vibepaper 0.1.0__py3-none-any.whl

vibepaper/__init__.py

@@ -0,0 +1 @@
+version = "0.1.0"

vibepaper/build.py

@@ -0,0 +1,294 @@
+"""Orchestration: Jinja2 pass → table pass → pandoc."""
+
+import json
+import logging
+import re
+import shutil
+import subprocess
+import sys
+import zipfile
+from datetime import date
+from pathlib import Path
+
+from lxml import etree
+
+log = logging.getLogger(__name__)
+
+try:
+    import tomllib
+except ImportError:
+    import tomli as tomllib  # type: ignore
+
+from .render import load_facts, make_jinja_env, render_file as render_jinja, sanity_check
+from .tables import process_file as render_tables
+
+W = "http://schemas.openxmlformats.org/wordprocessingml/2006/main"
+
+
+def load_config(config_path: Path) -> dict:
+    """Read paper.toml and return config dict with defaults filled in.
+
+    Required: paper.sections (list of str).
+    Optional with defaults:
+      paper.supplementary   = []
+      paper.name            = parent directory name
+      paper.facts_dir   = "output/facts"
+      paper.output_dir      = "output"
+      paper.build_dir       = "build"
+      paper.reference_doc   = "paper/reference.docx"  (only used if file exists)
+    """
+    with open(config_path, "rb") as fh:
+        raw = tomllib.load(fh)
+
+    paper = raw.get("paper", {})
+
+    if "sections" not in paper:
+        raise ValueError(f"{config_path}: [paper] must contain a 'sections' list")
+
+    return {
+        "sections":      paper["sections"],
+        "supplementary": paper.get("supplementary", []),
+        "name":          paper.get("name", config_path.parent.name),
+        "facts_dir":     paper.get("facts_dir", "output/facts"),
+        "output_dir":    paper.get("output_dir", "output"),
+        "build_dir":     paper.get("build_dir", "build"),
+        "reference_doc": paper.get("reference_doc", "paper/reference.docx"),
+        "bibliography":  paper.get("bibliography", None),
+        "csl":           paper.get("csl", None),
+    }
+
+
+def minimal_config(sections: list[str], name: str | None = None) -> dict:
+    """Return a default config for use without a paper.toml."""
+    return {
+        "sections":      sections,
+        "supplementary": [],
+        "name":          name or Path.cwd().name,
+        "facts_dir":     "output/facts",
+        "output_dir":    "output",
+        "build_dir":     "build",
+        "reference_doc": "paper/reference.docx",
+        "bibliography":  None,
+        "csl":           None,
+    }
+
+
+def load_sections_file(path: Path) -> list[str]:
+    """Read a plain text file and return a list of section paths.
+
+    Format: one file path per line. Blank lines and lines starting with
+    '#' are ignored. Paths are relative to the sections file's directory.
+
+    Example sections.txt:
+        # Main manuscript
+        paper/abstract.md
+        paper/introduction.md
+        paper/methods.md
+        paper/results.md
+        paper/discussion.md
+    """
+    base = path.parent
+    sections = []
+    for line in path.read_text().splitlines():
+        stripped = line.strip()
+        if stripped and not stripped.startswith("#"):
+            p = Path(stripped)
+            if not p.is_absolute():
+                p = base / p
+            sections.append(str(p))
+    if not sections:
+        raise ValueError(f"{path}: no sections found")
+    return sections
+
+
+def load_json_data(source: str) -> dict:
+    """Load template context data from a JSON file path or inline JSON string.
+
+    Inline:    --data '{"n_samples": 100, "stats": {"pvalue": 0.001}}'
+    From file: --data results.json
+
+    Top-level keys become template namespaces:
+      {"n_samples": 100}           → {{ n_samples }}
+      {"stats": {"pvalue": 0.05}}  → {{ stats.pvalue }}
+    """
+    source = source.strip()
+    if source.startswith("{"):
+        data = json.loads(source)
+    else:
+        data = json.loads(Path(source).read_text())
+    if not isinstance(data, dict):
+        raise ValueError("JSON data must be a top-level object (dict)")
+    return data
+
+
+def strip_bookmarks(docx: Path):
+    """Remove Word bookmark elements that pandoc inserts for heading anchors.
+    These cause warnings when opening in Google Docs / older Word versions.
+    """
+    log.debug("Stripping bookmarks from %s", docx)
+    tmp = docx.with_suffix(".tmp.docx")
+    shutil.copy(docx, tmp)
+
+    with zipfile.ZipFile(tmp, "r") as zin, \
+         zipfile.ZipFile(docx, "w", zipfile.ZIP_DEFLATED) as zout:
+        for item in zin.infolist():
+            data = zin.read(item.filename)
+            if item.filename == "word/document.xml":
+                tree = etree.fromstring(data)
+                for tag in (f"{{{W}}}bookmarkStart", f"{{{W}}}bookmarkEnd"):
+                    for el in tree.iter(tag):
+                        el.getparent().remove(el)
+                data = etree.tostring(
+                    tree, xml_declaration=True, encoding="UTF-8", standalone=True
+                )
+            zout.writestr(item, data)
+
+    tmp.unlink()
+
+
+def pandoc_args(
+    sections: list[str],
+    output: Path,
+    reference_doc: Path | None,
+    bibliography: Path | None = None,
+    csl: Path | None = None,
+) -> list[str]:
+    """Build pandoc command."""
+    args = [
+        "pandoc",
+        *sections,
+        "--from", "markdown",
