[gentoo-commits] proj/devmanual:master commit in: bin/

["Ulrich Müller" <ulm@gentoo.org>]

commit:     1bfa9bda3f10627a9798edfc65472d59bc9ffeba
Author:     Göktürk Yüksek <gokturk <AT> gentoo <DOT> org>
AuthorDate: Sat Dec 21 04:36:03 2019 +0000
Commit:     Ulrich Müller <ulm <AT> gentoo <DOT> org>
CommitDate: Sat Dec 21 05:18:16 2019 +0000
URL:        https://gitweb.gentoo.org/proj/devmanual.git/commit/?id=1bfa9bda

bin/build_search_documents.py: fix aggressive whitespace stripping

In stringify_node(), we aggressively strip the whitespaces around
children nodes. This results in something like
"<c>SLOT</c>, <c>:SLOT<c/>" being parsed as "SLOT,:SLOT",
removing the white space between ',' and ':'.

Signed-off-by: Göktürk Yüksek <gokturk <AT> gentoo.org>
Signed-off-by: Ulrich Müller <ulm <AT> gentoo.org>

 bin/build_search_documents.py | 18 ++++++++++++++----
 1 file changed, 14 insertions(+), 4 deletions(-)

diff --git a/bin/build_search_documents.py b/bin/build_search_documents.py
index 3816fdb..e19dce6 100755
--- a/bin/build_search_documents.py
+++ b/bin/build_search_documents.py
@@ -18,22 +18,32 @@ def stringify_node(parent: ET.Element) -> str:
     parent -- the node to convert to a string
 
     """
+    # We usually have something like:
+    #   <p>\nText
+    # Left strip the whitespace.
     if parent.text:
         text = parent.text.lstrip()
     else:
         text = str()
 
+    # For each child, strip the tags and append to text
+    # along with the tail text following it.
+    # The tail may include '\n' if it spans multiple lines.
+    # We will worry about those on return, not now.
     for child in parent.getchildren():
         # The '<d/>' tag is simply a fancier '-' character
         if child.tag == 'd':
             text += '-'
         if child.text:
-            text += child.text.lstrip()
+            text += child.text
         if child.tail:
-            text += child.tail.rstrip()
+            text += child.tail
 
-    text += parent.tail.rstrip()
-    return text.replace('\n', ' ')
+    # A paragraph typically ends with:
+    #   Text\n</p>
+    # Right strip any spurious whitespace.
+    # Finally, get rid of any intermediate newlines.
+    return text.rstrip().replace('\n', ' ')
 
 
 def process_node(documents: list, node: ET.Element, name: str, url: str) -> None: