feat: report indexing progress with page counts in status panel

backend/app/indexer.py

@@ -38,6 +38,9 @@ class IndexStatus:
     model: str = ""
     last_indexed_at: str | None = None
     errors: list[str] = field(default_factory=list)
+    pages_processed: int = 0
+    pages_total: int = 0
+    current_file: str | None = None
 
 
 @dataclass
@@ -78,6 +81,10 @@ class Indexer:
         self._lock = threading.Lock()
         self._last_indexed_at: str | None = None
         self._errors: list[str] = []
+        self._indexing = False
+        self._pages_processed = 0
+        self._pages_total = 0
+        self._current_file: str | None = None
         self._load_manifest()
 
     # ------------------------------------------------------------------ #
@@ -154,6 +161,21 @@ class Indexer:
     # Indexação
     # ------------------------------------------------------------------ #
 
+    def mark_indexing(self) -> None:
+        """Sinaliza que uma indexação está em andamento."""
+        self._indexing = True
+        self._pages_processed = 0
+        self._pages_total = 0
+        self._current_file = None
+
+    def report_progress(
+        self, pdf_name: str, pages_done: int, pages_total: int
+    ) -> None:
+        """Registra o progresso da extração de um arquivo."""
+        self._current_file = pdf_name
+        self._pages_processed = pages_done
+        self._pages_total = pages_total
+
     def discover_pdfs(self) -> list[Path]:
         """Lista os PDFs do diretório configurado, ordenados por nome."""
         directory = self.settings.pdf_dir
@@ -165,8 +187,10 @@ class Indexer:
         """Indexa os PDFs novos ou alterados."""
         with self._lock:
             self._errors = []
+            self.mark_indexing()
             pdfs = self.discover_pdfs()
 
+            try:
                 for pdf in pdfs:
                     try:
                         self._index_file(pdf, force=force)
@@ -179,6 +203,9 @@ class Indexer:
 
                 self._save_manifest()
                 self._last_indexed_at = _now()
+            finally:
+                self._indexing = False
+
             return self.status()
 
     def _index_file(self, pdf: Path, force: bool) -> None:
@@ -190,7 +217,15 @@ class Indexer:
             return
 
         logger.info("Indexando %s", pdf.name)
-        pages = list(self.extractor.extract_document(pdf, file_hash=file_hash))
+        pages = []
+        page_count = _pdf_page_count(pdf)
+        for position, page in enumerate(
+            self.extractor.extract_document(pdf, file_hash=file_hash), start=1
+        ):
+            pages.append(page)
+            if position % 10 == 0 or position == page_count:
+                self.report_progress(pdf.name, position, page_count)
+
         failed_pages = [page.page for page in pages if page.failed]
 
         book_starts = self._book_starts(pages)
@@ -229,16 +264,36 @@ class Indexer:
             for entry in self._manifest.values()
             if entry.state in (STATE_COMPLETE, STATE_PARTIAL)
         )
+        if self._indexing:
+            state = "indexing"
+        elif self._errors and indexed == 0:
+            state = "error"
+        else:
+            state = "ready"
         return IndexStatus(
-            state="error" if self._errors and indexed == 0 else "ready",
+            state=state,
             pdfs_total=len(self.discover_pdfs()),
             pdfs_indexed=indexed,
             chunks=self.store.count(),
             model=self.settings.embedding_model,
             last_indexed_at=self._last_indexed_at,
             errors=list(self._errors),
+            pages_processed=self._pages_processed,
+            pages_total=self._pages_total,
+            current_file=self._current_file,
         )
 
 
+def _pdf_page_count(pdf: Path) -> int:
+    """Número de páginas do PDF, ou 0 se não for possível determinar."""
+    try:
+        import pymupdf
+
+        with pymupdf.open(pdf) as document:
+            return len(document)
+    except Exception:  # pragma: no cover - PDF malformado
+        return 0
+
+
 def _now() -> str:
     return datetime.now(timezone.utc).isoformat()

backend/app/main.py

@@ -182,6 +182,9 @@ def create_app(
             model=current.model,
             last_indexed_at=current.last_indexed_at,
             errors=current.errors,
+            pages_processed=current.pages_processed,
+            pages_total=current.pages_total,
+            current_file=current.current_file,
         )
 
