From e0bca34a257f4da33de98eccabea04e4c116867e Mon Sep 17 00:00:00 2001
From: Elron Bandel <elron.bandel@ibm.com>
Date: Sun, 28 Jan 2024 06:36:13 -0500
Subject: [PATCH] Add catalog summary printing functionality

Signed-off-by: Elron Bandel <elron.bandel@ibm.com>
---
 src/unitxt/catalog.py | 38 ++++++++++++++++++++++++++++++++++++--
 1 file changed, 36 insertions(+), 2 deletions(-)

diff --git a/src/unitxt/catalog.py b/src/unitxt/catalog.py
index 477df9bd3..e58467cc4 100644
--- a/src/unitxt/catalog.py
+++ b/src/unitxt/catalog.py
@@ -2,12 +2,13 @@
 import re
 from pathlib import Path
 from typing import Optional
-
+from collections import Counter
 import requests
 
-from .artifact import Artifact, Artifactory, reset_artifacts_cache
+from .artifact import Artifact, Artifactory, reset_artifacts_cache, Artifactories
 from .logging_utils import get_logger
 from .version import version
+from .text_utils import print_dict
 
 logger = get_logger()
 COLLECTION_SEPARATOR = "."
@@ -35,6 +36,7 @@ class Catalog(Artifactory):
 class LocalCatalog(Catalog):
     name: str = "local"
     location: str = default_catalog_path
+    is_local: bool = True
 
     def path(self, artifact_identifier: str):
         assert (
@@ -92,6 +94,7 @@ class GithubCatalog(LocalCatalog):
     repo = "unitxt"
     repo_dir = "src/unitxt/catalog"
     user = "IBM"
+    is_local: bool = False
 
     def prepare(self):
         tag = version
@@ -133,3 +136,34 @@ def add_to_catalog(
         artifact, name, overwrite=overwrite, verbose=verbose
     )  # remove collection (its actually the dir).
     # verify name
+
+
+def get_local_catalogs_paths():
+    result = []
+    for artifactory in Artifactories():
+        if isinstance(artifactory, LocalCatalog):
+            if artifactory.is_local:
+                result.append(artifactory.location)
+    return result
+
+def count_files_recursively(folder):
+    file_count = 0
+    for _, _, files in os.walk(folder):
+        file_count += len(files)
+    return file_count
+
+def local_catalog_summary(catalog_path):
+    result = {}
+
+    for dir in os.listdir(catalog_path):
+        if os.path.isdir(os.path.join(catalog_path, dir)):
+            result[dir] = count_files_recursively(os.path.join(catalog_path, dir))
+
+    return result
+
+def summary():
+    result = Counter()
+    for local_catalog_path in get_local_catalogs_paths():
+        result +=  Counter(local_catalog_summary(local_catalog_path))
+    print_dict(result)
+    return result