feat(pdf): Home-style file list + Clear-all button

User feedback: the standard file_uploader didn't visually match
the Home page, and there was no obvious way to clear out
uploaded files between scans (have to refresh the browser tab).

**Persistent stash + add-only sync.** Files captured into
``st.session_state["pdf_uploads"]`` (dict name → {bytes, size})
via an ``on_change`` callback on the file_uploader widget. The
callback is **add-only** — never removes files from the stash
based on widget state. Removal is owned by the custom X buttons
+ widget-counter bump (see below). This guarantees a hidden
native X click can't silently drop files behind the user's
back.

**Hidden native file list.** A small CSS block suppresses the
file_uploader's built-in file rows + their delete buttons
(``stFileUploaderFile`` + ``stFileUploaderDeleteBtn``), so the
custom list below is the single source of truth on screen.

**Custom file list (Home pattern).** Below the dropzone, every
uploaded file gets a row: ``✕ | 📄 filename | size``. Top of
section shows ``N files · 12.3 MB total``. Counts and sizes
update in real time as the user adds or removes files. The X
button per row calls ``log_event("upload", "PDF removed: …")``,
removes the entry from the stash, and bumps the widget counter
to clear the widget too.

**Clear-all button.** Sits next to the Scan button. Wipes the
stash, bumps the widget counter, drops any cached scan results
(``K_ROWS``, ``K_WARNINGS``, ``K_SOURCE_COUNT``). Audited via
``log_event("upload", "PDF list cleared", count=N)``.

**Widget reset via counter bump.** Streamlit disallows
programmatic mutation of widget session-state entries; the
standard workaround is to rotate the widget's ``key``. Page
maintains ``K_UPLOAD_COUNTER`` which gets incremented on
remove / clear-all, producing a fresh ``pdf_upload_v{N}`` key
and a freshly-instantiated empty widget. The stash retains any
unaffected files; on next upload, the add-only sync picks up
the new ones without re-adding the removed ones.

**Scan rewired to read the stash.** Instead of iterating the
widget's UploadedFile objects (which the previous code did and
which broke when the widget unmounted on remove), the scan
loop iterates ``pdf_uploads.items()`` and uses the cached
``bytes``. Diagnostic expander does the same — re-reads from
the stash, removing the need for a separate ``K_DIAGNOSTIC``
cache (deleted).

**``_format_size`` helper** ports the byte-formatting logic
from ``_home.py``'s pattern (KB / MB / GB rollover).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

src/gui/pages/10_PDF_Extractor.py

@@ -8,6 +8,7 @@ no coordinate picking.
 
 from __future__ import annotations
 
+import hashlib
 import sys
 from datetime import datetime
 from pathlib import Path
@@ -59,7 +60,30 @@ render_sticky_footer()
 K_ROWS = "pdf_scan_rows"
 K_WARNINGS = "pdf_scan_warnings"
 K_SOURCE_COUNT = "pdf_scan_source_count"
-K_DIAGNOSTIC = "pdf_scan_diagnostic"
+# ``pdf_uploads`` is the persistent stash of uploaded PDFs (dict
+# keyed by filename → {"bytes": ..., "size": ...}). It survives
+# Streamlit reruns and navigation away from the page. The
+# uploader widget feeds this stash via ``_sync_pdf_uploads`` and
+# the custom file list / Clear-all button operate on it.
+K_UPLOADS = "pdf_uploads"
+# Bumped to force the file_uploader to re-instantiate (clear its
+# internal state) when the user removes a file via the custom X or
+# clicks Clear-all. Streamlit's widget state is keyed on the widget
+# key, so changing the key resets the widget without us having to
+# touch its session-state directly (which Streamlit disallows).
+K_UPLOAD_COUNTER = "pdf_upload_counter"
+
+
+def _format_size(n_bytes: int) -> str:
+    """Human-friendly file size — KB / MB / GB."""
+    size = float(n_bytes)
+    for unit in ("B", "KB", "MB", "GB"):
+        if size < 1024:
+            if unit == "B":
+                return f"{int(size)} {unit}"
+            return f"{size:.1f} {unit}"
+        size /= 1024
+    return f"{size:.1f} TB"
 
 
 # ---------------------------------------------------------------------------
@@ -143,38 +167,151 @@ with st.expander("Scan options", expanded=False):
             ),
         )
 