+        "--to", "docx",
+        "--output", str(output),
+        "--resource-path", ".",
+    ]
+    if reference_doc and reference_doc.exists():
+        args += ["--reference-doc", str(reference_doc)]
+        log.debug("Using reference doc: %s", reference_doc)
+    elif reference_doc:
+        log.warning("reference_doc %s not found; output will use default Word formatting.", reference_doc)
+    if bibliography and bibliography.exists():
+        args += ["--bibliography", str(bibliography), "--citeproc"]
+        log.debug("Using bibliography: %s", bibliography)
+    elif bibliography:
+        log.warning("bibliography %s not found; citations will not be resolved.", bibliography)
+    if csl and csl.exists():
+        args += ["--csl", str(csl)]
+        log.debug("Using CSL: %s", csl)
+    elif csl:
+        log.warning("CSL file %s not found; pandoc will use its default citation style.", csl)
+    return args
+
+
+def build_docx(
+    sections: list[str],
+    output: Path,
+    reference_doc: Path | None,
+    bibliography: Path | None = None,
+    csl: Path | None = None,
+):
+    """Run pandoc to produce a .docx file."""
+    output.parent.mkdir(parents=True, exist_ok=True)
+    log.debug("Building %s", output)
+    subprocess.run(pandoc_args(sections, output, reference_doc, bibliography, csl), check=True)
+
+
+def run_build(
+    config: dict,
+    project_root: Path,
+    output_dir: Path,
+    combined: bool = False,
+    extra_context: dict | None = None,
+):
+    """Full pipeline: Jinja2 pass → table pass → pandoc.
+
+    All paths are resolved relative to project_root; no os.chdir is used.
+
+    extra_context is merged into the Jinja2 context on top of any facts
+    CSVs. Nested dicts are deep-merged at the namespace level; scalar values
+    are set directly.
+    """
+    today = date.today()
+    paper_name = config["name"]
+
+    facts_dir = _resolve(config["facts_dir"], project_root)
+    build_dir     = _resolve(config["build_dir"], project_root)
+    build_jinja   = build_dir / "jinja" / "paper"
+    build_paper   = build_dir / "paper"
+
+    reference_doc_path = _resolve(config["reference_doc"], project_root)
+    reference_doc = reference_doc_path if reference_doc_path.exists() else None
+
+    bibliography = _resolve(config["bibliography"], project_root) if config.get("bibliography") else None
+    csl = _resolve(config["csl"], project_root) if config.get("csl") else None
+
+    all_sections = config["sections"] + config["supplementary"]
+
+    _warn_stub_sections(all_sections, project_root)
+
+    # Build Jinja2 context: facts CSVs, then merge any extra JSON data
+    context: dict = {}
+    if facts_dir.exists():
+        context.update(load_facts(facts_dir))
+    else:
+        log.warning("%s not found — template context will be empty.", facts_dir)
+    if extra_context:
+        for key, value in extra_context.items():
+            if key in context and isinstance(context[key], dict) and isinstance(value, dict):
+                context[key].update(value)
+            else:
+                context[key] = value
+
+    jinja_env = make_jinja_env(project_root)
+
+    jinja_warnings = []
+    for section in all_sections:
+        section_path = _resolve(section, project_root)
+        out = render_jinja(section_path, build_jinja, context, jinja_env)
+        jinja_warnings.extend(sanity_check(out))
+    if jinja_warnings:
+        print("WARNING: suspicious content in rendered output:", file=sys.stderr)
+        for w in jinja_warnings:
+            print(w, file=sys.stderr)
+
+    for section in all_sections:
+        section_name = Path(section).name
+        render_tables(build_jinja / section_name, build_paper, project_root)
+
+    def build_paths(sections):
+        return [str(build_paper / Path(s).name) for s in sections]
+
+    # Main manuscript
+    main_sections = config["sections"] + (config["supplementary"] if combined else [])
+    main_docx = output_dir / f"{paper_name}_{today}.docx"
+    build_docx(build_paths(main_sections), main_docx, reference_doc, bibliography, csl)
+    strip_bookmarks(main_docx)
+    print(f"Done: {main_docx}")
+
+    # Supplementary (separate unless --combined)
+    if not combined and config["supplementary"]:
+        supp_docx = output_dir / f"{paper_name}_supplementary_{today}.docx"
+        build_docx(build_paths(config["supplementary"]), supp_docx, reference_doc, bibliography, csl)
+        strip_bookmarks(supp_docx)
+        print(f"Done: {supp_docx}")
+
+
+def _resolve(path_str: str, project_root: Path) -> Path:
+    """Return absolute Path: if already absolute use as-is, else join with project_root."""
+    p = Path(path_str)
+    return p if p.is_absolute() else project_root / p
+
+
+def _warn_stub_sections(sections: list[str], project_root: Path):
+    """Warn if any section files appear to be stubs (only comments/headings)."""
+    threshold = 100
+    for section in sections:
+        path = _resolve(section, project_root)
+        if not path.exists():
+            log.warning("section not found: %s", path)
+            continue
+        text = path.read_text()
+        stripped = re.sub(r"<!--.*?-->", "", text, flags=re.DOTALL)
+        stripped = re.sub(r"^#+.*$", "", stripped, flags=re.MULTILINE)
+        if len(stripped.strip()) < threshold:
+            log.warning("%s appears to be a stub", path)

