zik/lampac-ukraine
1
0
Fork 0
mirror of https://github.com/lampame/lampac-ukraine.git synced 2026-04-16 17:32:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

refactor(makhno): improve file array extraction logic

Browse Source 
Replace regex-based file array extraction with a more robust bracket
matching algorithm that properly handles nested structures. The new
implementation uses manual parsing to track bracket depth, ensuring
correct extraction of JSON arrays from HTML content.
This commit is contained in:
baliasnyifeliks 2026-02-03 20:01:22 +02:00
parent ed7bfa67de
commit 6208c14ef0
1 changed files with 36 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

47
Makhno/MakhnoInvoke.cs
View File

@ -366,20 +366,45 @@ namespace Makhno
if (string.IsNullOrEmpty(html))
return null;
var matches = new[]
{
Regex.Match(html, @"file\s*:\s*'(\[.*\])'", RegexOptions.Singleline),
Regex.Match(html, @"file\s*:\s*""(\[.*\])""", RegexOptions.Singleline),
Regex.Match(html, @"file\s*:\s*(\[[\s\S]*?\])", RegexOptions.Singleline)
};
var startIndex = FindFileArrayStart(html);
if (startIndex < 0)
return null;
foreach (var match in matches)
string jsonArray = ExtractBracketArray(html, startIndex);
if (string.IsNullOrEmpty(jsonArray))
return null;
return jsonArray
.Replace("\\'", "'")
.Replace("\\\"", "\"");
}
private int FindFileArrayStart(string html)
{
int fileIndex = html.IndexOf("file", StringComparison.OrdinalIgnoreCase);
if (fileIndex < 0)
return -1;
int bracketIndex = html.IndexOf('[', fileIndex);
return bracketIndex;
}
private string ExtractBracketArray(string text, int startIndex)
{
if (startIndex < 0 || startIndex >= text.Length || text[startIndex] != '[')
return null;
int depth = 0;
for (int i = startIndex; i < text.Length; i++)
{
if (match.Success)
char ch = text[i];
if (ch == '[')
depth++;
else if (ch == ']')
{
return match.Groups[1].Value
.Replace("\\'", "'")
.Replace("\\\"", "\"");
depth--;
if (depth == 0)
return text.Substring(startIndex, i - startIndex + 1);
}
}