JP2018163432A - Automatic correction device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an automatic correction device that automatically corrects a model so that a dead code is not generated in model base development.SOLUTION: An automatic correction device of a model in model base development comprises a dead code detection processing unit and a model automatic correction processing unit. The dead code detection processing unit detects a point that generates a dead code from an uncorrected model before correction based on a dead code pattern which is a description pattern of a model that generates a dead code. The model automatic correction processing unit corrects a point that generates the dead code detected by the dead code detection processing unit, and generates a model after correction. The model includes a plurality of blocks which perform signal processing, and connections which couple between blocks. The model automatic correction processing unit deletes a block of a point that generates a dead code from a model before correction, and generates a model after correction by reconnecting between blocks of a disconnection point occurred by deletion.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a model automatic correction apparatus in model-based development.

  In recent years, in program development, there is a “model-based development” method in which a developer develops a program using a “model” that is easy to create and visually expressed, rather than describing the source code of the program. It is taken. A developer generates a source code by using a workstation, a personal computer, or the like having a program development environment corresponding to this model. A program having a function of generating source code based on a model is called a code generation tool, and is incorporated in a model development environment.

  In model-based development, quality evaluation is performed to evaluate the quality of a model and source code as a program product in a development process based on a predetermined standard. An example of the quality evaluation standard is coverage.

  Coverage is an index indicating how many of the branch destinations in the conditional branch of the model and source code are not executed (not reached). The coverage of the model and the source code is specified by performing an execution simulation of the model and the source using the test input pattern and investigating the branch destination that has been executed and the branch destination that has not been executed. A process for specifying the coverage is called a coverage test. As a result of the coverage test, a conditional branch having a branch destination that is not executed is extracted as a code that is not executed, that is, a dead code.

  Patent Document 1 discloses a technique for extracting a code (dead code) that is not executed as a result of a coverage investigation on a source code generated from a model, and displaying a correspondence relationship between the model of the portion that becomes the dead code and the source code on a display. It is disclosed.

JP 2005-301568 A

  In Patent Document 1, it takes time because a developer manually corrects a model of a portion that becomes a dead code. There is also the risk of human error.

  In addition, examples of dead code include executable statements under conditions that do not hold, but if such dead code is included in the source code, the number of lines in the program increases, and the program is executed. Therefore, it is necessary to increase the storage capacity of the memory for temporarily storing the program. In addition, there is a problem that readability is reduced as the number of program lines increases.

  Furthermore, since the source code including dead code includes a possibility that the software operates unexpectedly depending on the value of a parameter input to the source code, there are cases where the code is restricted or prohibited by standards or guidelines. For example, in the case of conforming to “ASIL D” in ASIL (Automotive Safety Integrity Level), which is an index for evaluating hazards in automotive functional safety standard ISO26262, source code including dead code is prohibited. Thus, when a dead code is included in a source code, it becomes an obstacle at the time of showing that the said source code is based on a standard.

  The present invention has been made to solve the above-described problems, and an object thereof is to provide an automatic correction device that automatically corrects a model so that dead code is not generated in model-based development.

  An aspect of an automatic correction device according to the present invention is a model automatic correction device in model-based development, in which a dead code is converted from an uncorrected model based on a dead code pattern which is a description pattern of a model for generating a dead code. A dead code detection processing unit for detecting a location for generating a model automatic correction processing unit for generating a corrected model by correcting the location for generating the dead code detected by the dead code detection processing unit, The model has a plurality of blocks for performing signal processing and a connection connecting the blocks, and the model automatic correction processing unit deletes the block at a location where the dead code is generated from the pre-correction model, The corrected model is generated by re-connecting between the blocks at the disconnection location generated by the deletion.

  According to the automatic correction apparatus of the present invention, it is possible to create a model in which dead code is not generated before generating source code.