vibepaper/cli.py

@@ -0,0 +1,116 @@
+"""Command-line entry point for vibepaper."""
+
+import argparse
+import logging
+import sys
+from pathlib import Path
+
+from .build import load_config, load_json_data, load_sections_file, minimal_config, run_build
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        prog="vibepaper",
+        description=(
+            "Build a Word document from Markdown paper sections.\n\n"
+            "Three ways to specify which files to build:\n"
+            "  1. paper.toml config file (default, if present)\n"
+            "  2. --sections-file order.txt  (plain text list of .md files)\n"
+            "  3. vibepaper intro.md methods.md results.md  (positional args)"
+        ),
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+
+    # --- Input ---
+    parser.add_argument(
+        "sections", nargs="*", metavar="FILE.md",
+        help="Markdown files to build in order. Skips paper.toml sections list.",
+    )
+    parser.add_argument(
+        "--sections-file", metavar="FILE",
+        help="Plain text file listing .md paths in order, one per line.",
+    )
+    parser.add_argument(
+        "--config", default="paper.toml", metavar="FILE",
+        help="paper.toml config file (default: paper.toml in cwd).",
+    )
+
+    # --- Data ---
+    parser.add_argument(
+        "--data", metavar="JSON",
+        help=(
+            "Extra template context as a JSON file path or inline JSON dict. "
+            "Merged on top of facts CSVs. "
+            "Example: --data '{\"n\": 100}' or --data stats.json"
+        ),
+    )
+    parser.add_argument(
+        "--facts-dir", metavar="DIR",
+        help="Directory of 1-row facts CSVs (overrides paper.toml / default).",
+    )
+
+    # --- Output ---
+    parser.add_argument(
+        "--output-dir", metavar="DIR",
+        help="Directory for output .docx files.",
+    )
+    parser.add_argument(
+        "--name", metavar="NAME",
+        help="Output filename stem, e.g. 'my_paper' produces my_paper_2025-01-01.docx.",
+    )
+    parser.add_argument(
+        "--combined", action="store_true",
+        help="Append supplementary into the main document instead of a separate file.",
+    )
+    parser.add_argument(
+        "--verbose", "-v", action="store_true",
+        help="Print detailed progress (section rendering, pandoc invocation, etc.).",
+    )
+
+    args = parser.parse_args()
+    level = logging.DEBUG if args.verbose else logging.WARNING
+    logging.basicConfig(level=level, format="%(message)s")
+
+    # --- Determine sections and project_root ---
+    if args.sections:
+        # Mode: positional file arguments
+        sections = [str(Path(s).resolve()) for s in args.sections]
+        project_root = Path.cwd()
+        config = minimal_config(sections, name=args.name)
+
+    elif args.sections_file:
+        # Mode: plain text sections file
+        sections_path = Path(args.sections_file).resolve()
+        sections = load_sections_file(sections_path)
+        project_root = Path.cwd()
+        config = minimal_config(sections, name=args.name)
+
+    else:
+        # Mode: paper.toml
+        config_path = Path(args.config).resolve()
+        if not config_path.exists():
+            print(
+                f"error: {config_path} not found.\n"
+                "Pass .md files directly, use --sections-file, or create a paper.toml.",
+                file=sys.stderr,
+            )
+            sys.exit(1)
+        config = load_config(config_path)
+        project_root = config_path.parent
+        if args.name:
+            config["name"] = args.name
+
+    # Apply CLI overrides that work across all modes
+    if args.facts_dir:
+        config["facts_dir"] = args.facts_dir
+
+    output_dir = (
+        Path(args.output_dir).resolve()
+        if args.output_dir
+        else project_root / config["output_dir"]
+    )
+
+    extra_context = load_json_data(args.data) if args.data else None
+
+    run_build(config, project_root, output_dir, combined=args.combined,
+              extra_context=extra_context)

