[lucene-jira-archive] 01/01: fix stack overflow when parsing lists

[to...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

tomoko pushed a commit to branch fix-stackoverflow-in-list-parser
in repository https://gitbox.apache.org/repos/asf/lucene-jira-archive.git

commit a0cc0f5caa0541cd992af480fee8ad2a2978c76f
Author: Tomoko Uchida <to...@gmail.com>
AuthorDate: Tue Jul 12 22:23:16 2022 +0900

    fix stack overflow when parsing lists
---
 migration/src/markup/lists.py | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/migration/src/markup/lists.py b/migration/src/markup/lists.py
index 47853a2c..3ce03ca9 100644
--- a/migration/src/markup/lists.py
+++ b/migration/src/markup/lists.py
@@ -40,6 +40,11 @@ class TweakedList(AbstractMarkup):
 
         for line in tokens:
             # print(repr(line))
+            if line == "\n":
+                # can't really explain but if this is the first item, an empty string should be added to preserve line feed
+                if len(result) == 0:
+                    result.append("")
+                continue
             bullets, text = line.split(" ", maxsplit=1)
 
             nested_indent = 0
@@ -66,12 +71,11 @@ class TweakedList(AbstractMarkup):
         NL = LineEnd()
         LIST_BREAK = NL + Optional(White(" \t")) + NL | StringEnd()
         IGNORE = BlockQuote(**self.init_kwargs).expr | Panel(**self.init_kwargs).expr | Color(**self.init_kwargs).expr
-        ROW = Optional(LineStart()) + Combine(
-            Optional(White(" \t"))
+        ROW = (LineStart() ^ LineEnd()) + Combine(
+            Optional(NL)
             + Optional(self.nested_token, default="")
             + ListIndent(self.indent_state, self.tokens)
             + SkipTo(NL + Optional(White(" \t")) + Char(self.nested_token + self.tokens) | LIST_BREAK, ignore=IGNORE)
-            + Optional(NL),
         )
 
         return OneOrMore(ROW, stopOn=LIST_BREAK).setParseAction(self.action)