It is a functional block diagram which shows the structure of the automatic correction apparatus of embodiment which concerns on this invention. It is a block diagram which shows the structure in the case of implement | achieving the automatic correction apparatus of embodiment which concerns on this invention with a personal computer. It is a figure which shows an example of the model produced by the model editor. It is a figure which shows the pattern of a dead code pattern and a model after correction. It is a flowchart which shows operation | movement of a dead code detection process part. It is a figure which shows dead code pattern information. It is a figure which shows a test object block. It is a figure which shows a test object block. It is a figure which shows a test object block. It is a figure which shows the example of a storage of a dead code detection result. It is a flowchart which shows operation | movement of a model automatic correction process part. It is a figure which shows the example of a comment output to a model after automatic correction.

<Embodiment>
<Device configuration>
FIG. 1 is a functional block diagram showing a configuration of an automatic correction apparatus 100 according to an embodiment of the present invention. As shown in FIG. 1, the automatic correction apparatus 100 includes a processing unit 10 and a storage unit 20. Note that an input device such as a keyboard and a mouse for operating the automatic correction device 100 by a user is connected to the automatic correction device 100, but the illustration is omitted.

  The processing unit 10 includes a dead code detection processing unit 101 and a model automatic correction processing unit 102, and the storage unit 20 stores a dead code pattern that is a description pattern of a model in which dead code is generated when generating source code. A code pattern information storage unit 201; an uncorrected model storage unit 202 that stores an uncorrected model before correction so that a dead code pattern is not generated; a dead code detection result storage unit 203 that stores a dead code detection result; It has a corrected model storage unit 204 that stores a corrected model that has been corrected so that a dead code pattern is not generated.

  The dead code detection processing unit 101 reads the dead code pattern stored in the dead code pattern information storage unit 201 and the pre-correction model stored in the pre-correction model storage unit 202, and sets the dead code pattern in the read pre-correction model. A matching part, that is, a model description part that generates a dead code is detected. The detected model description location is stored in the dead code detection result storage unit as a detection result.

  The model automatic correction processing unit 102 reads out the pre-correction model from the pre-correction model storage unit 202, reads out the detection result from the dead code detection result storage unit 203, and prevents the dead code from being generated at the model description location where the dead code is generated. A model after correction is generated by automatically correcting the model before correction. The corrected model is stored in the corrected model storage unit 204.

  A model in model-based development is composed of a plurality of blocks expressing output of a signal based on a predetermined rule, a connection expressing a connection of signal input / output between the blocks, and the like. Such a model is created by using a model editor.

  As the model editor, for example, software called Simulink (registered trademark) operating on Matlab (registered trademark) can be used. Although the model editor is not limited to this, in the following description, Simlink (registered trademark) operating on Matlab (registered trademark) will be described as an example of the model editor.

  FIG. 2 is a block diagram showing a configuration in the case where the automatic correction apparatus 100 shown in FIG. As shown in FIG. 2, the personal computer 1 includes an input device 11, a processor 12, a storage device 13, a first memory 14, and a second memory 15. Although a display device such as a liquid crystal display is connected to the personal computer 1 as an output device, illustration is omitted.

  The input device 11 includes a keyboard, a mouse, and the like, and outputs a signal corresponding to the operation to the processor 12 when operated by the user.

  The first memory 14 is a volatile memory capable of reading and writing such as a RAM (Random Access Memory), and the second memory 15 is a read-only nonvolatile memory such as a ROM (Read Only Memory). . The storage device 13 is a storage medium that requires a drive device such as a hard disk, a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, and a DVD (Digital Versatile Disc). The first memory 14 may be a flash memory, an EPROM (Erasable Programmable Read Only Memory), an EEPROM (Electrically Erasable Programmable Read Only Memory), or the like.

  In the storage device 13 and the second memory 15, software such as a program read and executed by the processor 12 is stored in advance. The storage device 13 stores a model, model data information, and model description pattern information for generating a dead code. The first memory 14 is used as a storage area for temporarily storing the program when the processor 12 executes the program stored in the storage device 13 and the second memory 15.