-uploads = st.file_uploader(
+# Persistent stash + rotating widget key. See K_UPLOADS / K_UPLOAD_COUNTER
+# docstrings for why the counter exists.
+pdf_uploads: dict = st.session_state.setdefault(K_UPLOADS, {})
+upload_counter: int = st.session_state.setdefault(K_UPLOAD_COUNTER, 0)
+uploader_key = f"pdf_upload_v{upload_counter}"
+
+# Hide the file_uploader's built-in file list (Streamlit shows
+# tiny chips with X buttons under its dropzone). We render our own
+# Home-style list below, so suppressing the native one leaves a
+# single source of truth on screen.
+st.markdown(
+    """<style>
+[data-testid="stFileUploader"] [data-testid="stFileUploaderFile"] {
+    display: none !important;
+}
+[data-testid="stFileUploader"] [data-testid="stFileUploaderDeleteBtn"] {
+    display: none !important;
+}
+</style>""",
+    unsafe_allow_html=True,
+)
+
+
+def _sync_pdf_uploads() -> None:
+    """``on_change`` callback. Adds newly-uploaded files to the
+    persistent stash. **Add-only** — removal happens through the
+    custom X buttons + counter bump, NOT through this callback.
+    That way the widget's hidden native X buttons can't silently
+    drop files behind the user's back, and we can ignore them.
+    """
+    widget_files = st.session_state.get(uploader_key) or []
+    for f in widget_files:
+        if f.name not in pdf_uploads:
+            pdf_uploads[f.name] = {
+                "bytes": f.getvalue(),
+                "size": f.size,
+            }
+            log_event(
+                "upload",
+                f"PDF: {f.name}",
+                filename=f.name,
+                bytes=f.size,
+                page="10_PDF_Extractor",
+            )
+
+
+st.file_uploader(
     "PDF file(s)",
     type=["pdf"],
     accept_multiple_files=True,
+    key=uploader_key,
+    on_change=_sync_pdf_uploads,
     help="Drop one or more bank-statement PDFs. Multi-file batches "
     "are merged into a single table with a ``source_file`` column.",
 )
 
-scan_clicked = st.button(
+
-    "Scan", type="primary", disabled=not uploads,
+# ---------------------------------------------------------------------------
-)
+# Custom file list (Home-style: one row per file, X to remove)
+# ---------------------------------------------------------------------------
+
+if pdf_uploads:
+    n = len(pdf_uploads)
+    total = sum(m["size"] for m in pdf_uploads.values())
+    word = "file" if n == 1 else "files"
+    st.markdown(
+        f"**{n} {word}** · {_format_size(total)} total",
+    )
+    to_remove: str | None = None
+    with st.container(border=True):
+        for name, meta in pdf_uploads.items():
+            digest = hashlib.sha1(
+                name.encode("utf-8"), usedforsecurity=False,
+            ).hexdigest()[:10]
+            col_x, col_name, col_size = st.columns([0.55, 8, 1.6])
+            if col_x.button(
+                "✕",
+                key=f"pdf_rm_{digest}",
+                help=f"Remove {name}",
+                type="tertiary",
+            ):
+                to_remove = name
+            col_name.markdown(f"📄 **{name}**")
+            col_size.markdown(
+                f"<div style='text-align:right;color:#5a5f6e;'>"
+                f"{_format_size(meta['size'])}</div>",
+                unsafe_allow_html=True,
+            )
+
+    c_scan, c_clear = st.columns([1, 4])
+    with c_scan:
+        scan_clicked = st.button("Scan", type="primary")
+    with c_clear:
+        if st.button(
+            "Clear all files",
+            type="secondary",
+            help="Removes all uploaded files and the last scan result.",
+        ):
+            st.session_state[K_UPLOADS] = {}
+            st.session_state[K_UPLOAD_COUNTER] = upload_counter + 1
+            for k in (K_ROWS, K_WARNINGS, K_SOURCE_COUNT):
+                st.session_state.pop(k, None)
+            log_event(
+                "upload",
+                "PDF list cleared",
+                page="10_PDF_Extractor",
+                count=n,
+            )
+            st.rerun()
+
+    if to_remove is not None:
+        log_event(
+            "upload",
+            f"PDF removed: {to_remove}",
+            filename=to_remove,
+            page="10_PDF_Extractor",
+        )
+        del pdf_uploads[to_remove]
+        # Bump the uploader counter so the widget re-instantiates
+        # and forgets the removed file. Without this, the user
+        # would have to click the widget's own X (which is hidden)
+        # OR re-upload to refresh the state.
+        st.session_state[K_UPLOAD_COUNTER] = upload_counter + 1
+        st.rerun()
+else:
+    st.caption("No files uploaded yet.")
+    scan_clicked = False
 
