JP2003036170A - Device, method and program for cross reference construction - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a cross reference construction device capable of constructing precise cross reference. SOLUTION: The cross reference construction device is provided with a watched function designating means for designating a watched symbol to be a retrieving object, an inter-symbol relation analytic means for obtaining a symbol having a specific relation with this watched symbol, a alias analytic means for obtaining a symbol having a alias from that of the watched symbol and a connection means for connecting the retrieving results of the inter-symbol relation analytic means and the alias analytic means. Thus, the device is capable of not only showing a relation between symbols but also reflecting the result of the alias analysis when constructing the cross reference.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for analyzing a relation between symbols and constructing a cross reference for symbols such as variables and functions constituting a program, and in particular, by analyzing aliases of the symbols. The present invention relates to a cross reference construction device and a cross reference construction method with improved analysis accuracy.

[0002]

2. Description of the Related Art In an IT-based society in recent years, services provided by IT have become huge and complicated, and accompanying this, programs for realizing the IT-based society have also become huge and complicated. On the other hand, as many people and companies enter the IT industry globally, competition for program development is intensifying.

In such a situation, it is necessary to efficiently develop the program in a short period of time. One method for that purpose is to reuse a program created in the past or a program created by another person (other company) to develop a new program. When reusing an existing program in this way, there are a method of using the existing program as it is without any change, and a method of using the existing program with some change. It is easy if you can use an existing program without making any changes, but usually it is necessary to add a new function or change (customize) an existing function.

In order to change the existing program as described above, it is necessary to understand the program before the change. For that purpose, it is necessary to understand the program by referring to the specifications and design documents of the program. However, specifications and design documents are sometimes changed after the programs are written, and the description contents do not match the program, or the specifications and design documents are not written in so much detail. Sometimes, you need to read the program to understand it correctly.

However, as described above, the programs are becoming huge and complicated, and it is difficult to read all the programs. Therefore, it is very efficient if it is possible to focus on the points necessary for changing the program. In order to know the necessary parts to change the program, first of all, an outline of the whole structure is necessary, and it is necessary to find out where the part to be changed is. Next, it is necessary to confirm where the change will affect the change. To do so, analyze in advance what kind of symbols are described in the program, and the relationship between each symbol, and reuse the relationship between the symbols in the program. There is a need for an environment that can be presented to the trying user.

[0006]

As a technique for analyzing the relationship between symbols and constructing a cross reference as described above, conventionally, a program is analyzed to extract symbols which are constituent elements of the program, Some have analyzed the relationships between the extracted symbols. Examples of such cross-references are a call graph / cold graph showing a relation by function call, a relation diagram of a function and a global variable showing a relation by writing and referencing a function and a global variable, data from a certain symbol to a certain symbol. There were data flow charts, etc. showing the relationship due to delivery.

However, in the conventional cross reference as described above, no consideration has been given to a so-called "alias" in which the same entity is given a different name in the relationship between the symbols. Therefore, in constructing the cross reference, there is a case where the relation between the symbols having an indirect relation via the pointer function, the pointer variable, the variable passed by reference, or the like is lost.

The present invention has been proposed in order to solve the above-mentioned problems of the prior art, and its purpose is to reflect the result of alias analysis when presenting the relationship between symbols. Accordingly, it is an object of the present invention to provide a cross reference construction device and a cross reference construction method capable of constructing a more accurate cross reference.

[0009]

In order to achieve the above object, the cross reference construction apparatus according to claim 1 is
Attention function designating means for designating a target symbol to be searched, and relationship analysis means between symbols for acquiring a symbol having a specific relationship with the target symbol,
The present invention is characterized by comprising an alias analysis means for obtaining a symbol of an alias of the symbol of interest and a connection means for connecting the relation analysis means between the symbols and the search result of the alias analysis means.

A cross reference construction method according to a third aspect of the present invention is obtained by considering the invention according to the first aspect from the viewpoint of the method, and an attention function designation process for designating a notable symbol to be searched. And a relationship analysis process between the symbols that acquires a symbol having a specific relationship with the symbol of interest, an alias analysis process that acquires a symbol of an alias of the symbol of interest, and a relationship analysis process between the symbols. The present invention is characterized by including a concatenation process for connecting the search results obtained by executing the alias analysis process.

