Retrieve valid tags beforehand

tools/sep_test.sh

@@ -2,13 +2,13 @@
 
 # Function to print usage information
 usage() {
-    echo "Usage: $0 <image> <start_tag> <end_tag>"
-    echo "Example: $0 tabledevil/sep 230101 230916"
+    echo "Usage: $0 <image> <start_tag> <end_tag> <username> <token>"
+    echo "Example: $0 tabledevil/sep 230101 230916 username dckr_pat_8FEgaA5ovvL1V-UEWfV5T3jVABC"
     exit 1
 }
 
-# Check if correct number of arguments is provided
-if [ "$#" -ne 3 ]; then
+# Check if the correct number of arguments is provided
+if [ "$#" -ne 5 ]; then
     usage
 fi
 
@@ -16,6 +16,10 @@ fi
 image="$1"
 start_tag="$2"
 end_tag="$3"
+DOCKER_USERNAME="$4"
+DOCKER_PAT="$5"
+REPOSITORY=$(echo "$image" | cut -d'/' -f2)
+PATTERN="Threat Found!"
 
 # Validate that start_tag and end_tag are in the correct format
 if ! [[ "$start_tag" =~ ^[0-9]{6}$ ]] || ! [[ "$end_tag" =~ ^[0-9]{6}$ ]]; then
@@ -23,50 +27,117 @@ if ! [[ "$start_tag" =~ ^[0-9]{6}$ ]] || ! [[ "$end_tag" =~ ^[0-9]{6}$ ]]; then
     usage
 fi
 
-# Function to generate all possible tags between two dates
-generate_tags() {
-    local start_date=$(date -d "20$1" +%Y%m%d)
-    local end_date=$(date -d "20$2" +%Y%m%d)
-    local current_date="$start_date"
-    local tags=()
-
-    while [[ "$current_date" != "$end_date" ]]; do
-        if docker manifest inspect "${image}:${current_date:2}" > /dev/null 2>&1; then
-            tags+=("${current_date:2}")
+# Function to get Docker Hub token using PAT
+get_token() {
+    TOKEN=$(curl -s -H "Content-Type: application/json" -X POST -d '{"username": "'${DOCKER_USERNAME}'", "password": "'${DOCKER_PAT}'"}' https://hub.docker.com/v2/users/login/ | jq -r .token)
+    if [ "$TOKEN" == "null" ]; then
+        echo "Failed to get token. Please check your credentials."
+        exit 1
     fi
-        current_date=$(date -d "$current_date + 1 day" +%Y%m%d)
-    done
-    tags+=("${end_date:2}") # Include the end date as well
-    echo "${tags[@]}"
 }
 
-# Generate tags
-tags=($(generate_tags "$start_tag" "$end_tag"))
+# Function to get tags for a repository
+get_tags() {
+    PAGE=1
+    PAGE_SIZE=100
+    TAGS=()
 
-# Binary search setup
-low=0
-high=${#tags[@]}
+    while true; do
+        RESPONSE=$(curl -s -H "Authorization: JWT ${TOKEN}" "https://hub.docker.com/v2/repositories/${DOCKER_USERNAME}/${REPOSITORY}/tags/?page_size=${PAGE_SIZE}&page=${PAGE}")
+        TAGS_PAGE=$(echo $RESPONSE | jq -r '.results[].name')
+        TAGS+=($TAGS_PAGE)
 
-# Binary search to find the first tag detecting the virus
-while [ $low -lt $high ]; do
-    mid=$(((low + high) / 2))
-    tag=${tags[$mid]}
-
-    # Run the Docker command with the current tag
-    result=$(docker run -it --rm -v "$(pwd):/data:ro" --network=none "$image:$tag" scan)
-
-    if [[ $result == *"Threat Found!"* ]]; then
-        # Virus found, search in the lower half
-        high=$mid
-    else
-        # Virus not found, search in the upper half
-        low=$((mid + 1))
+        NEXT=$(echo $RESPONSE | jq -r '.next')
+        if [ "$NEXT" == "null" ]; then
+            break
         fi
-done
+        PAGE=$((PAGE + 1))
+    done
 
-# Output the tag of the first image that finds the malware
-if [ $high -lt ${#tags[@]} ]; then
-    echo "The first tag that detected the malware is: ${tags[$high]}"
-else
-    echo "No malware detected within the tag range."
-fi
+    echo "${TAGS[@]}"
+}
+
+# Function to filter tags based on date range
+filter_tags_by_date() {
+    local tags=("$@")
+    local filtered=()
+    local start="$start_tag"
+    local end="$end_tag"
+
+    for tag in "${tags[@]}"; do
+        if [[ "$tag" =~ ^[0-9]{6}$ ]]; then
+            if [[ "$tag" -ge "$start" && "$tag" -le "$end" ]]; then
+                filtered+=("$tag")
+            fi
+        fi
+    done
+
+    echo "${filtered[@]}"
+}
+
+# Function to check for malware in the given Docker tag
+check_malware() {
+    local tag=$1
+    echo -n "Checking $image:$tag "
+    result=$(docker run -it --rm -v "$(pwd):/data:ro" --network=none "$image:$tag" scan)
+    if echo "$result" | grep -q "$PATTERN"; then
+        echo "Pattern found"
+        return 0
+    else
+        echo "Pattern NOT found"
+        return 1
+    fi
+}
+
+# Function to run the binary search
+binary_search() {
+    local tags=("$@")
+    local low=0
+    local high=$(( ${#tags[@]} - 1 ))
+    local mid
+
+    # Determine the initial states for low and high
+    check_malware "${tags[$low]}"
+    local low_result=$?
+    
+    check_malware "${tags[$high]}"
+    local high_result=$?
+
+    # If the results for low and high are the same, there is no switch point in the range
+    if [ $low_result -eq $high_result ]; then
+        echo "No change in detection within the tag range."
+        return
+    fi
+
+    # Binary search to find the exact switching point
+    while [ $((low + 1)) -lt $high ]; do
+        mid=$(((low + high) / 2))
+        check_malware "${tags[$mid]}"
+        local mid_result=$?
+
+        if [ $mid_result -eq $low_result ]; then
+            low=$mid
+        else
+            high=$mid
+        fi
+    done
+
+    # Output the tag of the first image that finds the malware switch
+    echo "The detection changes between tags: ${tags[$low]} - ${tags[$high]}"
+}
+
+
+# Main script execution
+get_token
+
+# Retrieve all tags from Docker Hub
+echo "Retrieving all tags for $image from Docker Hub..."
+all_tags=($(get_tags))
+
+# Filter tags to include only those within the date range
+echo "Filtering tags from $start_tag to $end_tag..."
+filtered_tags=($(filter_tags_by_date "${all_tags[@]}"))
+
+# Run the binary search on the filtered tags
+echo "Running binary search on the filtered tags..."
+binary_search "${filtered_tags[@]}"