TWI498727B - Method and system for debugging - Google Patents

Description

Method and system for debugging

The present invention relates to a debugging method, and more particularly to a debugging method and system for automatically finding an error point and correcting it by means of parsing an error message.

The general software can only obtain the binary executable file that has been compiled. Usually, only the author or copyright owner of the software can own the source code of the program. Open Source software is free to redistribute, and can be derived from books, which facilitates communication and sharing of code.

In the development of open source software, since the open source code system may be modified by many people, the chances of debugging will be relatively large. Generally, when an error occurs in the compilation, the engineer needs to perform manual debugging to debug. This is not only time-consuming but also prone to spelling errors, etc. It is quite inconvenient to check the debugging line by line.

The invention provides a debugging method and system, which can automatically parse an error message to find an error and automatically correct it.

The invention provides a debugging method suitable for debugging a code file. First, an error message is retrieved from the message log file. Then, according to the matching rule, the error message is parsed, and the file name, the error type, and the error code segment in which the error occurred are obtained. The matching rules are established based on the compiler used by the code file. Then, based on the file name and the error type, the source code source is used to find the correct code segment corresponding to the error code segment. Then, the modification mode is determined by the error pattern. Finally, according to the modification mode, the code file corresponding to the file name is corrected by the correct code segment.

In an embodiment of the present invention, after the step of searching for the correct code segment corresponding to the error code segment from the source code, the method further includes opening the code file corresponding to the file name.

In an embodiment of the invention, the error type includes one of an undeclared type, an undefined type, and a path nonexistent type. The above modification modes include a replacement mode and an insertion mode. If the error type is a path non-existent type, it is determined that the modification mode is the replacement mode. In addition, if the error type is an undeclared type or an undefined type, it is determined that the modified mode is the insert mode.

In an embodiment of the present invention, in the step of modifying the code file corresponding to the file name by using the correct code segment according to the modification mode, when the modification mode is the insert mode, the correct code segment is inserted into the code file, where the correct The position of the code segment is before the position of the error code segment; on the other hand, when the modification mode is the replacement mode, the error code segment is replaced with the correct code segment.

In an embodiment of the present invention, when the modification mode is the replacement mode, if the error code segment is to be deleted, the empty set is used as the correct code segment, so that the error code is replaced by the empty set in the replacement mode. segment.

In an embodiment of the present invention, the foregoing matching rule includes a first keyword, a second keyword, a third keyword, and a fourth keyword. The step of parsing the error message according to the matching rule first identifies the first keyword to divide the error message into the first segment and the second segment according to the first keyword. Then, in the first segment, the file name is extracted according to the second keyword. Next, in the second segment, the error code segment and the error pattern are respectively extracted according to the third keyword and the fourth keyword.

The invention further provides a debugging system suitable for debugging a code file. Here, the debugging system includes a capture module, an analysis module, and a correction module. The capture module is used to retrieve an error message from the message log file. The parsing module is configured to parse the error message according to the matching rule, and obtain the file name, the error type, and the error code segment in which the error occurred. The matching rules are established based on the compiler used by the code file. Moreover, the parsing module searches for the correct code segment corresponding to the error code segment from the kernel source according to the file name and the error type. The correction module is used to determine the modification mode by the error type, and then correct the code file corresponding to the file name with the correct code segment according to the modification mode.

In an embodiment of the invention, the correction module further includes a code file corresponding to the file name. In addition, the correction module further includes determining whether the modification mode is an insertion mode or a replacement mode. If the error type is a path non-existent type, the correction module determines that the modification mode is the replacement mode; if the error type is an undeclared type or an undefined type, the correction module determines that the modification mode is the insertion mode.

In an embodiment of the invention, the correcting module inserts the correct code segment into the code file when the modified mode is the insert mode, wherein the correct code segment is located before the position of the error code segment. Moreover, when the modification mode is the replacement mode, the correction module replaces the error code segment with the correct code segment.

In an embodiment of the present invention, the foregoing matching rule includes a first keyword, a second keyword, a third keyword, and a fourth keyword. The parsing module identifies the first keyword in the error message to divide the error message into the first segment and the second segment according to the first keyword. Moreover, the parsing module extracts the file name according to the second keyword in the first segment, and extracts the error code segment and the error pattern according to the third keyword and the fourth keyword in the second segment.

Based on the above, the present invention establishes a set of matching rules based on the compiler used by the code file. Accordingly, once an error occurs during compilation, the error message can be retrieved from the message log file, and the error message can be further parsed by the above matching rule to find out where the error occurred, and the correction is further automatically performed. As a result, the time for modifying the code file is greatly saved, and the convenience of debugging is also increased.

The above described features and advantages of the present invention will be more apparent from the following description.

In order to clarify the content of the present invention, the following specific examples are given as examples in which the present invention can be implemented.

1 is a flow chart of a method of debugging in accordance with an embodiment of the present invention. Referring to FIG. 1, in step S105, an error message is retrieved from the message record file. When compiling, the generated message can be directed to a message log file. Accordingly, once an error occurs during compilation, the error message can be retrieved from the message log file for parsing to find out the error point.