  When the personal computer 1 is started by turning on the power, the processor 12 reads out and executes a predetermined boot program from the second memory 15, an operating system (hereinafter abbreviated as OS) defined by the boot program, and others Is read from the storage device 13 and executed to start the personal computer 1. The processor 12 executes various programs stored in the storage device 13 as processes on the OS based on a signal from the input device 11, a schedule predetermined by the OS, etc. until the power is turned off after the startup process. To do. In the activation process and process, the processor 12 receives a signal input from the input device 11 as necessary, and controls the reading and writing of data to the first memory 14 and the storage device 13.

  The dead code detection processing unit 101 and the model automatic correction processing unit 102 in the processing unit 10 of the automatic correction device 100 shown in FIG. 1 are realized by executing a program by the processor 12, and the user performs a predetermined operation on the input device 11. To begin execution, or as directed by other methods.

  In addition, the dead code pattern information storage unit 201, the pre-correction model storage unit 202, the dead code detection result storage unit 203, and the post-correction model storage unit 204 in the storage unit 20 of the automatic correction device 100 illustrated in FIG. It is a part of 13 storage areas.

<Example of model>
FIG. 3 is a diagram illustrating an example of a model created by the model editor. As shown in FIG. 3, rectangular or oval elements denoted by reference numerals 301, 302, 303, 304, and 305 are blocks constituting the model. The block has either or both of an input end and an output end, and performs predetermined signal processing. This signal processing is based on a signal received at the input end if the block has an input end. If the block has an output end, the result of signal processing performed in the block is output from the output end.

  Moreover, the arrow line which connects between blocks represents a connection (signal line). The connection is set so that the output end of the block is the start point and the input end of the block is the end point. The starting point and the ending point of the connection refer to an end point on the side without an arrow and an end point on the side with an arrow, respectively, representing the connection.

  Here, the connection by connection between the output end of a certain block A and the input end of a certain block B indicates that the output from the output end of the block A is input at the input end of the block B. That is, the connection represents the input / output relationship of signals between blocks.

  A block 301 in FIG. 3 is a constant block that outputs a constant (Param1) that does not change with time. The specific value of the constant can be set by the developer for each constant block. FIG. 3 shows an example in which Param1 = 50 is set.

  Further, the block 302 in FIG. 3 outputs the input value as it is if the input value is within a predetermined range, and if it is outside the predetermined range, a predetermined upper limit value or lower limit value is output. This is an output value restriction block that outputs a value. In FIG. 3, since the upper limit value of the block 302 is set to 500 and the lower limit value is set to 0, when the constant (Param1 = 50) output from the block 301 is input, the value is output as it is.

  Also, a block 303 in FIG. 3 is a block that performs four arithmetic operations. A block 303 shown in FIG. 3 adds the input values received at the two input terminals and outputs the result.

  A block 304 in FIG. 3 is an input block that receives data input from the outside of the model and outputs the data.

  A block 305 in FIG. 3 is an output block that outputs the input data from the block 303 received through the connection to the outside of the model.

  In the model of FIG. 3 configured by combining blocks 301 to 305 and connection, the addition result of the data of block 301 and the data input to block 304 is output to block 305. At this time, if the constant (Param1) of the block 301 is within the limit range of the block 302, the input value is output as it is, and if the constant (Param1) is equal to or greater than the upper limit, the upper limit (500) is output. If the value is less than or equal to the lower limit value, the lower limit value (0) is output and added to the data input to the block 304 in the block 303.

  In this manner, the model is configured by connecting a plurality of blocks by connection, and expresses data input / output with the outside by the connection structure. However, even in the model shown in FIG. 3, a dead code is generated depending on conditions. In the automatic correction device 100 according to the present embodiment, the model automatic correction processing unit 102 reads out the detection result from the dead code detection result storage unit 203 and automatically corrects the model description portion that becomes the dead code so that the dead code is not generated. Since the model after correction is generated, source code that does not include dead code can be obtained.

<Detection and correction of dead code>
FIG. 4 is a diagram illustrating a model pattern (dead code pattern) in which a dead code is generated at the time of generating a source code and a corrected model pattern corrected so that a dead code is not generated. FIG. 4 is represented as a 3 × 6 table.

  In FIG. 4, the first column at the left end indicates the pattern number, and the item name is “No.”. The second column describes a dead code pattern in which a dead code is generated, and the item name is “pattern”. The third column shows the block type of the dead code pattern in which the dead code is generated, and the item name is “block to be inspected”. It is the first condition for pattern determination that these match. The fourth column shows a model corresponding to a dead code pattern in which a dead code is generated, and the item name is “pre-correction model”. The fifth column shows the source code that may be actually generated, and the item name is “generated code”. The sixth column at the right end illustrates the model after automatic correction and the correction method, and the item name is “correction proposal”.

  Here, in the patterns of No. 1-1 and No. 1-2 in FIG. 4, the input value of the output range restriction process is a fixed value, and the input value is within the range between the lower limit value and the upper limit value (output It is a pattern that satisfies the two conditions of being within the limit range. Therefore, in any pattern, the inspection target block is a block that limits the output value.

  The pattern of No. 2 in FIG. 4 is a description pattern of determination processing in which an input value from one of two inputs is an output value based on a control input value as a determination condition, and the control input value is a fixed value. It is a pattern that satisfies the two conditions that the input block does not exist on the side where the condition is not satisfied (not selected) of the two inputs. Therefore, the inspection target block is a block for determining an output value according to a condition.

<No. 1-1 Pattern Model>
The model of the pattern No. 1-1 in FIG. 4 is the same as the model shown in FIG. 3, and the same components are denoted by the same reference numerals and redundant description is omitted. In the No. 1-1 pattern model in FIG. 4, since the input value of the block 302 is a constant (Param1 = 50) output by the block 301, the value is always within a predetermined range (0 to 500). It becomes. Therefore, in block 302, it is determined whether or not the input value is within the predetermined range. If the input value is outside the predetermined range, the output range limiting process for outputting the predetermined upper limit value or lower limit value is performed. The source code that is not executed and is generated from such a model includes dead code. Specifically, the 4th to 10th lines surrounded by a broken line of the source code shown in the item “generated code” are dead codes.

  As described above, in the model of the pattern No. 1-1 in FIG. 4, it is not necessary to limit the output range in the block 302, and there is no problem even if the block 302 that limits the output value is deleted. Therefore, by deleting the block that limits the output value by automatic correction, and reconnecting the part that was disconnected by the deletion, the block 302 is deleted as shown in the item “correction proposal” in FIG. An automatically corrected model in which the output end of the block 301 is connected to one input end of 303 is obtained.

  Since the dead code is not generated by deleting the process for limiting the output range, the dead code is not extracted from the coverage evaluation result, and the coverage is improved.

<No. 1-2 pattern model>
4 is a constant block that outputs a constant (Param1) that does not change with time, and FIG. 4 shows an example in which Param1 = 50 is set.

  A block 403 is a constant block that outputs a constant (Param2) that does not change with time. FIG. 4 shows an example in which Param2 = 200 is set.

  A block 405 is an input block that receives data input from outside the model and outputs the data.

  The block 402 is a switch block that receives the output of the block 401 and the output of the block 403 as data inputs, and outputs one of the two data inputs according to the output value of the block 405 serving as a control input.

  The block 404 outputs the input value as it is if the input value is within a predetermined range, and outputs a predetermined upper limit value or lower limit value if the input value is outside the predetermined range. This is an output value limiting block. In FIG. 4, the upper limit value of the block 402 is set to 500, and the lower limit value is set to 0.

  A block 406 is an output block that outputs the input data from the block 404 received through the connection to the outside of the model.

  In the model of the pattern No. 1-2 in FIG. 4, when the control input to the block 402 is other than 0, a constant (Param1 = 50) input from the block 401 is output, and the control input to the block 402 is In the case of 0, the constant (Param1 = 200) input from the block 403 is output. Therefore, since the value output from the block 402 is 50 or 200, the input value of the block 404 is always a value within a predetermined range (0 to 500). Therefore, in block 404, it is determined whether or not the input value is within the predetermined range. If the input value is outside the predetermined range, the output range limiting process for outputting the predetermined upper limit value or lower limit value is performed. The source code that is not executed and is generated from such a model includes dead code. Specifically, the 10th to 16th lines surrounded by a broken line of the source code shown in the item “generated code” are dead codes.

