docs: replace internal blob APIs with helper functions in notebook

Author: shuoweil

notebooks/multimodal/multimodal_dataframe.ipynb

@@ -157,7 +157,51 @@
         "    runtime = bbq.obj.get_access_url(s, mode=mode)\n",
         "    \n",
         "    # 3. Convert the runtime object to a JSON string\n",
-        "    return bbq.to_json_string(runtime)"
+        "    return bbq.to_json_string(runtime)\n",
+        "\n",
+        "def get_metadata(series):\n",
+        "    # Fetch metadata and extract GCS metadata from the details JSON field\n",
+        "    metadata_obj = bbq.obj.fetch_metadata(series)\n",
+        "    return bbq.json_query(metadata_obj.struct.field(\"details\"), \"$.gcs_metadata\")\n",
+        "\n",
+        "def get_content_type(series):\n",
+        "    return bbq.json_value(get_metadata(series), \"$.content_type\")\n",
+        "\n",
+        "def get_size(series):\n",
+        "    return bbq.json_value(get_metadata(series), \"$.size\").astype(\"Int64\")\n",
+        "\n",
+        "def get_updated(series):\n",
+        "    return bpd.to_datetime(bbq.json_value(get_metadata(series), \"$.updated\").astype(\"Int64\"), unit=\"us\", utc=True)\n",
+        "\n",
+        "def display_blob(series, n=3):\n",
+        "    import IPython.display as ipy_display\n",
+        "    import pandas as pd\n",
+        "    import requests\n",
+        "    \n",
+        "    # Retrieve access URLs and content types\n",
+        "    runtime_json = bbq.to_json_string(bbq.obj.get_access_url(series, mode=\"R\"))\n",
+        "    read_url = bbq.json_value(runtime_json, \"$.access_urls.read_url\")\n",
+        "    content_type = get_content_type(series)\n",
+        "    \n",
+        "    # Pull to pandas to display\n",
+        "    pdf = bpd.DataFrame({\"read_url\": read_url, \"content_type\": content_type}).head(n).to_pandas()\n",
+        "    \n",
+        "    width = bigframes.options.display.blob_display_width\n",
+        "    height = bigframes.options.display.blob_display_height\n",
+        "    \n",
+        "    for _, row in pdf.iterrows():\n",
+        "        if pd.isna(row[\"read_url\"]):\n",
+        "            ipy_display.display(\"<NA>\")\n",
+        "        elif pd.isna(row[\"content_type\"]):\n",
+        "            ipy_display.display(requests.get(row[\"read_url\"]).content)\n",
+        "        elif row[\"content_type\"].casefold().startswith(\"image\"):\n",
+        "            ipy_display.display(ipy_display.Image(url=row[\"read_url\"], width=width, height=height))\n",
+        "        elif row[\"content_type\"].casefold().startswith(\"audio\"):\n",
+        "            ipy_display.display(ipy_display.Audio(requests.get(row[\"read_url\"]).content))\n",
+        "        elif row[\"content_type\"].casefold().startswith(\"video\"):\n",
+        "            ipy_display.display(ipy_display.Video(row[\"read_url\"], width=width, height=height))\n",
+        "        else:\n",
+        "            ipy_display.display(requests.get(row[\"read_url\"]).content)"
       ]
     },
     {
@@ -461,9 +505,9 @@
       "source": [
         "# Combine unstructured data with structured data\n",
         "df_image[\"author\"] = [\"alice\", \"bob\", \"bob\", \"alice\", \"bob\"]  # type: ignore\n",
-        "df_image[\"content_type\"] = df_image[\"image\"].blob.content_type()\n",
-        "df_image[\"size\"] = df_image[\"image\"].blob.size()\n",
-        "df_image[\"updated\"] = df_image[\"image\"].blob.updated()\n",
+        "df_image[\"content_type\"] = get_content_type(df_image[\"image\"])\n",
+        "df_image[\"size\"] = get_size(df_image[\"image\"])\n",
+        "df_image[\"updated\"] = get_updated(df_image[\"image\"])\n",
         "df_image"
       ]
     },
@@ -536,7 +580,7 @@
       ],
       "source": [
         "# filter images and display, you can also display audio and video types\n",
-        "df_image[df_image[\"author\"] == \"alice\"][\"image\"].blob.display()"
+        "display_blob(df_image[df_image[\"author\"] == \"alice\"][\"image\"])"
       ]
     },
     {