CN1277213C - A flash memory file system management method - Google Patents

Abstract

The invention discloses a flash memory file system management method, which divides the flash memory into a FAT table area and a data storage area, and divides it into a plurality of block areas at the same time; after power-on, a space mapping table is established in the memory, and the data contained in each file is written. The information of the block area number where the block is located; when creating a new file, add a new record in the FAT table, write the new file ID, the number of blocks occupied, and the file index number into the record; then each data block of the file in turn Write free blocks, and write the block number of each data block into the corresponding position in the space mapping table; when updating file data, only operate on the file data storage area; when reading data, determine where the data to be read is After entering a data block, query the block number of the data block according to the file ID, and only search for the data block in the block area. The method of the invention can reduce the update times of the file index area, and at the same time can quickly retrieve and locate.

Description

A kind of flash memory file system management method

technical field

The invention relates to a flash memory (FLASH) file system, in particular to a method for managing a stored file system by using a FAT (file allocation table) table area.

Background technique

FLASH memory is a new type of non-volatile semiconductor memory gradually developed in the late 1980s. It combines the advantages of simple structure and high density of EPROM in the past and the electrical erasability of EEPROM in the system to achieve high density, low cost and high reliability. At present, more and more storage products use FLASH as a storage medium in order to reduce costs. However, FLASH has unique requirements in read and write operations. It can only be written from "1" to "0", and cannot be written in reverse. If you want to rewrite, you must first erase in units of sectors, and the life of the erased sectors is also limited. Restricted, the number of erases generally allowed is about 100,000 times.

In order to ensure the security and consistency of the data on the FLASH, and prolong the service life of the FLASH as much as possible, it is necessary to design a specific file management system for the FLASH to manage the file data on the FLASH chip. At present, many FLASH-based file management systems have been invented in the field of storage of computers and communication equipment.

The technical point of inventing the FLASH file system is that the file data is scattered and stored on the FLASH chip. When updating the file, it is not updated at the original location of the file data, but to find a new place to store the modified data or write log information. , this is done to ensure the security of the data, and the data will not be lost if an accident occurs during the update. In order to prolong the service life of FLASH, each erase sector on the FLASH chip should be erased as balanced as possible.

The application number is 01142240, and the patent application named "FLASH storage file management method" proposes a relatively complete FLASH file system management method. This method divides the FLASH storage space into two areas, the file index area and the file data storage area. The file index area is used to record the allocation information of the file directory and the basic allocation unit "block", and the file data storage area is used to store the specific content of the file. , update the file by rewriting the "block" of the file. The invention of the file system solves the problem of data security and consistency on the FLASH chip to a certain extent.

However, the existing FLASH file system records the information of the file data area (generally the position information of the first block of the file) in the file index area. If the file is updated, it also needs to update the corresponding content of the file index area, which increases the file index area. wear times. Damage to the file data area may only destroy the file stored in the block. If the file index area is damaged, the file system may crash.

In addition, the existing FLASH file system management method is not aimed at the smart card platform. Due to the particularity of the smart card platform and its applications, that is, the slow speed of the CPU, the file system is required to be able to quickly retrieve and locate; in addition, in order to reserve enough RAM space for the upper application, the design of the file system should use as little RAM space as possible, so The smart card file system has higher requirements on the retrieval speed and occupied space than the general file management system, so it needs special consideration.

Contents of the invention

The technical problem to be solved by the present invention is to provide a flash memory file system management method, which can reduce the number of updates of the file index area, and can quickly search and locate at the same time.

In order to solve the above technical problems, the present invention provides a flash file system management method, comprising the following steps:

(a) dividing the flash memory into a file allocation table area and a file data storage area by function, and simultaneously dividing it into a plurality of block areas containing the same number of blocks and numbering them;

(b) after power-on, set up a space mapping table in internal memory, read the valid records in the file allocation table and the data block in the flash memory, write in the space mapping table the location of the block number of the data block that each file contains information;

(c) When creating a new file, calculate how many data blocks need to be occupied according to the size of the new file, and judge whether the quantity of free blocks is enough, if not enough, first perform the erasing operation;

(d) Add a new record in the file allocation table, and write the identifier of the new file, the number of blocks occupied by the file, and the allocated file index number into the new record;

(e) each data block of this new file is written in the idle block successively, and inquire about the corresponding position of each data block place block area number in the space mapping table, the place block area number of each data block is written into this space mapping table, Complete the creation of a new file;

(f) When updating file data, only the file data storage area is operated, and new data is written into a new data block or added in the remaining space of the original data block to complete the update;

(g) When reading data, determine which data blocks in the file the data to be read is in according to the offset and length given by the user, and use the file identifier and concatenated block number to query the block number of these data blocks in the space mapping table location, and get the block number where it is located;

(h) Find the data block of the file in the found block area according to the file identifier and the concatenated block number, read the data into the memory, and hand it over to the user after processing.

Further, the above method can also have the following characteristics: in the step (a), the flash memory is divided into 16 or 256 block areas, and the block area number of each data block in the space mapping table occupies half a byte or a byte.

Further, the above method can also have the following characteristics: in the step (b), the block numbers of the data blocks contained in the files in the space mapping table are arranged continuously according to the concatenation order of the data blocks, and the files corresponding to the block numbers The order of the files matches the order in the file allocation table.

Further, the above method can also have the following characteristics: the step (b) is further divided into the following steps:

(b1) After power-on, read the effective file allocation table records, calculate the total number N of blocks occupied by all valid files, set up a space mapping table for the bytes used by N block numbers in the memory, and set up a space mapping table in the memory at the same time Temporary entries whose number of entries is equal to or greater than the number of valid files;

(b2) Scan the file allocation table one by one, and each time a valid record is retrieved, the sum of the number of data blocks in the valid records before the record is accumulated, and used as the starting address of the file in the space mapping table, and filled into the temporary entry middle;

(b3) scan the data block on the flash memory, read the file index and the concatenated block number of the valid data block;

(b4) find the starting address of the file in the space mapping table in the temporary entry according to the file index, then determine the position of the block area number where the data block is located in the space mapping table and write according to the concatenated block number;

(b5) After the processing of all data blocks in the file data storage area is completed, the space mapping table is established, and the space occupied by the temporary table items is released.

Further, the above method can also have the following characteristics: the allocation of the file index number is realized by the following steps:

(o) After power-on, open up a file index management area in memory, each binary bit of this management area is corresponding to a file index number, represents whether this index number already exists, initial value is 0;

(p) Scanning the file allocation table one by one, every time a valid record is retrieved, the binary position corresponding to the file index number in the record in the file index number management area is "1";

(q) When creating a new file, scan each bit in the file index management area, and assign the index number corresponding to the first "0" bit scanned to the new file, and then set the position to 1;

(r) After deleting a file, set the corresponding position of the file index number in the file index management area to 0.

Further, the above method can also have the following characteristics: in the steps (e) and (g), querying the position of the block number of each data block of the new file in the space mapping table is realized by the following steps:

(u) Retrieve the files one by one from the starting address of the file allocation table area according to the file identification;

(v) After the file to be searched is retrieved, the accumulative value of the block numbers of all valid files before the record is the starting address of the file in the space mapping table;

(w) The corresponding position of the starting address in the space mapping table is the position of the block number of the starting block of the file, and the position of the next block number is the position of the block number of the second block of the file, so as to determine the file The location of the block number where the data block of any concatenated block number is located.

Further, the above method can also have the following characteristics: in the step (f), for a large file whose file length exceeds half the data size, the update operation is performed according to the following steps:

(f1) The file system receives an external instruction to update file data, and reads the offset of the data to be updated in the file;

(f2) Determine the data block and its quantity where the data to be updated is located, if the current space is not enough, first perform the erasing operation;

(f3) find the same number of new blocks sequentially in the free sector, write data;

At the same time, for small files whose file length is less than half the data size of a block, log information is written directly at the end of the valid data of the block where the data is updated. If the log area of the block is full, follow the above steps (f1)~(f3) ) to perform an update operation.

Further, the above method can also have the following characteristics: in the step (h), first judge whether the file is a small file, if so, read the data of the file block into the memory, then scan the log area at the end of the file, and log The latest data in the area covers the memory, and then take out the part of the data that the user wants to read to the user according to the offset and length given by the user; if it is a large file, read the data blocks of the file in each block area, and splice the data to the user later.

As can be seen from the above, the information about the file data in the file FAT table record on the FLASH chip of the present invention only includes the number of blocks occupied by the file, and there is no relevant information of the block address. In this way, for the file system whose length does not change after most files are created, this method greatly reduces the number of erasing times of the file FAT table area and ensures the security of the file system. At the same time, the present invention adopts the method of combining the FAT table of the FLASH file and the RAM space mapping table, which greatly improves the retrieval efficiency of the file.

Furthermore, in order to reserve sufficient RAM space for upper-layer applications, the present invention also proposes a method for establishing an efficient space mapping table, which can occupy as little RAM space as possible. Therefore, the method of the invention is simpler and more feasible on the basis of ensuring the operating efficiency and performance of the file system, and reduces the product cost. It is a good solution for smart card products with less CPU and RAM resources. However, it should be emphasized that the present invention has universality and can also be used for other products.

Description of drawings

FIG. 1 is a flowchart of establishing a RAM space mapping table according to an embodiment of the present invention.

Fig. 2 is a flow chart of allocating file index numbers according to the embodiment of the present invention.

Fig. 3 is an operation flowchart of creating a file in the embodiment of the present invention.

Fig. 4 is a flow chart of reading file data according to an embodiment of the present invention.

Detailed ways

The method of the present invention will be described in detail below with an embodiment. The FLASH file system management method of this embodiment is applied to smart card products. The hardware platform characteristics of smart card products are: the CPU speed is relatively slow, requiring the file system to be able to quickly retrieve and locate; in addition, in order to reserve enough RAM space for upper-layer applications, the design of the file system should use as little RAM space as possible.

File system storage scheme

The FLASH chip is divided into a file FAT table area (namely, a file index area) and a file data storage area. The smallest unit of file storage is "block", and the file data storage area is composed of data "blocks". The block size is determined by the size of the file and the size of the erase sector (the smallest unit for erasing the FLASH). In order to ensure the processing speed of the file, it is best not to exceed 5 blocks for each operation. At the same time, try to ensure that the size of the erased sector is an integer multiple of the "block". The first n bytes of each "block" are used to store the information of the block, including: the current state of the block, the file index of the file to which the block data belongs and the concatenated block number (indicating which block is the first file data of the file) piece). The header information of the first block of the file also includes attribute information that may change during the update process of the file, such as the pointer of the circular file in the SIM card.

The content of the FAT table area of the present embodiment file is as shown in Table 1 below:

Table 1 FAT table status directory index file index file ID The number of blocks occupied (HEX) Other properties of the file FF 00 01 03 File 1 0004 00 01 ... FF 00 01 05 file 2 0009 00 01 ... FF 00 02 02 file 3 0170 00 0F ... ...

It can be seen that the FAT table area is composed of several FAT table records, each file corresponds to a FAT table record, and the record content includes: the state of the FAT table record, directory index, file index, file ID, and the number of "blocks" occupied by the file. Quantity and other basic properties of the file.

Different from the existing method, the FAT table record of the present embodiment does not include the location information of the block, but only records the number of blocks occupied by the file, so that only when the file is created, deleted and the file size is changed, the FAT table is involved. Update, if you just update the file data, you don't need to modify the FAT table record content of the file, so you can greatly reduce the erasing times of the FAT table area.

In the smart card, directory files and ordinary files belong to different file types. If it is used as a general file system, there may be no directory area, and no directory index is set.

Create a spatial map

According to the above storage method, there is no information about the location of the file data block in the FAT table record. To retrieve the data of a file, it is necessary to retrieve each block in the FLASH data storage area, and locate a piece of data through the file index and concatenated block number. The time consumed by the file is unbearable. In order to improve the retrieval speed, it is necessary to add an optimized query mechanism and use the RAM space to establish a FLASH space mapping table.

Firstly, the whole FLASH film is divided into several areas, which are called "block areas", and the number of blocks contained in each block area is the same. The number of blocks divided into the whole slice depends on the RAM space allowed to be consumed by the space mapping table. If the RAM space allowed to be consumed is small, the number of blocks contained in the divided block area may be larger; if the RAM space is large enough, then each The number of blocks contained in a block area is relatively small. In order to facilitate the operation of the space mapping table, the number of blocks can be expressed in nibbles or 1 byte as much as possible, for example, the whole film is divided into 16 blocks or 256 blocks.

Assumption: the hardware platform is 64Ksmart flash, and the erase sector size is 64 bytes. The size of the smallest unit "block" of file storage is also 64 bytes. There are 1024 data blocks on the entire FLASH, which are divided into 16 block areas. Each file is uniquely identified by a one-byte file index, and a maximum of 254 files can be stored on a FLASH chip.

The basic element in the space mapping table is the block area number where each data block is located. The 16 block area numbers can be distinguished by nibbles, and a space mapping table of 512 bytes is required for 1024 data blocks. The block number of each data block included in the file in the space mapping table is arranged continuously according to the concatenation order of the data block, and the order of the files corresponding to the block number is consistent with the order in the file allocation table. As shown in Figure 1, the specific process of establishing the RAM space mapping table in this embodiment is as follows:

Step 110, after power-on, read the effective FAT table records, calculate the total number N of blocks occupied by all valid files, and set up a space mapping table of N/2 bytes in RAM; meanwhile, set up a 512-byte space mapping table in RAM A total of 255 temporary entries (the number of entries can also be determined according to the calculated number of valid files), each two bytes, indicating the offset of the file represented by the entry in the space mapping table (in nibbles) ), the initial value is "FF FF";

Step 120, scan the FAT table of the file one by one, and every time a record in the "valid" state is retrieved, the sum of the number of data blocks in the valid records before the record is accumulated, and used as the nibble start address of the file in the space mapping table , fill in the temporary entry, as shown in Table 2:

Table 2 file index 01 02 03 04 05 06 ... Temporary entry content ... 0002 0001 ... 0000 FFFF ...

Step 130, scan the data block on the FLASH, if the current data block is valid, then read the file index and the concatenated block number;

Step 140, find the starting address of the file in the space mapping table in the temporary entry according to the file index read, then determine the corresponding nibble ( The first block is written in the nibble corresponding to the starting address, the second block is written in the next nibble, and so on), fill in the area number of the data block to which the block belongs;

Step 150, after all the data blocks in the data storage area are processed, the space mapping table is established, and the space occupied by the temporary entries is released. The arrangement of the obtained spatial mapping table is shown in Table 3 below:

table 3 significance Spatial map data (nibbles) The block area where the block of file 1 is located block 1 block number 1 block 2 block number 2 block 3 block number 3 block 4 block number 4 The block area where the block of file 2 is located block 1 block number 1 ... ... Block 9 block number 9 The block area where the block of file 3 is located block 1 block number 1 ... ... Block 368 block number 368 ... ...

According to the above method, an efficient space mapping table can be established, occupying little RAM space. However, an existing method for establishing a spatial mapping table may also be used.

Query the position of the block number of the file data block in the space mapping table

After the space mapping table is established, the location of the block number of the data block contained in the file can be queried through the following steps:

Step 1, according to the file mark, from the starting address of the FAT district, retrieve the file one by one to find, if not, then add up the file block quantity in the FAT record, and continue to search; if the file mark to be searched is retrieved, Execute the next step;

Step 2, the accumulative value of the block quantity of all files before this FAT record is the starting address of the file in the space mapping table, and the nibble corresponding to the starting address is the position of the block number where the starting block of the file is located, The next half byte is the position of the block number of the second block of the file, so that the position of the block number of any concatenated block number of the file can be determined, and then the corresponding block number can be read or written.

In this way, in this embodiment, when retrieving file data, only the block area where the file data block is located can be queried according to the file identification, and only the corresponding block area is searched, thus greatly speeding up the retrieval speed.

Assignment of file index numbers

For a newly created file, a file index number needs to be allocated, and the file index number is one byte. As shown in Figure 2, this embodiment is realized through the following steps:

Step 210, after power-on, open up a 32-byte file index management area in RAM, these bytes are regarded as a string of binary bits, the nth bit is used to indicate whether the index n has been used, and the initial value is 0;

Step 220, scanning the file FAT table one by one, each time a valid record is retrieved, the position corresponding to the file index number in the record in the file index number management area is "1", if the bit is 0, it means that the index number is free;

Step 230, when creating a new file, scan each bit of the 32 bytes in the file index management area, assign the index number corresponding to the first scanned bit that is "0" to the new file, and assign the BIT position 1;

Step 240, after deleting a file, set the BIT position corresponding to the index number of the file to 0.

The above allocation method can effectively manage file index numbers, and ensure that the index numbers are used sequentially without redundancy. But the present invention can also adopt other index number allocation methods.

Assuming that during the initialization process, the space mapping table and the file index number management area have been filled in according to the above method, based on the above schemes of file storage, index number allocation and block number location query, the operation process of the file system in this embodiment to create a file is as follows: As shown in Figure 3, it includes the following steps:

Step 310, the file system receives an external command to create a new file, and obtains the identification and file size information of the file to be read;

Step 320, calculating how many data blocks need to be occupied according to the size of the new file, and judging whether the number of free blocks is enough, if enough, directly execute the next step, if not enough, first perform the erasing operation, and then execute the next step;

Step 330, add a new record in the FAT table, the record status is set to "invalid", and the identification (ID) of the file, the quantity and other attributes of the blocks are written into the record;

Step 340, retrieve the file index number management area, assign the first free index number scanned to the new file, and write the FAT table record of the file;

Step 350, write the data blocks of the file in the free block in turn, and query the corresponding nibbles of each data block in the space mapping table, and write the block number of each data block into the corresponding nibbles;

Step 360, after all the blocks are written, change the state of the new record in the FAT table from "invalid" to "valid", and complete the creation of a new file.

If the power is lost when creating the file, the file status in the above FAT table area can be used to know, and power-off processing is performed.

Update of file data

Because updating file data does not change the size of the file, there is no need to modify the FAT table record content of the file.

Below, the file whose length is less than half the data size of the block (the space occupied by the header information of the block is removed) is called a small file, otherwise it is called a large file. For large files, follow the steps below to complete the update of file data:

Step 1, the file system receives an external instruction to update file data, and reads the offset of the data to be updated in the file;

Step II, determine the data block to which the data to be updated belongs, assuming that n new blocks need to be found, if the current space is not enough, first perform the erasing operation;

Step III, find n new blocks sequentially in the free sector, write data, and the data update process is completed.

When updating the data of a small file, you only need to write log information at the end of the valid data of the block where it is located, including the version number and the updated file content. You can write the updated file content first, and then write the version number (not "FF"). Once the log area of the block is full, another blank block is found to store the file, and the process is the same as that of updating a large file.

According to the file characteristics of smart card products, the above method adopts the combination of rewriting the file data block and writing the update log in the block, and reduces the number of erasing as much as possible on the basis of ensuring the erasure balance.

