Reduced the number of htmlization URLs visited further.

 - Legacy-Id: 17999

Author: levkowetz

bin/commitlog

@@ -1,15 +1,31 @@
 #!/bin/bash
-# -*- indent-with-tabs: 1 -*-
-date
 cur=$(svn info | awk '/^Revision:/ { print $2 }')
-svn log $PWD -r ${2:-HEAD}:${1:-$((cur-100))} \
+# List recent commits, extract revision numbers, stop at release, and grab the last revision
+#       - get svn log, looking backwards from HEAD
+#       - stop on 'Set version info ...'
+#       - filter out revision lines
+#       - grab the first field
+#       - grab the last line
+#       - remove the leading 'r'
+rev=$(svn log $PWD -r ${2:-HEAD}:${1:-$((cur-100))}  \
+	| sed -r -n -e '1,/^Set version info( and settings)?( back)? to (development|release)/p' \
+        | egrep '^r[0-9]+ \|' \
+        | cut -d ' ' -f 1 \
+        | tail -n 1 \
+        | sed 's/^r//')
+
+# Grab commit log lines from just after the start rev, going forwards, and reformat
+#       - Get svn log entries, starting with the earliest
+#       - Filter out dividing lines and revision/committer/date lines, keeping the messages
+#       - Insert leading asterisks: '  * ' at the front of the first line in each text block
+#       - Unwrap lines that start with unindented text
+#       - Do line folding at column 76
+#       - Indent any unindented lines 4 spaces
+#       - Add blank lines in front of log entries
+svn log $PWD -r $((rev+2)):${2:-HEAD} \
 	| sed -r 's/^(----------|r[0-9]+).*$/\n/'  \
-	| sed '/./,/^$/!d'  \
 	| sed '1,/./s/^/  * /;/^$/,/./s/^/  * /;/^  \* $/d'  \
 	| sed -e :a -e '$!N;s/\n([A-Za-z0-9])/ \\1/;ta' -e 'P;D'  \
-	| head -n -1  \
-	| fold -sw1000 \
-	| tac \
 	| fold -sw76  \
 	| sed -r 's/^([^ ].*)$/    &/'  \
 	| sed -r 's/^  \* /\n  * /'

bin/test-crawl

@@ -1,6 +1,8 @@
 #!/usr/bin/env python
+# -*- indent-with-tabs: 0 -*-
 # Copyright The IETF Trust 2013-2019, All Rights Reserved
 
+
 import os, sys, re, datetime, argparse, traceback, json, subprocess
 import html5lib
 import random
@@ -225,7 +227,8 @@ def skip_url(url):
             r"draft-touch-msword-template-v2\.0",
 
             # Skip most html conversions, not worth the time
-            "^/doc/html/draft-[0-9ac-z][0-9b-z].*",
+            "^/doc/html/draft-[0-9ac-z]",
+            "^/doc/html/draft-b[0-9b-z]",
             "^/doc/html/charter-.*",
             "^/doc/html/status-.*",
             "^/doc/html/rfc.*",