 
 # ---------------------------------------------------------------------------
 # Scan
 # ---------------------------------------------------------------------------
 
-if scan_clicked and uploads:
+if scan_clicked and pdf_uploads:
     all_rows: list[dict] = []
     all_warnings: list[str] = []
-    # Cache the raw bytes per file so the diagnostic expander can
+    n_files = len(pdf_uploads)
-    # re-extract lines without asking the user to re-upload.
-    cached_bytes: list[tuple[str, bytes]] = []
     with st.status(
-        f"Scanning {len(uploads)} file(s)…",
+        f"Scanning {n_files} file(s)…",
         expanded=True,
     ) as status:
-        for i, up in enumerate(uploads, start=1):
+        for i, (name, meta) in enumerate(pdf_uploads.items(), start=1):
-            st.write(f"**{i}/{len(uploads)}** · {up.name}")
+            st.write(f"**{i}/{n_files}** · {name}")
             try:
-                raw = up.read()
+                raw = meta["bytes"]
-                cached_bytes.append((up.name, raw))
                 rows, warns = scan_pdf_for_transactions(
                     raw,
                     negative_in_parens=negative_in_parens,
@@ -182,19 +319,19 @@ if scan_clicked and uploads:
                     output_date_format=output_date_format,
                 )
                 for r in rows:
-                    r["source_file"] = up.name
+                    r["source_file"] = name
                 all_rows.extend(rows)
-                all_warnings.extend(f"[{up.name}] {w}" for w in warns)
+                all_warnings.extend(f"[{name}] {w}" for w in warns)
             except PdfDependencyMissing as e:
-                all_warnings.append(f"[{up.name}] {e}")
+                all_warnings.append(f"[{name}] {e}")
             except Exception as e:
                 all_warnings.append(
-                    f"[{up.name}] scan failed: {type(e).__name__}: {e}"
+                    f"[{name}] scan failed: {type(e).__name__}: {e}"
                 )
         status.update(
             label=(
                 f"Found {len(all_rows):,} candidate transactions "
-                f"across {len(uploads)} file(s)"
+                f"across {n_files} file(s)"
             ),
             state="complete",
             expanded=False,
@@ -202,14 +339,13 @@ if scan_clicked and uploads:
 
     st.session_state[K_ROWS] = all_rows
     st.session_state[K_WARNINGS] = all_warnings
-    st.session_state[K_SOURCE_COUNT] = len(uploads)
+    st.session_state[K_SOURCE_COUNT] = n_files
-    st.session_state[K_DIAGNOSTIC] = cached_bytes
 
     log_event(
         "tool_run",
         "PDF scan",
         page="10_PDF_Extractor",
-        files=len(uploads),
+        files=n_files,
         rows=len(all_rows),
         warnings=len(all_warnings),
     )
@@ -229,7 +365,7 @@ if warnings:
             st.warning(w)
 
 if rows is None:
-    if uploads:
+    if pdf_uploads:
         st.info("Click **Scan** to detect transactions.")
     else:
         st.info("Upload one or more PDF files to begin.")
@@ -242,13 +378,13 @@ elif not rows:
         "``has_date`` and ``has_amount`` columns to spot which "
         "pieces are missing (usually one or the other)."
     )
-    cached_bytes = st.session_state.get(K_DIAGNOSTIC) or []
+    if pdf_uploads:
-    if cached_bytes:
         with st.expander(
             "Diagnostic: what the scanner saw",
             expanded=True,
         ):
-            for fname, raw in cached_bytes:
+            for fname, meta in pdf_uploads.items():
+                raw = meta["bytes"]
                 st.markdown(f"**{fname}**")
                 try:
                     lines, dwarns = diagnose_pdf_lines(