ThomasMoreAI/legal-skills-open / br/litigation/skills/juriscan-brunobracaioli
juriscan-brunobracaioli
VerifiedAnálise forense de processos judiciais brasileiros. Extrai, classifica e cruza peças processuais, detecta contradições, calcula prazos CPC e gera relatórios de risco com exportação para Obsidian. Use quando mencionar: processo judicial, petição, sentença, acórdão, contradições, prazos, timeline, análise forense, Obsidian.
tests/golden/_pdf_builder.py
View on GitHub"""Deterministic stdlib-only PDF builder for golden fixtures.
Emits a minimal PDF 1.4 file from plain text. Helvetica + WinAnsiEncoding —
covers all Portuguese accents we need for synthetic legal filings.
Usage:
from _pdf_builder import build_pdf
build_pdf(text, output_path)
Or as CLI:
python tests/golden/_pdf_builder.py source.txt output.pdf
"""
from __future__ import annotations
import sys
import textwrap
from pathlib import Path
PAGE_WIDTH = 612 # Letter size in points
PAGE_HEIGHT = 792
MARGIN_X = 50
MARGIN_TOP = 750
FONT_SIZE = 10
LINE_HEIGHT = 12
LINES_PER_PAGE = 60
CHARS_PER_LINE = 90
def _escape_pdf_string(s: str) -> bytes:
"""Escape a string for inclusion in a PDF (...) literal and encode to Latin-1.
WinAnsiEncoding covers Latin-1 range, so Portuguese accents survive.
Characters outside Latin-1 are replaced with '?'.
"""
# Escape backslash, parens
s = s.replace("\\", "\\\\").replace("(", "\\(").replace(")", "\\)")
return s.encode("latin-1", errors="replace")
def _wrap_text(text: str) -> list[str]:
"""Wrap text to CHARS_PER_LINE, preserving blank lines."""
out: list[str] = []
for raw_line in text.splitlines():
if not raw_line.strip():
out.append("")
continue
wrapped = textwrap.wrap(
raw_line,
width=CHARS_PER_LINE,
break_long_words=True,
break_on_hyphens=False,
) or [""]
out.extend(wrapped)
return out
def _paginate(lines: list[str]) -> list[list[str]]:
"""Split wrapped lines into pages.
A literal form-feed (\\f) character forces a page break.
"""
pages: list[list[str]] = []
current: list[str] = []
for line in lines:
if line == "\f":
if current:
pages.append(current)
current = []
continue
if len(current) >= LINES_PER_PAGE:
pages.append(current)
current = []
current.append(line)
if current:
pages.append(current)
if not pages:
pages = [[""]]
return pages
def _build_content_stream(page_lines: list[str]) -> bytes:
"""Emit a PDF content stream that draws the lines of one page."""
out = bytearray()
out += b"BT\n"
out += f"/F1 {FONT_SIZE} Tf\n".encode("ascii")
out += f"{LINE_HEIGHT} TL\n".encode("ascii")
out += f"{MARGIN_X} {MARGIN_TOP} Td\n".encode("ascii")
for line in page_lines:
out += b"("
out += _escape_pdf_string(line)
out += b") Tj T*\n"
out += b"ET\n"
return bytes(out)
def build_pdf(text: str, output: Path | str) -> Path:
"""Build a multi-page PDF from the given text and write to `output`.
Returns the output path.
"""
# Pre-process text: respect explicit \f as page break marker
raw_lines: list[str] = []
for raw in text.splitlines():
if raw.strip() == "\f" or raw == "\x0c":
raw_lines.append("\f")
else:
raw_lines.append(raw)
wrapped = _wrap_text("\n".join(raw_lines).replace("\n\f\n", "\n\f\n"))
# After wrapping, form-feed lines may have been stripped; re-insert markers
final_lines: list[str] = []
for line in wrapped:
final_lines.append(line)
pages = _paginate(final_lines)
# Object numbering:
# 1 = Catalog
# 2 = Pages
# 3 = Font
# 4..(4+N-1) = Page objects
# (4+N)..(4+2N-1) = Content streams
num_pages = len(pages)
page_obj_ids = list(range(4, 4 + num_pages))
content_obj_ids = list(range(4 + num_pages, 4 + 2 * num_pages))
body = bytearray()
offsets: list[int] = []
header = b"%PDF-1.4\n%\xe2\xe3\xcf\xd3\n"
body += header
def write_obj(obj_id: int, contents: bytes) -> None:
offsets.append(len(body))
body.extend(f"{obj_id} 0 obj\n".encode("ascii"))
body.extend(contents)
if not contents.endswith(b"\n"):
body.extend(b"\n")
body.extend(b"endobj\n")
# 1: Catalog
write_obj(1, b"<< /Type /Catalog /Pages 2 0 R >>\n")
# 2: Pages
kids = " ".join(f"{pid} 0 R" for pid in page_obj_ids)
pages_obj = f"<< /Type /Pages /Kids [{kids}] /Count {num_pages} >>\n".encode("ascii")
write_obj(2, pages_obj)
# 3: Font
font_obj = b"<< /Type /Font /Subtype /Type1 /BaseFont /Helvetica /Encoding /WinAnsiEncoding >>\n"
write_obj(3, font_obj)
# 4..: Page objects
for page_idx, page_obj_id in enumerate(page_obj_ids):
content_id = content_obj_ids[page_idx]
page_dict = (
f"<< /Type /Page /Parent 2 0 R "
f"/MediaBox [0 0 {PAGE_WIDTH} {PAGE_HEIGHT}] "
f"/Resources << /Font << /F1 3 0 R >> >> "
f"/Contents {content_id} 0 R >>\n"
).encode("ascii")
write_obj(page_obj_id, page_dict)
# Content streams
for page_idx, content_id in enumerate(content_obj_ids):
stream_bytes = _build_content_stream(pages[page_idx])
stream_obj = bytearray()
stream_obj += f"<< /Length {len(stream_bytes)} >>\n".encode("ascii")
stream_obj += b"stream\n"
stream_obj += stream_bytes
stream_obj += b"endstream\n"
write_obj(content_id, bytes(stream_obj))
# Xref
xref_offset = len(body)
total_objs = 3 + 2 * num_pages # 1,2,3 + pages + contents
body += f"xref\n0 {total_objs + 1}\n".encode("ascii")
body += b"0000000000 65535 f \n"
for off in offsets:
body += f"{off:010d} 00000 n \n".encode("ascii")
# Trailer
body += f"trailer\n<< /Size {total_objs + 1} /Root 1 0 R >>\nstartxref\n{xref_offset}\n%%EOF\n".encode("ascii")
output_path = Path(output)
output_path.parent.mkdir(parents=True, exist_ok=True)
output_path.write_bytes(bytes(body))
return output_path
def main() -> int:
if len(sys.argv) != 3:
print(f"usage: {sys.argv[0]} source.txt output.pdf", file=sys.stderr)
return 2
src = Path(sys.argv[1])
dst = Path(sys.argv[2])
text = src.read_text(encoding="utf-8")
out = build_pdf(text, dst)
print(f"wrote {out} ({out.stat().st_size} bytes)")
return 0
if __name__ == "__main__":
raise SystemExit(main())
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |