From 59994c085c0c9779135e39780dad4ab2463aa697 Mon Sep 17 00:00:00 2001
From: "Arik Jones (aider)" <github@tnypxl.com>
Date: Sun, 22 Sep 2024 16:58:53 -0500
Subject: [PATCH] fix: improve file ignore logic and preserve newlines in
 extracted content

---
 cmd/files.go                | 11 +++++++----
 internal/scraper/scraper.go | 11 ++++++++---
 2 files changed, 15 insertions(+), 7 deletions(-)

diff --git a/cmd/files.go b/cmd/files.go
index a0d49d1..d6e6a7a 100644
--- a/cmd/files.go
+++ b/cmd/files.go
@@ -87,11 +87,14 @@ func isIgnored(filePath string, patterns []string) bool {
 				return true
 			}
 		} else {
-			// Check if the pattern matches any part of the file path
+			// Check if the pattern matches the full path or any part of it
+			if matched, _ := filepath.Match(pattern, filePath); matched {
+				return true
+			}
 			pathParts := strings.Split(filePath, string(os.PathSeparator))
-			for _, part := range pathParts {
-				matched, err := filepath.Match(pattern, part)
-				if err == nil && matched {
+			for i := range pathParts {
+				partialPath := filepath.Join(pathParts[:i+1]...)
+				if matched, _ := filepath.Match(pattern, partialPath); matched {
 					return true
 				}
 			}
diff --git a/internal/scraper/scraper.go b/internal/scraper/scraper.go
index 3c4e50e..81f7731 100644
--- a/internal/scraper/scraper.go
+++ b/internal/scraper/scraper.go
@@ -564,10 +564,15 @@ func ExtractContentWithCSS(content, includeSelector string, excludeSelectors []s
 		return "", fmt.Errorf("error extracting content with CSS selector: %v", err)
 	}
 
-	// Trim whitespace and normalize newlines
+	// Trim leading and trailing whitespace, but preserve internal newlines
 	selectedContent = strings.TrimSpace(selectedContent)
-	selectedContent = strings.ReplaceAll(selectedContent, "\n", "")
-	selectedContent = strings.ReplaceAll(selectedContent, "\t", "")
+
+	// Normalize newlines
+	selectedContent = strings.ReplaceAll(selectedContent, "\r\n", "\n")
+	selectedContent = strings.ReplaceAll(selectedContent, "\r", "\n")
+
+	// Remove any leading or trailing newlines
+	selectedContent = strings.Trim(selectedContent, "\n")
 
 	logger.Printf("Extracted content length: %d\n", len(selectedContent))
 	return selectedContent, nil