vibepaper/render.py

@@ -0,0 +1,111 @@
+"""Jinja2 templating pass for paper markdown files.
+
+Replaces {{ namespace.field | filter }} references with values loaded from
+1-row facts CSVs in output/facts/.  Runs before tables.py so
+that inline prose values are resolved before table directives are expanded.
+"""
+
+import logging
+import re
+from pathlib import Path
+
+import pandas as pd
+from jinja2 import Environment, FileSystemLoader, StrictUndefined, UndefinedError
+
+log = logging.getLogger(__name__)
+
+# Regexes that indicate a render problem in the output file.
+# Use word boundaries so "nan" doesn't match "annotated", "None" doesn't
+# match prose like "None of the…" — only standalone tokens flag.
+_SANITY_RE = re.compile(
+    r"\bnan\b"            # pandas NaN rendered as "nan"
+    r"|\bundefined\b"     # Jinja undefined leak
+    r"|\{\{"              # unresolved template tag
+)
+
+
+def load_facts(facts_dir: Path) -> dict:
+    """Load all 1-row CSVs from facts_dir into a namespace dict.
+
+    Each file 'foo_bar.csv' becomes context['foo_bar'] = {col: value, ...}.
+    Raises ValueError if any CSV has more than one data row.
+    """
+    context = {}
+    for csv_path in sorted(facts_dir.glob("*.csv")):
+        df = pd.read_csv(csv_path)
+        if len(df) != 1:
+            raise ValueError(
+                f"{csv_path}: facts CSVs must have exactly 1 row, got {len(df)}"
+            )
+        namespace = csv_path.stem
+        context[namespace] = df.iloc[0].to_dict()
+        log.debug("Loaded facts: %s (%d fields)", namespace, len(context[namespace]))
+    return context
+
+
+def make_jinja_env(project_root: Path) -> Environment:
+    """Create a Jinja2 environment with custom filters and strict undefined."""
+    env = Environment(
+        loader=FileSystemLoader(str(project_root)),
+        undefined=StrictUndefined,
+        keep_trailing_newline=True,
+    )
+
+    def filter_commas(value) -> str:
+        """Integer with thousands separator: 254129 → '254,129'"""
+        return f"{int(float(value)):,}"
+
+    def filter_pct(value, decimals=1) -> str:
+        """Format a pre-computed percentage: 52.2 → '52.2%'"""
+        return f"{float(value):.{decimals}f}%"
+
+    def filter_fold(value, decimals=1) -> str:
+        """Fold change: 2.003 → '2.0-fold'"""
+        return f"{float(value):.{decimals}f}-fold"
+
+    def filter_dp(value, decimals=1) -> str:
+        """Decimal places only, no suffix: 9.177 → '9.2'"""
+        return f"{float(value):.{decimals}f}"
+
+    def filter_fmt(value, spec) -> str:
+        """Escape hatch: raw Python format spec. {{ v | fmt('+,.0f') }}"""
+        return format(float(value), spec)
+
+    env.filters["commas"] = filter_commas
+    env.filters["pct"] = filter_pct
+    env.filters["fold"] = filter_fold
+    env.filters["dp"] = filter_dp
+    env.filters["fmt"] = filter_fmt
+
+    return env
+
+
+def render_file(
+    input_path: Path,
+    build_dir: Path,
+    context: dict,
+    env: Environment,
+) -> Path:
+    """Render Jinja2 templates in a single markdown file and write output."""
+    content = input_path.read_text()
+    try:
+        rendered = env.from_string(content).render(**context)
+    except UndefinedError as exc:
+        raise RuntimeError(f"Template error in {input_path}: {exc}") from exc
+
+    output_path = build_dir / input_path.name
+    build_dir.mkdir(parents=True, exist_ok=True)
+    output_path.write_text(rendered)
+    log.debug("Rendered %s → %s", input_path, output_path)
+    return output_path
+
+
+def sanity_check(path: Path) -> list:
+    """Return a list of warning strings for suspicious content in a rendered file."""
+    warnings = []
+    content = path.read_text()
+    for i, line in enumerate(content.splitlines(), start=1):
+        m = _SANITY_RE.search(line)
+        if m:
+            warnings.append(f"  {path}:{i}: found '{m.group()}'")
+    return warnings