A cross reference construction program according to a fifth aspect of the present invention captures the invention according to the third aspect from a software point of view, and specifies a target function for designating a target symbol to be searched. Steps,
Relationship analysis step between symbols for obtaining a symbol having a specific relationship with the symbol of interest, alias analysis step for acquiring a symbol of an alias of the symbol of interest, relationship analysis step between the symbols and alias analysis And a connecting step for connecting the search results obtained by executing the steps.

According to the inventions of claim 1, claim 3 and claim 5 having the above-mentioned structure, not only the relation between the symbol of interest to be searched and other symbols but also the symbol of interest In addition to directly related symbols, it is possible to cover not only directly related symbols but also indirectly related symbols. Can be built.

According to a second aspect of the present invention, in the cross-reference construction apparatus according to the first aspect, the alias analysis means generates an alias directed graph indicating which symbol a symbol may indicate the same entity. With reference to the alias directed graph creating unit to create and the alias directed graph, when the alias directed graph is traced in the positive direction starting from the symbol of interest, all the traceable symbols are registered as aliases, and as aliases Aliases with each registered symbol as a starting point, when the directed graph is traced in the negative direction, all traceable symbols are registered as aliases, and the registered symbols are "aliases" of the noted symbol. It is characterized by including a search unit.

The invention according to claim 4 is the same as claim 3
In the cross-reference construction method according to claim 1, the alias analysis processing refers to the alias directed graph creation step of creating an alias directed graph that represents a certain symbol and which symbol may refer to the same entity, and the alias directed graph. , When the alias directed graph is traced in the positive direction with the noted symbol as a starting point, all the traceable symbols are registered as aliases, and each of the symbols registered as aliases is used as a starting point and the alias directed graph is negative. All of the traceable symbols are registered as aliases when tracing the direction, and an alias search step is performed in which the registered symbols are “alias” of the noted symbol.

According to a sixth aspect of the present invention, in the cross-reference construction program according to the fifth aspect, the alias analysis step represents an alias directed graph indicating which symbol a symbol may indicate the same entity. Create an alias directed graph creating step, and referring to the alias directed graph, register all the symbols that can be traced as an alias when the alias directed graph is traced in the positive direction from the target symbol as a starting point. Starting from each of the symbols registered as, the alias can be traced in the negative direction when the directed graph is traced in the negative direction, and all registered symbols are registered as aliases, and the registered symbols are defined as "aliases" of the noted symbol. It is characterized by having an alias search step.

According to the inventions of claim 2, claim 4 and claim 6 having the above-mentioned configuration, the symbols having the aliases can be efficiently used for the symbol of interest.
Since it can be obtained with high accuracy, a cross reference with higher accuracy can be constructed.

[0017]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be specifically described below with reference to the drawings. The embodiment described below is an example of constructing a call graph as an example of constructing a cross reference.

The following embodiments are realized on a computer, and each function of the embodiments is realized by a predetermined procedure (program) controlling this computer.
Further, each means in this specification is a conceptual one corresponding to each function of the embodiment, and does not necessarily correspond to a specific hardware or software routine on a one-to-one basis. The same hardware element may in some cases constitute different means. For example, a computer may be one means for executing one instruction and another for executing another instruction. Further, one means may be realized by only one instruction or may be realized by a large number of instructions. Therefore, in the present specification, the embodiments will be described below assuming a virtual circuit block (means) having each function of the embodiments. In addition, the execution order of each step of each procedure in the present embodiment may be changed, a plurality of steps may be executed at the same time, or may be executed in a different order for each execution, as long as the property is not violated. Such an order change can be realized by a menu-type interface method, such as a user selecting an executable process.

