JP3377836B2 - Business specification analyzer for source programs - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a business specification analyzing apparatus for a source program which analyzes the processing contents of an existing source program and outputs the analysis result.

[0002]

2. Description of the Related Art In a computer system, it is necessary to reconstruct the software in accordance with the specification change of the system and the replacement of the computer main body. For this reconstruction, it is necessary to grasp the processing contents of the existing software, that is, the existing program. In order to grasp this, it is necessary to organize and analyze the processing contents of the existing program as a whole. Generally, the program processing includes a "business specification part" for performing processing (input / output processing or check processing for the data item) related to the data item having an interface with the outside of the program.
It is roughly classified into a "computer control part" including file access related commands and other OS (operating system) dependent commands. Also, recently, when reconstructing a source program, CA
SE (Computer aided Software Engineering) tools are increasingly used.

[0003]

Conventionally, the extraction of the "business specification part" and the "computer control part" in the rebuilding work of the above source program has been performed by a human. That is, a person analyzes and judges the contents of processing while looking at the source program list for each program unit, and sorts and extracts information necessary for the reconstruction. For this reason, the reconstruction of the source program has a heavy work load on human beings, and the time required for the work has been great.

It is an object of the present invention to reduce the load in the rebuilding work of such a source program and to efficiently carry out the rebuilding work.

[0005]

FIG. 1 is a block diagram showing the principle of the present invention. The present invention relates to the rebuilding work of the source program 1 of the computer system.

The data item information extraction means 2 extracts from the source program 1 of the program executed by the computer system,
The data item information defined in the source program 1 is extracted.

The condition extracting means 4 extracts conditional expressions from the statements in the source program 1 and collects the same conditional expressions. The simple condition is, for example, one conditional expression described by a comparison operator, and the compound condition is a combination of a plurality of simple conditions.

The business term input means 5 is the condition extraction means 4
The business terms corresponding to the business are attached to the conditions extracted by. This business term is input by an operator, for example.

The record name file 6 stores the name of the record to which the data item to be extracted as the business specification belongs. The name of this record is also input by the operator, for example.

The control / business specification information extraction means 3 is based on the data item information extracted by the data item information extraction means 2 and is based on the source program 1 statement information relating to control and statement information relating to business specifications. Is extracted and output.

Also, for example, as described in claim 2, the source program 1 controls based on the data item information extracted by the data item information extracting means 2 and the record name stored in the record name file 6. The information of the statement related to and the information of the statement related to the business specification are extracted and output.

Further, as described in claim 4, the source program 1 is based on the data item information extracted by the data item information extracting means 2 and the record name stored in the record name file 6. A statement relating to control and a statement relating to business specifications are extracted from the statement, and for the statement using the conditional expression extracted by the conditional expression extracting means 4 among the statements relating to the business specification, the above-mentioned business It outputs in association with the business term given by the term input means 5.

[0013]

When reconstructing the source program 1, for example, the data item name defined in the source program 1 by the data item information extraction means 2, the name of the section to which it belongs, the name of the record to which it belongs, and its use Extract the usage, etc. Then, the control / business specification information extraction means 3 pays attention to a data item group having an external interface with another program or an external peripheral device in the source program 1,
The statement that is performing the input check processing or the setting processing at the time of output for these data item groups is extracted as the statement related to the business specification. Also, the statement that is performing the file access-related processing is extracted as a statement related to computer control. Then, these extracted statements are divided into “business specifications” and “control”, and are output to a file together with, for example, the line number and the record to which the data item used in the syntax belongs. Also, the extracted statement is output in a list in association with the record name and the business term given by the business term input means 5.

Therefore, by referring to the extracted information such as the file or the list output from the control / business specification information extracting means 3, the input work to the CASE tool is facilitated and the source program 1 is reconstructed. It is possible to reduce the work load on the system engineer.

[0015]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of an embodiment of the present invention.

In the figure, the source program 100
Is a source program for analysis of business specifications. In this embodiment, this program 100 is
OL: Common Business Oriented Language).

The data item information extraction section 112 analyzes the source program 100 and extracts the data items defined in the data section (DATA DIVISION) of the source program 100. This data item extraction information 1
14 is as follows. (a) Data item name (b) Name of section to which each data item is defined (SECTION name) (c) Name of record to which each data item belongs (Record record name) (d) Each data Use of the item. This is the type of area to which each data item belongs, database area, file input / output area, communication function message area, communication function related area,
And other areas. (e) FD of file section (FILE SECTION)
The file name defined in the (File Description) clause and the name of the record that belongs to the file with that file name (f) The open mode of each of the above files. This includes INPUT for specifying an input file, OUTPUT for specifying an output file, and the like.

The simple condition extraction unit 115 is a simple condition (simple condition expression) 1 described in the source program 100.
27 is extracted. At this time, each simple condition 127 described in the compound condition is also extracted. The simple condition 127 to be extracted is as follows. (a) IF, SEARCH, EVALUATE, PERF
Comparison conditions in each sentence of the ORM. (b) Condition name Extracted in the same form as the comparison condition for aggregation processing after the condition.

Condition Name Condition is a description unique to COBOL and is a condition described by a condition name.
An example of such a description is shown in FIG. The condition name is defined using the level number "88" in the data division (DATA DIVISION). In this figure, BBB and CCC correspond. In addition, the data item (A that is defined by the level number “01” immediately before the level number “88” (A
AA is applicable) is called a condition variable.

Then, the processing unit (PROCEDURE D
In IVISION), a condition name condition is a condition name specified as a condition in an IF statement or a PERFORM statement. In the example shown in the figure, BBB in the sentence of IF BBB is the condition name condition, which is AAA = 0.
Is equivalent to 01.

As described above, when the condition name condition is used, the conditional expression can be simply described and the program can be easily modified. Further, by setting a name suitable for the condition as the condition name, it is possible to instantaneously determine the meaning of the conditional expression.

The compound condition extracting unit 116 extracts the compound condition 128 described in the source program 100, including the logical operator connecting a plurality of simple conditions.
The compound condition 128 to be extracted is as follows. (a) IF, SEARCH, EVALUATE. PERF
Comparison conditions in each sentence of the ORM. (b) Condition name Extracted in the same form as the comparison condition for aggregation processing after the condition.

The simple condition aggregating unit 117 aggregates the same conditions of the simple condition 127 extracted by the simple condition extracting unit 115 into one piece of information by using the comparison operator and its left and right sides as keys. Further, at the time of this aggregation, the number of appearances of each condition is counted. Then, the aggregated information is added with the number of pieces of information before aggregation, that is, the number of appearances, as a post-aggregation simple condition 129.

With respect to the composite condition 128 extracted by the composite condition extracting unit 116, the composite condition aggregating unit 118 aggregates the same condition into one information by using the comparison operator, the left side and the right side thereof and the logical operator as keys. . Further, at the time of this aggregation, the number of appearances of each condition is counted. Then, the aggregated information is added with the number of pieces of information before aggregation, that is, the number of appearances, to create a post-aggregation compound condition 130.

The business term inputting section 119 manually inputs the corresponding business terms for the simple conditions 127 and the complex conditions 128 collected by the simple condition collecting section 117 and the complex condition collecting section 118, and outputs them. The business term dictionary 129 is created in which each of the conditions is associated with the business term indicating a specific meaning of each condition.

The record name file 123 stores the name (record name) of a record to which a data item which does not have an interface with the outside to be extracted as a "business specification" belongs. This record name is externally input by the user or the like.

The control / business specification information extraction unit 121, based on the data item information 114 extracted by the data item information extraction unit 112 and the record name stored in the record name file 123, the source program 100.
From the source program 100, and the line numbers of the statement performing the processing corresponding to “computer control” (hereinafter simply referred to as “control”) and the statement performing the processing corresponding to “business specification” To extract. This extraction is performed as follows. (a) The check process and update process for the items defined in the LINKAGE SECTION are extracted as "business specifications". (b) The check processing and update processing for the database input / output area are extracted as “business specifications”. (c) Check processing for the input / output area of files opened in input mode and update mode is "business specification"
To extract. (d) The update processing for the input / output area of the file opened in the output mode and update mode is extracted as "business specifications". (e) The check processing and update processing for the communication function message area are extracted as “control”. The “communication function message area” is an area in which input contents (data) from the display terminal screen are stored, and output contents (data) to the display terminal screen and contents (data) to be output to the printer terminal are set.

Since the input data from the terminal and the output data to the terminal are stored in this area in this way, it is appropriate to handle the check / update processing for this area as "business specifications". (f) The check processing and update processing for the communication function related area are extracted as “control”. "Communication function-related area"
Is an area where information for controlling communication functions is set by an online monitor or program. The information includes, for example, destination type (display, printer, etc.) and destination name (workstation name, logical destination name, etc.).

As described above, since the information for controlling the exchange of data with the terminal is stored in this area,
Treat as “control”. (g) File access related commands such as OPEN, READ, WRITE, and CLOSE, communication function related commands, and database access related commands are extracted as “control”. The communication function related command is a command for inputting a message from the terminal or outputting a message to the terminal. (h) The check process and the update process for the record having the record name stored in the record name file 123 and the data items belonging to the record are extracted as “business specifications”.

The check process and the update process in the above (a) to (h) represent the following processes, respectively. Check processing: IF, EVALUATE, PERFOR
Update process that compares the value of the data item using each command of M, SEARCH, etc. Update process: Process that sets the value of the data item using MOVE, COMPUTE command, etc. An information file 125 in which the control / business specification information extracted in this way is put together in a predetermined format is created and output. Further, the information list 126 in which the annotation information of "control" and "business specification" and the business term are displayed in the same line as the corresponding statement is also created and output on the list of the source program 100.

Next, the operation of the embodiment having the above configuration will be described with reference to the source program 100 described in COBOL shown in FIG. First, the data item information extraction unit 11
2 inputs the entire source program 100, and FIG.
The data item information 114 as shown in (b) is extracted.

That is, the data (DATA DIVIS
ION) file section (FILESECTION) and working section (WORKING-STORAGE SECTI)
Check each section of (ON). And, "FILE SECT
In the “ION”, the record “INREC” belonging to the file “INFILE” is extracted, and then the data item “INREC− belonging to this record“ INREC ”is extracted.
“A”, two data items “IN-ENTYEAR” and “IN-” that compose the data item “INREC-A”.
SERNO ”, data items“ IN-NAME ”and“ IN-SEC-CO ”that also belong to the record“ INREC ”
"DE" is extracted. Further, for these extracted data items, each information of “FILE” as the belonging section name and “File” as the usage purpose is set. Further, "INREC" is extracted as the belonging record name for the data items other than the record "INREC".

The same processing is performed for the file "OUTFIL
The record “OUTREC” belonging to “E” is also performed, and the data items “OUTREC” and “OUTRE” are set.
"CA", "OT-SERNO", "OT-NAM"
"E" and "OT-SEC-CODE" are extracted. Then, "FILE" is set as the belonging section name corresponding to these data items, and "file" is set as the usage purpose. Further, for data items other than "OUTREC", "OUTREC" is set as the record name to which they belong.

The data item information extraction unit 112 further analyzes "WORKING-STORAGE SECTION", and records "WK-XXX", the item "WKREC-A" belonging to the record "WK-XXX",
"WK-ENTYEAR", "WK-SERNO",
"WK-NAME", "WK-SEC-CODE", and "WK-ERRCODE" are extracted. In addition, "WK-
"ENTYEAR" and "WK-SERNO" are "WKRE"
C-A ".

Furthermore, "WORK" is set as the corresponding belonging section name for the extracted data item.
By the way, this "WORKING-STORAGE S
The "ECTION" is a work area provided in the main memory and does not correspond to any of the database input / output area, the file input / output area, the communication function message area, and the communication function related area. Set "other" in the intended use column corresponding to the item. And
Finally, for the above data items other than "WK-XXX", the record name to which they belong is "WK-XX".
X "is extracted and set.

By the above processing, the data item information 114 as shown in FIG. 5A is created. Furthermore, the data item information extraction unit 112 uses the file section (FILE SEC
File name “INFIL” defined in (TION)
“E” and “OUTFILE” are extracted, and as records (record names) belonging to these files, “I” and “OUTFILE” are extracted.
"NREC" and "OUTREC" are extracted. In addition, in order to check the open mode of these files, a processing unit (PROCEDURE DIVISION USI
NG LK-A), and from the description sentence “OPENINPUT FILE.” At line number “000500” to “IN” as the open mode of the file “INFILE”.
PUT "(input mode) is extracted. On the other hand, for the file “OUTFILE”, the line number “00051
0 "description" OPEN OUTPUT OUTPUT "
E. , "OUTPUT" (output mode) is extracted as an open mode. Then, based on these extracted information, the data item information 114 as shown in FIG. 5B is created.

Next, the operation of the simple condition extracting section 115 will be described in more detail with reference to the flowchart of FIG. First, the processing unit of the source program 100 (PROC
The top line (top record) of EDURE DIVISION) is read (S11), and it is determined whether this record is a record indicating the end of the source program 100 (S12).

If it is not the end record (S12,
No), and then it is determined whether or not an IF statement is described (S13). Then, if the IF statement is described, the conditional expression part is saved in a storage area (not shown) by referring to this syntax (S14), and the saved conditional expression is analyzed to extract a simple conditional expression. Formula division output processing is performed (S15).

The conditional expression division output processing S15 will be described later in detail. On the other hand, the above step S13
Then, if the IF statement is not described, next EVALUAT
It is checked whether the E sentence is described (S16). And E
If the VALUE statement is described (S16, Ye
s), "selection subject" and "selection subject" are taken out from this syntax, these are temporarily saved in a storage area (not shown) (S17), and then the "selection subject" and the "selection subject" Are combined and output to a simple conditional expression output file (not shown) (S18).

If the record read in step S16 is not an EVALUATE statement (S16, N
o) Then, it is determined whether or not it is a PERFORM statement (S
19). If it is a PERFORM statement (S19,
Yes), referring to the syntax of the PERFORM statement, UNT
The IL condition is saved in a storage area (not shown) (S20). Then, based on the saved UNTIL condition, the conditional expression division output process S15 is performed.

Further, in step S19, the PERF
If it is not an ORM sentence (S19, No), it is determined whether or not the record read next is a SEARCH sentence (S2).
1). If it is a SEARCH sentence (S22, Ye
s), referring to this syntax, the corresponding WHEN condition is saved in a storage area (not shown) (S22). Then, based on the saved WHEN condition, the conditional expression division output processing S15 is performed.

The above-mentioned processing is performed in the above step S12.
Then, the processing unit of the source program 100 (PROCEDUR
Until it is determined that all the records have been read from (E DIVISION) (S12, Yes), the process is performed on all the records in the order of line numbers.

As a result, in the case of the source program 100 shown in FIG. 4, simple conditional expressions as shown in FIG. 8 are extracted. By the way, a data name corresponds to the above-mentioned “selection subject” in the EVALUATE statement, and a condition value corresponds to the above-mentioned “selection target”. Therefore, in the case of the syntax EVALUATE WK-A WHEN '1', the selection subject = WK-A and the selection target = '1'.

Next, the details of the conditional expression division output processing S15 will be described with reference to the flowchart of FIG. In this process, first, the conditional expression (conditional expression of IF statement, UNTIL conditional expression of PERFORM statement, WHEN conditional expression of SEARCH statement) saved from the save storage area is read out, and logical operation is performed in the conditional expression. It is checked whether a child is included, that is, whether the conditional expression is a compound conditional expression (S151). If the logical operator is included (S151, YES), the compound conditional expression is divided into a plurality of simple conditional expressions using the logical operator as a delimiter (S152).

Next, the simple conditional expressions obtained by the processing of step S152 are output one by one to the simple conditional expression extraction file (S153). By repeating the process of step S153 by the number of conditional expressions obtained in step S152, the simple conditional expression extraction file stores one simple conditional expression in the case of a simple condition and a simple conditional expression in the case of a complex condition. Stores a corresponding number of simple conditional expressions.

This conditional expression division output processing S15 is executed by the line number (000690) of the source program 100 shown in FIG.
IF statement [IF WK-SEC-CODE
= '1' AND IN-NAME = SPACE] will be described as an example.

In the case of this IF statement, the conditional expression [WK-SEC
-CODE = '1' ANDIN-NAME = SP
ACE] is saved. Then, the above step S151
Then, it is determined that the conditional expression includes the logical operator [AND] (S151, Yes), and the next step S152.
With this logical operator [AND] as a delimiter, the saved conditional expression is [WK-SEC CODE =
It is divided into two conditional expressions, “1”] and [IN-NAME = SPACE].

Then, step S153 is repeated twice, and the two conditional expressions [WK-SEC-CODE =
'1'] and [IN-NAME = SPACE] are output to the simple conditional expression extraction file.

On the other hand, in the case of an IF statement in which the conditional expression is a simple condition like the syntax of line number (000530),
When it is determined No in step S151, step S153 is performed only once, and the simple conditional expression is output to the simple conditional expression extraction file.

In the case of the source program 100 shown in FIG. 4, the above-mentioned processing is performed by line numbers (000550) and (0005).
70), (000590), and (000690) I
This is also performed for the F statement, and the contents shown in FIG. 8 are written in the simple conditional expression extraction file. Also, although not particularly shown, an EVALUATE statement, a PERFORM statement,
Also for the SEARCH statement, the conditional expression divided output processing S15 writes the corresponding conditional statements in the simple conditional expression extraction processing file in the format shown in FIG.

In COBOL, conditional expressions may be hierarchized. A method of extracting a conditional expression in such a case will be described with reference to FIG. In the example shown in the figure, the SEARCH statement and EVA are included in the EVALUATE statement.
A LUATE statement is included as a nest.
In order to deal with such a description, a table (hierarchical storage table) 200 hierarchized as a first hierarchy, a second hierarchy, a third hierarchy, ... As shown in FIG. To prepare.