vibepaper/tables.py

@@ -0,0 +1,182 @@
+"""Preprocess markdown files: replace include-csv directives with rendered tables.
+
+Scans markdown for <!-- include-csv: path/to/file.csv ... --> directives,
+loads the referenced CSV/TSV, applies formatting options, and writes rendered
+markdown to a build directory.
+
+Directive syntax (all options are optional):
+
+    <!-- include-csv: output/exome_sizes.csv
+      columns: [source, feature, size_bp]
+      rename:
+        source: Source
+        size_bp: Size (bp)
+      format:
+        size_bp: ",d"
+      align: right
+      max_rows: 50
+      sort: [-size_bp]
+      filter: "feature == 'exon'"
+      na_rep: "—"
+    -->
+"""
+
+import logging
+import re
+from pathlib import Path
+
+import pandas as pd
+import yaml
+
+log = logging.getLogger(__name__)
+
+DIRECTIVE_RE = re.compile(
+    r"<!-- include-csv:\s*(\S+)(.*?)-->",
+    re.DOTALL,
+)
+
+
+def parse_options(yaml_text: str) -> dict:
+    """Parse YAML options from the directive body."""
+    text = yaml_text.strip()
+    if not text:
+        return {}
+    try:
+        opts = yaml.safe_load(text)
+    except yaml.YAMLError as exc:
+        log.warning("Failed to parse YAML options: %s", exc)
+        return {}
+    return opts if isinstance(opts, dict) else {}
+
+
+def load_table(path: Path, options: dict) -> pd.DataFrame:
+    """Load a CSV/TSV file and apply filter, sort, and row-limit options."""
+    sep = options.get("sep", "\t" if path.suffix in (".tsv",) else ",")
+    df = pd.read_csv(path, sep=sep)
+
+    if "filter" in options:
+        df = df.query(options["filter"])
+
+    if "sort" in options:
+        sort_cols = options["sort"]
+        if isinstance(sort_cols, str):
+            sort_cols = [sort_cols]
+        ascending = []
+        clean_cols = []
+        for col in sort_cols:
+            if col.startswith("-"):
+                ascending.append(False)
+                clean_cols.append(col[1:])
+            else:
+                ascending.append(True)
+                clean_cols.append(col)
+        df = df.sort_values(clean_cols, ascending=ascending)
+
+    if "max_rows" in options:
+        df = df.head(options["max_rows"])
+
+    return df
+
+
+def format_cell(value, fmt: str, na_rep: str) -> str:
+    """Format a single cell value using a Python format spec."""
+    if pd.isna(value):
+        return na_rep
+    try:
+        return format(value, fmt)
+    except (ValueError, TypeError):
+        return str(value)
+
+
+def render_markdown_table(df: pd.DataFrame, options: dict) -> str:
+    """Render a DataFrame as a markdown table string."""
+    na_rep = options.get("na_rep", "—")
+
+    # Select and order columns
+    if "columns" in options:
+        df = df[[c for c in options["columns"] if c in df.columns]]
+
+    # Rename columns for display
+    renames = options.get("rename", {})
+    display_cols = [renames.get(c, c) for c in df.columns]
+
+    # Determine alignment per column
+    align_opt = options.get("align", None)
+    alignments = []
+    for col in df.columns:
+        if isinstance(align_opt, dict):
+            a = align_opt.get(col, align_opt.get(renames.get(col, col), None))
+        elif isinstance(align_opt, str):
+            a = align_opt
+        else:
+            a = None
+        if a is None:
+            a = "right" if pd.api.types.is_numeric_dtype(df[col]) else "left"
+        alignments.append(a)
+
+    # Format cells
+    fmt_specs = options.get("format", {})
+    rows = []
+    for _, row in df.iterrows():
+        cells = []
+        for col in df.columns:
+            fmt = fmt_specs.get(col)
+            val = row[col]
+            if fmt:
+                cells.append(format_cell(val, fmt, na_rep))
+            elif pd.isna(val):
+                cells.append(na_rep)
+            else:
+                cells.append(str(val))
+        rows.append(cells)
+
+    # Build alignment markers
+    markers = []
+    for a in alignments:
+        if a == "right":
+            markers.append("---:")
+        elif a == "center":
+            markers.append(":---:")
+        else:
+            markers.append("---")
+
+    # Assemble table
+    lines = []
+    lines.append("| " + " | ".join(display_cols) + " |")
+    lines.append("| " + " | ".join(markers) + " |")
+    for cells in rows:
+        lines.append("| " + " | ".join(cells) + " |")
+
+    return "\n".join(lines)
+
+
+def process_content(content: str, project_root: Path) -> str:
+    """Replace all include-csv directives in markdown content."""
+    def replace_directive(match):
+        rel_path = match.group(1)
+        yaml_text = match.group(2)
+        csv_path = project_root / rel_path
+
+        if not csv_path.exists():
+            log.warning("CSV file not found: %s", csv_path)
+            return f"*[Table data not found: {rel_path}]*"
+
+        options = parse_options(yaml_text)
+        df = load_table(csv_path, options)
+        table = render_markdown_table(df, options)
+        log.debug("Rendered %s (%d rows, %d cols)", rel_path, len(df), len(df.columns))
+        return table
+
+    return DIRECTIVE_RE.sub(replace_directive, content)
+
+
+def process_file(input_path: Path, build_dir: Path, project_root: Path) -> Path:
+    """Process a single markdown file and write rendered output."""
+    content = input_path.read_text()
+    rendered = process_content(content, project_root)
+
+    output_path = build_dir / input_path.name
+    build_dir.mkdir(parents=True, exist_ok=True)
+    output_path.write_text(rendered)
+    log.debug("Wrote %s", output_path)
+    return output_path

