Add scripts

extract_version_diffs.sh

@@ -0,0 +1,131 @@
+#!/bin/bash
+#
+# note that this tool uses pdfresurrect -w, which creates version directories
+# and extracts version files. to avoid modifying your directory structure, it's
+# a good idea to run this script in a copy of your project directory.
+#
+# deps: qpdf, pdfresurrect
+
+LOG_FILE="./resurrection_log.txt"
+FAILURE_LOG="./resurrection_failures.log"
+
+echo "--- Starting PDF resurrection scan ---" | tee "$LOG_FILE"
+echo "Diffs of modified objects will be logged directly to: $LOG_FILE" | tee -a "$LOG_FILE"
+echo "--------------------------------------" | tee -a "$LOG_FILE"
+
+# reset failure log
+> "$FAILURE_LOG"
+
+find . -type f -iname "*.pdf" -not -iname "*_text.pdf" -not -path "*/*-versions/*" | while read -r FULL_PATH; do
+
+    FILE_DIR=$(dirname "$FULL_PATH")
+    FILE_NAME=$(basename "$FULL_PATH")
+    BASE_NAME="${FILE_NAME%.*}" # filename w/o ext
+
+    echo "Checking: $FULL_PATH"
+
+    (
+        cd "$FILE_DIR" || { echo "ERROR: Cannot access $FILE_DIR" >> "$FAILURE_LOG"; exit 1; }
+
+        VERSION_COUNT_OUTPUT=$(pdfresurrect -q "$FILE_NAME" 2>/dev/null)
+        EXIT_CODE=$?
+
+        if [ "$EXIT_CODE" -ne 0 ]; then
+            echo "Failed to run pdfresurrect -q on $FULL_PATH (Exit $EXIT_CODE)." >> "$FAILURE_LOG"
+            exit 1
+        fi
+
+        VERSION_COUNT=$(echo "$VERSION_COUNT_OUTPUT" | grep -oP ': \K\d+' | tail -n 1)
+
+        if [ "$VERSION_COUNT" -le 1 ]; then
+            echo "  -> No history found ($VERSION_COUNT version(s))."
+            exit 0
+        fi
+
+        echo "  -> History detected ($VERSION_COUNT versions). Extracting..."
+
+        pdfresurrect -w "$FILE_NAME" > /dev/null 2>&1
+
+        SUMMARY_FILE="$BASE_NAME-versions/$BASE_NAME-versions.summary"
+
+        if [ ! -f "$SUMMARY_FILE" ]; then
+            echo "  -> CRITICAL ERROR: Summary file not created at $SUMMARY_FILE, skipping diff."
+            exit 1
+        fi
+
+        RAW_MODIFIED_OBJECTS=$(awk '/: --[DM]--/ { print $4, $7 }' "$SUMMARY_FILE" | sort -u)
+
+        echo "  -> Raw D/M object list (V OBJ):"
+        echo "$RAW_MODIFIED_OBJECTS" | while read -r line; do
+            [ -n "$line" ] && echo "     - $line"
+        done
+
+        while IFS= read -r line ; do
+            CURRENT_VERSION=$(echo "$line" | awk '{print $1}')
+            OBJ_ID=$(echo "$line" | awk '{print $2}')
+            
+            if [ -z "$CURRENT_VERSION" ] || [ -z "$OBJ_ID" ]; then
+                [ -n "$line" ] && echo "   -> WARNING: Parsed empty version/object pair from line: \"$line\". Skipping."
+                continue
+            fi
+
+            if ! [[ "$CURRENT_VERSION" =~ ^[0-9]+$ ]] || ! [[ "$OBJ_ID" =~ ^[0-9]+$ ]]; then
+                echo "   -> ERROR: Parsed invalid numbers ($CURRENT_VERSION : $OBJ_ID) from summary. Skipping."
+                continue
+            fi
+
+            if [ "$CURRENT_VERSION" -eq 1 ]; then
+                 echo "   -> Skipping Object $OBJ_ID: Current version is V1, no V0 exists to diff against."
+                 continue
+            fi
+            
+            PREV_VERSION=$((CURRENT_VERSION - 1))
+
+            V1_FILE="$BASE_NAME-versions/$BASE_NAME-version-$PREV_VERSION.pdf"
+            V2_FILE="$BASE_NAME-versions/$BASE_NAME-version-$CURRENT_VERSION.pdf"
+    
+            TEMP_V1=$(mktemp)
+            TEMP_V2=$(mktemp)
+
+            qpdf --show-object="$OBJ_ID" --unfilter -- "$V1_FILE" > "$TEMP_V1" 2>/dev/null
+            qpdf --show-object="$OBJ_ID" --unfilter -- "$V2_FILE" > "$TEMP_V2" 2>/dev/null
+
+            if [ ! -s "$TEMP_V1" ] || [ ! -s "$TEMP_V2" ]; then
+                if [ ! -s "$TEMP_V1" ]; then
+                    qpdf --show-object="$OBJ_ID" -- "$V1_FILE" > "$TEMP_V1" 2>/dev/null
+                fi
+                
+                if [ ! -s "$TEMP_V2" ]; then
+                    qpdf --show-object="$OBJ_ID" -- "$V2_FILE" > "$TEMP_V2" 2>/dev/null
+                fi
+            fi
+            if [ ! -s "$TEMP_V1" ] || [ ! -s "$TEMP_V2" ]; then
+                echo "   -> NOTE: Object $OBJ_ID (V$PREV_VERSION or V$CURRENT_VERSION) extraction failed or produced empty output. Skipping diff."
+                rm "$TEMP_V1" "$TEMP_V2"
+                continue
+            fi
+
+            if diff -u "$TEMP_V1" "$TEMP_V2" >/dev/null ; then
+                echo "   -> Object $OBJ_ID: No difference detected (V$PREV_VERSION vs V$CURRENT_VERSION)."
+            else
+                echo -e "\n\n--- DIFF START: $FULL_PATH | Object $OBJ_ID (V$PREV_VERSION vs V$CURRENT_VERSION) ---"
+                echo "Note: Output extracted as raw PDF content for reliable comparison."
+                diff -u "$TEMP_V1" "$TEMP_V2"
+                echo "--- DIFF END: Object $OBJ_ID ---\n"
+                DIFF_FOUND=1
+            fi
+
+
+            rm "$TEMP_V1" "$TEMP_V2"
+        done < <(echo "$RAW_MODIFIED_OBJECTS")
+
+
+    ) | tee -a "$LOG_FILE"
+
+    if [ ${PIPESTATUS[0]} -ne 0 ]; then
+        echo "Failed to process $FULL_PATH (See $FAILURE_LOG and $LOG_FILE)"
+    fi
+done
+echo "----------------------------------------------------" | tee -a "$LOG_FILE"
+echo "Scan complete. Check $LOG_FILE for inline object diffs." | tee -a "$LOG_FILE"
+echo "Check $FAILURE_LOG for files that failed processing." | tee -a "$LOG_FILE"