(1) Configuration The cross reference construction apparatus 1 of the present embodiment has a configuration shown in FIG.
As shown in the functional block diagram of 1., the relationship acquisition unit 12 between symbols is provided. The relationship acquisition unit 12 between symbols sequentially reads the source files from the source file group 11 and performs lexical analysis, syntactic analysis,
By performing semantic analysis, symbols such as functions and variables are extracted from the source code, function declarations and function calls are found, and function call relationships such as "which function calls which function" are acquired. Then
The relationship between symbols is shown as FIG.

Further, the cross-reference construction apparatus 1 of this embodiment is provided with an alias directed graph creation unit 14. The alias directed graph creation unit 14 sequentially reads the source files from the source file group 11 and performs lexical analysis, syntactic analysis, and semantic analysis that are existing techniques to determine which symbol such as a function or a variable is a certain symbol. It is analyzed whether there is a possibility that it points to the same memory area as, and is represented as an alias directed graph 15. For example, when the sample program shown in FIG. 2 is represented by an alias directed graph, it becomes as shown in FIG. 3. With the sample program shown in FIG. 7, an “alias directed graph” as shown in FIG. 9 is obtained. However, in all examples, nodes without arcs are omitted.

As described above, in the alias directed graph 15, each symbol such as a function or a variable is represented by one node,
The possibility that one symbol may point to another symbol is expressed by an arc connecting nodes. This arc has a direction, and is "forward" when going from a potential symbol to a potential symbol.
Alternatively, it is called "positive direction" and the opposite is called "reverse direction" or "negative direction". In addition, as an example in which one symbol may point to another symbol,
As shown in FIG. 4, it means that there is a so-called “alias” relationship in which there are pointer assignments, assignments by reference, virtual functions, etc., and different names are given to the same entity.

In the cross-reference construction apparatus 1 of this embodiment, an attention function designating section 16 for designating the function at the head of the call graph to be displayed and the relationship diagram 13 between the symbols for the designated function are shown in FIG. Referring to a relation search unit 17 that searches for a relation between symbols, “which function calls which function”, and a symbol determined to have a function calling relation as a result of the search by the relation search unit 17. And a relationship registration unit 19 for registering as the relationship information data 18.

Further, in the cross-reference construction apparatus 1 of the present embodiment, for the function designated by the attention function designation unit 16, the alias directed graph 15 is referred to, and "the symbol designated as the search target is the symbol And the symbol determined to be an alias as a result of the search by this alias search unit 20 is registered as the alias information data 21. A registration unit 22 is provided.

Further, the cross-reference construction apparatus 1 of the present embodiment is provided with a connecting section 23 and an output section 24 for outputting the search result. Then, in the linking unit 23, the “function called by the function of interest” obtained by the relationship searching unit 17 is associated with the function of interest and linked, and the “attention is obtained” by the alias searching unit 20. The "alias of the function" is associated with and linked to the function of interest.

In the cross-reference construction apparatus 1 of this embodiment, after a series of search processing for a certain function is completed, the symbol of interest is sequentially replaced with the related symbol, and the processing described later is repeated, A call graph is built.

Further, the relationship acquisition unit 12 between the symbols,
The relationship diagram between symbols 13, the relationship retrieval unit 17, the relationship information data 18, and the relationship registration unit 19 constitute the “relationship analysis unit between symbols” described in the claims. The alias directed graph creation unit 14, the alias directed graph 15, and the alias Search unit 2
0, by the alias information data 21 and the alias registration unit 22,
The "aliasing analysis means" shown in the claims is configured. In the present embodiment, the alias search unit 20 searches the alias directed graph 15 in the positive direction (forward direction) and then
It is assumed that the search is set in the negative direction (reverse direction).

(2) Operation Next, the flow of call graph creation processing in the cross reference construction apparatus of the present embodiment having the above-mentioned configuration will be described with reference to the flow charts shown in FIGS.

(2-1) Call graph creation processing First, referring to the call graph construction flow chart shown in FIG. 5, the call graph creation procedure will be described with reference to the call graph of the sample program using the function pointer shown in FIG. Will be explained using an example of constructing. First, the user obtains the symbol of interest designated by the user via the attention function designation unit 16 (step 501) and adds this symbol to the call graph (step 502). Here, assuming that "main" is specified as the symbol of interest, "main" is added to the call graph. At first, nothing is registered in the call graph, so in this case, "main" is registered at the beginning of the call graph (upper left in FIG. 8). Then, step 503
In, it is determined whether or not there is a symbol that is not focused, and one symbol that is focused is selected (step 50).
4). Here, since it is the first time, "main" is selected as the symbol of interest.