In the case of the source program shown in FIG. 9,
At the time of analysis up to END-SEARCH, the end of the SEARCH statement located in the second layer is detected. During this time,
Find the WHEN conditions of ,,,, and, and FIG.
As shown in (a), WH of the EVALUATE statement in the first layer
The conditional expression [WK-A = '1'] of the EN condition is stored in the first tier storage area 201 of the tier-specific storage table 200. Furthermore, as shown in FIG. 3A, the EV of the first layer is
SEA of the second layer that is a nest of ALUATE statement
Conditional expression of three WHEN conditions of RCH sentence [KEY-A
= '1'], [KEY-A = '2'], and [KE
YA = '3'] is stored in the second tier storage area 202 of the tier-specific storage table 200. When the end of the SEARCH statement of the second layer is detected as described above, the three conditional expressions KEY-A′1 ′ and KEY-A of the SEARCH statement stored in the second layer storage area 202 are detected.
"2", KEY-A "3" are output to the simple conditional expression extraction file.

Subsequently, by reading out WHEN '2', it is judged that this is the condition of the EVALUATE statement of the first layer, and the conditional expression WK-A =, as shown in FIG. 10 (b).
"2" is stored following the WK-A "1" in the first tier storage area 201 of the tiered storage table 200. Further, the WHEN conditional expressions (WHEN'1 ', WHEN'2') of the EVALUATE statement of the second layer are sequentially detected, and the conditional expressions WW-B'1 'and WW-B =' 2 'are sequentially created, These are stacked from the top of the second tier storage area 202 of the tiered storage table 200. Then, the end of the second-level EVALUATE statement is detected by reading END-EVALUATE, and two conditional expressions WK-B of the second-level EVALUATE statement that are stacked in the second-tier storage area 202 are detected.
= '1', WK-B'2 'are output to the simple conditional expression extraction file.

Next, the outer 1 END-EVALUAT
E. By reading, E of the first layer

[0055]

[Outer 1]

The end of the VALUATE statement is detected, whereby the first tier storage area 201 of the tiered storage table 200 is detected.
The conditional expressions WK-A = '1' and WK-A = '2' of the EVALUATE statement of the first layer, which have been stored in the above, are output to the simple conditional expression extraction file.

Thus, the extraction of the simple conditional expression of the EVALUATE statement having the nested structure is completed.
Note that this extraction method is performed by the compound condition extraction unit 116 described below.
It also applies to the processing of.

Next, details of the operation of the composite condition extracting section 116 are shown in the flowchart of FIG. In this flowchart, the simple condition extracting unit 115 shown in FIG. 6 described above.
The same steps as those in the operation flowchart of FIG.

The operation of the composite condition extracting unit 116 shown in the flowchart of FIG. 11 is different from the operation of the simple condition extracting unit 115 shown in the flowchart of FIG.
F statement, EVALUATE statement, PERFORM statement, and S statement
There is a step S31 for judging whether or not the conditional expression appearing in the EARCH statement is a compound condition, and in the case of the compound condition, it is not necessary to divide this into a plurality of simple conditions.

Therefore, only different steps will be mainly described here to avoid duplication of description. First, E
In the case of a VALUE statement record (S16, Ye
s) With reference to the syntax, "selection subject" and "selection target" are extracted (S17). Then, these two are combined to create a conditional expression, and this conditional expression is saved in a storage area (not shown) (S19). Also, IF sentence, PERFORM
In the case of a statement and a SEARCH statement, the conditional expression and UNTIL are respectively set in the same manner as the simple condition extraction unit 115.
The conditional expression and the WHEN conditional expression are saved in the storage area (S14, S21, A23).

Subsequently, it is judged whether or not the saved conditional expression is a compound condition (S31), and if it is a compound condition (S31, YES), the saved conditional expression of the compound condition is invalidated. Output to the composite conditional expression extraction file shown (S
33).

Through the above processing, the IF statement [IF WK-SEC = '1' AND I described in the line number (000690) from the source program 100 shown in FIG.
N-NAME = SPACE] combined condition [WK-SE
C = '1' AND IN-NAME = SPACE] is extracted and written in the composite conditional expression extraction file in the format shown in FIG.

Next, the operation of the simple condition aggregating unit 117 will be described with reference to the flowchart of FIG. The simple condition aggregating unit 117 first sorts the conditional expressions (see FIG. 8) stored in the simple conditional expression extraction file using the unique name, the comparison operator, and the constant as keys (S41).

By this sorting, the same simple conditional expressions are continuously stored as one lump. Next, the data for one conditional expression is read from the beginning of the simple conditional expression extraction file after sorting (S42). There may be a case where no record, that is, simple conditional expression data is stored in the simple conditional expression extraction file. For this purpose, the above step S42
In step S43, it is determined whether the read data is EOF (End Of File), that is, the end of the simple conditional expression extraction file (S43). If the file is the end (S43, Yes), the process is immediately performed. finish.

On the other hand, if the data read in step S42 is the data of the simple conditional expression (S43, No),
The unique name, comparison operator, and constant in the conditional expression are saved in a storage area (not shown) (S44). Then, the aggregation number counter is initialized to "1". This aggregate number counter is a counter used to aggregate the number of appearances of each simple conditional expression.

Next, the data for one conditional expression stored immediately after the simple conditional expression read in step S42 is read from the sorted simple conditional expression extraction file (S46). At this time, EOF indicating the end of the file
Since (End Of File) may be read out, first,
It is determined whether the data is EOF, that is, the end of the simple conditional expression extraction file after sorting (S4).
7). If the read data is not EOF but data of a new simple conditional expression (S47, No), the saved unique name, comparison operator, constant, and the above step S
It is determined whether the unique name, the comparison operator, and the constant of the simple conditional expression read in 46 are all the same (S48).

