WO2003040956A1

WO2003040956A1 - Method for synchronizing files between computers

Info

Publication number: WO2003040956A1
Application number: PCT/FR2002/003773
Authority: WO
Inventors: Alain Nicolas Piaton
Original assignee: Alain Nicolas Piaton
Priority date: 2001-11-07
Filing date: 2002-11-05
Publication date: 2003-05-15
Also published as: FR2831972B1; FR2831972A1

Abstract

The invention concerns a method for synchronizing computer files in the form of directory tree structures, comprising the following steps: storing at time t1 a first state of a tree structure, comprising a first table whereof the lines concern the directories and the columns contain data concerning said directories, and a second table whereof the lines concern the files and the columns contain data concerning said files; storing at time t2 a second state of another tree structure; comparing the first and second states which consists in storing a selection of data respectively derived from the first and second states, said selection being performed on a file-by-file basis, and for each of the files compared on the basis of elements associated with each of said files in the first and second states, an assembling step which consists in generating from said storage created during the comparison phase, a third table comprising lines concerning the types of files, which types organize the files contained in said directories.

Description

METHOD FOR SYNCHRONIZING FILES BETWEEN COMPUTERS

FIELD OF THE INVENTION

With the development of remote transmission modes (Internet, telephone networks, wired or not, infrared, etc.), and the proliferation of microcomputers and telephones of all kinds (PC, Mac (registered trademark), Palm (registered trademark) ), organizers, WAP or UMTS mobile phones, ...) the interconnection between these machines poses synchronization problems that are more and more frequent and more and more complex. For some devices, a fully automatic mode may be suitable, on the other hand in some others, it may be useful to see the list of information that will be copied from one machine to another, in order to avoid false operations, such as for example, overwriting recent data with older data.

Unfortunately, these devices often only have a small screen, and without a synthetic and clear presentation of this information, it is difficult, if not impossible, to carry out this control.

PRESENTATION OF THE INVENTION

To overcome this drawback, it may be interesting to compare files between the two computers and then group the key information resulting from this comparison, so that the number of information to be displayed is reduced to a few summary lines by type of files, and that thanks to a simple command we can have the detail of all the key information for each of the summary lines.

To this end, the invention relates to a method of synchronizing computer files, stored on computer media and organized in the form of directory trees containing said files, comprising the following steps: - storage at time t1 of a first state of at least part of a computer medium, comprising on the one hand at least a first table whose rows relate to the directories and whose columns contain information supplied by the operating system of the computer on these directories, and on the other hand at least one second table whose rows relate to the files and whose columns contain information provided by the operating system of the computer on these files or by the content of the files themselves,

- storage at time t2 of a second state of at least part of an IT medium, comprising on the one hand at least a first table whose rows relate to the directories and whose columns contain information provided by the operating system of the computer on these directories, and on the other hand at least a second table whose rows relate to the files and whose columns contain information supplied by the operating system of the computer on these files or by the contents of the files themselves,

- comparison of a part which can constitute the totality of the first state and of a part which can constitute the totality of the second state, consisting in an at least temporary memorization of the union, on the one hand of a selection of the information concerning respectively the directories and files cataloged respectively in the first and second tables of the compared part of the first state, and on the other hand of a selection of, information relating respectively to the directories and files cataloged respectively in the first and second tables of the compared part of the second state, said selection being made, file by file, and for each of the files compared from the elements associated with each of said files in the first and second states,

- grouping consisting in generating a third table comprising lines relating to the types of files in a directory in a tree structure, which types order the files contained in said directory. Advantageously, the third table associates each of said types of files in the directory with information, obtained by operations of logical type, of arithmetic type and of operation type on character strings, executed on the information given by the operating system of the computer or read from the file itself, on each of the files of the type considered contained in said directory of the trees of the computer media, which information is then stored in the columns of said third table.

