feat(gui): hidden-char-aware preview tables in Text Cleaner

The Text Cleaner had two st.dataframe previews — the initial upload
preview ("Preview: filename") and the post-clean "Cleaned preview"
table — that both rendered cells with the same browser-collapses-
whitespace, hides-invisibles problem the analyzer findings panel had
before commit 1049c03.

components.render_hidden_aware_preview(df, n_rows, caption) renders a
DataFrame as an HTML table where:
  - every cell uses visualize_hidden_html(mark_outer_whitespace=True),
    so leading/trailing ASCII spaces appear as per-character "·" badges
  - white-space: pre-wrap on every cell preserves internal multi-space
    runs and embedded newlines visually
  - headers route through the same visualizer so dirty column names
    (NBSP padding, ZWSP, smart quotes) show their badges too
  - NaN cells render as a faint "NaN" placeholder
  - rows are sticky-headed and scrollable inside a 26rem capped
    container so a 10-row preview doesn't push the rest of the UI off
    screen

2_Text_Cleaner.py wires it into both previews:
  - The upload preview gains its own "Show hidden characters in preview"
    toggle (default on).
  - The cleaned preview reuses the existing show_hidden toggle that
    already governs the Examples changes table, so one switch controls
    the whole results section.

Either toggle off falls back to the original st.dataframe view.

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

src/gui/components.py

@@ -797,6 +797,111 @@ def render_findings_panel(findings, *, header: str = "Detected issues") -> None:
                 _render_one_finding(f)
 
 
+_PREVIEW_TABLE_CSS = """
+<style>
+.hidden-aware-preview {
+    width: 100%;
+    border-collapse: collapse;
+    font-size: 0.9em;
+}
+.hidden-aware-preview th,
+.hidden-aware-preview td {
+    padding: 4px 8px;
+    border: 1px solid #eee;
+    text-align: left;
+    vertical-align: top;
+    font-family: ui-monospace, SFMono-Regular, monospace;
+    /* pre-wrap so internal ASCII whitespace and embedded newlines render
+       as the user wrote them; otherwise browsers collapse adjacent spaces. */
+    white-space: pre-wrap;
+    word-break: break-word;
+    max-width: 32em;
+}
+.hidden-aware-preview thead th {
+    background: #f6f8fa;
+    position: sticky;
+    top: 0;
+}
+.hidden-aware-preview tbody tr:nth-child(even) { background: #fafafa; }
+.hidden-aware-preview .row-num {
+    color: #888;
+    font-family: inherit;
+    background: #f6f8fa;
+    text-align: right;
+}
+.hidden-aware-preview-wrap {
+    max-height: 26rem;
+    overflow: auto;
+    border: 1px solid #eee;
+    border-radius: 4px;
+}
+</style>
+"""
+
+
+def render_hidden_aware_preview(
+    df,
+    *,
+    n_rows: int = 10,
+    caption: str | None = None,
+) -> None:
+    """Render a DataFrame preview that shows hidden characters in every cell.
+
+    Used for the Text Cleaner's "before" and "after" previews so the user
+    can actually see the leading/trailing whitespace, NBSP padding,
+    zero-width characters, and smart punctuation that the cleaner is going
+    to remove (or just removed). A plain ``st.dataframe`` collapses outer
+    ASCII whitespace and renders invisibles as nothing, defeating the
+    point of a preview in a cleanup tool.
+
+    Headers and cell values are both routed through
+    :func:`visualize_hidden_html` with ``mark_outer_whitespace=True``.
+    """
+    import pandas as pd
+    from src.core.text_clean import hidden_char_css, visualize_hidden_html
+
+    if df is None or len(df) == 0:
+        st.info("No rows to preview.")
+        return
+
+    sliced = df.head(n_rows) if len(df) > n_rows else df
+
+    st.markdown(hidden_char_css() + _PREVIEW_TABLE_CSS, unsafe_allow_html=True)
+    if caption:
+        st.caption(caption)
+
+    header_cells = "".join(
+        f"<th>{visualize_hidden_html(str(c), mark_outer_whitespace=True)}</th>"
+        for c in sliced.columns
+    )
+
+    body_rows: list[str] = []
+    for row_idx, (orig_idx, row) in enumerate(sliced.iterrows(), start=1):
+        cells = ["<td class='row-num'>" + str(row_idx) + "</td>"]
+        for col in sliced.columns:
+            value = row[col]
+            if isinstance(value, str):
+                rendered = visualize_hidden_html(value, mark_outer_whitespace=True)
+            elif pd.isna(value):
+                rendered = "<span style='color:#aaa'>NaN</span>"
+            else:
+                # Non-string scalars (numerics, bools) just stringify; they
+                # won't have invisible chars but we still need html-escape.
+                rendered = visualize_hidden_html(str(value))
+            cells.append(f"<td>{rendered}</td>")
+        body_rows.append("<tr>" + "".join(cells) + "</tr>")
+
+    st.markdown(
+        "<div class='hidden-aware-preview-wrap'>"
+        "<table class='hidden-aware-preview'>"
+        f"<thead><tr><th class='row-num'>#</th>{header_cells}</tr></thead>"
+        f"<tbody>{''.join(body_rows)}</tbody>"
+        "</table>"
+        "</div>",
+        unsafe_allow_html=True,
+    )
+
+
 _SAMPLE_TABLE_CSS = """
 <style>
 .findings-sample-table {

src/gui/pages/2_Text_Cleaner.py

@@ -14,7 +14,11 @@ _project_root = Path(__file__).resolve().parent.parent.parent.parent
 if str(_project_root) not in sys.path:
     sys.path.insert(0, str(_project_root))
 
-from src.gui.components import hide_streamlit_chrome, pickup_or_upload
+from src.gui.components import (
+    hide_streamlit_chrome,
+    pickup_or_upload,
+    render_hidden_aware_preview,
+)
 from src.core.text_clean import (
     PRESETS,
     CleanOptions,
@@ -81,7 +85,16 @@ except Exception as e:
 
 st.subheader(f"Preview: {uploaded.name}")
 st.caption(f"{len(df)} rows, {len(df.columns)} columns")
-st.dataframe(df.head(10), use_container_width=True)
+preview_show_hidden = st.toggle(
+    "Show hidden characters in preview",
+    value=True,
+    help="Highlights NBSP, zero-width chars, smart quotes, and leading/trailing whitespace.",
+    key="textclean_preview_show_hidden",
+)
+if preview_show_hidden:
+    render_hidden_aware_preview(df, n_rows=10)
+else:
+    st.dataframe(df.head(10), use_container_width=True)
 
 st.divider()
 
@@ -257,7 +270,12 @@ if result.cells_changed:
         st.dataframe(examples, use_container_width=True, hide_index=True)
 
 st.markdown("**Cleaned preview (first 10 rows)**")
-st.dataframe(result.cleaned_df.head(10), use_container_width=True)
+# Reuse the same toggle the Examples table uses so the user controls both
+# the changes audit and the cleaned preview with one switch.
+if show_hidden:
+    render_hidden_aware_preview(result.cleaned_df, n_rows=10)
+else:
+    st.dataframe(result.cleaned_df.head(10), use_container_width=True)
 
 # ---------------------------------------------------------------------------
 # Downloads