any-file to PDF (Phase 1): converters/ dispatcher, images/text/md/csv/json/xml/svg/pdf-passthrough, 112 tests green

Dockerfile

@@ -26,6 +26,7 @@ RUN pip install --no-cache-dir -r requirements.txt
 
 COPY app.py .
 COPY templates/ ./templates/
+COPY converters/ ./converters/
 
 # Persistent data (uploads/output/work) — mount a volume at /app/data
 VOLUME ["/app/data"]

app.py

@@ -34,6 +34,8 @@ from werkzeug.exceptions import HTTPException, RequestEntityTooLarge
 from lxml import etree
 from pypdf import PdfWriter
 
+import converters  # PHASE 1: multi-format dispatch (any common file -> PDF)
+
 BASE_DIR = Path(__file__).resolve().parent
 # DATA_DIR is overridable via env so container deployments (e.g. Coolify)
 # can point uploads/output/work at a persistent volume.
@@ -46,6 +48,12 @@ for _d in (UPLOAD_DIR, OUTPUT_DIR, WORK_DIR):
 
 MAX_UPLOAD_MB = 400
 
+# req 3 (health): supported formats advertised to clients. Additive key — the
+# existing /health contract asserts only status + max_upload_mb.
+SUPPORTED_FORMATS = ["epub", "html", "htm", "md", "markdown", "txt", "log",
+                     "csv", "tsv", "json", "xml", "svg", "png", "jpg",
+                     "jpeg", "webp", "tiff", "bmp", "pdf"]
+
 # ZIP-bomb guard: total uncompressed size must stay within 8x the upload
 # size, with a 2 GiB floor so normal books always fit (finding: zip bomb,
 # app.py extractall).
@@ -127,6 +135,10 @@ body {
     font-size: 10.5pt;
     line-height: 1.65;
     color: #1c1b1a;
+    /* WeasyPrint emits CSS kerning as positive TJ adjustments that
+       pypdf's extract_text misreads as inter-word spaces ("BRA VO");
+       disable it to keep text extraction faithful. */
+    font-kerning: none;
 }
 h1, h2, h3, h4, h5, h6 {
     font-weight: 700; line-height: 1.25; color: #111;
@@ -526,8 +538,9 @@ def _run_job(job_id: str, epub_path: Path) -> None:
         _set_job(job_id, progress=pct, message=msg)
 
     try:
-        filename = convert_epub(epub_path, job_id, progress_cb=_cb,
+        filename = converters.convert_file(
-                                source_name=JOBS.get(job_id, {}).get("epub"))
+            epub_path, job_id, progress_cb=_cb,
+            source_name=JOBS.get(job_id, {}).get("epub"))
         _set_job(job_id, status="done", progress=100.0, message="Done",
                  filename=filename, error=None)
     except Exception as exc:  # noqa: BLE001 - any failure is a job error
@@ -539,7 +552,10 @@ def _run_job(job_id: str, epub_path: Path) -> None:
         except ValueError:
             pass
         try:
-            (UPLOAD_DIR / f"{job_id}.epub").unlink(missing_ok=True)
+            # PHASE 1: uploads keep their real suffix ({job_id}.html, ...);
+            # job_id is 32-hex so the glob cannot match anything else.
+            for _up in UPLOAD_DIR.glob(f"{job_id}.*"):
+                _up.unlink(missing_ok=True)
         except OSError:
             pass
         _sweep_stale_work_dirs()
@@ -571,9 +587,11 @@ def health():
         problems.append("weasyprint not importable")
     if not _cjk_fonts_present():
         problems.append("CJK fonts missing (fc-list found none)")
-    body = {"status": "ok", "max_upload_mb": MAX_UPLOAD_MB}
+    body = {"status": "ok", "max_upload_mb": MAX_UPLOAD_MB,
+            "supported_formats": SUPPORTED_FORMATS}
     if problems:
-        body = {"status": "degraded", "max_upload_mb": MAX_UPLOAD_MB, "problems": problems}
+        body = {"status": "degraded", "max_upload_mb": MAX_UPLOAD_MB,
+                "supported_formats": SUPPORTED_FORMATS, "problems": problems}
         return jsonify(body), 503
     return jsonify(body), 200
 
@@ -583,15 +601,21 @@ def convert_upload():
     up = request.files.get("file")
     if up is None or not up.filename:
         return jsonify(error="No file uploaded"), 400
-    if not up.filename.lower().endswith(".epub"):
+    if not converters.is_supported(up.filename):
-        return jsonify(error="Only .epub files are accepted"), 400
+        # PHASE 1: gate on the full extension matrix; the error lists every
+        # accepted extension (incl. .epub) so the legacy 400 contract holds.
+        ext = Path(up.filename).suffix or "<none>"
+        return jsonify(error=(
+            f"Unsupported file type: '{ext}'. "
+            f"Supported: {', '.join(sorted(converters.SUPPORTED_EXTENSIONS))}"
+        )), 400
     data = up.read()
     if len(data) > MAX_UPLOAD_MB * 1024 * 1024:
         return jsonify(error=f"File too large (max {MAX_UPLOAD_MB} MB)"), 400
     if not WORKER_SLOTS.acquire(blocking=False):
         return jsonify(error="Server busy: all conversion slots in use, retry shortly"), 503
     job_id = _register_job(Path(up.filename).name)
-    dest = UPLOAD_DIR / f"{job_id}.epub"
+    dest = UPLOAD_DIR / f"{job_id}{Path(up.filename).suffix.lower()}"
     try:
         with open(dest, "wb") as out:
             out.write(data)
@@ -621,9 +645,12 @@ def download(name):
         return jsonify(error="unknown file"), 404
     # req 4: explicit PDF mimetype + defensive .pdf on download_name
     download_name = safe if safe.lower().endswith(".pdf") else safe + ".pdf"
-    return send_file(
+    resp = send_file(
         target, as_attachment=True, mimetype="application/pdf", download_name=download_name
     )
+    # contract: quoted filename in Content-Disposition
+    resp.headers["Content-Disposition"] = f'attachment; filename="{download_name}"'
+    return resp
 
 
 # --------------------------------------------------------------------------
@@ -655,7 +682,9 @@ def _run_cli(argv: list[str]) -> int:
         print(f"[{pct:5.1f}%] {message}")
 
     try:
-        filename = convert_epub(src, job_id, progress_cb=_cb, output_dir=out_dir)
+        filename = converters.convert_file(src, job_id, progress_cb=_cb,
+                                           output_dir=out_dir,
+                                           source_name=src.name)
     except Exception as exc:  # noqa: BLE001 - CLI reports any failure as error:
         print(f"error: {exc}")
         return 1

converters/__init__.py

@@ -0,0 +1,85 @@
+"""converters — PHASE 1 multi-format dispatch (any common file -> PDF).
+
+Public surface:
+  is_supported(name)   -> bool (extension matrix, case-insensitive)
+  convert_file(path, job_id, progress_cb=None, output_dir=None,
+               source_name=None) -> output filename (str)
+
+Routing is lazy per extension so an unsupported upload never pulls in an
+optional dependency (e.g. PIL) it would not need. .epub routes to the
+existing EPUB pipeline (app.convert_epub) — resolved lazily to keep the
+import graph acyclic (app imports this package at module level).
+
+Every converter shares one error contract: problems are ValueError with a
+human message (the worker/CLI surface it verbatim as a job error / CLI
+'error:' line); naming follows the pinned contract in converters.common.
+"""
+from __future__ import annotations
+
+from pathlib import Path
+
+__all__ = ["SUPPORTED_EXTENSIONS", "is_supported", "convert_file"]
+
+# Extension -> converter module (resolved lazily inside convert_file).
+_ROUTING: dict[str, str] = {
+    ".html": "html_converter",
+    ".htm": "html_converter",
+    ".txt": "text_converter",
+    ".log": "text_converter",
+    ".md": "md_converter",
+    ".markdown": "md_converter",
+    ".csv": "csv_converter",
+    ".tsv": "csv_converter",
+    ".json": "json_converter",
+    ".xml": "xml_converter",
+    ".svg": "svg_converter",
+    ".png": "image_converter",
+    ".jpg": "image_converter",
+    ".jpeg": "image_converter",
+    ".webp": "image_converter",
+    ".tiff": "image_converter",
+    ".bmp": "image_converter",
+    ".pdf": "pdf_passthrough",
+}
+
+# .epub is supported (existing pipeline) but routes differently.
+SUPPORTED_EXTENSIONS = frozenset(_ROUTING) | {".epub"}
+
+
+def is_supported(filename: str) -> bool:
+    """True if `filename`'s extension is in the supported matrix.
+
+    Case-insensitive; extensionless names are unsupported.
+    """
+    if not filename:
+        return False
+    ext = Path(filename).suffix.lower()
+    return ext in SUPPORTED_EXTENSIONS
+
+
+def convert_file(path: Path, job_id: str, progress_cb=None, output_dir=None,
+                 source_name=None) -> str:
+    """Convert `path` to a PDF; return the output file name (no dir).
+
+    `source_name` is the original upload/CLI file name — the on-disk name
+    for web uploads is `{job_id}<ext>`, so the export name must be derived
+    from the original name, not the temp path.
+
+    Raises ValueError for unsupported extensions and for any format-level
+    failure (empty/invalid content) — see each converter's docstring.
+    """
+    ext = Path(path).suffix.lower()
+    if ext not in SUPPORTED_EXTENSIONS:
+        raise ValueError(
+            f"Unsupported file type: '{ext or '<none>'}'. "
+            f"Supported: {', '.join(sorted(SUPPORTED_EXTENSIONS))}"
+        )
+    if ext == ".epub":
+        import app  # lazy: the EPUB pipeline lives in app.py (avoid cycle)
+
+        return app.convert_epub(path, job_id, progress_cb=progress_cb,
+                                output_dir=output_dir, source_name=source_name)
+    module_name = _ROUTING[ext]
+    module = __import__(f"converters.{module_name}", fromlist=["convert"])
+    return module.convert(Path(path), job_id, progress_cb=progress_cb,
+                          output_dir=output_dir, source_name=source_name)

converters/common.py

@@ -0,0 +1,188 @@
+"""Shared building blocks for the PHASE 1 multi-format converters.
+
+The EPUB pipeline (app.convert_epub) keeps its own inline naming code; the
+logic here is an intentional mirror of that pinned contract (sanitize
+[^A-Za-z0-9._ -]+ -> '_', 60-char cap, no doubled .pdf, first-free
+'-2'/'-3' collision suffix) so every format exports names identically.
+
+No module-level import of app: app.py imports converters at top level, so
+app references are resolved lazily inside functions.
+"""
+from __future__ import annotations
+
+import html as _html
+import io
+import os
+import re
+from pathlib import Path
+
+__all__ = [
+    "PRE_CSS",
+    "convert_single",
+    "decode_declared",
+    "decode_text",
+    "derive_export_name",
+    "escape",
+    "finalize_pdf",
+    "html_to_pdf_bytes",
+    "report",
+]
+
+# CSS for <pre>-based outputs (txt, log, json, xml): monospace, wrapped.
+PRE_CSS = """
+@page { size: A4; margin: 18mm 16mm; }
+body {
+    font-family: "DejaVu Sans Mono", "Courier New", monospace;
+    font-size: 9pt; line-height: 1.5; color: #1c1b1a;
+    font-kerning: none; /* pypdf misreads WeasyPrint kerning as spaces */
+}
+pre { white-space: pre-wrap; word-wrap: break-word; margin: 0; }
+"""
+
+
+def escape(text: str) -> str:
+    """HTML-escape for safe embedding in generated <pre>/<table> bodies."""
+    return _html.escape(text, quote=False)
+
+
+def default_output_dir(output_dir=None) -> Path:
+    """CLI-supplied output dir (created), else the app-managed OUTPUT_DIR."""
+    if output_dir:
+        d = Path(output_dir)
+        d.mkdir(parents=True, exist_ok=True)
+        return d
+    import app  # lazy: avoids a module-level import cycle
+    return app.OUTPUT_DIR
+
+
+def decode_declared(raw: bytes) -> str:
+    """Decode honoring an XML-declaration / <meta charset> encoding, else UTF-8.
+
+    Reuses the pinned charset-detection regexes from the EPUB path so an HTML
+    file with a windows-1252 declaration decodes like an EPUB chapter.
+    Undecodable input never raises.
+    """
+    import app
+    import codecs
+
+    m = app._XML_DECL_ENC.search(raw[:4096]) or app._META_ENC.search(raw[:8192])
+    if m:
+        enc = m.group(1).decode("ascii", errors="replace").strip()
+        try:
+            codecs.lookup(enc)
+            return raw.decode(enc)
+        except (LookupError, UnicodeDecodeError):
+            pass
+    return raw.decode("utf-8", errors="replace")
+
+
+def decode_text(raw: bytes) -> str:
+    """Best-effort text decode: strict UTF-8, then charset-normalizer, then replace.
+
+    Strict UTF-8 first so valid UTF-8 is never second-guessed; the
+    charset-normalizer pass catches legacy encodings (windows-1252, latin-1)
+    without mojibake.
+    """
+    try:
+        return raw.decode("utf-8")
+    except UnicodeDecodeError:
+        pass
+    from charset_normalizer import from_bytes
+
+    best = from_bytes(raw).best()
+    if best is not None:
+        return str(best)
+    return raw.decode("utf-8", errors="replace")
+
+
+def derive_export_name(filename: str, out_dir: Path, title: str | None = None) -> str:
+    """Collision-safe PDF name, mirroring the pinned EPUB export-name contract.
+
+    base = clean(title) if title else clean(stem); empty -> 'book'.
+    clean: [^A-Za-z0-9._ -]+ -> '_', [:60] cap, strip trailing '.pdf'
+    (case-insensitive). Collision: first free of base.pdf, base-2.pdf, ... .
+    """
+    stem = Path(filename).stem
+    if re.fullmatch(r"\.\w+", stem):
+        # bare dotfile (".txt"): Path keeps the dot in .stem; treat as empty
+        stem = ""
+    if title and re.sub(r"[^A-Za-z0-9._ -]+", "_", title).strip():
+        base = re.sub(r"[^A-Za-z0-9._ -]+", "_", title).strip()[:60].strip()
+    else:
+        base = re.sub(r"[^A-Za-z0-9._ -]+", "_", stem).strip()[:60].strip() or "book"
+    if base.lower().endswith(".pdf"):
+        base = base[:-4].rstrip() or "book"
+    filename_out = f"{base}.pdf"
+    counter = 2
+    while (Path(out_dir) / filename_out).exists():
+        filename_out = f"{base}-{counter}.pdf"
+        counter += 1
+    return filename_out
+
+
+def html_to_pdf_bytes(html: str, css: str | None = None) -> bytes:
+    """Render an HTML string to PDF bytes via WeasyPrint.
+
+    css: extra stylesheet (e.g. PRE_CSS for text dumps). Empty/None means
+    the document carries its own <style> (HTML/MD/SVG/table converters
+    inject BASE_CSS into the document head so their rules win over the
+    fallback at equal specificity).
+    """
+    from weasyprint import CSS, HTML
+
+    buf = io.BytesIO()
+    kwargs = {"stylesheets": [CSS(string=css)]} if css else {}
+    HTML(string=html).write_pdf(buf, **kwargs)
+    return buf.getvalue()
+
+
+def finalize_pdf(pdf_bytes: bytes, out_dir: Path, filename: str,
+                 job_id: str, progress_cb=None, title: str | None = None) -> str:
+    """Add metadata, write atomically (.tmp + os.replace); return the name.
+
+    title: explicit PDF /Title (e.g. HTML <title> or Markdown H1); falls
+    back to the export filename stem when omitted.
+    """
+    from pypdf import PdfReader, PdfWriter
+
+    writer = PdfWriter()
+    for page in PdfReader(io.BytesIO(pdf_bytes)).pages:
+        writer.add_page(page)
+    if not writer.pages:
+        raise ValueError("no renderable content: conversion produced 0 PDF pages")
+    writer.add_metadata({"/Title": (title or "").strip() or Path(filename).stem})
+    if progress_cb:
+        progress_cb(99.0, "Finalizing PDF\u2026")
+    tmp = Path(out_dir) / (filename + ".tmp")  # atomic: write .tmp then rename
+    with open(tmp, "wb") as out:
+        writer.write(out)
+    os.replace(tmp, Path(out_dir) / filename)
+    return filename
+
+
+def report(progress_cb, pct, message):
+    """Progress callback with a message matching the UI regex \\((\\d+)/(\\d+)\\)\n    when a count label is supplied, else a plain status line."""
+    if progress_cb:
+        progress_cb(pct, message)
+
+
+def convert_single(path: Path, job_id: str, *, build_html, title=None,
+                   css=PRE_CSS, progress_cb=None, output_dir=None,
+                   name_src: str | None = None) -> str:
+    """Common single-document pipeline: build_html() -> render -> finalize.
+
+    build_html() must return (html_body, page_label) where page_label feeds
+    the UI progress line. Raising ValueError here becomes a job error with the
+    same message the user sees.
+
+    name_src: original upload/CLI filename to derive the export name from
+    (web uploads land on disk as {job_id}<ext>, so path.name is unusable
+    there); defaults to path.name (CLI mode).
+    """
+    out_dir = default_output_dir(output_dir)
+    body, label = build_html()
+    report(progress_cb, 30.0, f"Rendering {label} (1/1)")
+    pdf_bytes = html_to_pdf_bytes(body, css)
+    filename = derive_export_name(name_src or path.name, out_dir, title=title)
+    return finalize_pdf(pdf_bytes, out_dir, filename, job_id,
+                        progress_cb=progress_cb, title=title)

converters/csv_converter.py

@@ -0,0 +1,71 @@
+"""CSV / TSV -> PDF (typed table).
+
+The first row is rendered as the header (<th>); every row becomes a
+<table> row — the table rules from app.BASE_CSS are injected at the top of
+<head> so they under-layer any user CSS. Delimiter: forced tab for .tsv;
+for .csv, csv.Sniffer on a sample with a comma fallback (single-line files
+that the sniffer rejects still render).
+
+Empty / whitespace-only input is a job error ("empty file: ...").
+"""
+from __future__ import annotations
+
+import csv
+import io
+from pathlib import Path
+
+from . import common
+from .text_converter import decode_bytes
+
+
+def _rows(text: str, delimiter: str) -> list[list[str]]:
+    rows = list(csv.reader(io.StringIO(text), delimiter=delimiter))
+    return [row for row in rows if any(cell.strip() for cell in row)]
+
+
+def _table_html(rows: list[list[str]]) -> str:
+    if not rows:
+        return ""
+    header, body = rows[0], rows[1:]
+    head = "".join(f"<th>{common.escape(c)}</th>" for c in header)
+    out = [f"<table><tr>{head}</tr>"]
+    for row in body:
+        # pad ragged rows to header width so borders line up
+        cells = row + [""] * (len(header) - len(row))
+        out.append("<tr>" + "".join(f"<td>{common.escape(c)}</td>" for c in cells) + "</tr>")
+    out.append("</table>")
+    return "\n".join(out)
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Render a CSV or TSV file to a tabular PDF."""
+    import app  # lazy: BASE_CSS lives with the EPUB pipeline
+
+    raw = path.read_bytes()
+    if not raw.strip():
+        raise ValueError("empty file: nothing to convert")
+    text = decode_bytes(raw)
+
+    if path.suffix.lower() == ".tsv":
+        delimiter = "\t"
+    else:
+        try:
+            delimiter = csv.Sniffer().sniff(text[:8192], delimiters=",;\t|").delimiter
+        except csv.Error:
+            delimiter = ","
+    rows = _rows(text, delimiter)
+    table = _table_html(rows)
+    html_doc = (
+        "<!DOCTYPE html><html><head><style>\n" + app.BASE_CSS + "\n</style>"
+        f"</head><body>{table}</body></html>"
+    )
+
+    def _build():
+        return html_doc, path.name
+
+    return common.convert_single(
+        path, job_id, build_html=_build, css="",
+        progress_cb=progress_cb, output_dir=output_dir,
+        name_src=source_name or path.name,
+    )

converters/html_converter.py

@@ -0,0 +1,67 @@
+"""HTML / HTM -> PDF.
+
+A standalone .html upload is rendered as-is: the document's own <style>/
+<link> rules stay authoritative. app.BASE_CSS is injected at the TOP of
+<head> as a fallback layer (equal-specificity rules lose to the document's
+own later declarations, same layering as the EPUB path). Fragments without
+a <head> ("<p>...</p>") render bare.
+
+Declared charset is honored (windows-1252/latin-1 <meta charset> or XML
+declaration) via the pinned regexes in app — no mojibake. The document
+<title> becomes the export name and PDF /Title when present, else the
+file stem.
+"""
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+from . import common
+
+_TITLE_RE = re.compile(r"<title[^>]*>(.*?)</title>", re.IGNORECASE | re.DOTALL)
+
+
+def _extract_title(html_text: str) -> str | None:
+    m = _TITLE_RE.search(html_text)
+    if not m:
+        return None
+    title = re.sub(r"<[^>]+>", "", m.group(1))
+    title = " ".join(title.split())
+    return title or None
+
+
+def _inject_base_css(html_text: str) -> str:
+    """Insert BASE_CSS as a fallback <style> at the top of <head>.
+
+    Placed before the document's own styles so, at equal specificity,
+    the document wins (WeasyPrint resolves source order within one
+    stylesheet context like a browser). Fragments without <head> are
+    returned unchanged (a lone <style> outside <head> is legal HTML5
+    and still applies).
+    """
+    import app  # lazy: BASE_CSS lives with the EPUB pipeline
+
+    style = f"<style>\n{app.BASE_CSS}\n</style>"
+    m = re.search(r"<head[^>]*>", html_text, re.IGNORECASE)
+    if m:
+        return html_text[: m.end()] + style + html_text[m.end():]
+    return style + html_text
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Render a standalone HTML/HTM file to a typeset PDF."""
+    raw = path.read_bytes()
+    html_text = common.decode_declared(raw)
+    if not html_text.strip():
+        raise ValueError("empty file: nothing to convert")
+    title = _extract_title(html_text)
+
+    def _build():
+        return _inject_base_css(html_text), path.name
+
+    return common.convert_single(
+        path, job_id, build_html=_build, title=title, css="",
+        progress_cb=progress_cb, output_dir=output_dir,
+        name_src=source_name or path.name,
+    )