If all of these elements are the same,
The aggregate number counter is incremented by "1" (S4
9) and returns to step S46 again. As a result, if the same simple conditional expression is continuously stored from the beginning of the sorted simple conditional expression extraction file, the above steps S46 to S49 are repeated by the number obtained by subtracting 1 from the continuous number. Then, in step S48, the simple conditional expression read in step S46 is the same as in step S4.
If it is determined that it is different from the first simple conditional expression saved in step 4 (S48, No), the saved simple conditional expressions are aggregated based on the value of the aggregation number counter, and then output to the simple condition output file. Yes (S50).

As a result, the simple conditional expressions continuously stored from the beginning of the sorted simple conditional expression extraction file are stored in the simple conditional expression aggregation file together with the number of appearances thereof. On the other hand, in the first step S48, if it is determined that all the elements of both simple conditional expressions do not match, that is, the two simple conditional expressions are different (S48, No), the saved simple condition after sorting is stored. The simple conditional expression stored at the head of the expression extraction file is stored in the simple conditional expression aggregation file with the number of appearances set to "1" (S50).

Following step S50, the unique name, comparison operator, and constant of the simple conditional expression read in step S46 are saved in the storage area (S51). Subsequently, the aggregation number counter is initialized to "1" (S5
2) Then, the process returns to step S46 again.

As a result, the same consolidation processing as above is performed for the remaining simple conditional expressions stored in the sorted simple conditional expression extraction file. When the EOF is read in step S47 (S47, Yes), the simple conditional expressions saved in the previous step S51 are aggregated based on the current value of the aggregation counter and output to the simple conditional expression aggregation file. Yes (S53).

Through the above processing, a simple conditional expression aggregate file having the contents shown in FIG. 14 is created from the simple conditional expression extraction file having the contents shown in FIG. By the way, a data item with a subscript may be used on the right side or the left side of the conditional expression. In such a case, at the time of the above simple condition integration, the integration is performed by the method shown in FIG. In other words, if the subscript consists of only constants, the subscript will be considered as the key and aggregated. AAA (1), AAA
(2) and BBB (11) are applicable. If the subscript is a unique name or only a partial constant,
After replacing the subscript with “*” once, this (*) is included in the key and aggregated. AAA (IX-Y), BBB (1
IX-2) and BBB (IX-1, IX-2) are applicable.

This aggregation method is also applied when aggregating complex conditions, which will be described later. Next, FIG. 16 is a flowchart for explaining the operation of the compound condition aggregation unit 118. As shown in this figure, the operation of the composite condition aggregation unit 118 is almost the same as the operation of the simple condition aggregation unit 117 described above. That is, the aggregation target is replaced with a simple conditional expression, and the same composite conditional expression is determined by determining whether all the elements of the composite conditional expression match (S66). ). Then, the number of appearances of various complex conditional expressions is aggregated by the aggregate number counter, and different complex conditional expressions are output together with the number of appearances to the complex conditional expression aggregation file (S70).

As a result, a composite conditional expression aggregate file having the contents shown in FIG. 17 is created from the composite conditional expression extraction file having the contents shown in FIG. In the case of this example, since there is only one compound conditional expression, aggregation is not performed. However, when the source program has a plurality of identical compound conditional expressions, the conditional expressions are aggregated. .

Next, the contents of the simple conditional expression aggregating file and the complex conditional expression aggregating file created by the simple condition aggregating section 117 and the complex condition aggregating section 118 as described above are input via the business term inputting section 119. Enter the business term corresponding to each conditional expression while referring to the business term book 1
Create 29.

The business terms are IF, SEARCH,
It is a term that specifically expresses the check processing (conditional expression) for a certain data item in the EVALUATE, PERFORM statement, etc. so that the actual work content becomes clear. This business term is manually input while looking at the display screen or the like. In the example of the simple conditional expression shown in FIG. 7A, IN-ENTYEAR IS NOT NUMERIC is a data item "IN-ENTY" in which the year is input.
The business term "year input error" is entered in the conditional expression for checking whether or not the numerical data is correctly entered in "EAR".

Similarly, with respect to the other four simple conditional expressions, "name input error", "department code input error", "employee number input error", and "Tokyo" corresponding to the respective jobs are similarly performed. The business term “district” is entered.

That is, for example, in the conditional expression WK-SEC-CODE = "1", the code "1" corresponds to "Tokyo area". Therefore, the conditional expression "Tokyo area" is used as the conditional expression. You are typing.

Further, WK-SEC-CODE = '1' AND IN-NA shown in FIG.
For the compound conditional expression ME = SPACE, the business term "Tokyo office" is entered.

In the record name file 123,
Although it does not have an interface with the outside through an information processing device (not shown), a process of writing the name of the record to be extracted as the "business specification" is performed. For example, as shown in FIG. 19, the source program 100 shown in FIG. Working department (WO
Record "WK-XXX" (line number 0003) defined in RKING STORAGE SECTION
20) is written in the record name file 123.

Subsequently, the control / business specification information extraction unit 1
The operation of 21 will be described with reference to the flowcharts of FIGS. First, the statement described in the processing unit (PROCEDURE DIVISION) of the source program 100 is decomposed into one statement unit (S101), and the file access related command (OPEN, READ, W) is first executed for the first statement.
RITE, CLOSE, etc.) (S102), a communication function related command (S104), or a database access related command (S106).
If any of the above statement statements is applicable, the statement is extracted as "control" (S103, S
105, S107).

If none of the above statement statements is true, then it is determined whether or not the statement is a check processing (S108). Then, if the statement is (S108, YES), it is determined whether the check target corresponds to the data item of or shown below (S109, S111, S113, S).
115). A data item (S109) in the record (WK-XXX) designated by the record name file 123. That is, in the source program 100 of FIG. 4, WKREC-
A, WK-ENTYEAR, WK, SERNO, WK-
NAME, WK-SEC-CODE, or WK-ER
RCODE is applicable. Data item (S111) in the record of the contact area (LINKAGE SECTION). In the source program 100 of FIG. 25, LK-01, LK-01-01, L
K-01-02, ... Applies. Data items in the database input / output area (S11
3). In the source program 100 of FIG. 25, RDBRE
C is applicable. Data item in the input / output area of the file opened in the input mode or the update mode (S115). In the source program 100 of FIG. 4 and FIG.
C, INREC-A, IN-ENTYEAR, IN-S
ERNO, IN-NAME, IN-SEC-CODE are applicable. Data items in the communication function message area (S11
7). In the source program 100 of FIG. 25, MENU
Corresponding to EC, MENUREC-01, ....

Steps S109 and S11 as described above
The determination of 1, S113, S115, and S117 is performed by referring to the source program 100, the data item information 114, and the record name file 123. Then, the statement for which the check processing is performed on any of the above-mentioned data items is extracted as a "business specification" (check processing) (S110, S112, S1).
14, S116, S118).

On the other hand, if the data item to be checked does not correspond to any of the above items (S117,
No), and then it is determined whether or not the check target is a data item in the communication function related area (S119). Then, the data item (in the case of the source program 100 of FIG. 25,
If DSP-DEST-1, ... Applies), the statement checking this data item is extracted as "control" (S120).

If the input statement is not checked in step S108 (S10).
8, No), then update processing (MOVE command or COMPU
It is determined whether or not a TE command or the like is performed (S12).
1). Then, when the update process is being performed (S121,
YES), it is determined whether or not the data item to be updated corresponds to the following, or, (S122, S124, S126, S128, S13).
0). A data item in the record (WK-XXX) specified in the record name file 123. That is, in the source program 100 of FIG. 4, WKREC-A, WK-EN
TYEAR, WK-SERNO, WK-NAME, WK
-Corresponds to SEC-CODE or WK-ERRCODE. A data item in a record in the LINKAGE SECTION. In the case of the source program 100 shown in FIG. 26, LK-01, LK-01-01, LK-
01-02 is applicable. A data item in the database I / O area. In the case of the source program 100 shown in FIG. 26, RDBREC is applicable. A data item in the I / O area of a file that is open in output or update mode. In the case of the source program 100 shown in FIG. 26, OUTREC, OUTR
EC-A, OT-ENTYEAR, OT-SERNO,
Applicable to OT-NAME and OT-SEC-CODE. A data item in the communication function message area. In the case of the source program 100 shown in FIG. 26, MENU,
MENUREC-01, ...

The above data item discrimination is also performed by referring to the source program 100, the data item information 114, and the record name file 123. Then, the relevant data items are extracted as "business specifications" (update processing) (S123, S125, S127, S129, S1).
31).

On the other hand, even in the case of a statement that is being updated, if the object to be updated does not correspond to any of the above ,, or data items (S130, NO), then the update is performed. It is determined whether the target is a data item in the communication function related area (S132). If the data item is (S132, YES),
The statement that is performing the update process for this item is extracted as "control" (S133).

No extraction is performed for statements that do not correspond to any of the above. Then, after the statements corresponding to the “business specification” or the “control” are extracted as described above (S103, S105, S10).
7, S110, S112, S114, S116, S11
8, S120, S123, S125, S127, S12
9, S131, S133), business term dictionary 129 (FIG. 1)
8), the corresponding business term is also output for the statement in the dictionary 129 in which the business term is set for the conditional expression in its own syntax (S134).

The control / business specification information extraction unit 121 performs the above-described processing of steps S101 to S134 by the processing unit (PROCEDURE DIVI) in the source program 100.
SION) for all statements.

As a result, in the case of the source program 100 shown in FIG. 4, the following statements (1) to (4) are extracted. The statements with line numbers (000500) and (000510) are extracted as "control" because they are OPEN instructions (file access related instructions). The statement at line number (000520) is RE
"Control" because it is an AD command (file access related command)
Is extracted as. Line numbers (000530), (000550), (0
00570), (000590), and (00069)
0) statements are input files (INFIL
E) data item (IN-ENTYE) belonging to the record
AR, IN-SERNO, IN-NAME, IN-SE
Since it is a check process (IF statement) for C-CODE), it is extracted as a "business specification". The statement (000690) is also a check process for the data item (WK-SEC-CODE) belonging to the record (WK-XXX) specified in the record name file 123. The statements of line numbers (000700), (000710), and (000720) are data items (O) belonging to the record of the output file (OUTFILE).
T-ENTYEAR, OT-SERNO, OT-NAM
Since it is an update process (MOVE command) for E), it is extracted as a "business specification" (update process). The statements with line numbers (000800), (000810), and (000820) are records (WK-XXX) specified by the record name file 123.
Data items belonging to (WK-ENTYEAR, WK-S
Update processing (MOV for ERNO, WK-NAME)
Since it is an E command), it is extracted as a "business specification" (update process). The statement at line number (000900) is WR
Since it is an ITE command, it is extracted as "control".

The control / business specification information extraction section 121 performs the above-described extraction processing from the source program 100 to create an information file 125 in the format shown in FIG.

In this information file 125, the statements of the "control" portion and the statements of the "business specification" are listed together with the corresponding line numbers in the source program 100 in the order of the line numbers. Further, in order to clearly indicate whether the "business specification" is the "check processing" or the "update processing", (check) or (update) is added after the description of the "business specification". Further, for the statement of "business specification", the record name to which the data item name used in the conditional expression in the statement belongs is also output. Extracting this record name
This is performed by referring to the data item information 114.

Further, the control / business specification information extraction unit 121
Also creates and outputs the information list 126 as shown in FIG. In the information list 126, the source program 100 is output as a list, and the annotation information 1001 is also output for the statements corresponding to the “control” portion and the “business specification” portion. Ie "control"
Partial statement (line number 000500-0005
10 and 000900), the symbol "*" is output.

On the other hand, the statement of the "business specification" part (line number 000530,000550,570,
000590600,690-0007
20, and 000800-000820),
Whether the check process or the update process is being performed is indicated by the symbols "C" (check mark) and "U", respectively.
(Update process) At this time, the name of the record to which the data item used in the conditional expression in the statement belongs is also specified. That is, referring to the data item information 114 shown in FIG. 5, “INREC” and “OUTREC” are assigned as the names of the records to which the data items used in the source program 100 shown in FIG. 25 belong.
And "WK-XXX". Then, these record names are assigned to the annotation information 10 on the left side of the information list 126.
Output at the top of column 01. Then, at the intersection of the vertical column of these record names and the line in which the statement of "business specification" is described, the corresponding symbol of "C" or "U" is output.