vibepaper-0.1.0.dist-info/METADATA

@@ -0,0 +1,343 @@
+Metadata-Version: 2.4
+Name: vibepaper
+Version: 0.1.0
+Summary: Data-bound Markdown-to-Word builder for scientific papers
+Project-URL: Repository, https://github.com/PLACEHOLDER/vibepaper
+License: MIT
+License-File: LICENSE
+Keywords: jinja2,manuscript,markdown,pandoc,paper,reproducible-research,scientific-writing
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering
+Classifier: Topic :: Text Processing :: General
+Classifier: Topic :: Text Processing :: Markup :: Markdown
+Requires-Python: >=3.10
+Requires-Dist: jinja2>=3.0
+Requires-Dist: lxml>=4.0
+Requires-Dist: pandas>=1.3
+Requires-Dist: pyyaml>=6.0
+Requires-Dist: tomli>=2.0; python_version < '3.11'
+Description-Content-Type: text/markdown
+
+# vibepaper
+
+Build scientific papers from Markdown where every number traces back to the analysis that produced it.
+
+## The problem
+
+Numbers in scientific prose go stale. You finish the analysis, write up the results, and six months later a reviewer asks you to rerun with a corrected dataset. Now you have updated CSVs and a paper full of hardcoded figures: "the mean increased from 4.6 to 9.2", "of the 7,318 variants that lost HIGH impact". Finding every number, checking which is still current, updating without introducing new errors — this is tedious, error-prone, and nearly impossible to review.
+
+## The solution
+
+vibepaper separates computation from communication. Analysis scripts write their key results to named CSV files. The paper references those values by name using Jinja2 template syntax. The build pipeline substitutes every reference before passing the document to pandoc for final Word output.
+
+```markdown
+Mean transcripts per variant doubled from
+{{ vep_impact.giab_mean_v112 | dp(1) }} to
+{{ vep_impact.giab_mean_v115_full | dp(1) }} on upgrading to Ensembl v115.
+```
+
+When you rerun the analysis, you rerun the build. The numbers update everywhere, simultaneously, with a loud error if any reference is missing.
+
+**Three design principles:**
+
+1. **Templates express intent; scripts express computation.** No arithmetic in templates. If you need a percentage increase, the analysis script computes and writes it. The template formats it.
+2. **Loud failures over silent omissions.** A missing or renamed CSV column is a build error, not an empty string in the output.
+3. **Every number is traceable.** Any figure in the rendered paper can be grepped back to the template reference and the script that wrote the CSV.
+
+---
+
+## Installation
+
+```bash
+pip install vibepaper
+```
+
+Requires [pandoc](https://pandoc.org/installing.html) as a system dependency:
+
+```bash
+# macOS
+brew install pandoc
+
+# Debian/Ubuntu
+sudo apt-get install pandoc
+```
+
+---
+
+## Quick start
+
+### Option 1 — paper.toml (recommended for full papers)
+
+Create `paper.toml` in your project root:
+
+```toml
+[paper]
+sections = [
+    "paper/abstract.md",
+    "paper/introduction.md",
+    "paper/methods.md",
+    "paper/results.md",
+    "paper/discussion.md",
+    "paper/references.md",
+    "paper/figures.md",
+]
+supplementary = ["paper/supplementary.md"]
+name = "my_paper"
+```
+
+Then build:
+
+```bash
+vibepaper
+# outputs: output/my_paper_2025-06-01.docx
+#          output/my_paper_supplementary_2025-06-01.docx
+```
+
+### Option 2 — sections file
+
+Create a plain text file listing your sections in order:
+
+```
+# order.txt
+paper/abstract.md
+paper/methods.md
+paper/results.md
+paper/discussion.md
+```
+
+Then:
+
+```bash
+vibepaper --sections-file order.txt --name my_paper
+```
+
+Lines starting with `#` and blank lines are ignored. Paths are relative to the sections file's location.
+
+### Option 3 — direct file list
+
+```bash
+vibepaper paper/abstract.md paper/results.md paper/discussion.md --name my_paper
+```
+
+---
+
+## Template syntax
+
+vibepaper uses [Jinja2](https://jinja.palletsprojects.com/) for template substitution. References follow the pattern `{{ namespace.field | filter }}`.
+
+### Number formatting filters
+
+| Filter | Example | Output |
+|---|---|---|
+| `\| commas` | `{{ n \| commas }}` | `254,129` |
+| `\| dp(n)` | `{{ mean \| dp(1) }}` | `9.2` |
+| `\| pct(n)` | `{{ rate \| pct(1) }}` | `52.2%` |
+| `\| fold(n)` | `{{ ratio \| fold(1) }}` | `2.0-fold` |
+| `\| fmt(spec)` | `{{ v \| fmt('+.1f') }}` | `+3.7` |
+
+`dp` (decimal places) is for numbers that will have surrounding text (e.g. "mean TPV was 9.2"). `pct` appends the `%` sign. Use `fmt` for any format string Python's `format()` accepts.
+
+### Examples
+
+```markdown
+Of {{ clinvar.total_variants | commas }} ClinVar variants, {{ clinvar.gained_high_count | commas }}
+({{ clinvar.gained_high_pct | dp(2) }}%) gained HIGH impact on upgrading to v115.
+
+Mean transcripts per variant increased {{ vep.mean_fold | fold }} from
+{{ vep.mean_v112 | dp(1) }} to {{ vep.mean_v115 | dp(1) }}.
+```
+
+---
+
+## Data sources
+
+### Facts CSVs (primary)
+