DESCRIPTION OF THE EXPLANATORY DIAGRAMS

FIG. 1 is a description of the various hardware and software components used in the context of the invention.

Figure 2 is a description of the different phases of the process. DESCRIPTION OF THE IMPLEMENTATION OF THE PROCESS ACCORDING TO THE INVENTION

In what follows, the implementation of the method for synchronization between two machines M1 (see FIG. 1) and M2 of the organizer type is described with an operating system of the Windows CE (registered trademark) type, and linked together by a LAN link of the local network type (workgroup for example).

The user data of each machine M1 and M2 are stored respectively on a part V1D1 (or all) of the volume V1 (for example a hard disk, a CD-ROM, a flash memory ... also called hereinafter, computer support ) and on part V2D2 (or all) of volume V2.

Except in the case where a database is used, in general, the information is structured in the form of a tree of information blocks commonly called files.

In machine 1, the tree branch V1D1 which starts on the directory V1 D1 itself comprises sub-directories V1D1-1, V1D1-2, themselves comprising sub-directories V1D1-11, V1D1-13 ( for the V1 D1-1 subdirectory), V1D1-21, V1D1-22 subdirectories (for the V1D1-2 subdirectory), etc. The same applies to the V2D2 tree branch of the WI2 machine. For the sake of simplification, it will be assumed that, in the file management system (of Windows (registered trademark)) the name of the directory V1D1-ϋ is "ϋ".

In the same way we will assume that the directories, sub-directories ... are sorted in alphabetical order, which will subsequently facilitate comparisons between branches, but with a slightly more complicated algorithm we could use the process with directories not sorted alphabetically; same remark as regards the respect (UNIX mode) or non respect (DOS mode) of the case.

Each directory, for example the V1D1-11 directory, contains files F1, F2, F3, F4, etc. Regarding the file F1, L1 will represent the length, and D1 the date (and time) of last update expressed preferably in UTC time (universal time convention, also called GMT).

These files can be sorted alphabetically, which will make it easier to compare file lists between directories, but with a slightly more complicated algorithm, the process could be used with files not sorted alphabetically. It is assumed that the machine M1 can directly access the various elements describing the tree V2D2 of the machine M2, as if it were the tree V1 D1 of the same machine M1, thanks to the functionalities offered by the network manager local. In the case where this facility is not available, it suffices that a protocol between the two machines M1 and M2 makes it possible to recover a set of (or only one) table containing the image of the tree V2D2.

Synchronization between the WI1 and M2 machines can be carried out entirely by one of the two machines, or even more complicated, by means of a cooperative mode between the two machines.

Figure 2 describes the different phases of the process.

In a first phase P1, we start by building a set of two tables Td1 (hereinafter called the directory table) and Tfi (hereinafter called the file table) which contain an image, or state, at the moment t1 of the V1D1 branch, or else the image of the V1 disk, but ignoring the branches that do not contain V1 D1, such as the System branch, ...)

Thereafter the word double-table will designate the set of two tables of directories and files representing a tree structure. The table of directories Td1 contains for each directory of the branch V1D1 a station called thereafter station-directory, containing: the information necessary to evolve in the tree structure, for example a system of pointers built on the same principle as the "FAT »Of DOS, namely a pointer to the father directory directory, a pointer to the 1st child directory directory, a pointer to the next directory with the same father, a pointer to the 1st file station belonging to said directory station.

- as well as information relating to each directory, namely the name of the directory, the creation date, the number of files in the directory, the number of sub-directories, and the Li level, to which this directory belongs in relation to the disc root: level 0 for the root, level L1 (equal to 1) for V1D1, level L2 (equal to 2) for V1D1-1 ... etc

This information is easily obtained by simple commands (findfirst, findnext) from the file management system. There are many other conventional ways of representing a tree, for example a system with several subtables, with one subtable per level, the level being the concept defined above; One can also, in this table as in all those which will be described hereafter, replace the system of pointers by an index system in a table, that is to say designate a row in a table, not by its address in memory, but by its rank in said table.

The table of files Tf1 contains for each file of the branch V1 D1 a station, hereinafter called post-file, containing: the information necessary to evolve in the tree structure, for example a set of pointers, namely a pointer to the workstation to which the file belongs, a pointer to the next workstation in the same directory. as well as information relating to each file, namely the name of the file, the dates of creation, update, and last reading, size, extension, attributes, etc.

Thereafter, and by analogy with the notion of table, we will say that the table Td1 is composed of rows at the rate of one row per post-directory, and of columns, each column comprising the information relating to each row also called post -répertoire. Same remark for the table Ff1 which is composed of lines, each line corresponding to a file item, and of columns containing the information relating to said lines.

In the same way, in a second phase P2, we build another set of two tables Td2 and Tf2 which contain an image, or state, at time t2 of the directory V2D2.

These two sets of tables can be entirely created before the start of the next phase of comparison P3, or else be partially created as and when required according to the needs of the comparison.

The comparison P3 essentially consists in scanning in parallel the two tables Td1 and Td2 in order to detect the directories common to the two trees, as well as the directories which belong only to the tree V1D1 or those which belong only to the V2D2 tree structure, and when a V1D1-i directory is encountered in V1 D1 and V2D2-i in V2D2, common to the two trees, to explore in parallel the list of Fj files belonging to each of the two directories V1D1-i and V2D2- i, in order to detect those which only belong to the V1 D1 tree or to the V2D2 tree, and to construct 6 new trees, namely:

1 - a tree structure which includes only the files present only in the tree V1DÎ of the machine M1, and which will be represented by a double-table composed of the two tables TRd-1 and TRf-1 built on the same model as the tables Td1 (table of directories) and Tf1 (table of files) described above. 2 - a tree structure which includes only the files present only in the tree V2D2 of the machine M2, and which will be represented by a double table composed of the two tables TRd-2 and TRf-2,

3 - a tree structure which includes only the files present both in the tree structure V1D1 of the machine M1, and in the tree structure V2D2 of the machine M2 and which are deemed to be identical, that is to say which have the same length , the same date of last update, ..., and possibly the same checksum (allowing to ensure the identity of the contents of the files) and which will be represented by the two tables TRd-1 = 2 and TRf- 1 = 2,

4 - a tree structure which includes only the files present both in the tree structure V1D1 of the machine W11, and in the tree structure V2D2 of the machine M2 and which are deemed to be different, (by length or date or another given), which will be represented by the two tables TRd-1 # 2 and TRf-1 # 2,

5 - a tree structure which results from the union of the tree structure V1D1 of the machine M1 and the files present only in the tree structure V2D2 of the machine M2, and which will be represented by the two tables TRd-1 + e2 and TRf-1 + e2.

At the start of the comparison process, these two tables are initialized so that they are identical to Td1 and Tfl.

6 - a tree structure which results from the union of the tree structure V2D2 of the machine M2 and the files present only in the tree structure V1D1 of the machine WI1, and which will be represented by the two tables TRd-2 + e1 and TRf-2 + e1.

At the start of the comparison process, these two tables are initialized so that they are identical to Td2 and Tf2.

More generally, creation of a tree structure which is a union, on the one hand from a selection of information relating respectively to the directories and files cataloged respectively in the first and second tables of the first state, and on the other hand a selection of information relating respectively to the directories and files cataloged respectively in the first and second tables of the second state, said selection being made, file by file, and for each of the files compared from the elements associated with each of said files in the first and second states,

It should be noted in passing that the directory stations and file stations of these 6 new double tables are not strictly identical to those of the two starting double tables Td1 (and Tf1) and Td2 (and Tf2); indeed, the system of pointers to evolve in the tree structure, (or an equivalent system) maybe identical, on the other hand the data of each of these new stations must include simultaneously: a set of data which corresponds to the data coming from V1D1 (and therefore from Td1 and Tfl), a set of data which corresponds to the data coming from V2D2 (and therefore from Td2 and Tf2). It may be interesting to add to it a so-called occurrence variable which specifies whether a position includes information coming from both the two starting trees, or only from one and only one.

To make this comparison, by means of an iterative mode, the procedure is as follows from the data present in the example of FIG. 1:

Exploration is started with the positions in the two tables Td1 and Td2 corresponding respectively to the V1D1 and V2D2 level 1 directories; thanks to the system of pointers making it possible to browse the list of level 2 child sub-directories, one arrives at the first two sub-directories V1D1-1 and V2D2-1 which bear the same name, namely “1”. We perform the processing (see below) at the level of the files in this common directory, V1D1-1 and V2D2-1, then we launch a new exploration for the level 3 child subdirectories of V1D1-1 and V2D2-1. which will allow by performing a comparison on the only name of the directory, to successively meet the following 4 typical configurations

Sub-directory V1D1-11 on M1 and having the name "11"; no corresponding sub-directory on M2; We insert into the two tables TRd-1 and TRd-2 + e1 a new directory item with the path V1D1 \ V1D1-1 W1D1-11, and in the two tables TRf-1 and TRf-2 + e1, a series of File stations identical to those which were attached to the starting directory directory V1D1-11, these file stations being linked to this new directory station by the same system of pointers as that described above. V2D2-12 sub-directory on M 2 and having the name "12"; no corresponding sub-directory on M1;

We insert in the two tables TRd-2 and TRd-1 + e2 a new directory item with the path V1D1 \ V1D1-1 \ V1D1-12, and in the two tables TRf-2 and TRf-1 + e2, a series file stations identical to those which were attached to the starting directory station V2D2-12.

V1D1-13 subdirectory on M1 and V2D2-13 on M2 with the name "13";

In this case, we must start by analyzing the list of file stations attached to the two starting directory stations, and we therefore meet different cases:

F1 file present simultaneously on WI1 and M2, and deemed identical, with the same T1 update dates and same L1 lengths: Insertion in table TRd-1 = 2 of a new directory item with the path V1D1 \ V1D1-1 \ V1D1-13, and in the table and TRf-1 = 2, of file item F1.

F2 file present simultaneously on M1 and M2, but with an update date T2 (or T2 ') or a different length L2 (or L2'):

Insertion into table TRd-1 # 2 of a new directory item with the path V1 D1 \ V1D1-1 \ V1D1-13, and into table and TRf-1 # 2, of file extension F2.

File F3 present only on M1: Insertion in the two tables TRd-1 and TRd-2 + e1 of a new directory directory with the path V1D1 \ V1 D1-1 \ V1 D1-13, and in the two tables TRf- 1 and TRf-2 + e1 from file position F3.

F4 file present only on M2:

Insertion in the two tables TRd-2 and TRd-1 + e2 of a new directory directory with the path V1D1 \ V1D1-1 \ V1 D1-13, and in the two tables TRf-2 and TRf-1 + e2 of the file extension F4.

Obviously, when a directory extension contains several file extensions linked to the same directory extension, the latter is only inserted once, the objective being that we can evolve and obtain the same information as with Td1 starter sets

(and Tf1) and Td2 (and Tf2).

During phase P4, all or part of the 6 double tables (TRd -... and TRf- ...) resulting from the comparison phase P3 are taken up, with a view to generating a third table for each of them. TRt -..., hereinafter called the table of types (of files), and composed of stations called thereafter standard-posts, the type being a concept making it possible to group files by nature; this may be the extension of a file such as "doc", "xls", "html", etc., a range of last access dates, for example files accessed before or after January 1, a notion of right of access or security, which can be provided by the file management system.

In some type of file organization, it is necessary to read the beginning of the file (commonly called "file header") to recover certain notions such as software version, program executable only with a 32-bit machine, etc. .

