-1">т</a></p>', [], [LINK, dict(LINK, code="DOC-9")])
    assert "(page:DOC-1)" in md


def test_external_href_spaces_and_parens_encoded():
    md, _ = convert.to_markdown('<p><a href="http://ex.com/a b(c).html">док</a></p>', [], [])
    assert md == "[док](http://ex.com/a%20b%28c%29.html)\n"


def test_external_href_already_encoded_untouched():
    md, _ = convert.to_markdown('<p><a href="http://ex.com/a%20b.html">док</a></p>', [], [])
    assert "(http://ex.com/a%20b.html)" in md and "%2520" not in md


def test_trailing_backslash_in_page_link_text():
    md, _ = convert.to_markdown('<p><a href="project/Document/DOC-1">путь C:\\</a></p>', [], [LINK])
    assert md == "[путь C:\\\\](page:DOC-1)\n"


def test_trailing_backslash_in_external_link_text():
    md, _ = convert.to_markdown('<p><a href="http://ex.com/">путь C:\\</a></p>', [], [])
    assert md == "[путь C:\\\\](http://ex.com/)\n"


def test_external_link_keeps_edge_spaces():
    """Финальное ревью I-1: пробелы на краях текста внешней ссылки выносятся наружу, слова не склеиваются."""
    md, _ = convert.to_markdown('<p>слово<a href="http://ex.com/"> док </a>дальше</p>', [], [])
    assert md == "слово [док](http://ex.com/) дальше\n"


def test_empty_external_link_dropped():
    """Финальное ревью M-1: пустая внешняя ссылка без вложения внутри не выводится адресом."""
    md, _ = convert.to_markdown('<p>a<a href="#top"></a>b<a href="http://x"><img src="static/i.png"></a>c</p>', [], [])
    assert md == "abc\n"


def test_version_markdown_attachment_as_text():
    md, files = convert.to_markdown(_html("page_file.html"), [THUMB, FILE], [], file_links=False)
    assert "📎 model.xlsx" in md and "attachment:" not in md and files == []


def test_version_markdown_page_links_kept():
    md, _ = convert.to_markdown('<p><a href="project/Document/DOC-1">т</a></p>', [], [LINK], file_links=False)
    assert md == "[т](page:DOC-1)\n"