This is an automated email from the ASF dual-hosted git repository.
gopidesupavan pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/airflow.git
The following commit(s) were added to refs/heads/main by this push:
new a4b6b77e683 add .md file support in LLMFileAnalysisOperator (#71611)
a4b6b77e683 is described below
commit a4b6b77e6832a0047d6857544a927b3108e7ed94
Author: Jeff(Wei-Hao) Lu <[email protected]>
AuthorDate: Sun Aug 23 18:54:59 2026 +0800
add .md file support in LLMFileAnalysisOperator (#71611)
---
.../common/ai/docs/operators/llm_file_analysis.rst | 4 ++--
.../providers/common/ai/utils/file_analysis.py | 5 +++--
.../unit/common/ai/utils/test_file_analysis.py | 24 ++++++++++++++++++++++
3 files changed, 29 insertions(+), 4 deletions(-)
diff --git a/providers/common/ai/docs/operators/llm_file_analysis.rst
b/providers/common/ai/docs/operators/llm_file_analysis.rst
index 8b2733ab05a..c60435aa906 100644
--- a/providers/common/ai/docs/operators/llm_file_analysis.rst
+++ b/providers/common/ai/docs/operators/llm_file_analysis.rst
@@ -153,10 +153,10 @@ This operator also inherits ``LLMOperator``'s HITL review
parameters --
Supported Formats
-----------------
-- Text-like: ``.log``, ``.json``, ``.csv``, ``.parquet``, ``.avro``
+- Text-like: ``.log``, ``.txt``, ``.md``, ``.json``, ``.csv``, ``.parquet``,
``.avro``
- Multimodal: ``.png``, ``.jpg``, ``.jpeg``, ``.pdf`` when ``multi_modal=True``
- Gzip-compressed text inputs are supported for ``.log.gz``, ``.json.gz``, and
- ``.csv.gz``.
+ ``.csv.gz``, ``.txt.gz``, and ``.md.gz``.
- Gzip is not supported for ``.parquet``, ``.avro``, image, or PDF inputs.
Parquet and Avro readers require their corresponding optional extras:
diff --git
a/providers/common/ai/src/airflow/providers/common/ai/utils/file_analysis.py
b/providers/common/ai/src/airflow/providers/common/ai/utils/file_analysis.py
index 38c3d6f1492..6f6366b515e 100644
--- a/providers/common/ai/src/airflow/providers/common/ai/utils/file_analysis.py
+++ b/providers/common/ai/src/airflow/providers/common/ai/utils/file_analysis.py
@@ -49,13 +49,14 @@ SUPPORTED_FILE_FORMATS: tuple[str, ...] = (
"jpg",
"json",
"log",
+ "md",
"parquet",
"pdf",
"png",
"txt",
)
-_TEXT_LIKE_FORMATS = frozenset({"csv", "json", "log", "avro", "parquet",
"txt"})
+_TEXT_LIKE_FORMATS = frozenset({"csv", "json", "log", "avro", "parquet",
"txt", "md"})
_MULTI_MODAL_FORMATS = frozenset({"jpeg", "jpg", "pdf", "png"})
_COMPRESSION_SUFFIXES = {
"bz2": "bzip2",
@@ -64,7 +65,7 @@ _COMPRESSION_SUFFIXES = {
"xz": "xz",
"zst": "zstd",
}
-_GZIP_SUPPORTED_FORMATS = frozenset({"csv", "json", "log", "txt"})
+_GZIP_SUPPORTED_FORMATS = frozenset({"csv", "json", "log", "txt", "md"})
_TEXT_SAMPLE_HEAD_CHARS = 8_000
_TEXT_SAMPLE_TAIL_CHARS = 2_000
_MEDIA_TYPES = {
diff --git
a/providers/common/ai/tests/unit/common/ai/utils/test_file_analysis.py
b/providers/common/ai/tests/unit/common/ai/utils/test_file_analysis.py
index 2b524998b60..bba7f56111a 100644
--- a/providers/common/ai/tests/unit/common/ai/utils/test_file_analysis.py
+++ b/providers/common/ai/tests/unit/common/ai/utils/test_file_analysis.py
@@ -90,6 +90,28 @@ class TestBuildFileAnalysisRequest:
assert "first line" in request.user_content
assert "format=txt" in request.user_content
+ def test_markdown_file_analysis(self, tmp_path):
+ path = tmp_path / "notes.md"
+ path.write_text("# Notes\n\nFirst finding.\n", encoding="utf-8")
+
+ request = build_file_analysis_request(
+ file_path=str(path),
+ file_conn_id=None,
+ prompt="Summarize the notes",
+ multi_modal=False,
+ max_files=20,
+ max_file_size_bytes=1024,
+ max_total_size_bytes=2048,
+ max_text_chars=500,
+ sample_rows=10,
+ )
+
+ assert isinstance(request, FileAnalysisRequest)
+ assert request.resolved_paths == [str(path)]
+ assert "Summarize the notes" in request.user_content
+ assert "First finding" in request.user_content
+ assert "format=md" in request.user_content
+
def test_file_conn_id_overrides_embedded_connection(self, tmp_path):
path = tmp_path / "data.json"
path.write_text('{"status": "ok"}', encoding="utf-8")
@@ -360,6 +382,8 @@ class TestFileAnalysisHelpers:
("app", "log", None),
("notes.txt", "txt", None),
("notes.txt.gz", "txt", "gzip"),
+ ("notes.md", "md", None),
+ ("notes.md.gz", "md", "gzip"),
],
)
def test_detect_file_format(self, tmp_path, filename, expected_format,
expected_compression):
