Add junk label, clear-all button, and previous sheet navigation

- Junk label (J key) for marking useless content discarded from chunks
- Clear All button to reset all labels on current sheet (with undo)
- Previous sheet navigation with history dropdown for re-editing
- Backend endpoints: /api/history, /api/load_labeled_sheet
- Updated .gitignore for deployment safety

Made-with: Cursor

.gitignore

@@ -1,19 +1,10 @@
 credentials/
+data/
 *.pyc
 __pycache__/
 .env
-data/labeled/
-data/tmp/
-data/raw/
-.DS_Store
-*.egg-info/
-dist/
-build/
 .sfdx/
+.cursor/
 spreadsheet-labeller-*.json
+.DS_Store
 docs/
-*.pt
-data/experiments/
-data/models/
-data/splits/
-data/spreadsheet_gnn.pt

app.py

@@ -318,6 +318,114 @@ def get_progress():
     return metadata_client.get_progress()
 
 
+@app.get("/api/history")
+def get_labeling_history(
+    labeler: str = Query("anonymous"),
+    limit: int = Query(20),
+    _=Depends(_require_auth),
+):
+    """Return recently labeled sheets by this labeler, newest first."""
+    rows = metadata_client.get_all_rows()
+    labeled = [
+        {
+            "file_name": r["file_name"],
+            "drive_file_id": r["drive_file_id"],
+            "sheet_name": r["sheet_name"],
+            "timestamp": r["timestamp"],
+            "num_labeled_cells": r["num_labeled_cells"],
+            "meta_row_index": r["_row_index"],
+        }
+        for r in rows
+        if r["status"] == "labelled" and r["labeler"] == labeler
+    ]
+    labeled.sort(key=lambda r: r.get("timestamp", ""), reverse=True)
+    return {"sheets": labeled[:limit]}
+
+
+@app.get("/api/load_labeled_sheet")
+def load_labeled_sheet(
+    drive_file_id: str = Query(...),
+    sheet_name: str = Query(...),
+    meta_row_index: int = Query(...),
+    _=Depends(_require_auth),
+):
+    """Load a previously labeled sheet with its saved labels for re-editing."""
+    # Load the grid from xlsx
+    try:
+        local_path = str(config.TEMP_DIR / f"{drive_file_id}.xlsx")
+        if not Path(local_path).exists():
+            drive_client.download_file(drive_file_id, local_path)
+        grid = xlsx_client.fetch_xlsx_sheet(local_path, sheet_name)
+    except Exception as e:
+        logger.exception("Error reading sheet for re-edit: %s/%s", drive_file_id, sheet_name)
+        raise HTTPException(status_code=502, detail=f"Sheet read error: {e}")
+
+    grid["drive_file_id"] = drive_file_id
+    grid["meta_row_index"] = meta_row_index
+
+    # Find saved labels: try local file first, then Drive
+    saved_labels = _load_saved_labels(drive_file_id, sheet_name, grid["spreadsheet_id"])
+
+    return {
+        "sheet": grid,
+        "saved_labels": saved_labels,
+    }
+
+
+def _load_saved_labels(
+    drive_file_id: str, sheet_name: str, spreadsheet_id: str,
+) -> list[dict] | None:
+    """Load previously saved labels from local JSON or Drive."""
+    safe_name = f"{spreadsheet_id}_{sheet_name}".replace("/", "_").replace(" ", "_")
+    local_path = config.LABELED_DIR / f"{safe_name}.json"
+
+    data = None
+    if local_path.exists():
+        with open(local_path) as f:
+            data = json.load(f)
+    else:
+        # Try finding it on Drive via metadata
+        rows = metadata_client.get_all_rows()
+        match = next(
+            (r for r in rows
+             if r["drive_file_id"] == drive_file_id
+             and r["sheet_name"] == sheet_name
+             and r["labels_file_id"]
+             and r["labels_file_id"] != "local-only"),
+            None,
+        )
+        if match:
+            try:
+                tmp = str(config.TEMP_DIR / f"labels_{safe_name}.json")
+                drive_client.download_file(match["labels_file_id"], tmp)
+                with open(tmp) as f:
+                    data = json.load(f)
+            except Exception:
+                logger.warning("Could not download labels from Drive for %s", safe_name)
+
+    if not data or "cells" not in data:
+        return None
+
+    result = []
+    for c in data["cells"]:
+        label = c.get("label", "unlabeled")
+        if label == "unlabeled":
+            continue
+        entry = {
+            "row": c["row"],
+            "col": c["col"],
+            "label": label,
+            "table": c.get("table", 0),
+        }
+        target = c.get("comment_target")
+        if label == "comment" and target:
+            entry["comment_target_row"] = target["row"]
+            entry["comment_target_col"] = target["col"]
+        result.append(entry)
+
+    return result
+
+
 @app.post("/api/ingest")
 def ingest_data(
     source: str = Query("sheetpedia"),

static/index.html

@@ -23,6 +23,26 @@
     border:2px solid #dc2626;border-radius:8px;cursor:pointer;transition:all .15s;white-space:nowrap}
   #clear-all-btn:hover{background:#dc2626;color:#fff}
 