  As described above, in the model of the pattern No. 1-2 in FIG. 4, it is not necessary to limit the output range in the block 404, and there is no problem even if the block 404 that limits the output value is deleted. Therefore, the block that restricts the output value is deleted by automatic correction, and the part disconnected by the deletion is reconnected, so that the block 403 is deleted and the block 403 is deleted as shown in the item “correction proposal” in FIG. A model after automatic correction in which the input end of the block 406 is connected to the output end of 402 is obtained.

  Since the dead code is not generated by deleting the process for limiting the output range, the dead code is not extracted from the coverage evaluation result, and the coverage is improved.

<No.2 pattern model>
The model block 501 of the pattern No. 2 in FIG. 4 is an input block that receives data input from the outside of the model and outputs the data.

  A block 502 is a constant block that outputs a constant (Param1) that does not change with time. FIG. 4 shows an example in which Param1 = 50 is set.

  The block 503 is a constant block that outputs a constant (Param0) that does not change with time. FIG. 4 shows an example in which Param0 = 0 is set.

  A block 504 is an absolute value block that outputs the absolute value of the input.

  The block 505 is a switch block that receives the output of the block 504 and the output of the block 503 as data inputs, and outputs one of the two data inputs based on the output value of the block 502 serving as a control input.

  A block 506 is an output block that outputs data received through the connection to the outside of the model.

  In the model of the pattern No. 2 in FIG. 4, the control input of the block 505 is the output of the block 502, and the block 502 outputs a constant (Param1 = 50). Therefore, the control input of the block 505 has Param1 = 50, That is, since a value other than 0 is given, the block 505 always outputs the input from the block 504. For this reason, when the control input to the block 505 is 0, the process of outputting the constant (Param0 = 0) input from the block 503 is not executed, and the dead code is included in the source code generated from such a model. Will be included. Specifically, the 7th to 9th lines surrounded by a broken line of the source code shown in the item “generated code” are dead codes.

  As described above, in the model of the pattern No. 2 in FIG. 4, it is not necessary to determine the output value according to the condition in the block 505, and the block 503 is connected to an input block that receives data input from outside the model. Not. Accordingly, there is no problem even if the block 505 for determining the output value, the block 502 for outputting the determination condition, and the block 503 for outputting the data input on which the condition is not satisfied are deleted. Therefore, by automatic correction, by deleting the block that determines the output value according to the condition, the block that outputs the determination condition, and the block that outputs the data input on the side where the condition is not satisfied, and reconnecting the location disconnected by the deletion, As shown in the item “Correction proposal” in FIG. 4, the blocks 502, 503, and 505 are deleted, and an automatically corrected model in which the input end of the block 506 is connected to the output end of the block 504 is obtained.

  Since the dead code is not generated by deleting the process for determining the output value, the dead code is not extracted from the coverage evaluation result, and the coverage is improved.

<Dead code detection processing>
Next, the operation of the dead code detection processing unit 101 will be described with reference to FIGS. 1 and 2 and the flowchart shown in FIG. When the dead code detection process is started by operating the input device (not shown), first, the dead code pattern information stored in the dead code pattern information storage unit 201 is read (step S1). The dead code pattern information is information on a pattern for generating a dead code, and an example thereof is shown in FIG.

  The dead code pattern information shown in FIG. 6 is set for two types of dead code patterns, and is represented as a 2 × 5 table.

  In FIG. 6, the first column at the left end indicates the classification of the inspection target block. In FIG. 6, the first column is classified as Type 1 and Type 2, and the item name is “Type of inspection target block”. The second column defines the connection source block to which the inspection target block is connected. In FIG. 6, only Type 3 is defined, and the item name is “type of connection source block”. The third column shows the port number when acquiring the connection source block, and the item name is “connection source port”. The fourth column defines a first condition for a dead code pattern, and the item name is “condition 1”. The fifth column defines the second condition for the dead code pattern, and the item name is “condition 2”.