Reading of file data

After the file is created, the process of reading file data is shown in Figure 4, including the following steps:

Step 410, the file system receives an external instruction to read file data, and obtains the identifier of the file to be read and the offset and length of the data to be read within the file;

Step 420, determine which data blocks in the file the data to be read is in according to the given offset and length, query the position of the block number of these data blocks in the space mapping table according to the file identifier and the concatenated block number, and Get the block number where it is located;

Step 430, search the data block of the file in the block area found according to the file identification and the concatenated block number, and judge whether it is a small file stored in a block, if yes, execute the next step, otherwise execute step 450;

Step 440, read the data of the file block into the memory, then scan the log area at the end of the file, overwrite the memory with the latest data in the log area, and then take out the part of the data that the user wants to read according to the offset and length given by the user. user, end;

Step 450, read out the data blocks of the file in each block area, splice the data and send it to the user, and end.

Performance analysis of the file system in this embodiment

As mentioned above, the hardware platform of this embodiment is 64Ksmart flash, and the erase sector size is 64 bytes. The size of the smallest unit "block" of file storage is also 64 bytes. There are 1024 data blocks in the whole FLASH, which are divided into 16 block areas, and each block area has 64 data blocks. The performance analysis is as follows:

A. Utilization of file storage methods

In this embodiment, one block can store 4 FAT table records, the FAT table area of 255 files occupies 64 blocks, and the FAT table area only occupies 1 block area.

Each file is uniquely identified by a one-byte file index, and a maximum of 254 files can be stored on a FLASH chip. Assuming the most extreme case, one file wastes one block, and the storage utilization ratio of FLASH is (1024-254)/1024≈75 %. B. The size of RAM space occupied by file system management:

Space mapping table (the most cases): 1024/2=512 bytes

File index number management area: 32 (file index number)+1 (directory index number)=33 bytes

A total of 545 bytes of RAM space is occupied

C, system running speed

The file system consumes the most time during power-on system initialization and power-off data recovery. The initialization process needs to scan 1024 data blocks in the flash medium and fill the file block information of the space mapping table, and the required time is about 2ms.

As can be seen from the above, the information about the file data in the file FAT table record on the FLASH chip of the present invention only includes the number of blocks occupied by the file, and there is no relevant information of the block address. In this way, for the file system whose length does not change after most files are created, this method greatly reduces the number of erasing times of the file FAT table area and ensures the security of the file system. At the same time, the present invention adopts the method of combining the FAT table of the FLASH file and the mapping table of the RAM space to improve the retrieval efficiency of the file.

The invention also proposes a method for establishing an efficient space mapping table, which can occupy as little RAM space as possible. Therefore, the method of the invention is simpler and more feasible on the basis of ensuring the operating efficiency and performance of the file system, and reduces the product cost. It is a good solution for smart card products with less CPU and RAM resources. However, it should be emphasized that the present invention has universality and can also be used for other products.

Claims (9)

1, a kind of flash memory file system management method may further comprise the steps:

(a) flash memory is divided into file allocation list area and file data storage region by function, simultaneously it is divided into a plurality of piece district and numberings that comprise the piece of equal number;

(b) power on after, in internal memory, set up the spatial mappings table, read effective record in the described file allocation table and the data block in the flash memory, in the spatial mappings table, write the information of the data block place piece area code that each file comprises;

When (c) creating new file, calculate needs according to the size of new file and take how many data blocks, and judge whether the quantity of free block is enough, if inadequately, wipe housekeeping operation earlier;

(d) in file allocation table, append a new record, the file index of the quantity of shared of sign, the file of this new file, distribution number is written in this new record;

(e) each data block that successively will this new file writes in the free block, and inquires about this each data block place piece area code corresponding position in the spatial mappings table, and each data block place piece area code is write this spatial mappings table, finishes the establishment of a new file;

(f) during the updating file data, only file data storage region is operated, be written to new data in the new data block or be added in the remaining space of former data block, finish renewal;

(g) during reading of data, the data that skew that provides according to the user and length are determined to read number are inquired about these position of piece area code in the spatial mappings table, data block place by file identification and cascade piece, and are obtained its place piece area code in which data block of file;

(h) data block of number searching this document in the piece district that finds according to file identification and cascade piece is read in internal memory with data, gives the user after the processing.