+  /* Prev button & history dropdown */
+  .prev-wrapper{position:relative}
+  #prev-btn{padding:8px 16px;font-size:13px;font-weight:600;background:#fff;color:#555;
+    border:2px solid #ccc;border-radius:8px;cursor:pointer;transition:all .15s;white-space:nowrap}
+  #prev-btn:hover{border-color:#2563eb;color:#2563eb}
+  .history-popup{position:absolute;top:100%;left:0;margin-top:6px;background:#fff;
+    border:1px solid #ddd;border-radius:8px;box-shadow:0 4px 16px rgba(0,0,0,.12);
+    z-index:200;min-width:340px;max-height:360px;overflow-y:auto;display:none}
+  .history-popup.open{display:block}
+  .history-popup .history-header{padding:10px 14px;font-size:13px;font-weight:600;color:#333;
+    border-bottom:1px solid #eee}
+  .history-popup .history-empty{padding:16px 14px;font-size:13px;color:#888;text-align:center}
+  .history-popup .history-item{padding:8px 14px;cursor:pointer;border-bottom:1px solid #f0f0f0;
+    transition:background .1s;display:flex;flex-direction:column;gap:2px}
+  .history-popup .history-item:hover{background:#f0f7ff}
+  .history-popup .history-item:last-child{border-bottom:none}
+  .history-popup .history-item .hi-name{font-size:13px;font-weight:500;color:#222;
+    overflow:hidden;text-overflow:ellipsis;white-space:nowrap}
+  .history-popup .history-item .hi-meta{font-size:11px;color:#888}
+
   /* Selector bars (table & level) */
   .selector-bar{display:flex;align-items:center;gap:6px;padding:0 24px;background:#fff;
     border-bottom:1px solid #eee;height:34px}
@@ -219,6 +239,10 @@
     <button data-lang="pl">PL</button>
   </div>
   <button id="help-btn" title="Labeling guide / Instrukcja">?</button>
+  <div class="prev-wrapper" id="prev-wrapper">
+    <button id="prev-btn" data-i18n="prev_btn">&larr; PREV</button>
+    <div class="history-popup" id="history-popup"></div>
+  </div>
   <button id="clear-all-btn" data-i18n="clear_all_btn">CLEAR ALL</button>
   <button id="save-btn" disabled data-i18n="save_btn">SAVE &amp; NEXT</button>
 </div>
@@ -271,8 +295,12 @@
       loading_sheet: "Loading first sheet\u2026",
       username_placeholder: "Your name",
       save_btn: "SAVE & NEXT",
+      prev_btn: "\u2190 PREV",
       clear_all_btn: "CLEAR ALL",
       clear_all_confirm: "Remove all labels from this sheet? This can be undone with Ctrl+Z.",
+      history_title: "Recently labeled sheets",
+      history_empty: "No labeled sheets yet.",
+      history_cells: "cells",
       saving: "Saving\u2026",
       done: "Done!",
       all_done: "All sheets have been labeled!",
@@ -314,8 +342,12 @@
       loading_sheet: "\u0141adowanie pierwszego arkusza\u2026",
       username_placeholder: "Twoje imi\u0119",
       save_btn: "ZAPISZ I DALEJ",
+      prev_btn: "\u2190 WRÓĆ",
       clear_all_btn: "WYCZYŚĆ WSZYSTKO",
       clear_all_confirm: "Usunąć wszystkie etykiety z tego arkusza? Można cofnąć przez Ctrl+Z.",
+      history_title: "Ostatnio oznaczone arkusze",
+      history_empty: "Brak oznaczonych arkuszy.",
+      history_cells: "komórek",
       saving: "Zapisywanie\u2026",
       done: "Gotowe!",
       all_done: "Wszystkie arkusze zosta\u0142y oznaczone!",
@@ -1226,6 +1258,118 @@
     drawCommentLinks();
   });
 