+The main data binding mechanism. Analysis scripts write **1-row CSVs** to `output/facts/`. The filename stem becomes the template namespace; column names become field names.
+
+```
+output/facts/
+    transcript_growth.csv      → {{ transcript_growth.v112_count | commas }}
+    vep_impact.csv             → {{ vep_impact.giab_mean_v115_full | dp(1) }}
+    clinvar_reclassification.csv → {{ clinvar_reclassification.total_variants | commas }}
+```
+
+A CSV named `transcript_growth.csv` with columns `v112_count, v115_count`:
+
+```csv
+v112_count,v115_count
+254129,509650
+```
+
+Is referenced as:
+
+```markdown
+Transcripts grew from {{ transcript_growth.v112_count | commas }}
+to {{ transcript_growth.v115_count | commas }}.
+```
+
+vibepaper raises a hard error if a referenced column doesn't exist. It warns if the rendered output contains literal `nan`, `None`, or unresolved `{{`.
+
+### JSON data (supplemental)
+
+Pass additional values directly without creating a CSV file:
+
+```bash
+# Inline dict
+vibepaper --data '{"cohort_size": 412, "stats": {"pvalue": 0.003}}'
+
+# From file
+vibepaper --data results.json
+```
+
+Top-level keys become namespaces:
+
+```markdown
+Cohort: {{ cohort_size }} participants (p = {{ stats.pvalue | dp(3) }}).
+```
+
+JSON is merged on top of facts CSVs. Nested dicts are deep-merged at the namespace level; scalar values override directly.
+
+---
+
+## Table directives
+
+For supplementary tables, embed CSVs directly into the Markdown with a directive comment:
+
+```markdown
+<!-- include-csv: output/consequence_changes.csv
+  columns: [consequence, v112_count, v115_count, pct_change]
+  rename:
+    v112_count: v112
+    v115_count: v115
+    pct_change: Change (%)
+  format:
+    v112_count: ",d"
+    v115_count: ",d"
+    pct_change: ".1f"
+  sort: [-pct_change]
+  max_rows: 20
+-->
+```
+
+Directive options:
+
+| Option | Description |
+|---|---|
+| `columns` | List of columns to include, in order |
+| `rename` | Dict mapping column names to display names |
+| `format` | Dict mapping column names to Python format specs |
+| `align` | `left`, `right`, `center`, or per-column dict |
+| `sort` | List of column names; prefix `-` for descending |
+| `filter` | pandas `query()` expression |
+| `max_rows` | Truncate to this many rows |
+| `na_rep` | String to use for missing values (default: `—`) |
+
+---
+
+## paper.toml reference
+
+```toml
+[paper]
+# Manuscript sections in order (paths relative to paper.toml)
+sections = [
+    "paper/title.md",
+    "paper/abstract.md",
+    "paper/introduction.md",
+    "paper/methods.md",
+    "paper/results.md",
+    "paper/discussion.md",
+    "paper/references.md",
+    "paper/figures.md",
+]
+
+# Built as a separate .docx unless --combined is passed
+supplementary = ["paper/supplementary.md"]
+
+# Output filename stem: {name}_{date}.docx
+# Default: parent directory name
+name = "my_paper"
+
+# Directory of 1-row facts CSVs
+# Default: "output/facts"
+facts_dir = "output/facts"
+
+# Output directory for .docx files
+# Default: "output"
+output_dir = "output"
+
+# Intermediate build directory
+# Default: "build"
+build_dir = "build"
+
+# Word reference document for custom formatting (double spacing, line numbers, etc.)
+# Only used if the file exists; silently skipped otherwise.
+# Default: "paper/reference.docx"
+reference_doc = "paper/reference.docx"
+```
+
+### Word reference document
+
+To apply journal-specific formatting (e.g. double line spacing, continuous line numbering):
+
+1. Open a blank Word document
+2. Set paragraph spacing to Double and enable Layout → Line Numbers → Continuous
+3. Save as `paper/reference.docx`
+
+vibepaper will use it automatically if it exists at the configured path.
+
+---
+
+## CLI reference
+
+```
+vibepaper [FILE.md ...] [options]
+
+Input (choose one):
+  FILE.md ...           Markdown files in order (no paper.toml needed)
+  --sections-file FILE  Plain text file with one .md path per line
+  --config FILE         paper.toml config file (default: paper.toml)
+
+Data:
+  --data JSON           JSON file path or inline dict for template context
+  --facts-dir DIR       Override facts CSV directory
+
+Output:
+  --output-dir DIR      Output directory for .docx files
+  --name NAME           Output filename stem
+  --combined            Merge supplementary into main document
+```
+
+---
+
+## Project layout convention
+
+```
+my_paper/
+├── paper.toml
+├── paper/
+│   ├── abstract.md
+│   ├── introduction.md
+│   ├── methods.md
+│   ├── results.md
+│   ├── discussion.md
+│   ├── references.md
+│   ├── figures.md
+│   ├── supplementary.md
+│   └── reference.docx        ← optional Word formatting template
+├── output/
+│   ├── facts/
+│   │   ├── cohort.csv        ← 1-row: n_patients, n_controls, ...
+│   │   ├── model_results.csv ← 1-row: auc, pvalue, effect_size, ...
+│   │   └── ...
+│   └── tables/
+│       └── full_results.csv  ← multi-row: used in include-csv directives
+└── scripts/
+    └── run_analysis.py       ← writes to output/facts/
+```

