Correct notebook tag logic (#317)

docs/tutorials/pipelines/creating_pipelines.ipynb

@@ -2,7 +2,9 @@
  "cells": [
   {
    "cell_type": "markdown",
-   "metadata": {},
+   "metadata": {
+    "tags": []
+   },
    "source": [
     "# Creating Custom Pipelines\n",
     "\n",
@@ -21,7 +23,9 @@
   },
   {
    "cell_type": "markdown",
-   "metadata": {},
+   "metadata": {
+    "tags": []
+   },
    "source": [
     "#### Model Components\n",
     "\n",
@@ -105,7 +109,6 @@
    "execution_count": 1,
    "metadata": {
     "tags": [
-     "remove-cell",
      "hide-input"
     ]
    },
@@ -132,7 +135,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "tags": []
+   },
    "outputs": [],
    "source": [
     "class CustomDataloaderConfig(DataloaderConfig):\n",
@@ -240,7 +245,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "tags": []
+   },
    "outputs": [],
    "source": [
     "dataloader = Dataloader()\n",
@@ -250,7 +257,9 @@
   },
   {
    "cell_type": "markdown",
-   "metadata": {},
+   "metadata": {
+    "tags": []
+   },
    "source": [
     "# Creating pipelines from a config\n",
     "\n",
@@ -272,7 +281,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "tags": []
+   },
    "outputs": [],
    "source": [
     "import pprint\n",
@@ -295,13 +306,17 @@
   },
   {
    "cell_type": "markdown",
-   "metadata": {},
+   "metadata": {
+    "tags": []
+   },
    "source": []
   },
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "tags": []
+   },
    "outputs": [],
    "source": [
     "pipeline.get_train_loss_dict()"
@@ -310,7 +325,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "tags": []
+   },
    "outputs": [],
    "source": []
   }

scripts/docs/add_nb_tags.py

@@ -35,19 +35,30 @@ def main(check: bool = False):
     for ipath in notebooks:
         ntbk = nbf.read(ipath, nbf.NO_CONVERT)
 
-        missing_metadata = False
+        incorrect_metadata = False
         for cell in ntbk.cells:
             cell_tags = cell.get("metadata", {}).get("tags", [])
+            found_keys = []
+            found_tags = []
             for key, val in text_search_dict.items():
-                if key in cell["source"]:
-                    if val not in cell_tags:
-                        missing_metadata = True
-                        cell_tags.append(val)
-            if len(cell_tags) > 0:
-                cell["metadata"]["tags"] = cell_tags
-        if missing_metadata:
+                if key in cell.source:
+                    found_keys.append(key)
+                    found_tags.append(val)
+
+            if len(found_keys) > 1:
+                print_yellow(f"Found multiple tags {found_keys} for {ipath}")
+                sys.exit(1)
+
+            if len(cell_tags) != len(found_tags):
+                incorrect_metadata = True
+            elif len(cell_tags) == 1 and len(found_keys) == 1:
+                if found_tags[0] != cell_tags[0]:
+                    incorrect_metadata = True
+
+            cell["metadata"]["tags"] = found_tags
+        if incorrect_metadata:
             if check:
-                print_yellow(f"{ipath} is missing metadata. Call `python scripts.docs.add_nb_tags.py` to add it.")
+                print_yellow(f"{ipath} has incorrect metadata. Call `python scripts.docs.add_nb_tags.py` to add it.")
                 any_missing = True
             else:
                 print(f"Adding metadata to {ipath}")