Mermutef/NLP
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Добавлены итоговый отчёт и архив лабораторной работы

Browse Source
This commit is contained in:
Vladislav Kopnin
2026-11-01 01:25:15 +03:00
parent 09fc05c327
commit 32b61750bf
26 changed files with 2081 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

570
lab1/scripts/build_report.py Normal file
View File

@@ -0,0 +1,570 @@
#!/usr/bin/env python3
"""Сборка отчёта: Markdown -> DOCX -> PDF.
Скрипт рассчитан на запуск из каталога lab1:
python scripts/build_report.py
Требуются pandoc, python-docx и libreoffice. PDF собирается через
LibreOffice, чтобы не зависеть от LaTeX-шаблона и корректно передать
кириллицу, таблицы и рисунки.
"""
from __future__ import annotations
import argparse
import shutil
import subprocess
import sys
from pathlib import Path
import docx
from docx.enum.section import WD_SECTION
from docx.enum.table import WD_TABLE_ALIGNMENT
from docx.enum.text import WD_ALIGN_PARAGRAPH, WD_BREAK, WD_LINE_SPACING
from docx.oxml import OxmlElement
from docx.oxml.ns import qn
from docx.shared import Cm, Pt, RGBColor
ROOT = Path(__file__).resolve().parents[1]
REPORT_MD = ROOT / "report" / "REPORT.md"
BUILD = ROOT / "report" / "build"
DOCX_OUT = ROOT / "report" / "REPORT.docx"
PDF_OUT = ROOT / "report" / "REPORT.pdf"
BODY_FONT = "Times New Roman"
HEAD_FONT = "Times New Roman"
MONO_FONT = "Liberation Mono"
INK = RGBColor(0x1A, 0x1A, 0x1A)
MUTED = RGBColor(0x55, 0x55, 0x55)
ACCENT = RGBColor(0x5B, 0x3A, 0x1E)
TABLE_WIDTHS = [
(2.8, 13.8),
(11.6, 5.0),
(4.6, 12.0),
(3.4, 13.2),
(5.6, 11.0),
(7.6, 3.4, 3.4, 2.2),
(9.6, 7.0),
(10.0, 6.6),
(11.6, 5.0),
]
TABLE_CAPTIONS = [
"Гипотезы исследования и способ их проверки",
"Характеристики корпуса",
"Группы ручных стилометрических признаков",
"Варианты латентно-семантического анализа",
"Метрики качества и их назначение",
"Сравнение конфигураций классификаторов",
"Вклад групп признаков",
"Контроль содержания: исходный и маскированный текст",
"Качество классификации по длине реплики",
]
def run(cmd: list[str], cwd: Path | None = None) -> None:
print("+", " ".join(cmd))
subprocess.run(cmd, cwd=cwd, check=True)
def set_run_font(run, name: str, size: float | None = None, bold: bool | None = None,
italic: bool | None = None, color: RGBColor | None = None) -> None:
run.font.name = name
run._element.rPr.rFonts.set(qn("w:eastAsia"), name)
if size is not None:
run.font.size = Pt(size)
if bold is not None:
run.bold = bold
if italic is not None:
run.italic = italic
if color is not None:
run.font.color.rgb = color
def set_style_font(style, name: str, size: float, bold: bool = False,
color: RGBColor = INK, italic: bool | None = None) -> None:
style.font.name = name
style.font.size = Pt(size)
style.font.bold = bold
style.font.color.rgb = color
if italic is not None:
style.font.italic = italic
rpr = style.element.get_or_add_rPr()
rfonts = rpr.find(qn("w:rFonts"))
if rfonts is None:
rfonts = OxmlElement("w:rFonts")
rpr.append(rfonts)
for attr in ("w:ascii", "w:hAnsi", "w:eastAsia", "w:cs"):
rfonts.set(qn(attr), name)
def set_paragraph_spacing(style, before: float = 0, after: float = 0,
line: float = 1.5) -> None:
pf = style.paragraph_format
pf.space_before = Pt(before)
pf.space_after = Pt(after)
pf.line_spacing = line
def shade(cell, fill: str) -> None:
tc_pr = cell._tc.get_or_add_tcPr()
shd = OxmlElement("w:shd")
shd.set(qn("w:val"), "clear")
shd.set(qn("w:color"), "auto")
shd.set(qn("w:fill"), fill)
tc_pr.append(shd)
def set_cell_margins(table, top=60, bottom=60, left=90, right=90) -> None:
tbl_pr = table._tbl.tblPr
mar = OxmlElement("w:tblCellMar")
for edge, value in (("top", top), ("left", left), ("bottom", bottom), ("right", right)):
el = OxmlElement(f"w:{edge}")
el.set(qn("w:w"), str(value))
el.set(qn("w:type"), "dxa")