Next, referring to FIG. 13 which shows the relationship between symbols, it is searched whether or not there is a symbol called by the noted symbol "main", and "* fpLocalFunc" is determined as the function called by "main". Is acquired (step 505). Therefore, this "* fpLocalFunc" is connected as a "calling symbol" below the symbol of interest (step 506). As a result, a call graph obtained by registering "* fpLocalFunc" under "main" is obtained (upper right of FIG. 8).

Next, in step 507, it is judged whether or not the symbol of interest has an alias, and if there is an alias, an alias search process (flowchart in FIG. 6) described later is performed. On the other hand, if the symbol of interest has no alias, the process returns to step 503. Here, since there is no alias for "main", the process proceeds to the next process.

That is, in step 503, it is determined by referring to the call graph whether or not there is a symbol that has not been paid attention to among the registered symbols. One symbol of interest is selected from among the symbols to be a new symbol of interest (step 504). Here, "* fpLocalFunc" registered under the "main" is the next symbol of interest.

[* F] which became the symbol of this new attention
For "pLocalFunc", it is judged whether or not there is a calling symbol (step 505).
Since there is no calling symbol in ocalFunc ”, the process directly proceeds to step 507 and it is determined whether or not there is an alias. Specifically, it is determined whether or not this“ * fpLocalFunc ”is registered in the alias directed graph 15. Here, "* fpLocalFunc" is shown in the alias directed graph 15 in FIG.
, The alias search processing (step 508) shown in FIG. 6 is performed.

(2-2) Alias Search Process If it is determined in step 507 of FIG. 5 that the symbol of interest has an alias, the alias search process shown in the flowchart of FIG. 6 is performed. That is, first, the alias information data 21 is made empty (step 601). Then, the symbol of interest is acquired as a search target symbol (step 602), and the symbol itself is registered as the alias information data 21 (step 603). Here, “* fpLocalFunc” is registered as the alias information data 21.

Next, one registered alias is selected from the alias information data 21 (step 604). Here, "* fpLocalFunc" is found as a symbol already registered as an alias. For the "* fpLocalFunc", see the alias directed graph (Fig. 9).
"* func_table" as a positive arc from "alFunc"
Therefore, "* func_table" is added as the alias information data 21 (steps 605 and 606).

On the other hand, if it is determined in step 605 that there is no arc in the positive direction, the process directly proceeds to step 607. In step 607, it is determined whether or not the forward search has been completed for all aliases registered as the alias information data 21, and if not completed, the process returns to step 604 and the alias information data 2
Select one of the unprocessed aliases from step 1, and then select step 605.
~ Repeat step 607. Here, since "* func_table" has already been added as an alias, this "* fuc_table"
nc_table "is selected (step 604), and it is determined whether or not there is an arc in the forward direction (step 605).

That is, referring to the alias directed graph (FIG. 9), since there are "funcA" and "funcB" as arcs in the more positive direction from "* func_table", these are added to the alias (steps 605 and 606). ). Since there is no more arc in the positive direction, the search in the positive direction ends there (step 607). at this point,
The symbol registered as the alias information data 21 is "* f
pLocalFunc ”,“ * func_table ”,“ funcA ”,“ func
B ".

In this way, when it is determined in step 607 that the forward search has been completed for all aliases registered as the alias information data 21,
Proceeding to step 608, a search in the reverse direction is performed. That is, one registered alias is selected from the alias information data 21. For this alias, it is judged whether or not there is an arc in the negative direction (reverse direction) from the symbol of the alias by referring to the alias directed graph (FIG. 9) (step 60).
9) If there is an arc in the negative direction, all the symbols of nodes at the tip of the arc (base end of the arc) are added as aliases to the alias information data 21 (step 610), and the process proceeds to step 611. On the other hand, when it is determined in step 609 that there is no negative arc, the process proceeds to step 611. Here, as shown in the alias directed graph (FIG. 9), there is no arc in the negative direction, so the process proceeds to step 611.

