From b81f98b8a66999117246fbc134fc09d64a04e230 Mon Sep 17 00:00:00 2001
From: Gardner Bickford <gbickford@gmail.com>
Date: Sun, 4 Jun 2023 09:57:25 +1200
Subject: [PATCH] Update confluence.py to return spaces between elements
 (#5383)

# Update confluence.py to return spaces between elements like headers
and links.

Please see
https://stackoverflow.com/questions/48913975/how-to-return-nicely-formatted-text-in-beautifulsoup4-when-html-text-is-across-m

Given:

```html
<address>
        183 Main St<br>East Copper<br>Massachusetts<br>U S A<br>
        MA 01516-113
    </address>
```

The document loader currently returns:

```
'183 Main StEast CopperMassachusettsU S A        MA 01516-113'
```

After this change, the document loader will return:

```
183 Main St East Copper Massachusetts U S A MA 01516-113
```


@eyurtsev would you prefer this to be an option that can be passed in?
---
 langchain/document_loaders/confluence.py | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/langchain/document_loaders/confluence.py b/langchain/document_loaders/confluence.py
index 2f3dde60..80e1ed12 100644
--- a/langchain/document_loaders/confluence.py
+++ b/langchain/document_loaders/confluence.py
@@ -347,15 +347,17 @@ class ConfluenceLoader(BaseLoader):
             attachment_texts = self.process_attachment(page["id"])
         else:
             attachment_texts = []
-        text = BeautifulSoup(
-            page["body"]["storage"]["value"], "lxml"
-        ).get_text() + "".join(attachment_texts)
+        text = BeautifulSoup(page["body"]["storage"]["value"], "lxml").get_text(
+            " ", strip=True
+        ) + "".join(attachment_texts)
         if include_comments:
             comments = self.confluence.get_page_comments(
                 page["id"], expand="body.view.value", depth="all"
             )["results"]
             comment_texts = [
-                BeautifulSoup(comment["body"]["view"]["value"], "lxml").get_text()
+                BeautifulSoup(comment["body"]["view"]["value"], "lxml").get_text(
+                    " ", strip=True
+                )
                 for comment in comments
             ]
             text = text + "".join(comment_texts)