+  /* ── Previous sheet / history ──────────────────────────────────────── */
+  const prevBtn = document.getElementById("prev-btn");
+  const historyPopup = document.getElementById("history-popup");
+
+  prevBtn.addEventListener("click", async () => {
+    if (historyPopup.classList.contains("open")) {
+      historyPopup.classList.remove("open");
+      return;
+    }
+    const username = usernameInput.value.trim() || "anonymous";
+    historyPopup.innerHTML = `<div class="history-header">${t("history_title")}</div>
+      <div class="history-empty">${t("loading")}</div>`;
+    historyPopup.classList.add("open");
+
+    try {
+      const resp = await fetch(`/api/history?labeler=${encodeURIComponent(username)}`);
+      const data = await resp.json();
+      const sheets = data.sheets || [];
+      if (sheets.length === 0) {
+        historyPopup.innerHTML = `<div class="history-header">${t("history_title")}</div>
+          <div class="history-empty">${t("history_empty")}</div>`;
+        return;
+      }
+      let html = `<div class="history-header">${t("history_title")}</div>`;
+      for (const s of sheets) {
+        const ts = s.timestamp ? new Date(s.timestamp).toLocaleString() : "";
+        const cells = s.num_labeled_cells ? `${s.num_labeled_cells} ${t("history_cells")}` : "";
+        const meta = [ts, cells].filter(Boolean).join(" · ");
+        html += `<div class="history-item" data-dfid="${s.drive_file_id}"
+                      data-sn="${escHtml(s.sheet_name)}" data-mri="${s.meta_row_index}">
+          <span class="hi-name">${escHtml(s.file_name)} / ${escHtml(s.sheet_name)}</span>
+          <span class="hi-meta">${meta}</span>
+        </div>`;
+      }
+      historyPopup.innerHTML = html;
+
+      historyPopup.querySelectorAll(".history-item").forEach(item => {
+        item.addEventListener("click", () => {
+          historyPopup.classList.remove("open");
+          loadPreviousSheet(
+            item.dataset.dfid,
+            item.dataset.sn,
+            parseInt(item.dataset.mri),
+          );
+        });
+      });
+    } catch (err) {
+      historyPopup.innerHTML = `<div class="history-empty" style="color:#dc2626">${err.message}</div>`;
+    }
+  });
+
+  document.addEventListener("click", (e) => {
+    if (!e.target.closest("#prev-wrapper")) historyPopup.classList.remove("open");
+  });
+
+  async function loadPreviousSheet(driveFileId, sheetName, metaRowIndex) {
+    saveBtn.disabled = true;
+    tableWrap.innerHTML = `<p style="padding:40px;color:#888">${t("loading")}</p>`;
+
+    try {
+      const url = `/api/load_labeled_sheet?drive_file_id=${encodeURIComponent(driveFileId)}`
+        + `&sheet_name=${encodeURIComponent(sheetName)}`
+        + `&meta_row_index=${metaRowIndex}`;
+      const resp = await fetch(url);
+      if (!resp.ok) throw new Error(`HTTP ${resp.status}`);
+      const data = await resp.json();
+
+      currentSheet = data.sheet;
+      cellLabels = {};
+      commentLinks = {};
+      exitLinkMode();
+      selectedCells.clear();
+      lastClickedKey = null;
+      maxTableId = 1;
+      maxLevel = 3;
+      activeLevel = 1;
+      clearUndoRedo();
+
+      if (data.saved_labels) {
+        prefillFromSavedLabels(data.saved_labels);
+      }
+
+      refreshDynamicCSS();
+      renderTableBar();
+      renderLevelBar();
+      renderLegend();
+      setActiveTable(1);
+      sheetInfoEl.textContent = `${currentSheet.spreadsheet_id} / ${currentSheet.sheet_name}`;
+      renderTable();
+      saveBtn.disabled = false;
+    } catch (err) {
+      tableWrap.innerHTML = `<p style="padding:40px;color:red">${t("error_prefix")}${err.message}</p>`;
+    }
+  }
+
+  function prefillFromSavedLabels(labels) {
+    if (!labels || !labels.length) return;
+    for (const c of labels) {
+      const key = `${c.row},${c.col}`;
+      cellLabels[key] = { label: c.label, table: c.table || 0 };
+      if (c.table > maxTableId) maxTableId = c.table;
+      const m = c.label.match(/^(?:row_header|col_header)_(\d+)$/);
+      if (m) {
+        const lv = parseInt(m[1]);
+        if (lv > maxLevel) maxLevel = lv;
+      }
+      if (c.label === "comment" && c.comment_target_row != null) {
+        commentLinks[key] = `${c.comment_target_row},${c.comment_target_col}`;
+      }
+    }
+  }
+
   /* ── Save & Next ───────────────────────────────────────────────────── */
   saveBtn.addEventListener("click", doSave);