derive snippet type from extension

docs/website/docs/getting-started.md

@@ -41,7 +41,7 @@ Make sure that your `dlt` version is **0.3.15** or above. Check it in the termin
 Let's load a list of Python objects (dictionaries) into `duckdb` and inspect the created dataset:
 
 <!--@@@DLT_SNIPPET_START ./getting-started-snippets.py::start-->
-```python
+```py
 import dlt
 
 data = [
@@ -128,7 +128,7 @@ The library will create/update tables, infer data types and deal with nested dat
   <TabItem value="json">
 
 <!--@@@DLT_SNIPPET_START json-->
-```python
+```py
 import dlt
 
 from dlt.common import json
@@ -158,7 +158,7 @@ We import **json** from `dlt` namespace. It defaults to `orjson`(otherwise `simp
   Pass anything that you can load with Pandas to `dlt`
 
 <!--@@@DLT_SNIPPET_START csv-->
-```python
+```py
 import dlt
 import pandas as pd
 
@@ -181,7 +181,7 @@ print(load_info)
   <TabItem value="api">
 
 <!--@@@DLT_SNIPPET_START api-->
-```python
+```py
 import dlt
 from dlt.sources.helpers import requests
 
@@ -212,7 +212,7 @@ to sync your databases with warehouses, data lakes, or vector stores.
 :::
 
 <!--@@@DLT_SNIPPET_START db-->
-```python
+```py
 import dlt
 from sqlalchemy import create_engine
 
@@ -257,7 +257,7 @@ For example, if the CSV file is updated, how we can refresh it in the database?
 One method is to tell `dlt` to replace the data in existing tables by using `write_disposition`:
 
 <!--@@@DLT_SNIPPET_START replace-->
-```python
+```py
 import dlt
 
 data = [
@@ -301,7 +301,7 @@ Instead of using `replace` write_disposition and downloading all issues each tim
 
 
 <!--@@@DLT_SNIPPET_START incremental-->
-```python
+```py
 import dlt
 from dlt.sources.helpers import requests
 
@@ -377,7 +377,7 @@ To get always fresh content of all the issues you combine incremental load with
 like in the script below.
 
 <!--@@@DLT_SNIPPET_START incremental_merge-->
-```python
+```py
 import dlt
 from dlt.sources.helpers import requests
 
@@ -433,7 +433,7 @@ This is a fun but practical example that reads GitHub events from **dlt** reposi
 Each event type is sent to a different table in `duckdb`.
 
 <!--@@@DLT_SNIPPET_START table_dispatch-->
-```python
+```py
 import dlt
 from dlt.sources.helpers import requests
 
@@ -511,7 +511,7 @@ Learn more:
 Below we extract text from PDFs and load it to [Weaviate](dlt-ecosystem/destinations/weaviate) vector store.
 
 <!--@@@DLT_SNIPPET_START pdf_to_weaviate-->
-```python
+```py
 import os
 
 import dlt
@@ -591,7 +591,7 @@ python pdf_to_weaviate.py
 
 Now it is time to query our documents.
 <!--@@@DLT_SNIPPET_START pdf_to_weaviate_read-->
-```python
+```py
 import weaviate
 
 client = weaviate.Client("http://localhost:8080")

docs/website/docs/intro.md

@@ -22,7 +22,7 @@ There's no need to start any backends or containers. Import `dlt` in your Python
 <!--AUTO-GENERATED-CONTENT:END-->
 
 <!--@@@DLT_SNIPPET_START index-->
-```python
+```py
 import dlt
 from dlt.sources.helpers import requests
 # Create a dlt pipeline that will load

docs/website/tools/update_snippets.js

@@ -83,7 +83,8 @@ function getSnippet(fileName, snippetName) {
     let result = lines.slice((snippetMap[snippetName]["start"]+1), snippetMap[snippetName]["end"]);
     // dedent works on strings, not on string arrays, so this is very ineffective unfortunately...
     result = dedent(result.join("\n")).split(/\r?\n/);
-    result.unshift("```python");
+    const codeType = path.extname(snippetsFileName).replace(".", "");
+    result.unshift(`\`\`\`${codeType}`);
     result.push("```");
     return filterDirectives(result);
 }