For example, in line number 000530, "IF I
The statement "N-ENTYEAR IS NOT NUMERIC" indicates that the data item "IN-ENTYEA
In the check process for “R”, the data item “IN-E
"NTYEAR" belongs to the record "INREC".
Therefore, for the statement with the line number “000530”, the record “INR” of the annotation information 1001 is displayed.
The symbol "C" is output in the field "EC". Do the same for the other "business specification" statements.
A symbol indicating the processing is output to the record column to which the data item in the statement belongs.

Further, in this information list 126,
The business term manually input via the business term input unit 119 is output to the same output line as the statement of the portion performing the check process (symbol “C”). The output of the business term is performed by referring to the business term dictionary 129 in step S134 of the flowchart of FIG.

As described above, the information file 125 and the information list 126 output from the control / business specification information extraction unit 121 rebuild the existing source program 100 in daily maintenance work (maintenance work). It is effectively used when

That is, when reconstructing the processing of the source program 100 by using a conventional language such as COBOL or a CASE tool, it is necessary to divide the source program 100 into a "control" portion and a "business specification" portion. It is effective in understanding the contents. For this,
The information list 126 is an effective means. The information file 125 also provides intermediate information for creating input data to the CASE tool.

With respect to a plurality of source programs 100 having a file or a transfer parameter as an interface, the plurality of source programs 100 can be referred to by referring to only the “business specification” part for the related record from each information list 126. It becomes easy to understand the entire business specification described.

Next, FIG. 24 shows the above information file 125.
3 is a schematic diagram illustrating a method of using the information list 126 for reconstructing the source program 100 in a daily maintenance work (maintenance work).

First, as shown on the left side of the figure, information regarding the "business specification" portion 1101 and the "computer control" portion 1102 of the existing source program 100 is extracted from the information file 125 and the information list 126. It

Of these extracted information, the information of the "computer control" portion 1102 is used as an input to the CASE tool 1300 for the downstream process. On the other hand, the information of the “business specification” part 1101 is the CASE tool 13
Used for input to 00. As a result, the source program 1400 is created via the CASE tool 1300.

That is, in general, a program is intricately intricately involved with a check / update process (business specification) for a record and a part related to computer control (corresponding to the above "control"). In this embodiment, by focusing on a record (data item group) having an interface with the outside and analyzing it, the above-mentioned “business specification” portion 1101 and “computer control” portion 1102 are clearly separated and extracted.

Most of the daily maintenance work (maintenance work) of the source program 100 is for the "business specification" part, and there are few maintenance requirements for "computer control". It is extremely effective for reducing the power consumption and improving the efficiency of system maintenance work.

Further, in the information list 126, the conditional expression in the statement of “business specification” is output in association with the “business term” indicating the specific processing content of the conditional expression, so the source program 100 It becomes very easy to understand the processing contents of. Further, in the information list 126, the analysis result of the source program 100 and the list of the source program 100 itself are output, so that the load of human organization and reconstruction work of the source program is reduced.

Here, the data section (DATA DIVIS
ION), communication function message area, communication function related area, database input / output area, and contact area (L
25 and 26 show an example of a source program 100 in which INKAGE SECTION) is defined.

In this source program 100,
The following statement is used for each of the above areas in the processing section (PROCEDURE DIVISION). Database access related commands (READY, etc.) Extracted as “control” Communication function related commands (OPEN I-O DSP-FILE, etc.) Extracted as “control” Syntax for checking the communication area (IF LK-01-01 =…) Extracted as “business specification” (check) Syntax that performs check processing for communication function message area (IF MENU-01 =…) Extracted as “business specification” (check) Communication function related area Check syntax for (IF DSP-DEST 1 = ...) Extracted as "control" Syntax for updating contact area (MOVE ZERO TO LK-01-01) "Business specification" Extract as (update) Syntax for updating the database input / output area (MOVE ZERO T DBREC01) Extracted as "business specification" (update) Syntax for updating communication function message area (MOVE ZERO TO MENU-REC-01) Extracted as "business specification" (update) Update processing for communication function related area Extracted as syntax (MOVE ZERO TO DSP-DEST 1) “business specifications” (update) As described above, file access related commands (OPEN, R)
(EAD, WRITE, etc.) and instructions and syntax other than the check processing and update processing for the data items in the record specified in the input / output file or the record name file 123 are the same as in the source program 100 shown in FIG. Then, the information file 125 and the information list 1
It is possible to create 26.

In the above embodiment, the information file 125 and the information list 126 are created and output with respect to the source program described in COBOL, but the present invention is not limited to this, and FORTRAN is not limited thereto. And PL
It is also applicable to the source program 100 written in another programming language such as / I. Also, the formats of the information file 125 and the information list 126 are not limited to those shown in FIGS. 22 and 23, and other formats may be used according to the input specifications of the CASE tool.

[0108]

As described above, according to the present invention,
A part that is likely to be modified in daily maintenance work and a part that is dependent on the OS (operating system) that is the main modification target when the system is updated are automatically separated and extracted from the source program. The load is reduced.

Further, for the conditional expression for which the check processing is being performed, the business term indicating the specific content of the check processing is also output, so that the content of the check processing can be easily determined.

Further, since the analysis result is automatically output on the list of source programs, the load of manual organization / reconstruction work of the source programs is reduced. In addition, information that helps input work to the CASE tool when rebuilding the source program is output.
Input to the ASE tool can be performed efficiently.

[Brief description of drawings]

FIG. 1 is a principle block diagram of the present invention.

FIG. 2 is a block diagram showing a system configuration of an embodiment of the present invention.

FIG. 3 is a diagram illustrating a condition name condition.

FIG. 4 is a diagram showing an example of a source program analyzed according to the present embodiment.

FIG. 5 is a diagram showing a format of data item information extracted by a data item information extraction unit.