     @app.post("/api/reindex", response_model=ReindexResponse, status_code=202)

backend/app/models.py

@@ -38,6 +38,9 @@ class StatusResponse(BaseModel):
     model: str = ""
     last_indexed_at: str | None = None
     errors: list[str] = Field(default_factory=list)
+    pages_processed: int = 0
+    pages_total: int = 0
+    current_file: str | None = None
 
 
 class HealthResponse(BaseModel):

backend/tests/test_indexer.py

@@ -207,3 +207,30 @@ def test_non_pdf_files_are_ignored(settings):
 
     assert status.pdfs_total == 0
     assert extractor.calls == []
+
+
+def test_status_reports_indexing_while_work_is_in_progress(settings):
+    """Durante a indexação o status deve mostrar progresso, não zero."""
+    write_pdf(settings.pdf_dir, "biblia.pdf")
+    extractor = FakeExtractor({"biblia.pdf": [PageText(page=6, text=GENESIS_PAGE)]})
+    indexer, _ = build_indexer(settings, extractor)
+
+    indexer.mark_indexing()
+    status = indexer.status()
+
+    assert status.state == "indexing"
+    assert status.pdfs_total == 1
+
+
+def test_status_counts_pages_processed_during_indexing(settings):
+    write_pdf(settings.pdf_dir, "biblia.pdf")
+    extractor = FakeExtractor({"biblia.pdf": [PageText(page=6, text=GENESIS_PAGE)]})
+    indexer, _ = build_indexer(settings, extractor)
+
+    indexer.mark_indexing()
+    indexer.report_progress(pdf_name="biblia.pdf", pages_done=120, pages_total=3364)
+
+    status = indexer.status()
+
+    assert status.pages_processed == 120
+    assert status.pages_total == 3364

frontend/src/api.ts

@@ -19,6 +19,9 @@ export interface Status {
   model: string;
   last_indexed_at: string | null;
   errors: string[];
+  pages_processed: number;
+  pages_total: number;
+  current_file: string | null;
 }
 
 export interface ChatMessage {

frontend/src/components/StatusPanel.tsx

@@ -44,6 +44,33 @@ export function StatusPanel({ status, onReindex, reindexing }: StatusPanelProps)
         </div>
       </dl>
 
+      {status?.state === "indexing" && (
+        <div className="status__progress">
+          <div className="status__progress-label">
+            {status.current_file ? `Lendo ${status.current_file}` : "Lendo PDFs"}
+            {status.pages_total > 0 && ` · ${status.pages_processed}/${status.pages_total} páginas`}
+          </div>
+          {status.pages_total > 0 && (
+            <div
+              className="status__progress-bar"
+              role="progressbar"
+              aria-valuenow={status.pages_processed}
+              aria-valuemin={0}
+              aria-valuemax={status.pages_total}
+            >
+              <span
+                style={{
+                  width: `${Math.min(
+                    100,
+                    (status.pages_processed / status.pages_total) * 100,
+                  )}%`,
+                }}
+              />
+            </div>
+          )}
+        </div>
+      )}
+
       {status && status.errors.length > 0 && (
         <ul className="status__errors">
           {status.errors.map((error) => (

frontend/src/styles.css

@@ -212,6 +212,31 @@ body {
   font-size: 12px;
 }
 
+.status__progress {
+  margin-top: 14px;
+}
+
+.status__progress-label {
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.status__progress-bar {
+  margin-top: 7px;
+  height: 5px;
+  border-radius: 999px;
+  background: var(--bg);
+  overflow: hidden;
+}
+
+.status__progress-bar > span {
+  display: block;
+  height: 100%;
+  border-radius: 999px;
+  background: var(--accent);
+  transition: width 0.4s ease;
+}
+
 .status__reindex {
   width: 100%;
   margin-top: 16px;