mar.append(el)
tbl_pr.append(mar)
def set_repeat_header(row) -> None:
tr_pr = row._tr.get_or_add_trPr()
el = OxmlElement("w:tblHeader")
el.set(qn("w:val"), "true")
tr_pr.append(el)
cant_split = OxmlElement("w:cantSplit")
cant_split.set(qn("w:val"), "true")
tr_pr.append(cant_split)
def set_table_borders(table, color: str = "B8A88F", size: int = 4) -> None:
tbl_pr = table._tbl.tblPr
borders = OxmlElement("w:tblBorders")
for edge in ("top", "left", "bottom", "right", "insideH", "insideV"):
el = OxmlElement(f"w:{edge}")
el.set(qn("w:val"), "single")
el.set(qn("w:sz"), str(size))
el.set(qn("w:space"), "0")
el.set(qn("w:color"), color)
borders.append(el)
tbl_pr.append(borders)
def add_page_field(paragraph, total: bool = False) -> None:
run = paragraph.add_run()
fld_char_begin = OxmlElement("w:fldChar")
fld_char_begin.set(qn("w:fldCharType"), "begin")
instr_text = OxmlElement("w:instrText")
instr_text.set(qn("xml:space"), "preserve")
instr_text.text = "NUMPAGES" if total else "PAGE"
fld_char_end = OxmlElement("w:fldChar")
fld_char_end.set(qn("w:fldCharType"), "end")
run._r.append(fld_char_begin)
run._r.append(instr_text)
run._r.append(fld_char_end)
def add_bottom_border(paragraph, color: str = "5B3A1E", size: int = 6) -> None:
p_pr = paragraph._p.get_or_add_pPr()
borders = OxmlElement("w:pBdr")
bottom = OxmlElement("w:bottom")
bottom.set(qn("w:val"), "single")
bottom.set(qn("w:sz"), str(size))
bottom.set(qn("w:space"), "4")
bottom.set(qn("w:color"), color)
borders.append(bottom)
p_pr.append(borders)
def add_left_border(paragraph, color: str = "B8A88F", size: int = 18) -> None:
p_pr = paragraph._p.get_or_add_pPr()
borders = OxmlElement("w:pBdr")
left = OxmlElement("w:left")
left.set(qn("w:val"), "single")
left.set(qn("w:sz"), str(size))
left.set(qn("w:space"), "8")
left.set(qn("w:color"), color)
borders.append(left)
p_pr.append(borders)
def add_page_break(paragraph) -> None:
run = paragraph.add_run()
run.add_break(WD_BREAK.PAGE)
def style_document(doc) -> None:
section = doc.sections[0]
section.page_width = Cm(21.0)
section.page_height = Cm(29.7)
section.left_margin = Cm(2.6)
section.right_margin = Cm(1.8)
section.top_margin = Cm(2.0)
section.bottom_margin = Cm(2.0)
section.header_distance = Cm(1.0)
section.footer_distance = Cm(1.0)
styles = doc.styles
normal = styles["Normal"]
set_style_font(normal, BODY_FONT, 12)
normal.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.JUSTIFY
normal.paragraph_format.first_line_indent = Cm(1.25)
normal.paragraph_format.line_spacing = 1.5
normal.paragraph_format.space_before = Pt(0)
normal.paragraph_format.space_after = Pt(0)
for name in ("Body Text", "First Paragraph"):
st = styles[name]
set_style_font(st, BODY_FONT, 12)
st.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.JUSTIFY
st.paragraph_format.first_line_indent = Cm(1.25)
st.paragraph_format.line_spacing = 1.5
st.paragraph_format.space_before = Pt(0)
st.paragraph_format.space_after = Pt(6)
styles["First Paragraph"].paragraph_format.space_before = Pt(0)
title = styles["Title"]
set_style_font(title, HEAD_FONT, 24, bold=True, color=ACCENT)
title.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.CENTER
title.paragraph_format.space_before = Pt(150)
title.paragraph_format.space_after = Pt(12)
title.paragraph_format.line_spacing = 1.1
subtitle = styles["Subtitle"]
set_style_font(subtitle, HEAD_FONT, 15, bold=False, color=INK)
subtitle.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.CENTER
subtitle.paragraph_format.space_before = Pt(0)
subtitle.paragraph_format.space_after = Pt(28)
subtitle.paragraph_format.line_spacing = 1.2
author = styles["Author"]
set_style_font(author, BODY_FONT, 13, color=INK)
author.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.CENTER
author.paragraph_format.space_before = Pt(0)
