Fix seed URL input: textbox + accurate stats parsing

app.py

@@ -278,22 +278,13 @@ THEME_JS = """
 """
 
 SEED_WIDGET_JS = """
-(seed_rows) => {
-  const parseRows = (rows) => {
-    if (!Array.isArray(rows)) return [];
-    const out = [];
-    for (const row of rows) {
-      let value = "";
-      if (Array.isArray(row)) {
-        value = String(row[0] ?? "").trim();
-      } else if (row && typeof row === "object") {
-        value = String(Object.values(row)[0] ?? "").trim();
-      } else if (row !== null && row !== undefined) {
-        value = String(row).trim();
-      }
-      if (value) out.push(value);
-    }
-    return out;
+(seed_text) => {
+  const parseSeedText = (value) => {
+    if (typeof value !== "string") return [];
+    return value
+      .split(/\\r?\\n/)
+      .map((line) => line.trim())
+      .filter((line) => line.length > 0);
   };
 
   const dedupe = (values) => {
@@ -323,7 +314,7 @@ SEED_WIDGET_JS = """
     .replaceAll('"', """)
     .replaceAll("'", "'");
 
-  const seeds = dedupe(parseRows(seed_rows));
+  const seeds = dedupe(parseSeedText(seed_text));
   const domainSet = new Set(seeds.map(domainOf).filter(Boolean));
   const chips = seeds.length
     ? seeds.slice(0, 12).map((url) => `<span class=\"seed-chip\">${escapeHtml(url)}</span>`).join("")
@@ -331,8 +322,9 @@ SEED_WIDGET_JS = """
   const overflow = seeds.length > 12
     ? `<span class=\"seed-overflow\">+${seeds.length - 12} more</span>`
     : "";
+  const firstUrlChars = seeds.length ? seeds[0].length : 0;
 
-  return `<div class=\"seed-widget\"><div class=\"seed-stats\"><span><strong>${seeds.length}</strong> seeds</span><span><strong>${domainSet.size}</strong> domains</span><span><strong>${seeds.slice(0, 1).join("").length || 0}</strong> first-url chars</span></div><div class=\"seed-chip-wrap\">${chips}${overflow}</div></div>`;
+  return `<div class=\"seed-widget\"><div class=\"seed-stats\"><span><strong>${seeds.length}</strong> seeds</span><span><strong>${domainSet.size}</strong> domains</span><span><strong>${firstUrlChars}</strong> first-url chars</span></div><div class=\"seed-chip-wrap\">${chips}${overflow}</div></div>`;
 }
 """
 
@@ -348,32 +340,40 @@ def safe_queue_size(queue: Any) -> int:
         return -1
 
 
-def parse_seed_url_rows(rows: Any) -> list[str]:
-    if rows is None:
+def parse_seed_url_rows(seed_urls_input: Any) -> list[str]:
+    if seed_urls_input is None:
         return []
 
-    if isinstance(rows, (list, tuple)):
-        rows_iterable: list[Any] = list(rows)
-    elif hasattr(rows, "values"):
+    if isinstance(seed_urls_input, str):
+        return [line.strip() for line in seed_urls_input.splitlines() if line.strip()]
+
+    if isinstance(seed_urls_input, (list, tuple)):
+        rows_iterable: list[Any] = list(seed_urls_input)
+    elif hasattr(seed_urls_input, "values"):
         try:
-            rows_iterable = rows.values.tolist()  # pandas.DataFrame path
+            rows_iterable = seed_urls_input.values.tolist()  # pandas.DataFrame path
         except Exception:
             rows_iterable = []
     else:
-        rows_iterable = [rows]
+        rows_iterable = [seed_urls_input]
 
     items: list[str] = []
     for row in rows_iterable:
-        value = ""
+        value_sources: list[Any]
         if isinstance(row, dict):
-            value = str(next(iter(row.values()), "") or "").strip()
+            value_sources = [next(iter(row.values()), "")]
         elif isinstance(row, (list, tuple)):
-            value = str(row[0] if row else "").strip()
-        elif row is not None:
-            value = str(row).strip()
-
-        if value:
-            items.append(value)
+            value_sources = [row[0] if row else ""]
+        else:
+            value_sources = [row]
+
+        for source in value_sources:
+            if source is None:
+                continue
+            for line in str(source).splitlines():
+                value = line.strip()
+                if value:
+                    items.append(value)
     return items
 
 
@@ -388,12 +388,12 @@ def unique_preserve_order(values: list[str]) -> list[str]:
     return out
 
 
-def collect_seed_urls(seed_urls_table: Any) -> list[str]:
-    return unique_preserve_order(parse_seed_url_rows(seed_urls_table))
+def collect_seed_urls(seed_urls_input: Any) -> list[str]:
+    return unique_preserve_order(parse_seed_url_rows(seed_urls_input))
 
 
-def render_seed_widget_html(seed_urls_table: Any) -> str:
-    seeds = collect_seed_urls(seed_urls_table)
+def render_seed_widget_html(seed_urls_input: Any) -> str:
+    seeds = collect_seed_urls(seed_urls_input)
     domains = {(urlsplit(u).hostname or "").lower().strip(".") for u in seeds}
     domains = {d for d in domains if d}
 
