File Processing

1
File Processing

• Assembly Language Programming
• University of Akron
• Dr. Tim Margush

2
File Handle

• An integer assigned to a file by DOS
• Used to refer to the file in programs
• DOS provides file handle functions to manipulate
  files
• These are available through the DOS interrupt
  (21h)

3
Predefined File Handles

• DOS manages standard input output devices through
  these handles
• 0 keyboard (stdin)
• 1 screen (stdout)
• 2 error output (stderr)
• 3 auxiliary device
• 4 printer

4
File Errors

• Many file handling routines return an error code
• carry clear indicates no error
• carry set indicates an error
• AX will contain an error number

• 1 invalid function number
• 2 file not found
• 3 path not found
• 4 no free handles
• 5 access denied
• 6 invalid handle
• C invalid access code
• F invalid drive
• 10 can't remove current dir
• 11 not the same device
• 12 no more files to be found

5
Opening a File

• Open new/Rewrite
• AH is 3Ch
• DSDX is the address of filename (ASCIIZ)
• CL is attribute
• Returns file handle in AX or error code 3, 4, or 5

• Open Existing
• AH is 3Dh
• DSDX is the address of filename (ASCIIZ)
• AL is access code
• 0 read
• 1 write
• 2 read/write
• Returns file handle in AX or error code 2,4,5,12

6
Attribute Byte

• Each entry in a disk directory contains an
  attribute byte identifying the entry
• ?? ?? ar sd vl sy hd ro
• archive set until file is backed up
• subdirectory entry is a subdirectory name
• volume label entry is disk volume label
• system a system file
• hidden file does not appear in directory list
• read only cannot change or delete

7
Changing a File's Attributes

• Get/Change File Attribute
• AH is 43h
• AL is change code
• 0 get attribute
• 1 set attribute
• DSDX is the address of filename (ASCIIZ)
• CX is attribute if setting

• Returns file attribute in CX or error 2, 3, or 5
• You cannot change bits 3 or 4 (volume label and
  subdirectory settings)

8
Disk Organization

• Each byte of data stored on a disk has a four
  part address
• Surface - 0 through n (floppies have 0 and 1)
• Track - 0 through m (concentric circles on data
  on a surface)
• Sector - 1 through k (512 byte section of a
  track)
• Offset - 0 through 511 (location within sector)

9
Special Sectors

• Boot record
• A small program used to boot an operating system
• File Allocation Table
• A map showing where files are stored on the disk
• Root Directory File Entries
• sequence of 32 byte records, each representing a
  directory or file name
• Data Area for File Data and Subdirectory
  information

10
File Directory Entry

• Bytes Contents
• 0-7 filename (byte 0 is also a status byte)
• 00 (never used), E5 (deleted), 2E (directory)
• 8-10 extension
• 11 attribute
• 22-25 timestamp
• 26-27 starting cluster number
• 28-31 file size (bytes)

11
FAT

• A map of disk storage
• This is an array based linked list
• FATi tells what cluster (if any) follows
  cluster i (-1 is end of chain)
• Each entry is a fixed number of bits (12, 16, or
  32)

• FAT cell contents
• 000 available
• FF0-FF7 reserved
• FF7 means bad
• FF8-FFF chain end
• 002-FEF next cluster
• The cluster code for cluster n is at FATn

12
Finding A File's Data

• The file directory entry for a file contains the
  starting cluster number, say n
• This cluster contains the first x bytes of the
  file
• FATn will tell the next cluster number

• By following the chain if clusters, all of the
  file's data may be located.

13
Closing a File

• Closing a file frees the file handle and updates
  file info (size, timestamp) in directory
• AH is 3Eh
• BX is the file handle
• Returns error code 6 in AX if unsuccessful

14
Reading a File (or keyboard)

• Read a File
• AH is 3Fh
• BX is file handle
• CX is number of bytes to read
• DSDX is address of buffer
• Returns number of bytes read in AX or error code
  5 or 6

• If a file is open for reading, this function will
  copy a specified number of bytes from the file to
  a memory buffer
• EOF is detected by comparing CX and AX after the
  read

15
Writing to a File (or screen or printer)

• Write File
• AH is 40h
• BX is file handle
• CX is number of bytes to write
• DSDX is address of bytes to be written
• Returns number of bytes written in AX or error
  code 5 or 6

• This function sends one or more bytes to a file
  that is open for output
• A full disk will result in less than the
  specified number of bytes being written

16
File Pointer

• The position in a file where the next I/O
  operation will take place
• A byte offset from the beginning of the file
• File pointers are 32-bit integers

• Move File Pointer
• AH is 42h
• Al is movement code
• 0 relative to start
• 1 relative to current
• 2 relative to eof
• BX is file handle
• CXDX is the number of bytes to move (signed)
• Returns new file pointer in DXAX or error code 1
  or 6