author.paragraph_format.space_after = Pt(6)
date = styles["Date"]
set_style_font(date, BODY_FONT, 13, color=MUTED)
date.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.CENTER
date.paragraph_format.space_before = Pt(0)
date.paragraph_format.space_after = Pt(0)
for name, size, before, after in (
("Heading 1", 17, 0, 10),
("Heading 2", 14, 14, 6),
("Heading 3", 12.5, 12, 4),
):
st = styles[name]
set_style_font(st, HEAD_FONT, size, bold=True, color=INK)
st.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.LEFT
st.paragraph_format.first_line_indent = Cm(0)
st.paragraph_format.space_before = Pt(before)
st.paragraph_format.space_after = Pt(after)
st.paragraph_format.line_spacing = 1.15
st.paragraph_format.keep_with_next = True
styles["Heading 1"].paragraph_format.page_break_before = False
styles["Heading 1"].paragraph_format.space_before = Pt(20)
block = styles["Block Text"]
set_style_font(block, BODY_FONT, 11.5, italic=False)
block.paragraph_format.left_indent = Cm(1.0)
block.paragraph_format.right_indent = Cm(0.5)
block.paragraph_format.first_line_indent = Cm(0)
block.paragraph_format.line_spacing = 1.3
block.paragraph_format.space_before = Pt(6)
block.paragraph_format.space_after = Pt(8)
block.paragraph_format.keep_together = True
source = styles["Source Code"]
set_style_font(source, MONO_FONT, 9)
source.paragraph_format.left_indent = Cm(0.6)
source.paragraph_format.first_line_indent = Cm(0)
source.paragraph_format.line_spacing = 1.0
source.paragraph_format.space_before = Pt(4)
source.paragraph_format.space_after = Pt(4)
source.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.LEFT
source.paragraph_format.keep_together = True
try:
verbatim = styles["Verbatim Char"]
set_style_font(verbatim, MONO_FONT, 9.5)
except KeyError:
pass
for name in ("Image Caption", "Caption"):
st = styles[name]
set_style_font(st, BODY_FONT, 10, italic=True, color=MUTED)
st.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.CENTER
st.paragraph_format.first_line_indent = Cm(0)
st.paragraph_format.line_spacing = 1.15
st.paragraph_format.space_before = Pt(4)
st.paragraph_format.space_after = Pt(12)
for name in ("Table Caption",):
st = styles[name]
set_style_font(st, BODY_FONT, 10, italic=False, color=INK)
st.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.CENTER
st.paragraph_format.first_line_indent = Cm(0)
st.paragraph_format.line_spacing = 1.15
st.paragraph_format.space_before = Pt(4)
st.paragraph_format.space_after = Pt(12)
st.paragraph_format.alignment = WD_ALIGN_PARAGRAPH.LEFT
st.paragraph_format.first_line_indent = Cm(0)
st.paragraph_format.line_spacing = 1.15
st.paragraph_format.space_before = Pt(10)
st.paragraph_format.space_after = Pt(4)
st.paragraph_format.keep_with_next = True
def add_header_footer(doc) -> None:
section = doc.sections[0]
header = section.header
header.is_linked_to_previous = False
hp = header.paragraphs[0]
hp.alignment = WD_ALIGN_PARAGRAPH.RIGHT
hp.paragraph_format.space_after = Pt(0)
run = hp.add_run("Стилометрия реплик Stardew Valley · отчёт по лабораторной работе №1")
set_run_font(run, BODY_FONT, 8.5, color=MUTED)
add_bottom_border(hp, color="D8CBB8", size=4)
footer = section.footer
footer.is_linked_to_previous = False
fp = footer.paragraphs[0]
fp.alignment = WD_ALIGN_PARAGRAPH.CENTER
fp.paragraph_format.space_before = Pt(0)
fp.paragraph_format.space_after = Pt(0)
run = fp.add_run("Стр. ")
set_run_font(run, BODY_FONT, 9, color=MUTED)
add_page_field(fp)
for r in fp.runs:
set_run_font(r, BODY_FONT, 9, color=MUTED)
run = fp.add_run(" из ")
set_run_font(run, BODY_FONT, 9, color=MUTED)
add_page_field(fp, total=True)
for r in fp.runs:
set_run_font(r, BODY_FONT, 9, color=MUTED)
def add_title_page_notes(doc) -> None:
"""Добавляет подпись направления под титульным блоком и разрыв страницы."""