This third table is also composed of rows, each row corresponding to a standard item, and of columns, each column comprising the information relating to this standard item described below. To generate this third table using the pointer system described above, we scan all the directories in the tree structure, and for each of the directory stations we proceed as follows:

We start by initializing at 0, a certain number of new zones, or columns, of the directory-directory, these zones being:

- pointer to the first standard extension attached to this directory extension (0 = empty string)

- number of standard stations attached to this directory station and containing data relating to the V1D1 tree structure

- number of standard stations attached to this directory station and containing data relating to the V2D2 tree structure.

We also initialize to 0 a temporary table which, thereafter, will comprise as many lines as there will be different types of files in said directory-directory, each line comprising the name of the type (for example “doc”, “xls », ...), as well as two counters which will contain the number of files which have this type of file in the said directory directory, and this in the tree V1D1, (same for V2D2); it is advantageously possible to add to it a system of pointers making it possible to quickly scan the file stations of said directory station having the same type name, that is to say at the level of said temporary table, a pointer to the first file station, and in each file item, a pointer to the next file item of the same type. It is obviously possible to replace the system of pointers with an index system as specified above.

Then, thanks to the pointer system described above, we browse all the file stations of said directory station. Each time a file position includes a type already existing in said temporary table, the counter (s) of the temporary table are incremented according to whether it belongs to V1 D1 or V2D2.

Each time a file extension includes a type which does not yet exist in said temporary table, a new line with the name of the type is added to the temporary table, the counter or counters of the temporary table are incremented according to its belonging to V1 D1 or V2D2 and the counters described above are incremented at the level of the directory directory, which will subsequently make it possible to immediately know the number of file types present in said directory.

When we have explored all the file-stations attached to said directory-station, it suffices to create in the table TRt -... a set of standard-stations, linked together by a system of pointers, each containing the information of a row of said temporary table, then. in the home directory station, fill in the pointer to the first standard station as described above. In the previous description, we looked at two different trees, compared to two very close moments t1 and t2, but we could also have compared the photographs at two different times of the same tree, or two photographs at two different times of two different trees.

In a different mode, thanks to a pointer system of the same type as that described above, it is possible to ensure that the double-table grouping the two tables known as the Td1 directories and the Tf1 files, is merged into a single table , like the “File Allocation Table - FAT” table in the Microsoft DOS system, each line corresponding to a record item, the content of which varies according to a code which specifies that it is a directory or file.

By way of example, and in a nonlimiting manner, according to the invention, software can be produced which makes it possible to synchronize two pocket microcomputers each comprising a tree structure containing only files of type “text” and of type “spreadsheet” , and which works as follows:

The list of directories which only contain files present in the first microcomputer is scrolled on the screen, one line per directory, each line comprising the full path of the directory, as well as the number of file of type "text" and type "spreadsheet" as in the example below.

"My documents \ legal file \ Company _1: 2 texts + 1 spreadsheet" In general, this type of information is sufficient for the user to ensure that he will not make a false move; if in doubt, just click with the mouse on the word "texts" to display just below the line described above, 2 new lines giving details of the 2 files of type "text" , to know :

"Contract of 10/11 / 2001.doc, 5643 bytes maj on 15/11/2001" "contract of 01/01 / 2001.doc, 1444 bytes maj on 10 / O2 / 2001"

Claims

CLAIMS:

1 - Method for synchronizing computer files, stored on computer media and organized in the form of directory trees containing said files, comprising the following steps:

- storage at time t1 of a first state of at least part of an IT medium, comprising on the one hand at least a first table whose rows relate to the directories and whose columns contain information provided by the operating system of the computer on these directories, and on the other hand at least a second table whose rows relate to the files and whose columns contain information supplied by the operating system of the computer on these files or by the contents of the files themselves,

