JP2005115696A - File access method and information processing apparatus - Google Patents

Abstract

[Objective] To provide a file access method capable of efficiently accessing a data file that is divided and recorded.
[Configuration] In a FAT file system that manages access to the hard disk 10, when recording the song 1 data file 11A, the degree of fragmentation is measured by measuring how far the recording positions of the cluster 13 group constituting the song 1 data file 11A are distributed. If the file exceeds the predetermined threshold, a cluster link file 60 that displays a list of identification numbers of the blocks constituting the song 1 data file 11A is created, and if the cluster link file 60 exists when the song 1 data file 11A is read, To identify a block group constituting the music 1 data file 11A.
[Selection] Figure 1

Description

  The present invention relates to a file access method and an information processing apparatus, and more particularly to a file access method in a device that reproduces digital data.

  A disk-type recording medium such as a hard disk (HDD) is used as a recording medium in various information processing apparatuses including computers due to its large capacity and high-speed accessibility. In recent years, it has begun to spread as a medium for recording real-time streams of moving images and audio.

  A disk-type recording medium includes a disk that can record information as an optical signal or a magnetic signal, and a recording head that scans the disk to write and read information. Therefore, in order to access desired information, the recording head needs to move on the disk to a position where this information is recorded. The time required for the recording head to move (hereinafter referred to as seek time) varies depending on the arrangement of information recorded on the disk. Generally, the seek time is orders of magnitude longer than the time required to actually transfer information (hereinafter referred to as transfer time). This is because the transfer time depends on the rotational speed of the disk, but the seek time depends on the speed at which the head moves in the radial direction of the disk. This is because it is difficult to increase the speed of movement in the direction.

  When a series of data is distributed and recorded on the disc, the seek time becomes long, so that the data access performance is extremely lowered. This degradation in access performance may cause an underflow of content reading during playback of a real-time stream, and may cause a fatal quality defect such as jumping sound. In order to prevent content underflow, the buffer for temporarily storing the content in the playback device needs to have an excessive capacity, which increases the cost of the playback device. Therefore, a method for preventing the data from being divided on the recording medium and a method for improving the access performance to the divided and recorded data have been devised. The details of the prior art will be described below with reference to the drawings.

  First, a method for recording data on a hard disk as a recording medium will be described with reference to FIG. FIG. 13 is a diagram showing a method of recording data on the hard disk HDD # 3 (10). The data recording method is controlled by the file system, but there are various data recording methods depending on the type of the file system. FIG. 13 shows a data recording method using a FAT (File Allocation Table) file system widely used in personal computers.

  Data is stored as a file 11, but the data itself is recorded in the data area 12 in units of fixed-length data blocks called clusters 13. A cluster constituting one file may be a continuous cluster group such as a file A (11A) or a discontinuous cluster group such as a file B (11B).

  A directory area 14 and a FAT area 15 exist as areas including file management information for associating a file with a cluster group constituting the file. In the directory area 14, there is a table for each file whose entry is information including its name, its attribute, start cluster number, and file size. An entry in this table is called a directory entry 16 and exists for each recorded file. For example, the file name corresponding to the directory entry DENT # 1 shown in FIG. 13 is “file A”, the corresponding data is recorded from the cluster # 6, and the data size is 45 Kbytes. ing.

  The FAT area 15 has information for listing a cluster group constituting the file. The FAT entry 17 determines the number of the corresponding cluster depending on the number of the entry, but the value of the FAT entry itself indicates the number of the cluster following the corresponding cluster. For example, the sixth FAT entry (FAT # 6) has a value of 7, indicating that cluster # 6 is followed by cluster # 7.

  As described above, the cluster group constituting the file can be identified from the directory entry 16 and the FAT entry 17. For example, as shown in FIG. 13, it can be seen from directory entry DENT # 2 that file B starts from cluster # 13, followed by FAT # 13 followed by cluster # 21, followed by FAT # 21 followed by cluster # 31. Since the end mark EOF is recorded in FAT # 31, it can be seen that cluster # 31 is the end of the file.

  FIG. 14 is a diagram illustrating the access timing to the hard disk 10 (HDD # 3) at the time of file reading. Hereinafter, the influence of the division of the recording data on the access performance will be described with reference to FIG. Arrows and squares in the figure respectively represent a time segment during which the recording head is moved and a time segment during which data is actually transferred. In FIG. 14, the length of the arrow indicating the seek time 20 is equal to the length of the square indicating the transfer time 21, but in actuality, the arrow is 10 times longer.

  FIG. 14A is a diagram showing access timing to the hard disk 10 (HDD # 3) when reading a file (file A in FIG. 13) recorded in a continuous cluster. In file access, first, the directory entry DENT # 1 of the file A is read, but the directory entry 16 is read from the hard disk HDD # 3 by collecting several entries. As shown in FIG. 13, the access unit to the directory area 14 is called a directory access unit 22. That is, first, the directory access unit DENT_ACCESS # 1 is read.

  Next, the FAT entry (FAT # 21) corresponding to the starting cluster number of the read directory entry (DENT # 1) is read. The FAT entry 17 is also read out from the hard disk 10 (HDD # 3) as several entries together. As shown in FIG. 13, the unit of access to the FAT area 15 is referred to as a FAT access unit 23. That is, the FAT access unit FAT_ACCESS # 1 is read.

  Since the read FAT entry 17 indicates that the file A is composed of continuous clusters from cluster # 6 to cluster # 8, these three clusters are read at a time. Assuming that the hard disk specifications have an average seek time of 20 milliseconds, a transfer rate of 10 megabytes / second, and a cluster size of 16 kilobytes, the total access time is 48 milliseconds.

  FIG. 14B is a diagram showing the access timing to the hard disk HDD # 3 when reading a file in which the cluster is divided (file B in FIG. 13). In reading the directory entry 16 and the FAT entry 17, DENT_ACCESS # 1 and FAT_ACCESS # 1 are read as in the case of FIG. However, since the data area is discontinuous, a seek occurs before each cluster is accessed. For this reason, compared with FIG. 14A, for the same hard disk specification, the total access time is increased by 40 milliseconds to 88 milliseconds.

  Further, as data segmentation proceeds, only one FAT entry 17 for the corresponding file exists among the entries that can be read in one FAT access unit 23. For example, in the music A.aud file shown in FIG. 16 to be described later, the only FAT entry valid in the FAT access unit (FAT_ACCESS # 2) is FAT # 1160, and the valid FAT entry in FAT_ACCESS # 3 is FAT # 2170. It has become only. FIG. 14C is a diagram showing the access timing to the hard disk HDD # 3 when reading a file in which the cluster is extremely divided as described above. In this case, every time data for one cluster is read, the head moves back and forth between the FAT area and the data area and seeks twice, so the time required to read the data for three clusters is as shown in FIG. ) Is increased by 40 milliseconds to 128 milliseconds. This value corresponds to about three times that in the case of FIG. 14A in which the file is composed of continuous clusters.

  As described above, when data is divided and recorded on a disk-type recording medium, the time required for the access greatly increases. Hereinafter, two conventional techniques devised to solve this problem will be described.

(Prior art 1)
First, a recording management method (see, for example, Patent Document 1) that prevents data from being divided and recorded and guarantees constant data access performance will be described with reference to FIGS. To do.

  FIGS. 15A and 15B are diagrams illustrating a data fragmentation prevention procedure of the recording management method described in Patent Document 1. FIG. FIG. 15A shows a situation where data for one cluster is newly added to the file C composed of data of one cluster length (cluster # 64). At this time, the cluster that is continuous with the cluster # 64, the cluster # 65 and the cluster # 66 are used as the file D, and the subsequent cluster # 67 is also used as the file E. Therefore, when data is added by a normal method, the file C is divided and recorded in discontinuous clusters such that cluster # 68 follows cluster # 64.

  For this reason, in the recording management method disclosed in Patent Document 1, prior to the addition of data to a file, the clusters in use are rearranged to create a free cluster sequence that can continuously store the file. Then, the recording position of the data divided and recorded on the recording medium is moved, and the process of continuously re-recording on the recording medium is performed. This process is called a defragmentation process 24, and the record management method of Patent Document 1 is characterized in that the defragmentation process 24 is executed every time a cluster is assigned as a file is written.

  In FIG. 15A, the cluster # 65 and the cluster # 66 are used by the file D, but they are exchanged with the unused cluster # 68 and cluster # 69. As a result, empty clusters capable of continuously storing the cluster # 64, the cluster # 65, and the file C have been secured. Thereafter, data recording for these clusters is executed. As a result, as shown in FIG. 15B, the file C is recorded without being divided.

(Prior art 2)
Prior to reading a file, a file system that speeds up access to actual file data by reading a necessary FAT entry group and holding it in a playback device (see, for example, Patent Document 2). Has also been devised.

  FIG. 16 is a diagram illustrating an audio playback device to which the file system of Patent Document 2 is applied. In FIG. 16, the audio playback device 30 accepts a song that the user desires to play as a key input 31, reads the corresponding song data from the hard disk HDD # 3 (10), and outputs it as an analog audio signal 32.

  The audio playback device 30 includes a processor C (33) that controls the operation of the entire device, a non-volatile memory ROM # 3 (34) in which an instruction code executed by the processor C (33) is recorded, and a processor C (33 ) Main memory C (35) which is a working memory, hard disk HDD # 3 (10) in which digital music data is recorded as a file, and an analog audio signal 32 corresponding to the encoded digital audio stream as an input. The audio controller 37 outputs a keyboard interface 38 that receives a user key input 31, and a system bus 39 that couples these components.

  FIG. 16 shows a case where one file “Song A.aud” is recorded on the hard disk HDD # 3 (10). As described with reference to FIG. 14C, this file corresponds to a case where the file is extremely divided and recorded. That is, since the clusters constituting the file are largely separated from the cluster # 1160 to the cluster # 2170, the FAT entry 17 (FAT # 1160, FAT # 2170, etc.) corresponding to each cluster has different FAT access units 23 ( FAT_ACCESS # 2, FAT_ACCESS # 3, etc.).

  The processor C (33) executes the file system 44, which is a software module, and controls access to data on the hard disk HDD # 3 (10). On the main memory C (35), management information used by the file system 44 exists as file system management information C (45). In the file system management information C (45), a FAT buffer 46 for temporarily recording data in the FAT area 15 of the hard disk, a directory buffer 47 for temporarily recording data in the directory area 14, and data in the data area 12 are temporarily recorded. A data buffer 48 exists.

  The file system management information C (45) includes file information 50 that holds information about individual files. The file information 50 exists one by one for the currently accessed file, that is, the currently opened file. The file information 50 includes a file name 51, an attribute 52, a file size 53, a buffer address 54 indicating where the target file is held on the data buffer 48, and which data byte of the target file is currently being accessed. Has a file pointer 55.

  The file system described in Patent Document 2 is further characterized by having a cluster link table 56 in the file information 50. In the cluster link table 56, the numbers of clusters constituting the target file are recorded in accordance with the configuration order. In FIG. 16, it can be seen from the cluster link table 56 that the first cluster of the file “Song A.aud” is cluster # 1160, followed by cluster # 2170 and cluster # 3290.

  The cluster link table 56 is generated when the file is opened prior to accessing the target file. The file system 44 searches the FAT area 15 for a list of all cluster numbers constituting the file when the file is opened, and holds it as a cluster link table 56. This eliminates the need to read the FAT entry 17 when accessing the target file.

When the extremely fragmented music A.aud file shown in FIG. 16 is accessed by a normal file system, it is necessary to access the FAT entry 17 for each cluster as shown in FIG. On the other hand, when accessing using the cluster link table 56, it is not necessary to read the FAT entry at the time of data access. As in FIG. 14B, the transfer time of the cluster and the seek when moving to the next cluster are obtained. Accessible in time. That is, according to the file system of Patent Document 2, the same data access performance as that obtained when the recording data is divided to a medium level can be realized even for the recording data that is divided to a high level.
Japanese Patent Laid-Open No. 2001-202274 (page 7, FIG. 1) JP 2002-163136 A (page 10, FIG. 1) JP-A-8-241230 (page 8, FIG. 1) Japanese Patent Laid-Open No. 2001-249669 (page 55, FIG. 20)

Hereinafter, the above-described problems of the conventional techniques 1 and 2 will be described.
First, in Prior Art 1, the defragmentation process 24 is performed when data is added to a file, and a continuous cluster that can store the entire file is generated. However, this defragmentation process 24 may take a long time. As described with reference to FIGS. 15A and 15B, the defragmentation process 24 requires data exchange between the clusters 13 in the data area 12. The cluster that moves in the defragmentation process 24 is not only the cluster that moves to generate a new continuous cluster. When the continuity of an existing file is lost along with the movement, the cluster needs to be moved to restore the continuity.