date_para = None
for p in doc.paragraphs:
if p.style.name == "Date":
date_para = p
break
if date_para is None:
return
anchor = date_para
note = anchor.insert_paragraph_before()
note.alignment = WD_ALIGN_PARAGRAPH.CENTER
note.paragraph_format.first_line_indent = Cm(0)
note.paragraph_format.space_before = Pt(18)
run = note.add_run("Направление: автоматический анализ текстов\n")
set_run_font(run, BODY_FONT, 11, color=MUTED)
run = note.add_run("Корпус: Stardew Valley · 2005 реплик · 34 персонажа")
set_run_font(run, BODY_FONT, 11, color=MUTED)
spacer = doc.add_paragraph()
spacer.paragraph_format.space_after = Pt(0)
add_page_break(spacer)
date_para._p.addnext(spacer._p)
def number_figures(doc) -> None:
n = 0
for p in doc.paragraphs:
if p.style.name == "Image Caption" and p.text.strip():
n += 1
p.text = f"Рисунок {n}. {p.text.strip()}"
for run in p.runs:
set_run_font(run, BODY_FONT, 10, italic=True, color=MUTED)
def add_table_captions(doc) -> None:
for idx, table in enumerate(doc.tables):
if idx >= len(TABLE_CAPTIONS):
break
para = doc.add_paragraph(style="Table Caption")
table._tbl.addprevious(para._p)
run = para.add_run(f"Таблица {idx + 1}. {TABLE_CAPTIONS[idx]}")
set_run_font(run, BODY_FONT, 10, bold=True, color=INK)
para.paragraph_format.first_line_indent = Cm(0)
para.paragraph_format.space_before = Pt(10)
para.paragraph_format.space_after = Pt(4)
para.paragraph_format.keep_with_next = True
def set_column_widths(table, widths_cm: tuple[float, ...]) -> None:
table.autofit = False
tbl_pr = table._tbl.tblPr
layout = OxmlElement("w:tblLayout")
layout.set(qn("w:type"), "fixed")
tbl_pr.append(layout)
grid = table._tbl.tblGrid
for idx, width in enumerate(widths_cm):
if idx < len(grid.gridCol_lst):
grid.gridCol_lst[idx].set(qn("w:w"), str(int(Cm(width).twips)))
for row in table.rows:
for idx, width in enumerate(widths_cm):
if idx < len(row.cells):
row.cells[idx].width = Cm(width)
def style_tables(doc) -> None:
for t_idx, table in enumerate(doc.tables):
table.alignment = WD_TABLE_ALIGNMENT.CENTER
table.autofit = True
set_table_borders(table)
set_cell_margins(table)
for r_idx, row in enumerate(table.rows):
if r_idx == 0:
set_repeat_header(row)
else:
tr_pr = row._tr.get_or_add_trPr()
cant_split = OxmlElement("w:cantSplit")
cant_split.set(qn("w:val"), "true")
tr_pr.append(cant_split)
for cell in row.cells:
for p in cell.paragraphs:
p.paragraph_format.first_line_indent = Cm(0)
p.paragraph_format.space_before = Pt(1)
p.paragraph_format.space_after = Pt(1)
p.paragraph_format.line_spacing = 1.0
if r_idx < len(table.rows) - 1:
p.paragraph_format.keep_with_next = True
for run in p.runs:
set_run_font(
run, BODY_FONT, 9.5 if len(table.columns) >= 5 else 10,
bold=(r_idx == 0),
)
if r_idx == 0:
shade(cell, "F1E8D8")
# Ширина таблицы по ширине полей.
tbl_pr = table._tbl.tblPr
tbl_w = OxmlElement("w:tblW")
tbl_w.set(qn("w:type"), "pct")
tbl_w.set(qn("w:w"), "100%")
tbl_pr.append(tbl_w)
if t_idx < len(TABLE_WIDTHS) and len(TABLE_WIDTHS[t_idx]) == len(table.columns):
set_column_widths(table, TABLE_WIDTHS[t_idx])
# Небольшой отступ между таблицей и следующим абзацем.
spacer = doc.add_paragraph()
spacer.paragraph_format.space_before = Pt(0)
spacer.paragraph_format.space_after = Pt(0)
spacer.paragraph_format.line_spacing = 1.0
sr = spacer.add_run("")
sr.font.size = Pt(4)
table._tbl.addnext(spacer._p)
def style_blockquotes(doc) -> None:
for p in doc.paragraphs:
if p.style.name == "Block Text":
add_left_border(p)
def add_math_font_note(doc) -> None:
"""OMML-формулы наследуют Cambria Math; явно задаём её для совместимости."""
for r in doc.element.body.iter(qn("m:r")):
rpr = r.find(qn("w:rPr"))
if rpr is None:
rpr = OxmlElement("w:rPr")
r.insert(0, rpr)
rfonts = rpr.find(qn("w:rFonts"))
if rfonts is None:
rfonts = OxmlElement("w:rFonts")
rpr.append(rfonts)
rfonts.set(qn("w:ascii"), "Cambria Math")
rfonts.set(qn("w:hAnsi"), "Cambria Math")
def build_docx() -> None:
BUILD.mkdir(parents=True, exist_ok=True)
raw = BUILD / "REPORT_raw.docx"
run([
"pandoc", str(REPORT_MD),
"-f", "markdown+tex_math_dollars+tex_math_single_backslash+raw_tex",
"-t", "docx",
"-o", str(raw),
f"--resource-path={REPORT_MD.parent}",
])
doc = docx.Document(raw)
style_document(doc)
section = doc.sections[0]
section.different_first_page_header_footer = True
add_header_footer(doc)
# Титульный лист без колонтитулов.
for part in (section.first_page_header, section.first_page_footer):
part.is_linked_to_previous = False
for paragraph in part.paragraphs:
paragraph.text = ""
add_title_page_notes(doc)
number_figures(doc)
add_table_captions(doc)
style_tables(doc)
style_blockquotes(doc)
add_math_font_note(doc)
doc.core_properties.title = "Отчёт по лабораторной работе №1"
doc.core_properties.subject = "Классификация текстов на основе стилометрических характеристик"
doc.core_properties.author = "А. А. Аладьина, В. Н. Копнин, М. В. Савостиков"
doc.core_properties.language = "ru-RU"
doc.save(DOCX_OUT)
print(f"DOCX: {DOCX_OUT}")
def build_pdf() -> None:
if shutil.which("libreoffice") is None:
raise SystemExit("libreoffice не найден: PDF не собран")
run([
"libreoffice", "--headless", "--convert-to", "pdf",
"--outdir", str(PDF_OUT.parent), str(DOCX_OUT),
])
produced = PDF_OUT.parent / (DOCX_OUT.stem + ".pdf")
if produced != PDF_OUT and produced.exists():
produced.replace(PDF_OUT)
print(f"PDF: {PDF_OUT}")
def check_pdf() -> None:
try:
import fitz # PyMuPDF
except ImportError:
print("PyMuPDF не установлен: пропускаю проверку PDF")
return
doc = fitz.open(PDF_OUT)
print(f"PDF страниц: {doc.page_count}")
text = "\n".join(page.get_text() for page in doc)
required = [
"Аннотация", "Постановка задачи", "Корпус", "Стилометрические признаки",
"Классификаторы", "Схема верификации", "Результаты", "Обсуждение",
"Ограничения", "Выводы", "Воспроизводимость",
]
missing = [item for item in required if item.lower() not in text.lower()]
if missing:
print("Не найдены разделы:", ", ".join(missing))
raise SystemExit(1)
print("Все основные разделы найдены в PDF")
def main() -> int:
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("--skip-pdf", action="store_true", help="только DOCX")
args = parser.parse_args()
if not REPORT_MD.exists():
raise SystemExit(f"нет файла {REPORT_MD}")
build_docx()
if not args.skip_pdf:
build_pdf()
check_pdf()
return 0
if __name__ == "__main__":
sys.exit(main())