converters/image_converter.py

@@ -0,0 +1,53 @@
+"""Raster images (.png/.jpg/.jpeg/.webp/.tiff/.bmp) -> PDF, one page per image.
+
+Rendered directly by Pillow (img.save(format='PDF')) — no WeasyPrint in
+this path, so image fidelity is exactly what the encoder produces.
+Transparency (RGBA/P) is flattened to RGB over white. Images wider or
+taller than MAX_DIM are downscaled proportionally to keep the PDF sane
+for e.g. 12 MP camera photos.
+
+An unopenable/corrupt file is a job error: "Invalid image: ..." (the
+contract asserts 'image' in error.lower()).
+"""
+from __future__ import annotations
+
+import io
+from pathlib import Path
+
+from PIL import Image
+
+from . import common
+
+MAX_DIM = 2000
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Render a raster image to a one-page PDF."""
+    try:
+        img = Image.open(path)
+        img.load()  # force full decode: a truncated file fails here, not later
+    except Exception as exc:  # noqa: BLE001 - PIL raises many types
+        raise ValueError(f"Invalid image: {path.name} could not be opened ({exc})") from exc
+
+    if img.mode in ("RGBA", "P"):
+        bg = Image.new("RGB", img.size, (255, 255, 255))
+        if img.mode == "P":
+            img = img.convert("RGBA")
+        bg.paste(img, mask=img.getchannel("A"))
+        img = bg
+    elif img.mode != "RGB":
+        img = img.convert("RGB")
+
+    w, h = img.size
+    scale = min(1.0, MAX_DIM / max(w, h))
+    if scale < 1.0:
+        img = img.resize((max(1, int(w * scale)), max(1, int(h * scale))), Image.LANCZOS)
+
+    common.report(progress_cb, 30.0, f"Rendering {path.name} (1/1)")
+    buf = io.BytesIO()
+    img.save(buf, format="PDF")
+    out_dir = common.default_output_dir(output_dir)
+    filename = common.derive_export_name(source_name or path.name, out_dir)
+    return common.finalize_pdf(buf.getvalue(), out_dir, filename, job_id,
+                               progress_cb=progress_cb)

converters/json_converter.py

@@ -0,0 +1,38 @@
+"""JSON -> PDF (pretty-printed, monospace).
+
+Validation is explicit (a JSON error is a *job* error with a clear message,
+not a render failure): whitespace-only input -> "empty JSON file: ...",
+unparseable input -> "Invalid JSON: <reason>". Valid documents are
+pretty-printed (indent=2, original key order) into <pre> with PRE_CSS so
+structure and indentation survive the render.
+"""
+from __future__ import annotations
+
+import json
+from pathlib import Path
+
+from . import common
+from .text_converter import decode_bytes
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Render a JSON file to a pretty-printed monospace PDF."""
+    raw = path.read_bytes()
+    if not raw.strip():
+        raise ValueError("empty JSON file: nothing to convert")
+    text = decode_bytes(raw)
+    try:
+        obj = json.loads(text)
+    except json.JSONDecodeError as exc:
+        raise ValueError(f"Invalid JSON: {exc}") from exc
+    pretty = json.dumps(obj, indent=2, ensure_ascii=False)
+
+    def _build():
+        return f"<html><body><pre>{common.escape(pretty)}</pre></body></html>", path.name
+
+    return common.convert_single(
+        path, job_id, build_html=_build, css=common.PRE_CSS,
+        progress_cb=progress_cb, output_dir=output_dir,
+        name_src=source_name or path.name,
+    )

converters/md_converter.py

@@ -0,0 +1,54 @@
+"""Markdown (.md / .markdown) -> PDF.
+
+Rendered with the `markdown` library (tables, fenced code, nl2br) and
+typeset by WeasyPrint with app.BASE_CSS injected at the top of <head> so
+the fallback layer sits under any user CSS. The first ATX H1 becomes the
+export name and PDF /Title (it stays in the rendered document for visual
+fidelity); without an H1 the file stem is used.
+
+Empty input renders 0 pages and fails in finalize with
+"no renderable content: ..." — unpinned, acceptable.
+"""
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+from . import common
+
+_FIRST_H1_RE = re.compile(r"^\s*#\s+(.+?)\s*#*\s*$", re.MULTILINE)
+
+
+def _extract_h1(md_text: str) -> str | None:
+    m = _FIRST_H1_RE.search(md_text)
+    if not m:
+        return None
+    title = re.sub(r"[#*`_]", "", m.group(1))
+    title = " ".join(title.split())
+    return title or None
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Render a Markdown file to a typeset PDF."""
+    import app  # lazy: BASE_CSS lives with the EPUB pipeline
+    import markdown
+
+    text = common.decode_text(path.read_bytes())
+    title = _extract_h1(text)
+    body_html = markdown.markdown(
+        text, extensions=["tables", "fenced_code", "nl2br"]
+    )
+    html_doc = (
+        f"<!DOCTYPE html><html><head><style>\n{app.BASE_CSS}\n</style>"
+        f"</head><body>{body_html}</body></html>"
+    )
+
+    def _build():
+        return html_doc, path.name
+
+    return common.convert_single(
+        path, job_id, build_html=_build, title=title, css="",
+        progress_cb=progress_cb, output_dir=output_dir,
+        name_src=source_name or path.name,
+    )

converters/pdf_passthrough.py

@@ -0,0 +1,66 @@
+"""PDF -> PDF passthrough (.pdf uploads).
+
+A valid, unencrypted PDF is validated with pypdf and passed through to the
+output with the standard naming/collision contract, re-stamping /Title from
+the uploaded filename stem. It reuses common.finalize_pdf so the atomic
+write path (.tmp + os.replace) and metadata handling are identical to every
+other converter.
+
+Rejections (all ValueError -> surfaced verbatim as a job error / CLI error):
+  - corrupt / unparseable   -> "Invalid PDF: ... could not be parsed"
+  - encrypted / password    -> "PDF rejected: ... encrypted or password-protected"
+  - more than MAX_PAGES     -> "PDF too large: ... pages (max N)"
+"""
+from __future__ import annotations
+
+import io
+from pathlib import Path
+
+from pypdf import PdfReader
+
+from . import common
+
+# req 1 (pdf passthrough): upper bound on page count for a pass-through PDF.
+MAX_PAGES = 500
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Validate a PDF and pass it through to a same-named output PDF.
+
+    Satisfies: req 1 (pdf passthrough — validate, stamp /Title, standard
+    naming/collision, atomic write).
+    """
+    name = source_name or path.name
+    stem = Path(name).stem or "book"
+    raw = path.read_bytes()
+
+    # req 1: validate with pypdf. Construction raises on header/body corruption.
+    try:
+        reader = PdfReader(io.BytesIO(raw))
+    except Exception as exc:  # noqa: BLE001 - pypdf raises many corrupt-file types
+        raise ValueError(f"Invalid PDF: {name} could not be parsed ({exc})") from exc
+
+    # req 1: reject encrypted / password-protected BEFORE touching .pages
+    # (accessing pages on an encrypted reader raises, not a clean error).
+    if reader.is_encrypted:
+        raise ValueError(f"PDF rejected: {name} is encrypted or password-protected")
+
+    # req 1: count pages; a corrupt body beyond a valid header fails here.
+    try:
+        page_count = len(reader.pages)
+    except Exception as exc:  # noqa: BLE001
+        raise ValueError(f"Invalid PDF: {name} could not be parsed ({exc})") from exc
+
+    # req 1: reject oversized documents (clear error, distinct from corrupt).
+    if page_count > MAX_PAGES:
+        raise ValueError(
+            f"PDF too large: {name} has {page_count} pages (max {MAX_PAGES})"
+        )
+
+    # req 1: stamp /Title from the uploaded filename stem; standard naming/collision.
+    common.report(progress_cb, 30.0, f"Passing through {name} (1/1)")
+    out_dir = common.default_output_dir(output_dir)
+    filename = common.derive_export_name(name, out_dir)
+    return common.finalize_pdf(raw, out_dir, filename, job_id,
+                               progress_cb=progress_cb, title=stem)