  15 (a) and 15 (b), if one new free cluster is generated after cluster # 64, new writing of file C can be handled. However, in order to maintain the continuity of file D, cluster # 65 And cluster # 66 need to be moved together. As described above, depending on the recording state of the file, the number of clusters to be moved increases prior to the writing of the file, and the required time may become extremely long. For example, it may take several tens of minutes to continuously record several gigabytes of moving image data on the hard disk HDD # 3 (10) in which unused clusters are extremely divided. That is, the prior art 1 has a problem that it may take a long time to add new file data depending on the recording state of an existing file.

  In response to this problem, a method has been devised in which the defragmentation process 24 is efficiently executed by making the cluster 13 as a file recording unit a variable length instead of a fixed length. (For example, refer to Patent Document 3). However, in this method, the file format is incompatible with the FAT file system, so that even if the hard disk HDD # 3 (10) is a removable medium that can exchange the recording medium, it is widely used information. There existed a subject that it could not be used with a processing terminal (for example, personal computer).

(Prior art 2)
Further, as shown in FIG. 17, in the file system 44 of the prior art 2, it takes time to create the cluster link table 56 (“cluster link creation period” in FIG. 17) before accessing the data of the target file. It becomes. That is, when the cluster link table 56 is created, there is no need to access the FAT area 15, but considering the time for creating the cluster link, the total time from the start of opening the target file to the completion of reading all the data. The required time is the same value as in the case of the “high-level sector division degree” described in FIG. Therefore, the access performance is improved only when the file data is randomly accessed. That is, the related art 2 has a problem that the access performance is not improved when the file data is accessed only once in order from the front by opening the file once as in the case of reproducing a moving image file. .

  The present invention solves the above-mentioned problems, and can perform sequential access to a divided file at high speed without requiring a long time for file creation, and is compatible with a conventional FAT file system. An object is to provide a file access method and an information processing apparatus.

  In order to solve the above problems, a file access method according to claim 1 of the present invention is the file access method for recording and reading the data file to and from the recording medium for recording the data file in units of blocks. When recording a file, measure the fragmentation rate indicating how much the recording position of each block constituting the data file is distributed on the recording medium, if the fragmentation rate exceeds a predetermined threshold, Create an identification file that lists the identification numbers of the blocks that make up the data file, record it on the recording medium, and check whether the identification file exists when reading the data file. Reading the identification file and identifying the block group constituting the data file And butterflies.

  Also, the file access method according to claim 2 of the present invention has a one-to-one correspondence with the stream data file, the stream data file, and the corresponding stream for the recording medium for recording the data file in units of blocks. In a file access method for recording a stream management file holding information about a data file and reproducing the stream data file recorded on the recording medium, the stream data file is configured when the stream data file is recorded. Information that lists the identification number of each block to be written to the stream management file, and when the stream data file is read, the information held in the stream management file is read and the stream data file is Characterized by identifying the group of blocks to be formed.

  According to a third aspect of the present invention, there is provided an information processing apparatus for recording the data file in an information processing apparatus for recording and reading the data file with respect to a recording medium for recording the data file in units of blocks. , Measuring a fragmentation rate indicating how much the recording positions of each block constituting the data file are distributed on the recording medium, and if the fragmentation rate exceeds a predetermined threshold, the data file A file recording means for creating an identification file that displays a list of identification numbers of blocks constituting the data file, and when reading the data file, confirm whether the identification file exists, and if present, read the identification file, File reading means for identifying a block group constituting the data file. To.

  The information processing apparatus according to claim 4 of the present invention has a one-to-one correspondence with a stream data file, the stream data file, and a corresponding stream for a recording medium that records the data file in block units. In the information processing apparatus for recording a stream management file that holds information about a data file and reproducing the stream data file recorded on the recording medium, the stream data file is configured when the stream data file is recorded. A file recording means for writing information listing the identification number of each block to be written to the stream management file, and when the stream data file is read, the information held in the stream management file is read and the stream data file A file reading means for identifying the blocks constituting, characterized by having a.

  According to the file access method and the information processing apparatus of the present invention, the creation (recording) of the file does not require a long time, and the sequential access to the divided file can be performed at high speed. File access compatible with the FAT file system can be realized.

  That is, according to the file access method according to claim 1 of the present invention, the data file is recorded in the file access method for recording and reading the data file on the recording medium for recording the data file in units of blocks. When measuring, a division rate indicating how much the recording positions of the blocks constituting the data file are distributed on the recording medium is measured, and if the division rate exceeds a predetermined threshold, the data file An identification file that displays a list of identification numbers of the blocks that constitute the file is created and recorded on the recording medium, and when the data file is read out, it is confirmed whether or not the identification file exists. The block group constituting the data file is identified by reading. Because, without requiring a long time to create the file (recording), and an effect capable of performing a sequential access at a high speed for cutting files.

  Further, the recording medium on which the identification file is recorded can be accessed by a conventional procedure, and there is an effect that file access compatible with a conventional file system can be realized.

  According to the file access method of claim 2 of the present invention, there is a one-to-one correspondence between the stream data file and the stream data file for the recording medium on which the data file is recorded in block units. In the file access method for recording the stream data file recorded on the recording medium and recording the stream data file, the stream data file is recorded when the stream data file is recorded. The information that lists the identification numbers of the blocks that constitute the block is written to the stream management file, and when the stream data file is read, the information held in the stream management file is read and the stream data file is read out. Since the block group constituting the file is identified, sequential access to the stream data file can be executed at high speed, and the sequential access to the divided stream data file can also be executed at high speed. There is an effect that can.

  According to the information processing apparatus of claim 3 of the present invention, the data file is recorded in the information processing apparatus that records and reads the data file with respect to a recording medium that records the data file in units of blocks. When measuring a fragmentation rate indicating how much the recording position of each block constituting the data file is distributed on the recording medium, and if the fragmentation rate exceeds a predetermined threshold, A file recording means for creating an identification file for displaying a list of identification numbers of blocks constituting the data file, and when the data file is read, the existence of the identification file is checked, and if it exists, the identification file is read. And a file reading means for identifying a block group constituting the data file. Because characterized, without requiring a long time to create the file (recording), and an effect capable of performing a sequential access at a high speed for cutting files.

  Further, the recording medium on which the identification file is recorded can be accessed by a conventional procedure, and there is an effect that file access compatible with a conventional file system can be realized.

