From 4fa585aee375867c99787467a0258fe10a3d4bd5 Mon Sep 17 00:00:00 2001
From: Karl Ostmo <kostmo@fb.com>
Date: Mon, 14 Jan 2019 17:58:46 -0800
Subject: [PATCH] fix --list on truncated files

fseek() doesn't indicate when it moves past the end of a file.
Consequently, if a file is truncated within its last block, the error would't be detected.

This PR adds a test scenario that induces this situation using a small compressed file of only one block in size.
This test is added to tests/playTests.sh

Check is implemented by ensuring that the filehandle position is equal to the filesize upon exit.
---
 programs/fileio.c  | 52 +++++++++++++++++++++++++++++++++-------------
 tests/playTests.sh | 13 ++++++++++++
 2 files changed, 51 insertions(+), 14 deletions(-)

diff --git a/programs/fileio.c b/programs/fileio.c
index 9fb795ed..8538285b 100644
--- a/programs/fileio.c
+++ b/programs/fileio.c
@@ -237,10 +237,13 @@ void FIO_addAbortHandler()
 ***************************************************************/
 #if defined(_MSC_VER) && _MSC_VER >= 1400
 #   define LONG_SEEK _fseeki64
+#   define LONG_TELL _ftelli64
 #elif !defined(__64BIT__) && (PLATFORM_POSIX_VERSION >= 200112L) /* No point defining Large file for 64 bit */
 #  define LONG_SEEK fseeko
+#  define LONG_TELL ftello
 #elif defined(__MINGW32__) && !defined(__STRICT_ANSI__) && !defined(__NO_MINGW_LFS) && defined(__MSVCRT__)
 #   define LONG_SEEK fseeko64
+#   define LONG_TELL ftello64
 #elif defined(_WIN32) && !defined(__DJGPP__)
 #   include <windows.h>
     static int LONG_SEEK(FILE* file, __int64 offset, int origin) {
@@ -261,6 +264,7 @@ void FIO_addAbortHandler()
     }
 #else
 #   define LONG_SEEK fseek
+#   define LONG_TELL ftell
 #endif
 
 
@@ -2142,7 +2146,13 @@ typedef struct {
     U32 nbFiles;
 } fileInfo_t;
 
-typedef enum { info_success=0, info_frame_error=1, info_not_zstd=2, info_file_error=3 } InfoError;
+typedef enum {
+  info_success=0,
+  info_frame_error=1,
+  info_not_zstd=2,
+  info_file_error=3,
+  info_truncated_input=4,
+} InfoError;
 
 #define ERROR_IF(c,n,...) {             \
     if (c) {                           \
@@ -2164,6 +2174,12 @@ FIO_analyzeFrames(fileInfo_t* info, FILE* const srcFile)
               && (numBytesRead == 0)
               && (info->compressedSize > 0)
               && (info->compressedSize != UTIL_FILESIZE_UNKNOWN) ) {
+                unsigned long long file_position = (unsigned long long) LONG_TELL(srcFile);
+                unsigned long long file_size = (unsigned long long) info->compressedSize;
+                ERROR_IF(file_position != file_size, info_truncated_input,
+                  "Error: seeked to position %llu, which is beyond file size of %llu\n",
+                  file_position,
+                  file_size);
                 break;  /* correct end of file => success */
             }
             ERROR_IF(feof(srcFile), info_not_zstd, "Error: reached end of file with incomplete frame");
@@ -2332,20 +2348,28 @@ FIO_listFile(fileInfo_t* total, const char* inFileName, int displayLevel)
     fileInfo_t info;
     memset(&info, 0, sizeof(info));
     {   InfoError const error = getFileInfo(&info, inFileName);
-        if (error == info_frame_error) {
-            /* display error, but provide output */
-            DISPLAYLEVEL(1, "Error while parsing %s \n", inFileName);
-        }
-        else if (error == info_not_zstd) {
-            DISPLAYOUT("File %s not compressed by zstd \n", inFileName);
-            if (displayLevel > 2) DISPLAYOUT("\n");
-            return 1;
-        }
-        else if (error == info_file_error) {
-            /* error occurred while opening the file */
-            if (displayLevel > 2) DISPLAYOUT("\n");
-            return 1;
+        switch (error) {
+            case info_frame_error:
+                /* display error, but provide output */
+                DISPLAYLEVEL(1, "Error while parsing \"%s\" \n", inFileName);
+                break;
+            case info_not_zstd:
+                DISPLAYOUT("File \"%s\" not compressed by zstd \n", inFileName);
+                if (displayLevel > 2) DISPLAYOUT("\n");
+                return 1;
+            case info_file_error:
+                /* error occurred while opening the file */
+                if (displayLevel > 2) DISPLAYOUT("\n");
+                return 1;
+            case info_truncated_input:
+                DISPLAYOUT("File \"%s\" is truncated \n", inFileName);
+                if (displayLevel > 2) DISPLAYOUT("\n");
+                return 1;
+            case info_success:
+            default:
+                break;
         }
+
         displayInfo(inFileName, &info, displayLevel);
         *total = FIO_addFInfo(*total, info);
         assert(error == info_success || error == info_frame_error);
diff --git a/tests/playTests.sh b/tests/playTests.sh
index ef4861b5..8342455a 100755
--- a/tests/playTests.sh
+++ b/tests/playTests.sh
@@ -809,6 +809,19 @@ $ZSTD --list tmp* && die "-l must fail on non-zstd file"
 $ZSTD -lv tmp1* && die "-l must fail on non-zstd file"
 $ZSTD --list -v tmp2 tmp12.zst && die "-l must fail on non-zstd file"
 
+$ECHO "test : detect truncated compressed file "
+TEST_DATA_FILE=truncatable-input.txt
+FULL_COMPRESSED_FILE=${TEST_DATA_FILE}.zst
+TRUNCATED_COMPRESSED_FILE=truncated-input.txt.zst
+./datagen -g50000 > $TEST_DATA_FILE
+$ZSTD -f $TEST_DATA_FILE -o $FULL_COMPRESSED_FILE
+head -c 100 $FULL_COMPRESSED_FILE > $TRUNCATED_COMPRESSED_FILE
+$ZSTD --list $TRUNCATED_COMPRESSED_FILE && die "-l must fail on truncated file"
+
+rm $TEST_DATA_FILE
+rm $FULL_COMPRESSED_FILE
+rm $TRUNCATED_COMPRESSED_FILE
+
 $ECHO "\n===>  zstd --list/-l errors when presented with stdin / no files"
 $ZSTD -l && die "-l must fail on empty list of files"
 $ZSTD -l - && die "-l does not work on stdin"