xlsongc commented on code in PR #73926:
URL: https://github.com/apache/airflow/pull/73926#discussion_r4227256404
##########
providers/common/ai/docs/operators/document_loader.rst:
##########
@@ -73,14 +73,30 @@ Install the ``docx`` extra to parse Word documents via
pip install "apache-airflow-providers-common-ai[docx]"
-All non-empty paragraphs are concatenated into a single document per file.
+Paragraphs and tables in the document body are concatenated, in document
+order, into a single document per file. Empty paragraphs and table rows are
+skipped. Each table row becomes one line:
+
+.. code-block:: text
+
+ Quarterly results
+
+ | Region | Revenue |
+ | EMEA | 1.2M |
+
+A cell merged across columns appears once. A cell merged down several rows is
+repeated on each of those rows, so every row still reads on its own. A table
+nested inside a cell is flattened into that cell, with ``/`` between its cells
+and ``;`` between its rows. A table that ``python-docx`` cannot" read, such as
one
Review Comment:
Thanks, removed the stray `"`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