FIG. 6 is a flowchart illustrating an operation of a simple conditional expression extracting unit.

FIG. 7 is a flowchart illustrating details of conditional expression division output processing S15.

FIG. 8 is a diagram showing an extraction format of a simple conditional expression extracted by a simple condition extracting unit.

FIG. 9 is a diagram (part 1) explaining a method of extracting a hierarchical conditional expression.

FIG. 10 is a diagram (part 2) explaining a method of extracting a hierarchical conditional expression.

FIG. 11 is a flowchart illustrating an operation of a composite condition extraction unit.

FIG. 12 is a diagram showing an extraction format of a composite conditional expression extracted by a composite condition extracting unit.

FIG. 13 is a flowchart illustrating an operation of a simple conditional expression aggregating unit.

FIG. 14 is a diagram showing a format of aggregated information of simple conditional expressions aggregated by a simple conditional expression aggregating unit.

FIG. 15 is a diagram illustrating a method of collecting data items with subscripts.

FIG. 16 is a flowchart illustrating an operation of a complex conditional expression aggregating unit.

FIG. 17 is a diagram showing a format of aggregated information of complex conditional expressions aggregated by a complex conditional expression aggregating unit.

FIG. 18 is a diagram showing the contents of a business term dictionary created via a business term input unit.

FIG. 19 is a diagram showing an example of a record name file.

FIG. 20 is a flowchart (part 1) explaining the operation of the control / business specification information extraction unit.

FIG. 21 is a flowchart (part 2) explaining the operation of the control / business specification information extraction unit.

FIG. 22 is a diagram showing the contents of an information file.

FIG. 23 is a diagram showing the contents of an information list.

FIG. 24 is a diagram illustrating a method of using an information file and an information list for a CASE tool.

FIG. 25 is a diagram (No. 1) showing another example of the analysis-target source program of the present embodiment.

FIG. 26 is a diagram (No. 2) showing another example of the analysis-target source program of the present embodiment.

[Explanation of symbols]

1 source program 2 Data item information extraction means 3 Control / business specification extraction means 4 Condition extraction means 5 Business term input means 6 record name file

─────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A 5-40616 (JP, A) Masahiro Nakamura, Challenge to escape from maintenance hell Part 3: Maintenance tool trends, Nikkei Computer, Nikkei BP, 1990 1990 30th, p 82-89 (58) Fields investigated (Int.Cl. ⁷ , DB name) G06F 9/06

Claims (7)

(57) [Claims] 1. A data item information extracting means for extracting data item information defined in the source program of a program executed by a computer system, and a data item extracted by the data item information extracting means. Based on the information, there is a control / business specification extraction means for extracting / outputting information on control statements and statement information on business specifications from the source program, wherein the control / business specification extraction means is the data Among the data items extracted by the item information extraction means, the statement that performs the input check process or the setting process at the time of output for at least the data item having the interface with other programs and external peripheral devices is related to the business specification. Extract as statement and at least file access The apparatus for analyzing task specifications source program and extracting a statement that performs communication processing as a statement relating to the control. 2. A data item information extracting means for extracting data item information defined in the source program of a program executed by a computer system, and an external interface, but extracted as a business specification. A record name file that stores the name of the record to which the data item to be stored is stored, and control is performed from the source program based on the data item information extracted by the data item information extraction means and the record name stored in the record name file. and a control / business specifications extracting means for extracting and outputting the information of the statement relating to the statement of information and operational specifications related to the
The control / business specification extraction means has the data item information extraction means.
At least one of the data items extracted by the output means
Statements that are performing file access-related processing
Extract as a statement related to the control,
Records stored in the record name file
Process for the data item belonging to the record of the first name
Or the statement that is performing the update processing
A business specification analysis device for source programs, which is characterized by being extracted as statements related to specifications. 3. A data item information extracting means for extracting data item information defined in the source program of a program executed by a computer system, and a conditional expression from a statement in the source program. Then, the same conditional expression is aggregated by a condition extracting means, a business term inputting means for attaching a business term corresponding to the business to the condition extracted by the condition extracting means, and the data item information extracting means. Based on the extracted data item information, a statement related to control and a statement related to business specifications are extracted from the source program, and among the statements related to the business specifications, for statements using conditional expressions, Correlation with business terms associated by the business term input means And a control / business specifications extracting means for outputting Te, the control / operational specification extracting means, said data item information extraction
At least other of the data items extracted by the output means
Has an interface with programs and external peripheral devices
Input check processing or output setting for data items
The statement that is performing regular processing is related to the business specification
At least the file extracted as
The statement that performs access related processing is
A business specification analysis device for a source program, which is characterized by extracting as a statement related to control. 4. A data item information extracting means for extracting data item information defined in the source program of a program executed by a computer system, and a conditional expression from a statement in the source program. For the same conditional expression, the condition extracting means for aggregating, the business term inputting means for attaching the corresponding business term to the conditional expression extracted by the condition extracting means, and the interface having no external interface are provided as the business specification. A record name file that stores the name of the record to which the data item to be extracted belongs, and control from the source program based on the data item information extracted by the data item information extracting means and the record name stored in the record name file Statements related to business and state statements related to business specifications Among the statements relating to the business specification, the statement using the conditional expression extracted by the conditional expression extracting means is related to the business term attached by the business term inputting means. And a control / business specification extraction means for outputting the data item information.
At least one of the data items extracted by the output means
Statements that are performing file access-related processing
Extract as a statement related to the control,
Records stored in the record name file
Process for the data item belonging to the record of the first name
Or the statement that is performing the update processing
A business specification analysis device for source programs, which is characterized by being extracted as statements related to specifications. 5. The data item information extracting means extracts, for each data item defined in the source program, the name, the affiliation record name, and the intended use thereof. , 3, or 4, the business specification analysis device for the source program. 6. The control / business specification extraction means outputs the affiliation record name of a data item used in the statement in addition to the extracted statement information. Business specification analysis device for the listed source program. 7. The control / business specification extraction means outputs a list of the extracted statements while clearly specifying a record name of a data item used in the statement. Business specification analysis device for source programs.