  Also, according to the information processing apparatus of the fourth aspect of the present invention, there is a one-to-one correspondence between the stream data file and the stream data file on the recording medium on which the data file is recorded in units of blocks. When the stream data file is recorded in an information processing apparatus that records the stream management file that holds information about the stream data file to be recorded and reproduces the stream data file recorded on the recording medium, the stream data file The file recording means for writing information listing the identification number of each block constituting the stream management file to the stream management file, and reading the information held in the stream management file when reading the stream data file, Data File reading means for identifying a block group constituting a file, so that sequential access to the stream data file can be performed at high speed, and sequential access to the divided stream data file is possible. There is an effect that the access can be executed at high speed.

(Embodiment 1)
Hereinafter, the file access method and the information processing apparatus according to the first embodiment will be described with reference to the drawings. FIG. 1 is a diagram showing file management information used in the file access method in the information processing apparatus according to the first embodiment.

  In FIG. 1, the song data file “Song 1.audi” 11A is divided and recorded in cluster # 1160, cluster # 2170, etc., but the cluster group constituting these files is a list of FAT entries 17, That is, it is recorded as a unidirectional list of FAT # 1160, FAT # 2170, FAT # 3290,. In addition, cluster numbers are listed as file contents in the corresponding cluster link file “Song 1.fat” 60.

A file access method using the above file management information will be described.
First, how the file management information is constructed when a music data file is recorded will be described with reference to FIG.

  First, when the song data file “Song 1.audi” 11A is recorded, a fragmentation rate indicating how dispersed the recording positions of the cluster groups constituting the song data file are measured. If the fragmentation rate exceeds a predetermined threshold, a cluster link file “Song 1.fat” 60 is generated, and the cluster numbers constituting “Song 1.aud” 11A are listed as the file contents. . Thereby, file management information composed of the cluster link file is constructed. On the other hand, if the fragmentation rate does not exceed the threshold, the cluster link file 60 is not created.

  Next, a read operation of recorded data will be described. When reading the song data file “Song 1.aud” 11A, first, it is checked whether or not the corresponding cluster link file “Song 1. fat” 60 exists. If it exists, the cluster link file is read to identify the cluster group constituting “Song 1.aud” 11A, and then the identified cluster group (cluster # 1160, cluster # 2170, cluster # 3290,...) Is read. put out. If the cluster link file “Song 1.Fat” 60 does not exist, the list of FAT entries 17 is traced to identify the cluster group constituting “Song 1.aud” 11A.

  As described above, in the conventional file access method, when the data file is divided, the FAT entries 17 corresponding to the data file are distributed over a wide area of the FAT area 15. In order to specify the cluster group constituting the file from the list, the FAT area 15 must be accessed many times. In the first embodiment, the cluster group constituting the data file is divided into a wide range. In this case, the cluster link file corresponding to the data file is created, the cluster link file is read at the time of reading, and the cluster group constituting the data file is identified, thereby reading the divided data file. Reducing access to the FAT area 15 at the time It can be.

Next, a specific example of the information processing apparatus to which the file access method according to the first embodiment is applied will be described with reference to FIGS.
FIG. 2 is a diagram showing a configuration of an audio playback system to which the file access method according to the first embodiment is applied.

  The audio reproduction system shown in the figure is connected to a personal computer 90 as a sound source having an original song file 11B via a USB line 91 to create a song data file 11A that is a copy of the original song file 11B. A portable audio playback device 30A. Note that the portable audio player 30A operates as a portable audio player dedicated to playing the music data file 11A held therein when the USB connection with the personal computer 90 is released.

  The portable audio player 30A includes a system bus # 1 (39A) that connects main components, a peripheral circuit bus 92 that connects peripheral circuits, and a DMAC 93 that controls data transfer on these buses. .

  In the system bus # 1 (39A), when the data file 11A is recorded from the personal computer 90, how much the recording positions of the clusters 13 constituting the data file are dispersed on the HDD # 1 (10A). A file recording means (not shown) for creating a cluster link file 60 for displaying a list of identification numbers of the clusters 13 if the fragmentation rate indicated is measured and if the fragmentation rate exceeds a predetermined threshold; When reading the file 11A, it is confirmed whether the cluster link file 60 exists. If it exists, the cluster link file 60 is read to identify a cluster group constituting the data file 11A (not shown). Processor # 1 (33A) for controlling the operation of the entire device, and a processor Non-volatile memory ROM # 1 (34A) in which an instruction code executed by the server # 1 (33A) is recorded, a main memory # 1 (35A) which is a working memory of the processor # 1 (33A), and a power line The battery interface 95A that charges the power from # 1 (99A) and supplies the power stored in the device is connected to the keyboard interface # 1 (38A) that receives the user's key input 31A.

  The processor # 1 (33A) executes the file system 44A, which is a software module, to control access to data on the hard disk HDD # 1 (10A), but the file system 44A is on the main memory # 1 (35A). Management information used by is present as local file system management information # 1 (45A).

  The peripheral circuit bus 92 includes an HDD # 1 (10A) that records digital music data as a file 11A, a decoder 96 that outputs an encoded digital audio stream and outputs a corresponding analog audio signal 32A, and a hard disk. A stream buffer memory 98 that temporarily stores the digital audio stream read from 10A until it is input to the decoder 96, and a USB I / F # 1 (97A) that terminates communication data on the USB line 91 And are connected.

  The DMAC 93 includes four independent DMAC communication channels, that is, a communication channel CH1 between the hard disk # 1 (10A) and the main memory # 1 (35A), a USB I / F # 1 (97A), and a main memory # 1. (35A), a communication channel CH3 from the HDD # 1 (10A) to the stream buffer memory 98, and a communication channel CH4 from the stream buffer memory 98 to the decoder 96. Since CH3 and CH4 are closed transfers to the peripheral circuit bus 92, they can be executed simultaneously with data transfer closed to the system bus # 1 (39A) such as access to the main memory # 1 (35A) by the processor # 1 (33A). it can. Furthermore, since there are two stream buffers 101, stream buffer # 1 (101A) and stream buffer # 2 (101B) in the stream buffer memory 98, while CH4 is reading data from one stream buffer, CH3 can write data to the other stream buffer.

  Next, the configuration of the personal computer 90 will be described only with respect to differences from the portable audio playback device 30A. In the personal computer 90, the peripheral circuit bus 92 does not exist and all the components are connected to the system bus # 2 (39B). A decoder 96 for decoding the digital audio stream and a stream that is an input buffer of the decoder 96 There is no buffer memory 98.

