From b6068823271c96b4ab344dde03813aae8b1a72c3 Mon Sep 17 00:00:00 2001
From: Jarrod Johnson <jjohnson2@lenovo.com>
Date: Fri, 19 Apr 2024 08:22:32 -0400
Subject: [PATCH] Have collate preserve relative whitespace

The change to tolerate either a space or no space ended up
greedily consuming whitespace.  Do best possible in two cases:

For log, use the first line as a clue, and consistently pad or not pad
according to first line.  It won't catch different pad strategies, or
handle first line being indented but other lines not being indented.

For the textgroup variant, allow subsequent lines to revise the
pad downward, and accept any whitespace, not just space.
---
 confluent_client/bin/collate            | 11 +++++++++--
 confluent_client/confluent/textgroup.py | 11 +++++++++--
 2 files changed, 18 insertions(+), 4 deletions(-)

diff --git a/confluent_client/bin/collate b/confluent_client/bin/collate
index 07095901..2a086303 100755
--- a/confluent_client/bin/collate
+++ b/confluent_client/bin/collate
@@ -21,6 +21,7 @@
 
 import optparse
 import os
+import re
 import select
 import sys
 
@@ -84,6 +85,7 @@ fullline = sys.stdin.readline()
 printpending = True
 clearpending = False
 holdoff = 0
+padded = None
 while fullline:
     for line in fullline.split('\n'):
         if not line:
@@ -92,13 +94,18 @@ while fullline:
             line = 'UNKNOWN: ' + line
         if options.log:
             node, output = line.split(':', 1)
-            output = output.lstrip()
+            if padded is None:
+                if output.startswith(' '):
+                    padded = True
+                else:
+                    padded = False
+            if padded:
+                output = re.sub(r'^ ', '', output)
             currlog = options.log.format(node=node, nodename=node)
             with open(currlog, mode='a')  as log:
                 log.write(output + '\n')
             continue
         node, output = line.split(':', 1)
-        output = output.lstrip()
         grouped.add_line(node, output)
         if options.watch:
             if not holdoff:
diff --git a/confluent_client/confluent/textgroup.py b/confluent_client/confluent/textgroup.py
index cd35b6fa..e2f0dc7f 100644
--- a/confluent_client/confluent/textgroup.py
+++ b/confluent_client/confluent/textgroup.py
@@ -98,17 +98,24 @@ class GroupedData(object):
         self.byoutput = {}
         self.header = {}
         self.client = confluentconnection
+        self.detectedpad = None
 
     def generate_byoutput(self):
         self.byoutput = {}
+        thepad = self.detectedpad if self.detectedpad else ''
         for n in self.bynode:
-            output = '\n'.join(self.bynode[n])
+            output = ''
+            for ln in self.bynode[n]:
+                output += ln.replace(thepad, '', 1) + '\n'
             if output not in self.byoutput:
                 self.byoutput[output] = set([n])
             else:
                 self.byoutput[output].add(n)
 
     def add_line(self, node, line):
+        wspc = re.search(r'^\s*', line).group()
+        if self.detectedpad is None or len(wspc) < len(self.detectedpad):
+            self.detectedpad = wspc
         if node not in self.bynode:
             self.bynode[node] = [line]
         else:
@@ -219,4 +226,4 @@ if __name__ == '__main__':
         if not line:
             continue
         groupoutput.add_line(*line.split(': ', 1))
-    groupoutput.print_deviants()
\ No newline at end of file
+    groupoutput.print_deviants()