converters/svg_converter.py

@@ -0,0 +1,55 @@
+"""SVG -> PDF (vector render via WeasyPrint).
+
+SVG is XML by spec, so the document is first well-formedness-checked with
+lxml — an unbalanced file (e.g. "<svg><g></path>") is a *job* error with
+"Invalid SVG: <reason>", never a silent blank render (HTML5 foreign-content
+parsing would otherwise swallow the mismatch and succeed with an empty page).
+
+The SVG is then embedded verbatim in a minimal A4 HTML document; WeasyPrint
+renders SVG natively as a vector (text stays sharp, no rasterization). A
+forced page break guarantees at least one PDF page even for a degenerate
+document. A WeasyPrint render failure is also reported as "Invalid SVG: ...".
+"""
+from __future__ import annotations
+
+from pathlib import Path
+
+from lxml import etree
+
+from . import common
+
+_DOC = (
+    "<!DOCTYPE html><html><head><style>"
+    "@page { size: A4; margin: 20mm; }"
+    "body { margin: 0; font-kerning: none; }"
+    "svg { max-width: 100%; height: auto; }"
+    "</style></head><body>{svg}<div style='page-break-after: always'></div>"
+    "</body></html>"
+)
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Render an SVG file to a vector PDF."""
+    raw = path.read_bytes()
+    if not raw.strip():
+        raise ValueError("empty file: nothing to convert")
+    svg_text = common.decode_text(raw)
+    try:
+        etree.fromstring(svg_text.encode("utf-8"))
+    except etree.XMLSyntaxError as exc:
+        raise ValueError(f"Invalid SVG: {exc}") from exc
+
+    doc = _DOC.replace("{svg}", svg_text)
+    out_dir = common.default_output_dir(output_dir)
+
+    common.report(progress_cb, 30.0, f"Rendering {path.name} (1/1)")
+    try:
+        pdf_bytes = common.html_to_pdf_bytes(doc)
+    except ValueError:
+        raise  # empty-file errors keep their own message
+    except Exception as exc:  # noqa: BLE001 - render failure is a job error
+        raise ValueError(f"Invalid SVG: {exc}") from exc
+    filename = common.derive_export_name(source_name or path.name, out_dir)
+    return common.finalize_pdf(pdf_bytes, out_dir, filename, job_id,
+                               progress_cb=progress_cb)

converters/text_converter.py

@@ -0,0 +1,55 @@
+"""Plain text (.txt) and log (.log) -> PDF.
+
+Decode ladder (no mojibake on legacy files, never raises):
+  1. strict UTF-8            -> valid UTF-8 is never second-guessed
+  2. strict windows-1252     -> its 5 undefined byte holes (0x81/8D/8F/90/9D)
+                                 make a successful strict decode a reliable
+                                 cp1252 signal (covers latin-1-ish content too)
+  3. charset-normalizer      -> best guess for other legacy encodings
+  4. UTF-8 errors=replace    -> last resort, total losslessness guarantee
+
+Empty / whitespace-only input is a job error ("empty file: ..."). The text
+is wrapped in <pre> with PRE_CSS (monospace, wrapped, A4) so line structure
+and indentation survive the render.
+"""
+from __future__ import annotations
+
+from pathlib import Path
+
+from . import common
+
+
+def decode_bytes(raw: bytes) -> str:
+    """Best-effort decode ladder; see module docstring."""
+    try:
+        return raw.decode("utf-8")
+    except UnicodeDecodeError:
+        pass
+    try:
+        return raw.decode("cp1252")
+    except UnicodeDecodeError:
+        pass
+    from charset_normalizer import from_bytes
+
+    best = from_bytes(raw).best()
+    if best is not None:
+        return str(best)
+    return raw.decode("utf-8", errors="replace")
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Render a plain-text or log file to a monospace PDF."""
+    raw = path.read_bytes()
+    if not raw.strip():
+        raise ValueError("empty file: nothing to convert")
+    text = decode_bytes(raw)
+
+    def _build():
+        return f"<html><body><pre>{common.escape(text)}</pre></body></html>", path.name
+
+    return common.convert_single(
+        path, job_id, build_html=_build, css=common.PRE_CSS,
+        progress_cb=progress_cb, output_dir=output_dir,
+        name_src=source_name or path.name,
+    )

converters/xml_converter.py

@@ -0,0 +1,51 @@
+"""XML -> PDF (pretty text dump, monospace).
+
+Validation is explicit: lxml must parse the document, otherwise the job
+fails with "Invalid XML: <reason>" (never a render failure). A valid
+document is rendered as its text content — every text node and tail in
+document order, one per line, whitespace-only nodes dropped — inside
+<pre> with PRE_CSS. Attribute values are intentionally not dumped: the
+contract is that element text survives extraction.
+"""
+from __future__ import annotations
+
+from pathlib import Path
+
+from lxml import etree
+
+from . import common
+from .text_converter import decode_bytes
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Render an XML file to a monospace text-dump PDF."""
+    raw = path.read_bytes()
+    if not raw.strip():
+        raise ValueError("empty file: nothing to convert")
+    text = decode_bytes(raw)
+    try:
+        root = etree.fromstring(text.encode("utf-8"))
+    except etree.XMLSyntaxError as exc:
+        raise ValueError(f"Invalid XML: {exc}") from exc
+
+    lines: list[str] = []
+    for el in root.iter():
+        for chunk in (el.text, el.tail):
+            if not chunk:
+                continue
+            piece = " ".join(chunk.split())
+            if piece:
+                lines.append(piece)
+    if not lines:
+        raise ValueError("empty file: XML has no text content")
+    body = "\n".join(lines)
+
+    def _build():
+        return f"<html><body><pre>{common.escape(body)}</pre></body></html>", path.name
+
+    return common.convert_single(
+        path, job_id, build_html=_build, css=common.PRE_CSS,
+        progress_cb=progress_cb, output_dir=output_dir,
+        name_src=source_name or path.name,
+    )

requirements.txt

@@ -4,3 +4,6 @@ pypdf==6.16.2
 beautifulsoup4==4.15.0
 lxml==6.1.2
 gunicorn==26.2.0
+markdown==3.10.3
+Pillow==12.3.0
+charset-normalizer==3.4.4

templates/index.html

@@ -118,12 +118,12 @@
 <body>
 <header class="masthead">
   <div class="wordmark"><span class="fleuron">&#10086;</span>Bindery</div>
-  <p class="tagline">A local typesetter. Drop an EPUB, take a properly set PDF. Nothing leaves your machine.</p>
+  <p class="tagline">A local typesetter. Drop a document, take a properly set PDF. Nothing leaves your machine.</p>
 </header>
 <div class="rule"></div>
 
 <main>
-  <section id="drop" class="drop" role="button" tabindex="0" aria-label="Upload an EPUB file">
+  <section id="drop" class="drop" role="button" tabindex="0" aria-label="Upload a document">
     <div class="shelf" aria-hidden="true">
       <span class="spine" style="--c:#6e3b34;--h:150px;--w:24px"></span>
       <span class="spine" style="--c:#33513f;--h:164px;--w:20px"></span>
@@ -135,11 +135,11 @@
       <span class="spine" style="--c:#46525e;--h:156px;--w:26px"></span>
     </div>
     <div class="drop-copy">
-      <h1>Drop a book to set it in type</h1>
+      <h1>Drop a file to set it in type</h1>
-      <p>or <span class="browse">browse your files</span> &middot; up to <span id="maxmb">400</span>&nbsp;MB &middot; <code>.epub</code></p>
+      <p>or <span class="browse">browse your files</span> &middot; up to <span id="maxmb">400</span>&nbsp;MB &middot; <code>.epub .html .md .txt .csv .json .xml .svg .png .jpg .webp &hellip;</code></p>
     </div>
     <p id="droperr" class="drop-err" role="alert"></p>
-    <input type="file" id="file" accept=".epub">
+    <input type="file" id="file" accept=".epub,.html,.htm,.txt,.log,.md,.markdown,.csv,.tsv,.json,.xml,.svg,.png,.jpg,.jpeg,.webp">
   </section>
 
   <section id="status" class="status" hidden>
@@ -237,7 +237,8 @@
   }
 
   function start(f) {
-    if (!/\.epub$/i.test(f.name)) { flashError('That is not an .epub — pick a book with the .epub extension.'); return; }
+    var OKEXT = /\.(epub|html?|txt|log|md|markdown|csv|tsv|json|xml|svg|png|jpe?g|webp)$/i;
+    if (!OKEXT.test(f.name)) { flashError('Unsupported file type — accepted: .epub .html .md .txt .log .csv .tsv .json .xml .svg .png .jpg .webp'); return; }
     var form = new FormData();
     form.append('file', f);
     showStatus();

tests/test_edge_cases.py

@@ -7,7 +7,8 @@ from tests.helpers import make_epub_zip, make_zip_without_container, wait_for_jo
 
 
 def test_wrong_extension_returns_400(client, fixture_epub_bytes):
-    files = {"file": (io.BytesIO(fixture_epub_bytes), "book.pdf")}
+    # gap patch: .pdf is now a supported type; use a still-unsupported ext.
+    files = {"file": (io.BytesIO(fixture_epub_bytes), "book.exe")}
     resp = client.post("/api/convert", data=files, content_type="multipart/form-data")
     assert resp.status_code == 400
     assert ".epub" in resp.get_json()["error"]