  The personal computer 90 has the original music file 11B in the built-in HDD # 2 (10B), which is stored in the CD-ROM via a CD-ROM interface or network interface not shown in FIG. It was obtained from media or a network.

  The personal computer 90 mainly executes a process of copying the original music file 11B to the portable audio playback device 30A connected via the USB line 91 (referred to as checkout of music data). For this purpose, the personal computer 90 needs to access the HDD # 1 (10A) built in the portable audio playback device 30A, which is realized by using the remote file system # 2 management information 45C. The That is, the personal computer 90 has not only a file system for accessing the built-in hard disk HDD # 2, but also a remote HDD # 1 file system that can be accessed via the USB line 91.

  In the present embodiment, the case where the personal computer 90 performs the check-out of the song data has been described. However, when the portable audio playback device 30A performs the check-out, the remote file system is provided on the 30A side. If you can.

  A file is recorded in the internal hard disk HDD # 1 (10A) of the portable audio player 30A in the format shown in FIG. As described above, access to the data area 12 is executed in units of clusters 13, but similarly, it is more efficient to execute access to the directory area 14 and FAT area 15 in a certain unit. The directory area 14 is accessed in a directory access unit 22, and the FAT area 15 is accessed in a FAT access unit 23.

  In the attribute of the directory entry 16, “Song 1.aud” which is the data file of the song 1 is “Read Only”, but this is the fact that once this file is created, it is only read when the song is played back. Reflects. If the song data file is writable, the file configuration information is duplicated in the FAT area 15 and the cluster link file 60, so when the file size is changed, the two configuration information are changed in synchronization. This increases the processing cost of writing the file. Therefore, it contributes to improving the performance of a device that mainly reads a file such as the audio playback device 30A, but for a device that needs to read and write at random like a normal data file, Increase the cost of accessing the file.

  In addition, the attribute of the directory entry 16 is a system of the cluster link file “Song 1.fat”, which indicates that this file is used by the file system and that access by the user is prohibited. Represent.

  Next, file system management information will be described with reference to FIG. FIG. 3 shows file system information for managing the internal hard disk HDD # 1 (10A) of the portable audio playback device 30A. The local file system information # 1 management information 45A in FIG. Represents the data structure of the remote file system # 2 management information 45C. Here, only differences from the file system management information 45 shown in FIG. 16 will be described. Unlike the file system management information 45 shown in FIG. 16, the local file system information # 1 management information 45A and the remote file system # 2 management information 45C include a division degree 110 and a pointer 111 to the cluster link file in the file information 50. It is characterized by including.

  The division degree 110 represents how much the corresponding file is divided and recorded. The division degree 110 is the minimum value for a file composed of continuous clusters such as the file A (11A) in FIG. 13, and the cluster constituting the file as in the music 1 data file 11A in FIG. A large value is obtained when these are distributed. The division degree 110 is measured when a file is created or written to the file, and the cluster link file 60 is created when this value exceeds a predetermined threshold value.

  As the pointer 111 to the cluster link file in FIG. 3, when the corresponding file is accompanied by the cluster link file 60, a pointer to the file information 50 of the cluster link file is set. When a file is read, if the pointer 111 to the cluster link file is set in the file information 50, instead of searching the FAT entry 17, the instructed cluster link file 60 is read and the configuration information of the target file is read. To get.

  Next, software executed in the portable audio playback device 30A and the personal computer 90 will be described. FIG. 4 shows a module configuration of software executed by the two devices.

  The personal computer 90 includes a content checkout module 120 as an application, which accesses the internal HDD # 2 (10B) using the local file system # 2 (44B) and reads the original music file 11B. Thereafter, the read music data is stored in the portable audio playback device 30A using the remote file system # 2 (44C). The remote file system # 2 (44C) uses the remote media client 122 as a driver for the media. The remote media client 122 uses the USB driver # 2 (123B), and the remote media server 124 in the portable audio playback device 30A. Exchange information with. The remote media server 124 receives an instruction from the remote media client 122 via the USB driver # 1 (123A), and controls the HDD driver # 1 (121A) based on the instruction to the built-in HDD # 1 (10A). Is written, and the execution result is returned to the remote media client 122.

  The portable audio playback device 30A has a music playback module 125 as an application, which accesses the internal HDD # 1 (10A) using the local file system # 1 (44A) and streams the song data file 11A. Read to buffer 101. In addition, the music playback module 125 transfers the data in the stream buffer 101 to the decoder 96, and the decoder driver 126 activates music decoding.

Next, how each software module operates cooperatively will be described in detail.
FIG. 5 shows a procedure for constructing a remote file system # 2 (44C) for accessing a file in the portable audio playback device 30A from the personal computer 90. The content checkout module 120 receives a notification from the remote media detection 130 when the portable audio playback device 30A is connected via the USB line 91. In response to this, a device authentication process 131 for checking whether the device connected via USB is a correct device is executed. In the device authentication process 131, the remote media client 122 is requested to exchange an authentication key. The remote media client 122 sends a key exchange request to the remote media server 124 in the portable audio playback device 30A via the USB line 91. The remote media server 124 sends a device authentication key unique to the portable audio player 30A to the remote media client 122 as a response. When the content checkout module 120 receives the device authentication key from the remote media client 122, the content checkout module 120 checks whether the device authentication key is correct, and executes the mount process 132 if it is correct.

  In the mount process 132, a mount request is sent to the remote media server 124 via the remote media client 122. The remote media server 124 acquires information (referred to as device information) regarding the internal HDD # 1 (10A) using the HDD driver # 1 (121A), and notifies the remote media client 122 of this information. Upon acquiring this device information, the content checkout module 120 generates a remote file system # 2 (44C).

  Next, how the software module operates when the music file on the personal computer 90 is checked out to the portable audio playback device 30A will be described with reference to FIG.

  When the content checkout module 120 receives the checkout request 140 from the user, the content checkout module 120 first executes the file preparation processing 141 to open the transfer source file (for example, the original music file 11B), and the transfer destination file. Create a new one. The transfer destination file is composed of, for example, a plurality of clusters arranged in a divided manner. In FIG. 1, the plurality of clusters constitute the music data file 11A. The transfer source file can be opened directly by the local file system # 2 (44B), but the transfer destination file is created by using the remote file system # 2 (44C) constructed by the mount processing 132.