- storage at time t2 of a second state of at least part of an IT medium, comprising on the one hand at least a first table whose rows relate to the directories and whose columns contain information provided by the operating system of the computer on these directories, and on the other hand at least a second table whose rows relate to the files and whose columns contain information supplied by the operating system of the computer on these files or by the content of the files themselves, - comparison of a part which can constitute the totality of the first state and a part which can constitute the totality of the second state, consisting in an at least temporary memorization of the union, d 'a part of a selection of information concerning respectively the directories and files cataloged respectively in the first and second tables of the compared part of the first report, and on the other part of a selection of information concerning. respectively the directories and files cataloged respectively in the first and second tables of the compared part of the second state, said selection being made, file by file, and for each of the files compared from the elements associated with each of said files in the first and second states, characterized in that the memorizations resulting from the comparison phase is followed by a grouping phase consisting in generating a third table comprising lines relating to the types of files in a directory in a tree structure, which types order the files contained in said directory.

2 - comparison method according to claim 1 characterized in that the third table associates each of said types of files in the directory with information obtained by operations of logical type, of arithmetic type and of operation type on strings of characters, executed on the information given by the computer's operating system or read from the file itself, on each of the files of the type considered contained in said directory of the trees of the computer media, which information is then stored in the columns of said third table. 3 - comparison method according to claims 1 or 2 characterized in that the first table whose rows relate to the directories and the second table whose rows relate to the files, are grouped into a single table.

4 - comparison method according to claims 1 to 3, characterized in that the information which appears in the columns of said second and third tables are developed from data supplied by the file management system of the computer.

5 - comparison method according to claims 1 to 4, characterized in that the type of file used to constitute the third table is the extension of the file name.

6 - comparison method according to claims 1 to 5, characterized in that the at least temporary storage of the union, on the one hand of a selection of information respectively concerning the directories and files cataloged respectively in the first and second tables of the compared part of the first state, and on the other hand of a selection of information relating respectively to the directories and files cataloged respectively in the first and second tables of the compared part of the second state, said selection being made, file by file , and for each of the files compared from the elements associated with each of said files in the first and second states, consists of one of the following storage: a) at least temporary storage of information relating respectively to the directories and files cataloged respectively in the first and second tables of the compared part of the first state but not inventoried respectively in the first and second tables of the compared part of the second state, b) at least temporary storage of information relating respectively to the directories and files not inventoried respectively in the first and second tables of the compared part of the first state but cataloged respectively in the first and second tables of the compared part of the second state, c) at least temporary storage of information concerning the directories and files respectively in the first and second tables of the compared part of the first state, but also cataloged respectively in the first and second tables of the compared part of the second state, with information supplied by the operating system of the computer, stored in the columns of said first and second tables, different in the compared part of the second state from those which are associated to said directories and said files in the tables of the compared part of the first state, with regard to the rows of said first and second tables which are common to the two compared parts of said states, d) at least temporary storage of information, concerning the directories and files cataloged respectively in the first and second tables of the compared part of the first state, but also cataloged respectively in the first and second tables of the compared part of the second state, with information provided by the operating system of the computer, stored in the columns of said first and second tables, identical in the compared part of the second state to those associated with said directories and said files in the tables of the compared part of the first state, as regards the rows of said first and second tables which are common to both states, e) storage at m oins temporary of the meeting on the one hand of the information, respectively concerning the files and directories, stored in the compared part of the first state and on the other hand of the information stored in the compared parts of the reports, concerning the directories and files inventoried only in the first and second tables of the compared part of the second state f) at least temporary storage of the meeting on the one hand of the information, respectively concerning the files and directories, stored in the compared part of the second state and on the other hand of the information stored in the compared parts of the reports, concerning the directories and files inventoried only in the first and second tables of the compared part of the first report.

7 - comparison method according to claims 1 to 6, characterized in that each station of said second and third tables include information making it possible to link together the file stations depending on the same type station.