chg: more docker services

REFERENCES.md

@@ -160,3 +160,4 @@ source: <https://www.youtube.com/watch?v=AeASAsPp9LE>
 - https://github.com/langchain-ai/langchain/tree/master/cookbook
 - https://github.com/codingjoe/relint - `Write your own linting rules using regular expressions.`
 - https://github.com/ionelmc/python-manhole/ - `Debugging manhole for python applications.`
+- https://github.com/langchain-ai/langchain/blob/master/cookbook/Multi_modal_RAG.ipynb

src/goob_ai/utils/imgops.py

@@ -19,6 +19,7 @@
 import os
 import os.path
 import pathlib
+import re
 import sys
 import tempfile
 import time
@@ -103,6 +104,39 @@ def setup_model() -> torch.nn.Module:
 ###########################################################################3
 
 
+def looks_like_base64(sb):
+    """Check if the string looks like base64"""
+    return re.match("^[A-Za-z0-9+/]+[=]{0,2}$", sb) is not None
+
+
+def is_image_data(b64data):
+    """
+    Check if the base64 data is an image by looking at the start of the data
+    """
+    image_signatures = {
+        b"\xff\xd8\xff": "jpg",
+        b"\x89\x50\x4e\x47\x0d\x0a\x1a\x0a": "png",
+        b"\x47\x49\x46\x38": "gif",
+        b"\x52\x49\x46\x46": "webp",
+    }
+    try:
+        header = base64.b64decode(b64data)[:8]  # Decode and get the first 8 bytes
+        for sig, format in image_signatures.items():
+            if header.startswith(sig):
+                return True
+        return False
+    except Exception:
+        return False
+
+
+# SOURCE: https://github.com/langchain-ai/langchain/blob/master/cookbook/multi_modal_QA.ipynb
+def encode_image(image_path: str):
+    """Getting the base64 string"""
+
+    with open(image_path, "rb") as image_file:
+        return base64.b64encode(image_file.read()).decode("utf-8")
+
+
 def split_image_text_types(docs):
     """Split numpy array images and texts"""
     images = []