  In the remote file system # 2 (44C), the remote media client 122 is used as driver software for media managed by the remote file system # 2 (44C). The remote media client 122 transmits a request to the remote media server 124 in the portable audio playback device 30A via the USB line 91. Based on the received request, the remote media server 124 controls the HDD driver # 1 (121A) to access a file on the HDD # 1 (10A) built in the portable audio playback device 30A, and displays the access result. Responds to the remote media client 122. Through the above-described client-server type cooperative operation, the remote media client 122 provides the remote file system # 2 with an access function to a remote hard disk (HDD # 1 in this example).

  When the content checkout module 120 succeeds in preparing the transfer source file and the transfer destination file in the file preparation process 141, the content checkout module 120 executes the content transfer process 142 and transfers it to the transfer destination file (an empty file because it is immediately after new creation at this point). Transfer the contents of the original file. In reading the contents of the source file, the FAT entry 17 is read to identify the cluster in which the contents are stored, and the cluster is read. Even when the read content is written to the transfer destination file, first, an empty cluster from which the FAT entry 17 is read is secured, and the content is written to the cluster. However, since the transfer destination file is in the portable audio playback device 30A, the access to the file is realized by the above-described client-server type cooperative operation using the remote file system 44C.

  When the content transfer process 142 is completed, the “transfer destination file closing process” 143 is executed to confirm the update to the transfer destination file. The file size of the directory entry 16 is updated, and the updated entry is written to the directory area 14 of the remote HDD # 1 (10A) (directory entry update 144). Also for the FAT entry 17, a chain of FAT entries corresponding to the cluster storing the contents of the transfer destination file is created and written to the FAT area 15 of the remote HDD # 1 (10A). In this process, the division degree 110 of this file is measured. If the degree of division is high, “create cluster link file” 146 is executed to create the cluster link file 60 for this file.

  Next, how the software module operates when the portable audio playback device 100 plays the music file created by the above checkout procedure will be described with reference to FIGS. FIG. 8 shows a conventional access procedure in which the cluster link file 60 does not exist, and FIG. 9 shows an access procedure using the cluster link file 60.

  First, FIG. 7 will be described. When the music playback module 125 receives the playback request 147 from the user, the music playback module 125 first executes “file playback preparation 148” to open the corresponding music file in order to play back the music instructed there. Here, the directory entry 16 of the music file 11A is read, and then the directory entry is acquired in order to check whether the cluster link file 60 for the music file exists. FIG. 8 shows a case where the cluster link file 60 does not exist and acquisition of this directory entry fails.

  Thereafter, the music playback module 125 executes “load content 149 to the stream buffer 1”, and reads the content of the music file 11A to the stream buffer # 1 (101A). At this time, since the cluster link file 60 does not exist, it is necessary to read the FAT entry 17 and identify the cluster 13 constituting the music file 11A. If the song file 11A is divided and there is only one valid FAT entry 17 in one FAT access unit 23 as described with reference to FIG. 1, the FAT entry 17 is read each time content is read for one cluster. Will be read again. For this reason, this hard disk access corresponds to the case of “high-level sector division” in FIG. 14, and the transfer performance decreases.

  When the reading of the content to the stream buffer # 1 (101A) is completed, the music playback module 125 executes “decoding activation 150 of the stream buffer 1”, and the decoder 96 decodes the data on the stream buffer # 1 (101A). Start up. Once the decoder 96 is activated, it takes out data from the stream buffer until it becomes empty and continues the decoding operation.

  The music playback module 125 executes “load content 151 to the stream buffer 2” in parallel with the music data decoding process by the decoder 96, and reads the content of the music file 11A to the stream buffer # 2 (101B). This reading is performed in the same procedure as the “loading content 149 to the stream buffer 1” described above. Since the access speed to the HDD # 1 (10A) is one digit or more higher than the decoding speed by the decoder 96, “load content 151 to the stream buffer 2” is performed before the content of the stream buffer 1 is completely decoded. Complete. Otherwise, the stream buffer 101 becomes underflow and sound jump occurs.

  When the decoder 96 completes the decoding of the content on the stream buffer # 1 (101A), the music playback module 125 executes “decoding start 152 of the stream buffer 2”. Thereafter, the content is read out again to the stream buffer # 1 (101A). Thereafter, the contents are read out to one of the stream buffers until the end of the music data file 11A is reached, and the procedure for decoding the contents in the other stream buffer is repeated.

  Next, file access using the cluster link file 60 will be described with reference to FIG. For the description of FIG. 8, only the difference from FIG. 7 will be described.

  In “File Playback Preparation 148A”, an attempt is made to acquire the directory entry of the cluster link file 60 for the music file. In FIG. 8, since the cluster link file 60 exists, acquisition of this directory entry is successful.

  Thereafter, the content of the music file 11A is read out to the stream buffer # 1 (101A) by “load content 149A to the stream buffer 1”. However, prior to “load music content 154”, “read 153 of the cluster link file” To read the contents of the cluster link file 60 for only one cluster 13. Since the cluster 13 constituting the music file 11A can be identified from the read data, the “content loading 154” can read only the content without reading the FAT entry 17.

  If the size of the cluster 13 is 16 kilobytes, 512 cluster numbers are recorded in the cluster link file 60 for one cluster, so that the music file 11A for 8 megabytes can be read out. 8 megabytes of music content corresponds to about 8 minutes when the playback speed of the music is 16 kilobytes / second, so one song can be accommodated in a normal music. Therefore, in the procedure of FIG. 8, once the cluster link file 60 is read by “load content 149A to the stream buffer 1”, the subsequent “load content 151A to the stream buffer 2” or “stream buffer 1 again” is read. "Load content to 149B" allows the music content to be loaded without having to read the cluster link file 60.

  Next, FIG. 9 shows an access sequence to the hard disk for the procedure of FIG. As shown in FIG. 9, even if the song 1 data file is recorded as extremely divided such as cluster # 1160, cluster # 2170, and cluster # 3290, the cluster group constituting the song 1 data file is a cluster link file. Since a cluster link file is read only for one cluster, a song data file of several megabytes can be read, so the processing load of hard disk access to obtain the file structure is ignored. As low as possible. On the other hand, in the file access using the FAT entry 17 shown in FIG. 14C, it is necessary to read one FAT entry 17 in order to read the music 1 data file of one cluster. The load becomes high.

  The file access method according to the first embodiment of the present invention as described above creates a cluster link file 60 corresponding to a cluster group constituting a segmented file, lists the cluster number as the file content, Since the cluster to be read is specified when reading the divided file, the number of accesses to the HDD 10 is reduced, and as a result, the access performance to the divided file is improved.

