From 49188fc3c2d894c5e1cd909b612873f6ef60fb1d Mon Sep 17 00:00:00 2001
From: Kouame Behouba Manasse <behouba@gmail.com>
Date: Wed, 6 Mar 2024 07:46:13 -0500
Subject: [PATCH] feat(crit): add SearchPattern method on MemoryReader

This commit add a new method SearchPattern to MemoryReader to search
for patterns inside the process memory pages. This method accept regular
expressions for flexible pattern matching.

Signed-off-by: Kouame Behouba Manasse <behouba@gmail.com>
---
 crit/mempages.go | 52 ++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 52 insertions(+)

diff --git a/crit/mempages.go b/crit/mempages.go
index c823edbaa..4134fc6c7 100644
--- a/crit/mempages.go
+++ b/crit/mempages.go
@@ -6,6 +6,7 @@ import (
 	"fmt"
 	"os"
 	"path/filepath"
+	"regexp"
 
 	"github.com/checkpoint-restore/go-criu/v7/crit/images/mm"
 	"github.com/checkpoint-restore/go-criu/v7/crit/images/pagemap"
@@ -193,3 +194,54 @@ func (mr *MemoryReader) GetShmemSize() (int64, error) {
 
 	return size, nil
 }
+
+// SearchPattern searches for a pattern in the process memory pages.
+// If the pattern is found, it returns slices of strings containing the matched patterns
+// along with the specified context before and after each match.
+func (mr *MemoryReader) SearchPattern(pattern string, context int) ([]string, error) {
+	if context < 0 {
+		return nil, errors.New("context size cannot be negative")
+	}
+
+	var results []string
+	regexPattern, err := regexp.Compile(pattern)
+	if err != nil {
+		return nil, err
+	}
+
+	for _, entry := range mr.pagemapEntries {
+		start := entry.GetVaddr()
+		end := start + uint64(entry.GetNrPages())*uint64(mr.pageSize)
+
+		// Read the entire page content
+		buffer, err := mr.GetMemPages(start, end)
+		if err != nil {
+			return nil, err
+		}
+
+		// Replace null bytes with space
+		filteredBytes := bytes.ReplaceAll(buffer.Bytes(), []byte{0x0}, []byte{0x20})
+
+		// Search for all matching pattern using the regular expression
+		matches := regexPattern.FindAllIndex(filteredBytes, -1)
+
+		for _, match := range matches {
+			startPattern, endPattern := match[0], match[1]
+
+			// Calculate the start and end indices for the context surrounding the match
+			startContext := startPattern - context
+			if startContext < 0 {
+				startContext = startPattern
+			}
+
+			endContext := endPattern + context
+			if endContext > buffer.Len() {
+				endContext = endPattern
+			}
+
+			results = append(results, string(filteredBytes[startContext:endContext]))
+		}
+	}
+
+	return results, nil
+}