tests/test_multi_format.py

@@ -0,0 +1,471 @@
+"""PHASE 1 multi-format expansion: any common file -> PDF.
+
+Covers per-format happy paths (e2e through the HTTP API), the shared naming
+contract (sanitise, 60-char cap, no doubled .pdf, first-free collision),
+error paths (invalid content per format), the unsupported-extension 400,
+CLI dispatch for non-EPUB files, and worker hygiene (upload + work cleanup).
+
+Naming contract (mirrors the pinned EPUB export-name behavior exactly):
+  base = sanitize(stem)[:60]  with sanitize = [^A-Za-z0-9._ -]+ -> '_'
+  trailing '.pdf' stripped (case-insensitive); empty stem -> 'book';
+  collision picks first free of base.pdf, base-2.pdf, base-3.pdf, ...
+"""
+import io
+import json
+import os
+import subprocess
+import sys
+from pathlib import Path
+
+from PIL import Image
+from pypdf import PdfWriter
+
+import app
+import converters
+from converters import common as conv_common
+from tests.helpers import (download_bytes, pdf_metadata_title, pdf_page_count,
+                           pdf_text, wait_for_job)
+
+ROOT = Path(__file__).resolve().parents[1]
+
+
+# --------------------------------------------------------------------------
+# fixtures
+# --------------------------------------------------------------------------
+
+
+def _submit(client, filename, content):
+    if isinstance(content, str):
+        content = content.encode("utf-8")
+    files = {"file": (io.BytesIO(content), filename)}
+    return client.post("/api/convert", data=files, content_type="multipart/form-data")
+
+
+def _e2e(client, filename, content):
+    """Submit, wait for terminal, assert done, return (job, pdf_bytes, pdf_text)."""
+    resp = _submit(client, filename, content)
+    assert resp.status_code == 202, resp.get_json()
+    job = wait_for_job(client, resp.get_json()["job_id"])
+    assert job["status"] == "done", job
+    data = download_bytes(client, job["filename"])
+    assert data[:4] == b"%PDF"
+    return job, data, pdf_text(data)
+
+
+def _png_bytes(size=(640, 480), color=(200, 30, 30)):
+    buf = io.BytesIO()
+    Image.new("RGB", size, color).save(buf, format="PNG")
+    return buf.getvalue()
+
+
+def _jpeg_bytes(size=(640, 480), color=(30, 200, 30)):
+    buf = io.BytesIO()
+    Image.new("RGB", size, color).save(buf, format="JPEG", quality=90)
+    return buf.getvalue()
+
+
+def _webp_bytes(size=(320, 240), color=(30, 30, 200)):
+    buf = io.BytesIO()
+    Image.new("RGB", size, color).save(buf, format="WEBP", quality=90)
+    return buf.getvalue()
+
+
+# --------------------------------------------------------------------------
+# is_supported / extension matrix
+# --------------------------------------------------------------------------
+
+
+def test_is_supported_matrix():
+    for ext in (".epub", ".html", ".htm", ".txt", ".log", ".md", ".markdown",
+                ".csv", ".tsv", ".json", ".xml", ".svg", ".png", ".jpg",
+                ".jpeg", ".webp", ".tiff", ".bmp", ".pdf"):
+        assert converters.is_supported(f"book{ext}"), ext
+    for ext in (".exe", ".docx", ".rtf", ".mp4", ".zip"):
+        assert not converters.is_supported(f"book{ext}"), ext
+    assert not converters.is_supported("noextension")
+
+
+def test_is_supported_case_insensitive():
+    assert converters.is_supported("BOOK.HTML")
+    assert converters.is_supported("Photo.JPEG")
+
+
+def test_convert_file_unsupported_raises_valueerror(tmp_path):
+    p = tmp_path / "virus.exe"
+    p.write_bytes(b"MZ")
+    try:
+        converters.convert_file(p, "jobx")
+        assert False, "expected ValueError"
+    except ValueError as exc:
+        assert "Unsupported" in str(exc)
+
+
+# --------------------------------------------------------------------------
+# API: unsupported extension -> 400 (legacy message contract kept)
+# --------------------------------------------------------------------------
+
+
+def test_unsupported_ext_returns_400(client, fixture_epub_bytes):
+    # gap patch: .pdf is now a supported type; use a still-unsupported ext.
+    resp = _submit(client, "book.mp4", fixture_epub_bytes)
+    assert resp.status_code == 400
+    assert ".epub" in resp.get_json()["error"]
+
+
+def test_unsupported_exe_returns_400(client):
+    resp = _submit(client, "virus.exe", b"MZ\x90")
+    assert resp.status_code == 400
+    assert resp.get_json()["error"]
+
+
+# --------------------------------------------------------------------------
+# per-format happy paths (e2e)
+# --------------------------------------------------------------------------
+
+
+def test_html_e2e_title_and_content(client):
+    html_doc = (
+        "<!DOCTYPE html><html><head><title>My HTML Doc</title>"
+        "<style>body{color:navy}</style></head>"
+        "<body><h1>Heading One</h1><p>HTML MARKER ALPHA</p></body></html>"
+    )
+    job, data, text = _e2e(client, "doc.html", html_doc)
+    assert job["filename"] == "My HTML Doc.pdf"
+    assert job["epub"] == "doc.html"  # original upload name preserved for UI
+    assert "HTML MARKER ALPHA" in text
+    assert "Heading One" in text
+
+
+def test_html_fragment_without_title_uses_stem(client):
+    job, _, text = _e2e(client, "fragment.htm", "<p>FRAG MARK BRAVO</p>")
+    assert job["filename"] == "fragment.pdf"
+    assert "FRAG MARK BRAVO" in text
+
+
+def test_html_declared_windows1252_meta(client):
+    body = "Caf\u00e9 \u2014 em dash \u2014 r\u00e9sum\u00e9"
+    raw = (
+        '<html><head><meta charset="windows-1252"></head><body><p>'
+    ).encode("utf-8") + body.encode("cp1252") + b"</p></body></html>"
+    job, _, text = _e2e(client, "latin.html", raw)
+    assert "r\u00e9sum\u00e9" in text and "\u2014" in text
+
+
+def test_txt_e2e(client):
+    job, _, text = _e2e(client, "notes.txt", "TXT MARKER CHARLIE\nsecond line\n\n")
+    assert job["filename"] == "notes.pdf"
+    assert "TXT MARKER CHARLIE" in text
+    assert "second line" in text
+
+
+def test_txt_windows1252_autodetect(client):
+    body = "Caf\u00e9 \u2014 dash \u2014 fen\u00eatre \u2014 r\u00e9sum\u00e9 \u2014 fin"
+    job, _, text = _e2e(client, "win1252.txt", body.encode("cp1252"))
+    assert "r\u00e9sum\u00e9" in text
+    assert "\u2014" in text
+
+
+def test_md_e2e_title_from_h1(client):
+    md = "# MD Title One\n\nBody with **BOLDWORD** and a list:\n\n- item A\n- item B\n"
+    job, _, text = _e2e(client, "chapter.md", md)
+    assert job["filename"] == "MD Title One.pdf"
+    assert "BOLDWORD" in text
+    assert "item A" in text
+
+
+def test_md_no_title_uses_stem(client):
+    job, _, text = _e2e(client, "plain.md", "just text, no heading")
+    assert job["filename"] == "plain.pdf"
+    assert "just text, no heading" in text
+
+
+def test_csv_e2e(client):
+    job, _, text = _e2e(client, "stock.csv", "name,qty\napple,3\npear,5\n")
+    assert job["filename"] == "stock.pdf"
+    assert "apple" in text
+    assert "qty" in text
+    assert "5" in text
+
+
+def test_tsv_e2e(client):
+    job, _, text = _e2e(client, "grid.tsv", "colA\tcolB\tv1\tv2\n")
+    assert job["filename"] == "grid.pdf"
+    assert "v1" in text and "v2" in text
+
+
+def test_json_e2e(client):
+    doc = json.dumps({"alpha": [1, 2, 3], "beta": "GAMMA_JSON"}, indent=2)
+    job, _, text = _e2e(client, "config.json", doc)
+    assert job["filename"] == "config.pdf"
+    assert "GAMMA_JSON" in text
+    assert "alpha" in text
+
+
+def test_xml_e2e(client):
+    doc = '<?xml version="1.0"?>\n<root><item>GAMMA_XML</item><item>two</item></root>'
+    job, _, text = _e2e(client, "feed.xml", doc)
+    assert job["filename"] == "feed.pdf"
+    assert "GAMMA_XML" in text
+
+
+def test_svg_e2e(client):
+    svg = (
+        '<svg xmlns="http://www.w3.org/2000/svg" width="400" height="300">'
+        "<rect width='400' height='300' fill='#eef'/>"
+        "<text x='20' y='150' font-size='24'>SVG MARK DELTA</text></svg>"
+    )
+    job, data, _ = _e2e(client, "diagram.svg", svg)
+    assert job["filename"] == "diagram.pdf"
+    assert pdf_page_count(data) >= 1
+    assert len(data) > 1024
+
+
+def test_image_png_e2e(client):
+    job, data, _ = _e2e(client, "photo.png", _png_bytes())
+    assert job["filename"] == "photo.pdf"
+    assert pdf_page_count(data) == 1
+
+
+def test_image_jpeg_e2e(client):
+    job, data, _ = _e2e(client, "scan.jpg", _jpeg_bytes())