  Here, FIGS. 7 and 8 show blocks to be inspected corresponding to Type 1 and Type 2, respectively, where Type 1 is a block for limiting the output value, and Type 2 is a block for determining the output value. FIG. 9 shows Type 3 connection source blocks. The Type 3 connection source blocks are constant blocks that output constants that do not change with time.

  Next, the dead code detection processing unit 101 reads out the pre-correction model stored in the pre-correction model storage unit 202 (step S2). The pre-correction model is, for example, a model described in the item “Pre-correction model” in FIG.

  Next, the dead code detection processing unit 101 detects, from the pre-correction model, a block whose type of the inspection target block corresponds to TypeX (step S3). For example, when the inspection target block is Type 1, if the same block is included in the pre-correction model, it is detected. Since the inspection target block of Type 1 is a block that limits the output value, all blocks that limit the output value included in the pre-correction model are detected. Here, the detected block is referred to as a detected block. If the pre-correction model does not include a block corresponding to the inspection target block, the series of processing ends.

  The inspection target block is, for example, a block described in the item “inspection target block” in FIG. 4, and two types are assumed in the present embodiment.

  Here, the block to be inspected is classified as Type 1 and Type 2 in the item “Type of block to be inspected” in FIG. 6, and the processes in and after step S3 are repeated for both Type 1 and Type 2.

  Next, the dead code detection processing unit 101 determines whether or not the detected block detected in step S3 satisfies the first condition (condition 1) that becomes a dead code pattern (step S4). Here, Condition 1 is defined in the item of “Condition 1” in FIG. 6. For the inspection target block of Type 1, the inspection target block is connected to “the type of the connection source block matches Type 3. The condition is “Yes”.

  And when it determines with not satisfy | filling the conditions 1 (No), the process of step S4 is performed with respect to another detected block. If only one detected block is detected in step S3, the process waits until a block corresponding to the next inspection target block is detected in step S3. On the other hand, when it determines with satisfy | filling the conditions 1 (Yes), it transfers to step S5. The case where the condition 1 is satisfied is a case where the inspection target block is connected to a constant block that outputs a constant that does not change with time, and a case where a fixed value (constant) is input to the inspection target block. .

  In step S5, it is determined whether or not the detected block detected in step S3 satisfies a second condition (condition 2) that becomes a dead code pattern.

  Here, condition 2 is defined by the item “condition 2” in FIG. 6, and for the type 1 inspection target block, the input value of the inspection target block is within the range between the lower limit value and the upper limit value. Is a condition that holds. In FIG. 6, the condition is expressed as “lower limit value <output value from connection source <upper limit value is satisfied”. Condition 2 is a condition that “the connection source block on the side where the condition is not satisfied matches Type 3” for the Type 2 inspection target block.

  And when it determines with not satisfy | filling the conditions 2 (No), the process below step S4 is performed with respect to another detected block. If only one detected block is detected in step S3, the process waits until a block corresponding to the next inspection target block is detected in step S3. On the other hand, when it determines with satisfy | filling the conditions 2 (Yes), it transfers to step S6. In the case where the condition 2 is satisfied, for the Type 1 inspection target block, the inspection target block is an output value restriction block, and the output value from the connection source block is within the output restriction range of the inspection target block. This is the case. In addition, for the Type 2 inspection target block, the inspection target block is a block that determines an output value according to a condition, and the connection source block on which the condition is not satisfied outputs a constant that does not change with time. Is the case.

  As a result of the determination in steps S4 and S5, when the detected block detected in step S3 satisfies the condition 1 and the condition 2 that become a dead code pattern, it is determined that the block is a block that generates a dead code pattern. The block is stored in the dead code detection result storage unit 203 as a dead code detection result (step S6).

  FIG. 10 is a diagram illustrating a storage example of the dead code detection result. FIG. 10 shows a storage example for three types of patterns, which are represented as a table of 3 rows and 5 columns.

