From 61bc0bccf674169d6823c02c56e1f17c89604fbe Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?David=20Haj=C4=8Diar?= <david.hajciar@juicymo.cz>
Date: Fri, 30 May 2025 10:16:53 +0200
Subject: [PATCH 1/2] Use bytesize to get content size

---
 lib/azure_blob/client.rb   | 16 ++++++++++------
 test/client/test_client.rb | 23 ++++++++++++++++++++++-
 2 files changed, 32 insertions(+), 7 deletions(-)

diff --git a/lib/azure_blob/client.rb b/lib/azure_blob/client.rb
index 5f0da60..40588b5 100644
--- a/lib/azure_blob/client.rb
+++ b/lib/azure_blob/client.rb
@@ -49,7 +49,7 @@ def initialize(account_name:, access_key: nil, principal_id: nil, container:, ho
     # [+:block_size+]
     #   Block size in bytes, can be used to force the method to split the upload in smaller chunk. Defaults to +AzureBlob::DEFAULT_BLOCK_SIZE+ and cannot be bigger than +AzureBlob::MAX_UPLOAD_SIZE+
     def create_block_blob(key, content, options = {})
-      if content.size > (options[:block_size] || DEFAULT_BLOCK_SIZE)
+      if content_size(content) > (options[:block_size] || DEFAULT_BLOCK_SIZE)
         put_blob_multiple(key, content, **options)
       else
         put_blob_single(key, content, **options)
@@ -309,7 +309,7 @@ def append_blob_block(key, content, options = {})
       uri.query = URI.encode_www_form(comp: "appendblock")
 
       headers = {
-        "Content-Length": content.size,
+        "Content-Length": content_size(content),
         "Content-Type": options[:content_type],
         "Content-MD5": options[:content_md5],
       }.merge(additional_headers(options))
@@ -333,7 +333,7 @@ def put_blob_block(key, index, content, options = {})
       uri.query = URI.encode_www_form(comp: "block", blockid: block_id)
 
       headers = {
-        "Content-Length": content.size,
+        "Content-Length": content_size(content),
         "Content-Type": options[:content_type],
         "Content-MD5": options[:content_md5],
       }.merge(additional_headers(options))
@@ -361,7 +361,7 @@ def commit_blob_blocks(key, block_ids, options = {})
       uri.query = URI.encode_www_form(comp: "blocklist")
 
       headers = {
-        "Content-Length": content.size,
+        "Content-Length": content_size(content),
         "Content-Type": options[:content_type],
         "x-ms-blob-content-md5": options[:content_md5],
         "x-ms-blob-content-disposition": options[:content_disposition],
@@ -384,7 +384,7 @@ def generate_block_id(index)
     def put_blob_multiple(key, content, options = {})
       content = StringIO.new(content) if content.is_a? String
       block_size = options[:block_size] || DEFAULT_BLOCK_SIZE
-      block_count = (content.size.to_f / block_size).ceil
+      block_count = (content_size(content).to_f / block_size).ceil
       block_ids = block_count.times.map do |i|
         put_blob_block(key, i, content.read(block_size))
       end
@@ -398,7 +398,7 @@ def put_blob_single(key, content, options = {})
 
       headers = {
         "x-ms-blob-type": "BlockBlob",
-        "Content-Length": content.size,
+        "Content-Length": content_size(content),
         "Content-Type": options[:content_type],
         "x-ms-blob-content-md5": options[:content_md5],
         "x-ms-blob-content-disposition": options[:content_disposition],
@@ -407,6 +407,10 @@ def put_blob_single(key, content, options = {})
       Http.new(uri, headers, signer:, **options.slice(:metadata, :tags)).put(content.read)
     end
 
+    def content_size(content)
+      content.bytesize
+    end
+
     def host
       @host ||= "https://#{account_name}.blob.#{CLOUD_REGIONS_SUFFIX[cloud_regions]}"
     end
diff --git a/test/client/test_client.rb b/test/client/test_client.rb
index 73e88d3..bd7824e 100644
--- a/test/client/test_client.rb
+++ b/test/client/test_client.rb
@@ -247,7 +247,7 @@ def test_get_blob_properties
     blob = client.get_blob_properties(key)
 
     assert blob.present?
-    assert_equal content.size, blob.size
+    assert_equal content.bytesize, blob.size
   end
 
   def test_get_blob_properties_404
@@ -475,4 +475,25 @@ def test_create_append_blob_additional_headers
     dummy.expect :delete_blob, nil, [ key ]
     @client = dummy
   end
+
+  def test_append_blob_block_content_size
+    content = "Ň"
+    http_mock = Minitest::Mock.new
+    http_mock.expect :put, true, [ content ]
+
+    stubbed_new = lambda do |uri, headers = {}, signer: nil, **kwargs|
+      assert_equal "2", headers[:"Content-Length"]
+      http_mock
+    end
+
+    AzureBlob::Http.stub :new, stubbed_new do
+      custom_client = AzureBlob::Client.new(account_name: "foo", access_key: "bar", container: "cont")
+      custom_client.append_blob_block(key, content)
+    end
+
+    http_mock.verify
+    dummy = Minitest::Mock.new
+    dummy.expect :delete_blob, nil, [ key ]
+    @client = dummy
+  end
 end

From 5c55f19dcd7905321c3ed7e11310ebd87f03a6e5 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?David=20Haj=C4=8Diar?= <david.hajciar@juicymo.cz>
Date: Sat, 31 May 2025 10:41:36 +0200
Subject: [PATCH 2/2] Fallback to size if we can't use bytesize

---
 lib/azure_blob/client.rb | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/lib/azure_blob/client.rb b/lib/azure_blob/client.rb
index 40588b5..fee52bb 100644
--- a/lib/azure_blob/client.rb
+++ b/lib/azure_blob/client.rb
@@ -408,7 +408,11 @@ def put_blob_single(key, content, options = {})
     end
 
     def content_size(content)
-      content.bytesize
+      if content.respond_to?(:bytesize)
+        content.bytesize
+      else
+        content.size
+      end
     end
 
     def host