2, the method for claim 1 is characterized in that, this method is to be used for smart card document system is managed.

3, method as claimed in claim 1 or 2 is characterized in that, in the described step (a), is that flash memory is divided into 16 or 256 piece districts, and each data block place piece area code takies nybble or a byte in the described spatial mappings table.

4, method as claimed in claim 1 or 2, it is characterized in that, in the described step (b), the piece area code of each data block that file comprised is arranged continuously by the waterfall sequence of data block in the described spatial mappings table, the sequence consensus in the order of piece area code corresponding file and the file allocation table.

5, method as claimed in claim 1 or 2 is characterized in that, described step (b) is further divided into following steps:

(b1) power on after, read effective file allocation table record, calculate the total quantity N of the piece that all effective documents take, in internal memory, set up the spatial mappings table of N the used byte of piece area code, in internal memory, set up the interim list item that an item number is equal to or greater than the effective document number simultaneously;

(b2) scanning document allocation table one by one whenever retrieves an effective record, and data block quantity sum in the effective record before this record that adds up out plays the location as this document in the spatial mappings table, be filled in the interim list item;

(b3) data block of scanning on the flash memory reads the file index of valid data piece and cascade piece number;

(b4) in interim list item, finding this document in the spatial mappings table, to play the location according to file index, again according to the cascade piece number determine this data block place piece area code in the spatial mappings table the position and write;

(b5) all data blocks of file data storage region are finished dealing with after, the spatial mappings table is set up and to be finished, and discharges the shared space of interim list item.

6, method as claimed in claim 1 or 2 is characterized in that, the distribution of described file index number is realized by following steps:

(o) power on after, in internal memory, open up a file index directorial area, each binary digit of this directorial area represents corresponding to a file index number whether this call number existing, and initial value is 0;

(p) scanning document allocation table one by one whenever retrieves an effective record, will be in the file index directorial area the binary digit set of file index correspondence in should record;

When (q) creating a new file, each in the scanning document index management district is distributed to new file with first that scans for the corresponding call number in the position of " 0 ", then with this position 1;

(r) behind file of deletion, be this document call number corresponding position in the file index directorial area 0.

7, method as claimed in claim 4 is characterized in that, described step (e) and (g) in, inquire about each piece area code position in the spatial mappings table, data block place of described file and realize by following steps:

(u) begin to retrieve one by one the file that will look for according to file identification from the start address of file allocation list area;

(after v) retrieving the file that will search, the accumulated value of the number of blocks of all effective documents is the location of this document in the spatial mappings table before this record;

(w) this plays the position that corresponding position in the spatial mappings table, location is this document starting block place piece area code, the position of next piece area code is the position of second place of this document piece area code, determines the position of the data block place piece area code of the arbitrary cascade piece of this document number according to this.

8, method as claimed in claim 1 or 2 is characterized in that, in the described step (f), for the big file of file size above the half block size of data, upgrades operation according to the following steps:

(f1) file system is received the instruction of outside updating file data, reads more new data skew hereof;

(f2) determine to want the data block and the quantity thereof at data updated place, front space is not enough as if working as, and wipes housekeeping operation earlier;

(f3) order finds the new piece of equal number in free sector, writes data;

Simultaneously, for the small documents of file size less than the half block size of data, more during new data directly the end at the valid data of its place piece write log information, full if this piece log area is write, then upgrade operation by above-mentioned steps (f1)～(f3).

9, method as claimed in claim 8, it is characterized in that, in the described step (h), judge earlier whether small documents of this document, if the data of this document piece are read in internal memory, again the log area of scanning document afterbody, with data covering memory up-to-date in the log area, the skew and the length of giving according to the user is taken out those a part of data that the user will read to the user then; If big file, the data block of reading this document in each piece district is given the user after having spliced data.

Images