  In addition, for the construction of the cluster link file 60, it is not necessary to perform a heavy processing such as moving an existing file, and the processing load for tracing the list of the FAT entries 17 is dominant. Since this has already been performed at the time of file creation, and the result may remain in the working memory, the time required to construct the cluster link file 60 does not increase, so it takes a long time to create (record) the file. Sequential access to the divided data file can be executed at high speed without taking time.

  Further, since the file in which the cluster link file 60 exists can be accessed by the conventional procedure using the FAT entry 17, even when the HDD 10 is a removable medium, it can be accessed by a normal FAT file system. It is possible to maintain compatibility with popular file formats.

(Embodiment 2)
Hereinafter, a file access method and an information processing apparatus according to the second embodiment will be described with reference to the drawings. FIG. 10 is a diagram showing a file configuration used in the file access method in the information processing apparatus according to the second embodiment.

  First, the procedure for constructing the file shown in FIG. 10 will be described. On the recording medium of the information processing apparatus, as shown in FIG. 10, music data (song data 65) corresponding to one song is divided and recorded in a plurality of track data files 66. Note that the song data 65 has an upper limit (here, 7.68 megabytes) in file size. The recording medium has a one-to-one correspondence with the track data file 66, and track information 71 is recorded in the track management file 70 as information relating to the track data file 66.

When the data file is recorded, information regarding each track data file 66 is recorded in the track information 71 as the data file information 72. The data file information 72 includes a directory entry 74 and a cluster link 75. In the directory entry 74, the directory entry 16 of the corresponding track data file 66 in the directory area 14 is recorded. In the cluster link 75, the cluster numbers constituting the track data file 66 are recorded according to the construction order.

Each track information 71 corresponding to each track data file 66 in which one piece of music data 65 is divided and stored is combined by a next track 73 to form a one-side list. The order in which songs are played is expressed as a sequence of song entries 81 in the playlist information file 80. The song entry 81 includes a track information number 82 indicating the number of the track information 71 at the head of the corresponding song in the track management file 70.

  A procedure for reading a data file from the file configured as described above will be described. The playlist information file 80 designated by the user is read from the front, and the music corresponding to the read music entry 81 is sequentially reproduced. At this time, the file reading means reads the track information 71. Since the track information 71 has a fixed length (here, 16 kilobytes), it can be easily determined where in the track management file 70 the track information 71 indicated by the track identification number 82 in the song entry is recorded. The file reading unit identifies a cluster group constituting the track data file 66 from the data file information 72 in the track information 71.

  As described above, the information processing apparatus according to the second embodiment uses the data file information 72 in the track management file 70 to access the track data file 66 and reproduce the song data 65. When the reproduction of one track data file 66 is completed, the subsequent track information 71 is identified from the next track 73 in the corresponding track information 71, and the corresponding track data file 66 is reproduced. By executing this operation for all the song entries 81 in the playlist information file 80, the songs can be reproduced in a predetermined order.

  Next, an information processing apparatus to which the file access method according to the second embodiment is applied will be described.

  In addition, in order to explain the information processing apparatus to which the file access method according to the second embodiment is applied, it is compared with a conventional information processing apparatus (for example, see Patent Document 4). FIG. 11 is a diagram illustrating a configuration of a file used by the information processing apparatus described in Patent Document 4.

  As shown in FIG. 11, the track information 71A recorded in the data area 12 of the HDD by the file recording means (not shown) holds the data file name 160 as information regarding the corresponding track data file 66. . For this reason, when a transition is made from one track (T # 1) to another track (T # 2), the file reading means (not shown) changes the name of the track 2 data file from the data file name 160 of T # 2. And the directory entry 16 corresponding to the name needs to be read from the directory area 14. Thereafter, by reading the cluster 13 designated by the start cluster number in the read directory entry 16, the head content of the track 2 data file can be read. In order to read subsequent contents, it is necessary to trace the list of FAT entries 17 from the FAT area 15 and identify the cluster group constituting the track 2 data file. For this reason, hard disk access at the time of track switching is as shown in a sequence 171 shown in FIG.

  As shown in the sequence 171 of FIG. 12, when switching from the track # 1 to the track # 2, it is necessary to read the track 2 information (T # 2). For this purpose, the FAT entry (FAT # T2) for the track management file 70 is required. ) In advance. When the track 2 information T # 2 is read, the directory entry (DENT # 2) having the file name designated by the data file name 160 is read, but here the FAT entry (FAT # D2) for the directory area 12 is also read. It must be read before that. When the directory entry (DENT # 2) is read, the head cluster of the track 2 data file TD # 2 can be identified. However, for the subsequent clusters, it is necessary to read and identify the FAT entry (FAT # C2) for the track 2 data file. There is.

  On the other hand, in the information processing apparatus according to the second embodiment using the file information shown in FIG. 10, the track information 71 recorded by the file recording means (not shown) is transferred to the corresponding track data file 66. All the information necessary for access is retained. In other words, the directory entry 74 and the cluster link 75 in the track information 71 allow the file reading means (not shown) to read the track data file 66 without having to read the directory entry 16 or the FAT entry 17.

  Therefore, the access sequence to the hard disk in this case is as shown in the sequence 172 shown in FIG. In other words, since the track switching is completed by reading only the track 2 information (T # 2) and the FAT entry (FAT # T2) corresponding to the track switching, the time required for the track switching is halved compared with the conventional case. . In the track switching executed at the time of reproducing one piece of music, it is necessary to prevent underflow of input data to the decoder 96 in order to prevent sound skipping. For this purpose, since it is necessary to retain music data longer than the track switching time in the stream FIFO (see FIG. 3) in the decoder 96, the track switching time affects the hardware scale of the decoder 96. From this, in the information processing apparatus according to the second embodiment, the influence of file access on the hardware scale can be reduced.

  Further, since the cluster group constituting the track 2 data file TD # 2 is described in the cluster link 75 of the track 2 information, it is not necessary to read the FAT entry 17 when reading the music content data, and the track data file is divided. Even in the case of the conversion, it can be read out at high speed.

  As described above, in the file access method according to the second embodiment of the present invention, access to the track data file 66 holding the song data 65 is transferred to the directory entry 74 and the cluster link 75 in the track information 71. Therefore, there is no need to access the directory entry 16 in the directory area 14 or the FAT entry 17 in the FAT area 15, and if one track information 71 is read, the corresponding one track data file 66 is read. There is an effect that all information for access can be acquired. Therefore, the song data 65 can be read at a higher speed than the case of using the normal FAT file system method that needs to read the directory entry 16 and follow the list of the FAT entries 17.