  10, the first column at the left end shows the pattern number of the pre-correction model including the block in which the dead code is generated, and in FIG. 10, No. 1-1, No. 1-2, and No. 2 are shown. The item name is “No.”. This pre-correction model corresponds to the model read in step S2 in FIG. 5 and represents that a dead code is detected from the patterns of No. 1-1, No. 1-2, and No. 2.

  The second column shows the classification of the blocks to be inspected. In FIG. 10, No. 1-1 and No. 1-2 are both Type 1 and No. 2 is classified as Type 2. Block type ". The third column shows the path (absolute path) from the uppermost block of the model to the deletion target block, and the item name is “deletion target block path”. In the fourth column, the paths to the respective blocks on the input side and output side to be reconnected are shown in upper and lower stages, and the item name is “reconnection target block path”. The upper stage is a path to the input side block, and the lower stage is a path to the output side block. The fifth column specifies the pattern for deleting the inspection target block and reconnecting the disconnection location, and the pattern for deleting the block that becomes a dead code and reconnecting the disconnection location in addition to the inspection target block. The name is “deletion target option”. Note that a pattern for deleting only the inspection target block is “0”, and a pattern for deleting a plurality including the inspection target block is expressed by “1”.

  Returning to the flowchart of FIG. After the dead code detection result is stored in the dead code detection result storage unit 203 in step S6, it is determined whether or not the dead code detection processing in steps S4 to S6 has been executed for all the detected blocks detected in step S3. Determination is made (step S7).

  If it is determined that dead code detection processing has been executed for all detected blocks (Yes), the process proceeds to step S8, and it is determined that there are remaining detected blocks to be processed (No). Step S4 and subsequent steps are executed for the remaining detected blocks.

  In step S8, it is determined whether or not all inspection target blocks have been applied to the pre-correction model read in step S2. That is, as shown in FIG. 6, when there are two types of inspection target blocks, Type 1 and Type 2, the inspection target blocks of Type 1 and Type 2 are applied to the pre-correction model read in step S2. Check if it is determined whether the block is included or not.

  When it is determined that all the inspection target blocks have been applied to the pre-correction model (Yes), a series of processing ends, and when it is determined that the inspection target blocks to be applied remain (No) Repeats the processing from step S3.

<Automatic model correction processing>
Next, the operation of the model automatic correction processing unit 102 will be described using the flowchart shown in FIG. 11 with reference to FIGS. When the model automatic correction process is started by the user operating an input device (not shown), first, the dead code detection result stored in the dead code detection result storage unit 203 is read (step S11).

  Next, the model automatic correction processing unit 102 reads the pre-correction model stored in the pre-correction model storage unit 202 (step S12).

  Next, the model automatic correction processing unit 102 specifies a deletion target block (step S13). In the storage example of the dead code detection result shown in FIG. 10, the deletion target block is defined by information described in the item “deletion target block path”. That is, in the item “deletion target block path”, for example, in the case of the pattern No. 1-1, the deletion target is defined as “Sample1 / Target1”. The target block. “Target 1” defines a path from the top layer block of the model to the deletion target block.

  Next, the model automatic correction processing unit 102 determines a correction method (step S14). Specifically, it is determined whether or not the deletion target option is “0” by referring to the item “deletion target option” of the dead code detection result shown in FIG. 10, and in the case of “0” (Yes) Shifts to step S15, and if “1” (No), shifts to step S20. For example, in the case of the pattern No. 1-1, since the deletion target option is “0”, only the inspection target block is deleted in step S15.

  Next, the model automatic correction processing unit 102 reconnects a portion that is disconnected due to deletion of the deletion target block (step S16). The location of reconnection is defined by the information described in the item “reconnection target block path” of the dead code detection result shown in FIG. That is, in the item “reconnection target block path”, for example, in the case of the pattern No. 1-1, reconnection targets are defined as “Sample1 / Block1”, “Sample1 / Block2”, and “Target1”. Is a deleted block, “Block 1” indicates a path to a block on the input side of “Target 1”, and “Block 2” indicates a path to a block on the output side of “Target 1”. The model automatic correction processing unit 102 reconnects “Block 1” and “Block 2” based on this information. In the case of the pattern of No. 1-1, a model in which the block 302 is deleted and the output end of the block 301 is connected to one input end of the block 303 as shown in the item “Proposed amendment” in FIG. Become.