In step 611, it is judged whether or not the backward search has been completed for all aliases registered as the alias information data 21, and if not completed, the flow returns to step 608 to return the alias information data. One unprocessed alias is selected from 21 and steps 609 to 611 are repeated. On the other hand, when it is determined in step 611 that the backward search is completed for all the aliases registered as the alias information data 21, the alias search processing for the search target symbol is finished, and the alias search processing of FIG. Return to step 509.

Then, in step 509, the alias symbol is connected to the lower level of the symbol of interest in the call graph. Here, the symbol of interest is “* fpLoc
"alFunc" has aliases "* func_table" and "func
Since A "and" funcB "are obtained, these symbols are connected to the lower order of" * fpLocalFunc "(lower left of FIG. 8).

Next, returning to step 503, it is judged whether or not there are any symbols that are not of interest, in other words, whether or not all searches have been completed, and steps 504 to 509 are performed until there are no symbols of interest. The process of is repeated. When the call graph creation process and the alias search process as described above are all completed, the call graph with the alias information as shown in the lower right part of FIG. 8 is completed. In FIG. 8, in order to distinguish the “alias” from the “function call”, “≈” is added before the symbol name for the symbol registered by the alias.

(2-3) Specific Example of Alias Search In the above embodiment, since the negative search is not performed in the alias search processing, the alias directed graph shown in FIG. 3 is shown as an example in FIG. The flow of the alias search process will be described in more detail.

It is assumed that "p" is designated by the attention function designation section 16 (step 602). First, "p" itself is registered as an alias (step 603). Next, since the search is performed in the positive direction, the process proceeds to step 604, where "p" is found as a symbol already registered as an alias in the alias information data 21 (step 604).
Referring to the alias directed graph (FIG. 3) for the "p", since "a" is an arc having a positive direction from "p", "a" is added to the alias (steps 605 and 606).

Next, referring to the alias directed graph (FIG. 3), since there is no arc in the positive direction from "a", the search in the positive direction ends there. At this time, the symbols registered as the alias information data 21 are “p” and “a”. Next, the search is performed in the negative direction. Alias information data 21
, The symbol “p” is first extracted from the symbols “p” and “a” that have already been registered as aliases. Referring to the alias directed graph (FIG. 3), if the arc toward “p” proceeds in the negative direction, “r” is obtained, so “r” is added to the alias. Since there is no arc proceeding from “r” in the negative direction, the process ends.

Similarly, by referring to the alias information data 21, "a" is taken out as a symbol already registered as an alias. "P" when going from "a" to the arc in the negative direction
And "q", but since "p" has already been registered, only "q" is registered. Since there is no arc toward “q”, the alias search ends. As a result, "p, r, a, q" are obtained as aliases for "p". here,
“P, r, a, q” is a set in which the pointer may point to the same entity as “p”, but “p” is the acquired symbol itself, so it may be excluded from the aliases.

(3) Effects As described above, according to the cross-reference construction apparatus of this embodiment, not only the relation between the functions but also the relation between the symbols having an indirect relation by considering the aliases is covered. Therefore, a more accurate cross reference can be constructed.

(4) Other Embodiments The present invention is not limited to the above-described embodiments, but the number of layers can be limited when displaying a call graph. Further, in the completed call graph shown in the lower right of FIG. 8, by re-designating the “printf ()” hanging under the alias “funcA” as the leading symbol, the target symbol is acquired. Needless to say, the analysis can be performed again.

When the number of displayed layers is limited, there is a related layer below that, for example, printf
There are more related symbols under (), or the displayed symbol has an alias, for example printf
The fact that there is an alias in () can also be expressed using marks, colors, or other display methods. In that case, by clicking (instructing) the displayed symbol, it is possible to switch the display and display the lower hierarchy or the alias.

[0048]

As described above, according to the present invention, a cross-reference can be constructed with higher accuracy by reflecting the result of the alias analysis when presenting the relationship between symbols. A construction device and a cross-reference construction method can be provided.