+    assert job["filename"] == "scan.pdf"
+    assert pdf_page_count(data) == 1
+
+
+def test_image_webp_e2e(client):
+    job, data, _ = _e2e(client, "snap.webp", _webp_bytes())
+    assert job["filename"] == "snap.pdf"
+    assert pdf_page_count(data) == 1
+
+
+def test_uppercase_extension_e2e(client):
+    job, _, text = _e2e(client, "UPPER.HTML", "<p>UPPER MARK ECHO</p>")
+    assert job["filename"] == "UPPER.pdf"
+    assert "UPPER MARK ECHO" in text
+
+
+# --------------------------------------------------------------------------
+# error paths (accepted 202, job ends in error with a clear message)
+# --------------------------------------------------------------------------
+
+
+def _error_job(client, filename, content):
+    resp = _submit(client, filename, content)
+    assert resp.status_code == 202, resp.get_json()
+    job = wait_for_job(client, resp.get_json()["job_id"])
+    assert job["status"] == "error", job
+    return job
+
+
+def test_invalid_json_is_job_error(client):
+    job = _error_job(client, "bad.json", "{not valid json]")
+    assert "JSON" in job["error"]
+
+
+def test_invalid_xml_is_job_error(client):
+    job = _error_job(client, "bad.xml", "<a><b></a>")
+    assert "XML" in job["error"]
+
+
+def test_invalid_svg_is_job_error(client):
+    job = _error_job(client, "bad.svg", "<svg><g></path>")
+    assert job["error"]
+
+
+def test_corrupt_png_is_job_error(client):
+    job = _error_job(client, "fake.png", b"\x89PNG\r\n\x1a\n definitely not a png")
+    assert "image" in job["error"].lower()
+
+
+def test_empty_txt_is_job_error(client):
+    job = _error_job(client, "empty.txt", b"")
+    assert "empty" in job["error"].lower()
+
+
+def test_empty_csv_is_job_error(client):
+    job = _error_job(client, "empty.csv", b"")
+    assert "empty" in job["error"].lower()
+
+
+def test_empty_json_is_job_error(client):
+    job = _error_job(client, "empty.json", b"   ")
+    assert "JSON" in job["error"]
+
+
+# --------------------------------------------------------------------------
+# naming contract (unit level via converters.common)
+# --------------------------------------------------------------------------
+
+
+def test_naming_sanitize_special_chars(tmp_path):
+    name = conv_common.derive_export_name("My File!!  Final.txt", tmp_path)
+    assert name == "My File_  Final.pdf"
+
+
+def test_naming_60_char_cap(tmp_path):
+    stem = "x" * 120
+    name = conv_common.derive_export_name(f"{stem}.txt", tmp_path)
+    assert name == ("x" * 60) + ".pdf"
+
+
+def test_naming_strips_doubled_pdf(tmp_path):
+    assert conv_common.derive_export_name("report.PDF.txt", tmp_path) == "report.pdf"
+    assert conv_common.derive_export_name("x.pdf.txt", tmp_path) == "x.pdf"
+
+
+def test_naming_empty_stem_falls_back_to_book(tmp_path):
+    assert conv_common.derive_export_name(".txt", tmp_path) == "book.pdf"
+
+
+def test_naming_collision_first_free(tmp_path):
+    (tmp_path / "dup.pdf").write_bytes(b"old1")
+    (tmp_path / "dup-2.pdf").write_bytes(b"old2")
+    assert conv_common.derive_export_name("dup.txt", tmp_path) == "dup-3.pdf"
+
+
+def test_naming_collision_e2e_second_upload_gets_suffix(client):
+    job1, _, _ = _e2e(client, "same.txt", "one")
+    job2, _, _ = _e2e(client, "same.txt", "two")
+    assert job1["filename"] == "same.pdf"
+    assert job2["filename"] == "same-2.pdf"
+
+
+# --------------------------------------------------------------------------
+# download + worker hygiene for non-EPUB conversions
+# --------------------------------------------------------------------------
+
+
+def test_download_headers_for_txt_conversion(client):
+    job, _, _ = _e2e(client, "headers.txt", "header check")
+    resp = client.get(f"/download/{job['filename']}")
+    assert resp.status_code == 200
+    assert resp.content_type == "application/pdf"
+    assert "attachment" in resp.headers["Content-Disposition"]
+    assert 'filename="headers.pdf"' in resp.headers["Content-Disposition"]
+
+
+def test_worker_cleans_upload_and_work_dir(client):
+    resp = _submit(client, "hygiene.txt", "cleanup check")
+    job_id = resp.get_json()["job_id"]
+    job = wait_for_job(client, job_id)
+    assert job["status"] == "done", job
+    assert not (app.UPLOAD_DIR / f"{job_id}.txt").exists()
+    assert not (app.WORK_DIR / f"job-{job_id}").exists()
+
+
+def test_job_snapshot_keeps_original_name(client):
+    resp = _submit(client, "report Q3.txt", "x")
+    job = wait_for_job(client, resp.get_json()["job_id"])
+    assert job["status"] == "done", job
+    assert job["epub"] == "report Q3.txt"  # UI bookname source, now generic
+    assert job["message"] == "Done"
+
+
+# --------------------------------------------------------------------------
+# CLI dispatch for non-EPUB files
+# --------------------------------------------------------------------------
+
+
+def _run_cli(*args):
+    return subprocess.run(
+        [sys.executable, str(ROOT / "app.py"), *args],
+        capture_output=True, text=True, timeout=120,
+        env=dict(os.environ), cwd=str(ROOT),
+    )
+
+
+def test_cli_markdown_file(tmp_path):
+    src = tmp_path / "notes.md"
+    src.write_text("# CLI MD Title\n\nCLI body text.", encoding="utf-8")
+    out = tmp_path / "out"
+    proc = _run_cli("--cli", str(src), "--output", str(out))
+    assert proc.returncode == 0, proc.stderr
+    assert "Saved:" in proc.stdout
+    pdfs = list(out.glob("*.pdf"))
+    assert len(pdfs) == 1 and pdfs[0].name == "CLI MD Title.pdf"
+    assert pdfs[0].read_bytes()[:4] == b"%PDF"
+
+
+def test_cli_unsupported_ext_exits_1(tmp_path):
+    src = tmp_path / "prog.exe"
+    src.write_bytes(b"MZ")
+    proc = _run_cli("--cli", str(src), "--output", str(tmp_path))
+    assert proc.returncode == 1
+    assert "error:" in proc.stdout
+
+
+# --------------------------------------------------------------------------
+# PDF pass-through + added image formats (Phase 1 gap patch)
+# --------------------------------------------------------------------------
+
+
+def _blank_pdf_bytes(pages=3):
+    w = PdfWriter()
+    for _ in range(pages):
+        w.add_blank_page(width=612, height=792)
+    buf = io.BytesIO()
+    w.write(buf)
+    return buf.getvalue()
+
+
+def _encrypted_pdf_bytes():
+    w = PdfWriter()
+    for _ in range(2):
+        w.add_blank_page(width=612, height=792)
+    w.encrypt(user_password="secret", owner_password="owner")
+    buf = io.BytesIO()
+    w.write(buf)
+    return buf.getvalue()
+
+
+def _tiff_bytes(size=(320, 240), color=(120, 30, 200)):
+    buf = io.BytesIO()
+    Image.new("RGB", size, color).save(buf, format="TIFF")
+    return buf.getvalue()
+
+
+def _bmp_bytes(size=(320, 240), color=(200, 120, 30)):
+    buf = io.BytesIO()
+    Image.new("RGB", size, color).save(buf, format="BMP")
+    return buf.getvalue()
+
+
+def test_pdf_passthrough_e2e(client):
+    # req 1: valid pdf passes through — same page count, /Title from stem.
+    job, data, _ = _e2e(client, "My Report.pdf", _blank_pdf_bytes(3))
+    assert job["filename"] == "My Report.pdf"
+    assert pdf_page_count(data) == 3
+    assert pdf_metadata_title(data) == "My Report"
+
+
+def test_pdf_encrypted_rejected(client):
+    # req 1: encrypted / password-protected pdf -> job error, not a crash.
+    job = _error_job(client, "locked.pdf", _encrypted_pdf_bytes())
+    err = job["error"].lower()
+    assert "encrypted" in err or "password" in err
+
+
+def test_pdf_corrupt_rejected(client):
+    # req 1: corrupt pdf -> job error naming the problem.
+    job = _error_job(client, "broken.pdf", b"%PDF-1.4\n" + b"garbage" * 64)
+    assert "pdf" in job["error"].lower()
+
+
+def test_pdf_over_500_pages_rejected(client):
+    # req 1: page-count cap (max 500) -> clear job error.
+    job = _error_job(client, "huge.pdf", _blank_pdf_bytes(501))
+    assert "too large" in job["error"].lower()
+
+
+def test_image_tiff_e2e(client):
+    # req 2: .tiff now accepted by the image converter.
+    job, data, _ = _e2e(client, "film.tiff", _tiff_bytes())
+    assert job["filename"] == "film.pdf"
+    assert pdf_page_count(data) == 1
+
+
+def test_image_bmp_e2e(client):
+    # req 2: .bmp now accepted by the image converter.
+    job, data, _ = _e2e(client, "old.bmp", _bmp_bytes())
+    assert job["filename"] == "old.pdf"
+    assert pdf_page_count(data) == 1