Source file: Fix error on zipped files (airbytehq#39909)

Co-authored-by: Marcos Marx <[email protected]>

airbyte-integrations/connectors/source-file/metadata.yaml

@@ -10,7 +10,7 @@ data:
   connectorSubtype: file
   connectorType: source
   definitionId: 778daa7c-feaf-4db6-96f3-70fd645acc77
-  dockerImageTag: 0.5.3
+  dockerImageTag: 0.5.4
   dockerRepository: airbyte/source-file
   documentationUrl: https://docs.airbyte.com/integrations/sources/file
   githubIssueLabel: source-file

airbyte-integrations/connectors/source-file/pyproject.toml

@@ -3,7 +3,7 @@ requires = [ "poetry-core>=1.0.0",]
 build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
-version = "0.5.3"
+version = "0.5.4"
 name = "source-file"
 description = "Source implementation for File"
 authors = [ "Airbyte <[email protected]>",]
@@ -22,6 +22,7 @@ beautifulsoup4 = "==4.11.1"
 openpyxl = "==3.1.0"
 google-cloud-storage = "==2.5.0"
 pandas = "2.2.2"
+numpy = "<2"
 airbyte-cdk = "^0"
 paramiko = "==2.11.0"
 xlrd = "==2.0.1"

airbyte-integrations/connectors/source-file/source_file/client.py

@@ -75,7 +75,7 @@ def __init__(self, url: str, provider: dict, binary=None, encoding=None):
         self._file = None
         self.args = {
             "mode": "rb" if binary else "r",
-            "encoding": encoding,
+            "encoding": None if binary else encoding,
         }
 
     def __enter__(self):
@@ -452,7 +452,7 @@ def _unzip(self, fp):
         logger.info("Temp dir content: " + str(os.listdir(tmp_dir.name)))
         final_file: str = os.path.join(tmp_dir.name, os.listdir(tmp_dir.name)[0])
         logger.info("Pick up first file: " + final_file)
-        fp_tmp = open(final_file, "r")
+        fp_tmp = open(final_file, "rb")
         return fp_tmp
 
     def _cache_stream(self, fp):

airbyte-integrations/connectors/source-file/unit_tests/test_source.py

@@ -64,6 +64,31 @@ def test_csv_with_utf16_encoding(absolute_path, test_files):
     assert stream.json_schema == expected_schema
 
 
+def test_zipped_csv_with_utf16_encoding(absolute_path, test_files):
+    config_local_zipped_csv_utf16 = {
+        "dataset_name": "AAA",
+        "format": "csv",
+        "reader_options": '{"encoding":"utf_16", "parse_dates": ["header5"]}',
+        "url": f"{absolute_path}/{test_files}/test_utf16.csv.zip",
+        "provider": {"storage": "local"},
+    }
+    expected_schema = {
+        "$schema": "http://json-schema.org/draft-07/schema#",
+        "properties": {
+            "header1": {"type": ["string", "null"]},
+            "header2": {"type": ["number", "null"]},
+            "header3": {"type": ["number", "null"]},
+            "header4": {"type": ["boolean", "null"]},
+            "header5": {"type": ["string", "null"], "format": "date-time"},
+        },
+        "type": "object",
+    }
+
+    catalog = SourceFile().discover(logger=logger, config=config_local_zipped_csv_utf16)
+    stream = next(iter(catalog.streams))
+    assert stream.json_schema == expected_schema
+
+
 def get_catalog(properties):
     return ConfiguredAirbyteCatalog(
         streams=[

docs/integrations/sources/file.md

@@ -234,6 +234,7 @@ In order to read large files from a remote location, this connector uses the [sm
 
 | Version | Date       | Pull Request                                             | Subject                                                                                                 |
 | :------ | :--------- | :------------------------------------------------------- | :------------------------------------------------------------------------------------------------------ |
+| 0.5.4 | 2024-07-01 | [39909](https://github.com/airbytehq/airbyte/pull/39909) | Fix error with zip files and encoding. |
 | 0.5.3 | 2024-06-27 | [40215](https://github.com/airbytehq/airbyte/pull/40215) | Replaced deprecated AirbyteLogger with logging.Logger |
 | 0.5.2 | 2024-06-06 | [39192](https://github.com/airbytehq/airbyte/pull/39192) | [autopull] Upgrade base image to v1.2.2 |
 | 0.5.1 | 2024-05-03 | [37799](https://github.com/airbytehq/airbyte/pull/37799) | Add fastparquet engine for parquet file reader. |