[Brief description of drawings]

FIG. 1 is a functional block diagram showing a configuration of an embodiment of a cross reference construction device of the present invention.

FIG. 2 is a sample program for explaining an embodiment of alias analysis.

FIG. 3 is an alias directed graph created for the sample program of FIG.

FIG. 4 is a diagram showing an example of symbols having a relationship of “alias”.

FIG. 5 is a flowchart showing the flow of call graph creation processing in the cross-reference building device of the present invention.

FIG. 6 is a flowchart showing the flow of an alias search process in the cross reference construction device of the present invention.

FIG. 7 is a sample program for explaining the embodiment.

FIG. 8 is a call graph obtained based on the sample program of FIG.

9 is an alias directed graph created for the sample program of FIG.

[Explanation of symbols]

1 ... Cross-reference construction device 11 ... Source files 12 ... Relationship acquisition unit between symbols 13 ... Relationship between symbols 14 ... Aliased directed graph creation section 15 ... Aliased directed graph 16 ... Attention function specification section 17 ... Relational search section 18 ... Related information data 19 ... Relationship registration department 20 ... alias search section 21 ... Alias information data 22 ... alias registration section 23 ... Connection part 24 ... Output unit

Claims (6)

[Claims] 1. An attention function designating unit for designating a target symbol to be searched, a relation analysis unit between symbols for acquiring a symbol having a specific relation to the target symbol, and a target symbol for the target symbol. A cross-reference construction apparatus comprising: an alias analysis means for obtaining a symbol of an alias; and a connection means for connecting the relation analysis means between the symbols and the search result of the alias analysis means. 2. An alias directed graph creating unit that creates an alias directed graph showing that a certain symbol may point to the same entity as that of the symbol, and the alias analyzing unit refers to the alias directed graph to notice the attention. When tracing the aliased directed graph in the positive direction starting from the symbol,
All the traceable symbols are registered as aliases, and when each of the symbols registered as aliases is used as the starting point and the alias directed graph is traced in the negative direction,
The cross-reference construction apparatus according to claim 1, further comprising: an alias search unit that registers all traceable symbols as aliases and uses the registered symbols as "aliases" of the symbol of interest. . 3. A target function specifying process for specifying a target symbol to be searched, a relationship analysis process between symbols for acquiring a symbol having a specific relationship with the target symbol, and a target symbol for the target symbol. A cross-reference construction method comprising: an alias analysis processing for obtaining a symbol of an alias; and a connection processing for connecting a search result obtained by executing the relationship analysis processing between the symbols and the alias analysis processing. 4. An alias directed graph creating step of creating an alias directed graph showing that a certain symbol may point to the same entity as the symbol, and the alias directed processing is performed with reference to the alias directed graph. When tracing the aliased directed graph in the positive direction starting from the symbol,
All the traceable symbols are registered as aliases, and when each of the symbols registered as aliases is used as the starting point and the alias directed graph is traced in the negative direction,
4. The cross reference construction method according to claim 3, further comprising: an alias search step of registering all traceable symbols as an alias and using the registered symbol as an “alias” of the symbol of interest. 5. An attention function designating step of designating a symbol of interest to be searched, a step of analyzing a relation between symbols for obtaining a symbol having a specific relation to the symbol of interest, and a step of analyzing the symbol of interest. A cross reference construction program, comprising: an alias analysis step of obtaining a symbol of an alias, and a connection step of connecting a search result obtained by executing the relation analysis step between the symbols and the alias analysis step. 6. The alias directed graph creating step of creating an alias directed graph showing which symbol may point to the same entity as the symbol in the alias analyzing step, and the alias directed graph is referred to for the attention. When tracing the aliased directed graph in the positive direction starting from the symbol,
All the traceable symbols are registered as aliases, and when each of the symbols registered as aliases is used as the starting point and the alias directed graph is traced in the negative direction,
6. The cross reference construction program according to claim 5, further comprising: an alias search step in which all traceable symbols are registered as aliases, and the registered symbols are “alias” of the symbol of interest. .