vibepaper-0.1.0.dist-info/RECORD

@@ -0,0 +1,10 @@
+vibepaper/__init__.py,sha256=aOHawL1zuHMfBWKXqwUkXcW96oXLNCY-CXdHDqkz4g4,18
+vibepaper/build.py,sha256=aV5SsDJXfQkIFxe8f0C-ehxzBijbRix9a1ZRksJwXiM,10377
+vibepaper/cli.py,sha256=fUpQ4JK9ntx5uLUOSrDC5XLQwZ4lQHwfrhgrbptK7ko,3959
+vibepaper/render.py,sha256=CLL4s5PcoK0WTGtPiYBk-ZI4Cfav64aziWW2iNuPByc,3803
+vibepaper/tables.py,sha256=NGY1u-YmQTxrEOL6S0O39sFYI5h0Jgyshu-3bnBFp5A,5402
+vibepaper-0.1.0.dist-info/METADATA,sha256=uaxGCFAxjOcXuboKY8St9VH6W56LOX4taHWLD67XFbg,10157
+vibepaper-0.1.0.dist-info/WHEEL,sha256=QccIxa26bgl1E6uMy58deGWi-0aeIkkangHcxk2kWfw,87
+vibepaper-0.1.0.dist-info/entry_points.txt,sha256=vxSzwDcsA3BNhSRxJyc6i15RULosdkYhcD5xSMJ9Ve8,49
+vibepaper-0.1.0.dist-info/licenses/LICENSE,sha256=sC4Qc5X8JxIAWbDmnsQLQiiNS4iktXB2HzhvT02DV6A,1086
+vibepaper-0.1.0.dist-info/RECORD,,

vibepaper-0.1.0.dist-info/WHEEL

@@ -0,0 +1,4 @@
+Wheel-Version: 1.0
+Generator: hatchling 1.29.0
+Root-Is-Purelib: true
+Tag: py3-none-any

vibepaper-0.1.0.dist-info/entry_points.txt

@@ -0,0 +1,2 @@
+[console_scripts]
+vibepaper = vibepaper.cli:main

vibepaper-0.1.0.dist-info/licenses/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 ACRF Cancer Genomics Facility
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.