  Thereafter, the model automatic correction processing unit 102 outputs the fact that the model has been corrected as a comment in the corrected model, and leaves a history (step S17). FIG. 12 shows an example of a comment output in the model. As shown in FIG. 12, the comment describes the correction date, the corrector, and the details of the correction. Thereby, a user's convenience can be aimed at.

  Thereafter, in step S18, the model automatic correction processing unit 102 determines whether or not all deletion target blocks included in the dead code detection result have been corrected, and when all deletion target blocks have been corrected (Yes). In step S19, the corrected model is stored in the corrected model storage unit 204, and the series of processing ends. If there is an uncorrected deletion target block (No), the processing in step S13 and subsequent steps is repeated.

  If the deletion target option is “1” as a result of determining the correction method in step S14, since there are a plurality of deletion target blocks other than the inspection target block, all of them are deleted. For example, in the case of the pattern of No. 2, the “deletion target block path” item specifies deletion targets such as “Sample3 / Target3”, “Sample3 / Target4”, and “Sample3 / Target5”. With reference to “Sample3”, the block is a deletion target block. Further, since there are three absolute paths to the deletion target block such as “Target3”, “Target4”, and “Target5”, it indicates that the deletion target block is other than the inspection target block.

  In step S20, first, “Target3” that is a deletion target block is deleted. Next, in step S21, it is determined whether all deletion target blocks in the “deletion target block path” item have been deleted. When the deletion target block is deleted (Yes), the disconnection portion is deleted by deleting the deletion target block in step S22. In the item “block path for reconnection” in FIG. 10, for example, in the case of the pattern No. 2, reconnection targets are defined as “Sample3 / Block5” and “Sample3 / Block6”, and “Target3” In the deleted block, “Block 5” indicates a path to the block on the input side of “Target 3”, and “Block 6” indicates a path to the block on the output side of “Target 3”. The model automatic correction processing unit 102 reconnects between “Block 5” and “Block 6” based on this information.

  On the other hand, if it is determined in step S21 that there is an undeleted block to be deleted (No), the processing in step S20 and subsequent steps is repeated. Here, “Target4” and “Target5” in the “deletion target block path” field indicate that blocks other than “Target3” are included, but these blocks are “reconnection target block paths”. In the item “”, since it is not listed as a block to be reconnected, it is a deletion target in step S20. In the case of the pattern No. 2, as shown in the item “Proposed amendment” in FIG. Become.

  In the present invention, the embodiments can be appropriately modified and omitted within the scope of the invention.

  101 dead code detection processing unit, 102 model automatic correction processing unit.

Claims (4)

An automatic model correction device in model-based development,
A dead code detection processing unit that detects a location where a dead code is generated from an uncorrected pre-correction model based on a dead code pattern that is a description pattern of a model that generates a dead code;
A model automatic correction processing unit that generates a corrected model by correcting the location for generating the dead code detected by the dead code detection processing unit, and
The model includes a plurality of blocks that perform signal processing and a connection that connects the blocks.
The model automatic correction processing unit
An automatic correction device that generates the post-correction model by deleting the block at a location where the dead code is generated from the pre-correction model, and reconnecting between the blocks at the disconnection location generated by the deletion. The dead code pattern is:
The automatic correction device according to claim 1, wherein the automatic correction apparatus includes a description pattern of output range restriction processing, wherein the input is a fixed value, and the input satisfies a condition that is within a range between a lower limit value and an upper limit value. The dead code pattern is:
Including a description pattern of a determination process in which an input value from one of two inputs is an output value based on a control input serving as a determination condition, the control input value is a fixed value, and the condition of the two inputs is The automatic correction device according to claim 1 or 2, wherein a pattern that satisfies a condition that no data is input from the outside of the model is provided on the side that is not established. The model automatic correction processing unit
The automatic correction device according to claim 1, wherein a comment indicating the correction content is output and a history is left in the corrected model.

Images