@@ -454,7 +454,7 @@ def validate_hf_requirements(enable_hf_upload: bool, hf_repo_id: str, hf_token:
 
 def build_crawler_config(
     *,
-    seed_urls_table: Any,
+    seed_urls_input: Any,
     max_links_per_page: int,
     request_timeout_seconds: float,
     max_response_bytes: int,
@@ -469,7 +469,7 @@ def build_crawler_config(
     total_workers: int,
 ) -> CrawlerConfig:
     validate_hf_requirements(enable_hf_upload, hf_repo_id, hf_token)
-    seed_urls = collect_seed_urls(seed_urls_table)
+    seed_urls = collect_seed_urls(seed_urls_input)
 
     return CrawlerConfig(
         seed_urls=seed_urls,
@@ -692,7 +692,7 @@ def _format_dashboard_response(
 def _start_crawl(
     *,
     total_workers: int,
-    seed_urls_table: Any,
+    seed_urls_input: Any,
     max_links_per_page: int,
     request_timeout_seconds: float,
     max_response_bytes: int,
@@ -707,7 +707,7 @@ def _start_crawl(
 ) -> tuple[str, str, str, str]:
     try:
         config = build_crawler_config(
-            seed_urls_table=seed_urls_table,
+            seed_urls_input=seed_urls_input,
             max_links_per_page=max_links_per_page,
             request_timeout_seconds=request_timeout_seconds,
             max_response_bytes=max_response_bytes,
@@ -730,7 +730,7 @@ def _start_crawl(
 
 
 def start_crawl_standard(
-    seed_urls_table: Any,
+    seed_urls_input: Any,
     max_links_per_page: int,
     request_timeout_seconds: float,
     max_response_bytes: int,
@@ -745,7 +745,7 @@ def start_crawl_standard(
 ) -> tuple[str, str, str, str]:
     return _start_crawl(
         total_workers=NORMAL_TOTAL_WORKERS,
-        seed_urls_table=seed_urls_table,
+        seed_urls_input=seed_urls_input,
         max_links_per_page=max_links_per_page,
         request_timeout_seconds=request_timeout_seconds,
         max_response_bytes=max_response_bytes,
@@ -761,7 +761,7 @@ def start_crawl_standard(
 
 
 def start_crawl_super(
-    seed_urls_table: Any,
+    seed_urls_input: Any,
     max_links_per_page: int,
     request_timeout_seconds: float,
     max_response_bytes: int,
@@ -776,7 +776,7 @@ def start_crawl_super(
 ) -> tuple[str, str, str, str]:
     return _start_crawl(
         total_workers=SUPER_TOTAL_WORKERS,
-        seed_urls_table=seed_urls_table,
+        seed_urls_input=seed_urls_input,
         max_links_per_page=max_links_per_page,
         request_timeout_seconds=request_timeout_seconds,
         max_response_bytes=max_response_bytes,
@@ -823,7 +823,7 @@ def build_ui() -> gr.Blocks:
             "https://www.nasa.gov/",
         ]
     )
-    default_seed_rows = [[url] for url in defaults.seed_urls]
+    default_seed_text = "\n".join(defaults.seed_urls)
 
     with gr.Blocks(title="DataMuncherLabs AutoWS") as demo:
         gr.Markdown("# DataMuncherLabs AutoWS")
@@ -854,18 +854,16 @@ def build_ui() -> gr.Blocks:
 
         with gr.Row():
             with gr.Column(scale=2):
-                seed_urls_table = gr.Dataframe(
-                    headers=["seed_url"],
-                    datatype=["str"],
-                    type="array",
-                    row_count=(8, "dynamic"),
-                    value=default_seed_rows,
+                seed_urls_input = gr.Textbox(
+                    lines=10,
+                    value=default_seed_text,
                     interactive=True,
-                    label="Seed URL List (editable)",
+                    label="Seed URL List (one URL per line)",
+                    placeholder="https://example.com",
                 )
                 seed_widget_html = gr.HTML(
                     label="Seed URL Summary",
-                    value=render_seed_widget_html(default_seed_rows),
+                    value=render_seed_widget_html(default_seed_text),
                 )
                 token_widget_html = gr.HTML(
                     label="Live Tokenization",
@@ -953,7 +951,7 @@ def build_ui() -> gr.Blocks:
         logs_box = gr.Textbox(label="Run Log", lines=12, interactive=False)
 
         start_inputs = [
-            seed_urls_table,
+            seed_urls_input,
             max_links_per_page,
             request_timeout_seconds,
             max_response_bytes,
@@ -995,9 +993,9 @@ def build_ui() -> gr.Blocks:
             outputs=[incomplete_shard_flush_seconds],
         )
 
-        seed_urls_table.change(
+        seed_urls_input.change(
             fn=None,
-            inputs=[seed_urls_table],
+            inputs=[seed_urls_input],
             outputs=[seed_widget_html],
             js=SEED_WIDGET_JS,
         )
@@ -1011,7 +1009,7 @@ def build_ui() -> gr.Blocks:
         )
         demo.load(
             fn=None,
-            inputs=[seed_urls_table],
+            inputs=[seed_urls_input],
             outputs=[seed_widget_html],
             js=SEED_WIDGET_JS,
         )