  Further, as described above, the access to the data area 12 is executed in units of the cluster 13, but by setting the track information 71 to a size (16 kilobytes in the present embodiment) that can be accommodated in one cluster 13, one Management information necessary for reading the track data file 66 can be acquired by accessing the medium (for example, the hard disk 10) only once, and the song data 65 can be read at high speed.

  The file access method according to the present invention does not require a long time to create a file, can perform sequential access to a divided file at high speed, and is compatible with a conventional FAT file system. It provides access and is useful as a file system for information processing equipment, particularly for real-time stream playback equipment.

It is a figure showing the file management information used with the file access method concerning this Embodiment 1. FIG. It is a figure which shows the structure of the audio reproduction system using the file access method concerning this Embodiment 1. FIG. It is a figure which shows the file system management information used with the audio reproduction system using the file access method concerning this Embodiment 1. FIG. It is a figure which shows the module structure of the software performed with the audio reproduction system using the file access method concerning this Embodiment 1. FIG. It is a figure showing the file system initialization procedure performed with the audio reproduction system using the file access method concerning this Embodiment 1. FIG. It is a figure showing the file write-in procedure performed with the audio reproduction system using the file access method concerning this Embodiment 1. FIG. It is a figure showing the file read-out procedure performed with the conventional audio reproduction system. It is a figure showing the file read-out procedure performed with the audio reproduction system using the file access method concerning this Embodiment 1. FIG. It is a figure showing the access sequence to the hard disk at the time of the file read performed with the audio reproduction system using the file access method concerning this Embodiment 1. FIG. It is a figure which shows the file structure of the information processing apparatus using the file access method concerning this Embodiment 2. FIG. It is a figure which shows the structure of the file which the conventional audio reproduction apparatus uses. It is a figure showing the access sequence to the hard disk at the time of track switching in each of the conventional audio reproduction device and the information processing apparatus using the file access method according to the second embodiment. It is a figure which shows the file management method in a FAT file system. It is a figure showing the access timing to the hard disk at the time of reading the data file recorded on the continuous cluster in the conventional file system. It is a figure showing the access timing to the hard disk at the time of reading the data file divided in the conventional file system. It is a figure showing the access timing to the hard disk at the time of reading the data file extremely divided in the conventional file system. It is a figure which shows the structure of the file before the defragmentation process used with the conventional recording management method. It is a figure which shows the structure of the file after the defragmentation process used with the conventional recording management method. It is a figure which shows the structure of the conventional audio reproduction apparatus. It is a figure showing the access sequence to the hard disk at the time of the file reading of the conventional audio reproduction apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Hard disk 11 File 12 Data area 13 Cluster 14 Directory area 15 FAT area 16 Directory entry 17 FAT entry 20 Seek time 21 Transfer time 22 Directory access unit 23 FAT access unit 24 Defragmenting process 30 Audio playback device 31 Key input 32 Analog audio signal 33 Processor 34 ROM
35 Main memory 37 Audio controller 38 Keyboard I / F
39 System Bus 44 File System 45 File System Management Information 46 FAT Buffer 47 Directory Buffer 48 Data Buffer 50 File Information 51 File Name 52 Attribute 53 File Size 54 Buffer Address 55 File Pointer 56 Cluster Link Table 60 Cluster Link File 65 Song Data 66 Track Data file 70 Track management file 71 Track information 72 Data file information 73 Next track 74 Directory entry 75 Cluster link 80 Playlist information file 81 Song entry 82 Track information number 90 Personal computer 91 USB line 92 Peripheral circuit bus 93 DMAC
94 Power line 95 Battery circuit 96 Decoder 97 USB interface 98 Stream buffer memory 99 Power line 101 Stream buffer 110 Degree of division 111 Pointer to cluster link file 120 Content checkout module 121 HDD driver 122 Remote media client 123 USB driver 124 Remote media server 125 Music playback module 126 Decoder driver 130 Remote media detection 131 Device authentication process 132 Mount process 140 Checkout request 141 File preparation process 142 Content transfer 143 Transfer destination file closing process 144 Directory entry update 145 FAT update 146 Cluster link file update Create 147 Raw request 148 File playback preparation processing 149 Load content to stream buffer 150 150 Start decode of stream buffer 151 Load content to stream buffer 152 152 Start decode of stream buffer 153 Read cluster link file 154 Load song content 160 Data file name 171, 172 Access sequence

Claims (4)

In a file access method for recording and reading the data file with respect to a recording medium for recording the data file in block units,
When the data file is recorded, a fragmentation rate indicating how much the recording positions of the blocks constituting the data file are distributed on the recording medium is measured, and the fragmentation rate exceeds a predetermined threshold value. For example, an identification file that displays a list of identification numbers of blocks constituting the data file is created and recorded on the recording medium,
A file access method comprising: checking whether the identification file exists when reading the data file, and reading the identification file if it exists to identify a block group constituting the data file. A stream data file and a stream management file that has a one-to-one correspondence with the stream data file and holds information about the corresponding stream data file are recorded on a recording medium that records the data file in units of blocks. In the file access method for reproducing the stream data file recorded on the recording medium,
When recording the stream data file, information that lists the identification number of each block constituting the stream data file is written to the stream management file,
A file access method characterized in that when the stream data file is read, information held in the stream management file is read to identify a block group constituting the stream data file. In an information processing apparatus that records and reads the data file with respect to a recording medium that records the data file in block units,
When the data file is recorded, a fragmentation rate indicating how far the recording positions of the blocks constituting the data file are distributed on the recording medium is measured, and the fragmentation rate is a predetermined threshold value A file recording means for creating an identification file for displaying a list of identification numbers of blocks constituting the data file;
When reading the data file, confirm whether the identification file exists, if there is, read the identification file, a file reading means for identifying a block group constituting the data file;
An information processing apparatus comprising: A stream data file and a stream management file that has a one-to-one correspondence with the stream data file and holds information about the corresponding stream data file are recorded on a recording medium that records the data file in units of blocks. In the information processing apparatus for reproducing the stream data file recorded on the recording medium,
A file recording means for writing, in the stream management file, information listing the identification number of each block constituting the stream data file when recording the stream data file;
When reading the stream data file, file reading means for reading information held in the stream management file and identifying a block group constituting the stream data file;
An information processing apparatus comprising:

Images