Upgrade binary test fixtures management

.github/scripts/check_binary_fixture_size.sh

@@ -0,0 +1,35 @@
+#!/bin/bash
+
+# Check if a directory is provided as an argument
+if [ $# -eq 0 ]; then
+  echo "Usage: $0 <directory>"
+  exit 1
+fi
+
+directory="$1"
+
+# Check if the directory exists
+if [ ! -d "$directory" ]; then
+  echo "Directory not found: $directory"
+  exit 1
+fi
+
+# Use find to locate all files in the directory and its subdirectories
+found_large_files=0
+while IFS= read -r -d '' file; do
+  # Check if the file size is greater than 100 bytes
+  if [ $(wc -c < "$file") -gt 100 ]; then
+    echo "File $file is greater than 100 bytes."
+    found_large_files=1
+  fi
+done < <(find "$directory" -type f -print0)
+
+# Check if any large files were found
+if [ "$found_large_files" -eq 1 ]; then
+  echo "Script failed: Some files are greater than 100 bytes."
+  exit 1
+else
+  echo "All files in $directory and its subdirectories are 100 bytes or smaller. Script passed."
+  exit 0
+fi
+

Taskfile.yaml

@@ -54,6 +54,7 @@ tasks:
       - task: check-licenses
       - task: lint
       - task: check-json-schema-drift
+      - task: check-binary-fixture-size
 
   test:
     desc: Run all levels of test
@@ -169,6 +170,11 @@ tasks:
     cmds:
       - .github/scripts/json-schema-drift-check.sh
 
+  check-binary-fixture-size:
+    desc: Ensure that the binary test fixtures are not too large
+    cmds:
+      - .github/scripts/check_binary_fixture_size.sh syft/pkg/cataloger/binary/test-fixtures/classifiers/positive
+
 
   ## Testing tasks #################################
 

syft/pkg/cataloger/binary/test-fixtures/get-fixture-snippet.sh

@@ -0,0 +1,130 @@
+#!/bin/bash
+
+# Default values for length and prefix length
+LENGTH=100
+PREFIX_LENGTH=10
+
+# Function to show usage
+usage() {
+    echo "Usage: $0 <path-to-binary> <search-pattern> [--length <length>] [--prefix-length <prefix_length>]"
+    exit 1
+}
+
+# Parse arguments
+while [[ $# -gt 0 ]]; do
+    key="$1"
+
+    case $key in
+        --length)
+            LENGTH="$2"
+            shift # past argument
+            shift # past value
+            ;;
+        --prefix-length)
+            PREFIX_LENGTH="$2"
+            shift # past argument
+            shift # past value
+            ;;
+        *)
+            if [ -z "$BINARY_FILE" ]; then
+                BINARY_FILE="$1"
+            elif [ -z "$PATTERN" ]; then
+                PATTERN="$1"
+            else
+                echo "Unknown option: $1"
+                usage
+            fi
+            shift # past argument
+            ;;
+    esac
+done
+
+# check if binary file and pattern are provided
+if [ -z "$BINARY_FILE" ] || [ -z "$PATTERN" ]; then
+    usage
+fi
+
+# check if xxd is even installed
+if ! command -v xxd &> /dev/null; then
+    echo "xxd not found. Please install xxd."
+    exit 1
+fi
+
+PATTERN_RESULTS=$(strings -a -t d "$BINARY_FILE" | grep "$PATTERN")
+
+# if there are multiple matches, prompt the user to select one
+if [ $(echo "$PATTERN_RESULTS" | wc -l) -gt 1 ]; then
+    echo "Multiple string matches found in the binary:"
+    echo ""
+
+    # show result lines one at a time (in a numbered list)
+    # but only show everything after the first field (not the offset)
+    echo "$PATTERN_RESULTS" | cut -d ' ' -f 2- | nl -w 1 -s ') '
+
+
+    echo ""
+    read -p "Please select a match: " SELECTION
+
+    # if the selection is not a number, exit
+    if ! [[ "$SELECTION" =~ ^[0-9]+$ ]]; then
+        echo "Invalid selection."
+        exit 1
+    fi
+
+    # if the selection is out of bounds, exit
+    if [ "$SELECTION" -gt $(echo "$PATTERN_RESULTS" | wc -l) ]; then
+        echo "Invalid selection."
+        exit 1
+    fi
+
+    # select the line from the results
+    PATTERN_RESULTS=$(echo "$PATTERN_RESULTS" | sed -n "${SELECTION}p")
+fi
+
+# search for the pattern in the binary file and capture the offset
+OFFSET=$(echo "${PATTERN_RESULTS}" | cut -d ' ' -f 1)
+
+if [ -z "$OFFSET" ]; then
+    echo "Pattern not found."
+    exit 1
+fi
+
+# adjust the offset to capture prefix length before the match
+OFFSET=$(expr "$OFFSET" - "$PREFIX_LENGTH")
+
+# use xxd to capture the specified length from the calculated offset
+SNIPPET=$(xxd -l "$LENGTH" -s "$OFFSET" "$BINARY_FILE")
+
+# display the output and prompt the user
+echo ""
+echo "$SNIPPET"
+echo ""
+read -p "Does this snippet capture what you need? (Y/n) " RESPONSE
+RESPONSE=${RESPONSE:-y}
+
+if [ "$RESPONSE" != "y" ]; then
+    echo "Exiting with no action taken."
+    exit 1
+fi
+
+# generate a text file with metadata and the binary snippet
+SHA256=$(sha256sum "$BINARY_FILE" | cut -d ' ' -f 1)
+DATE=$(date)
+BASE64_PATTERN=$(echo -n "$PATTERN" | base64)
+FILENAME=$(basename "$BINARY_FILE")
+INFO=$(file -b "$BINARY_FILE")
+OUTPUT_FILE="classifiers/positive/$FILENAME-$SHA256-$OFFSET-$LENGTH.snippet"
+
+cat > "$OUTPUT_FILE" <<EOF
+### generated by script $(basename $0) at $DATE ###
+# filename:       $FILENAME
+# sha256:         $SHA256
+# file info:      $INFO
+# base64(search): $BASE64_PATTERN
+# start offset:   $OFFSET
+# length:         $LENGTH
+### start of binary snippet ###
+EOF
+echo "$SNIPPET" | xxd -r -s -"$OFFSET" >> "$OUTPUT_FILE"
+
+echo "Snippet written to $OUTPUT_FILE"