Next, in step S110, the error message is parsed according to the matching rule, and the file name, the error type, and the error code segment in which the error occurred are obtained. Here, since the error message has a fixed format, the matching rule can be established according to the compiler used by the code file. For example, the matching rule includes four sets of keywords (a first keyword, a second keyword, a third keyword, and a fourth keyword). The error message is divided into a first segment and a second segment by using the first keyword, and the second keyword, the third keyword, and the fourth keyword are respectively used to respectively extract the file name, the error code segment, and the error pattern.

Take the GNU Compiler Collection (GCC) as an example. The error message format is, for example, "<file name>:error:...<error type>...<error code segment>..." or "<file name>:error:...<error code segment>...<error type>... "." Accordingly, when parsing the error message, the first keyword ":error:" can be identified in the error message, and the left side of the first keyword ":error:" is set as the first segment, and ":error:" is set. The right side is set to the second segment. After the first segment and the second segment are divided, in the first segment, the file name is obtained by using the second keyword (for example, "/" and "."). In addition, in the second segment, the third keyword (for example, "", """, "/", and ".") and the fourth keyword (for example, "undeclared" or "undefined" or " No such") to get the error code segment and the error pattern.

For example, suppose the error message is "include/asm/irq.h:4:27:error:asm/arch/irqs.h:No such file or directory". By the above method, the file name "include/asm/irq.h", the error code segment "asm/arch/irqs.h", and the error type "No such file or directory" can be obtained.

Then, in step S115, according to the file name and the error type, the correct code segment corresponding to the error code segment is searched from the kernel source. In the above error message "include/asm/irq.h:4:27:error:asm/arch/irqs.h:No such file or directory", the error type is that the path does not exist, that is, it may be Asm/arch/ There is no irqs.h under this path, or there is no such path asm/arch/. According to this, the core to the source code of the compiled description file (Makefile), and the target platform architecture ARCH is set to the ARM platform, thereby searching for the correct placement of irqs.h and getting the correct code segment "asm/arch-s3c2410/ Irqs.h". After that, you can open the code file corresponding to the file name (that is, "include/asm/irq.h") for subsequent modification.

Then, in step S120, the modification mode (replacement mode or insertion mode) is determined by the error pattern. For example, the error pattern may be an undeclared type "undeclared" or an undefined type "undefined" or a path non-existent type "No such file or directory". If the error type is a path non-existent type, it is decided to modify the mode to the replacement mode. In addition, if the error type is undeclared or undefined, it is decided to modify the mode to insert mode.

Finally, in step S125, according to the modification mode, the code file corresponding to the file name is corrected by the correct code segment. That is, when the modification mode is the insert mode, the correct code segment is inserted before the position of the error code segment in the code file. When the modification mode is the replacement mode, the error code segment is replaced with the correct code segment. For example, in the above error message "include/asm/irq.h:4:27:error:asm/arch/irqs.h:No such file or directory", the correction mode is the replacement mode. In other words, replace the error code segment "asm/arch/irqs.h" in the code file with the correct code segment "asm/arch-s3c2410/irqs.h".

It is worth noting that if the error code segment is to be deleted, an empty set can be used as the correct code segment to replace the error code segment with the empty set in the replacement mode.

The following error message "networking/interface.c:822:error:`ARPHRD_INFINIBAND' undeclared here(not in a function)" is explained. After the error message is parsed by the matching rule, the file name "networking/interface.c", the error code segment "ARPHRD_INFINIBAND", and the error pattern "undeclared here (not in a function)" are obtained. Next, search for the source code based on "ARPHRD_INFINIBAND", and find the definition of "ARPHRD_INFINIBAND" in "include/linux/if_arp.h", and further get the correct code segment as "#define ARPHRD_INFINIBAND 32". Then, open the code file corresponding to the file name, which is "networking/interface.c". Since the error type is undeclared, it applies to the insert mode, so the correct code segment "#define ARPHRD_INFINIBAND 32" is inserted before "ARPHRD_INFINIBAND" in "networking/interface.c". Then close "networking/interface.c" and clear the file generated during the last compilation to recompile the modified code file.

The invention further provides a debugging system suitable for debugging a code file. 2 is a block diagram of a debug system in accordance with an embodiment of the present invention. Referring to FIG. 2 , the debug system 200 includes a capture module 210 , an analysis module 220 , and a correction module 230 .

The capture module 210 is configured to retrieve an error message from the message record file. The parsing module 220 is configured to parse the error message according to the matching rule, and obtain the file name, the error type, and the error code segment in which the error occurred. The matching rules are established based on the compiler used by the code file. Moreover, the parsing module 220 searches for the correct code segment corresponding to the error code segment from the core source code according to the file name and the error type.

Specifically, it is assumed that the matching rule includes a first keyword, a second keyword, a third keyword, and a fourth keyword. The parsing module 220 can identify the first keyword in the error message to divide the error message into the first segment and the second segment according to the first keyword. In addition, the parsing module 220 extracts the file name according to the second keyword in the first segment, and extracts the error code segment and the error pattern according to the third keyword and the fourth keyword in the second segment.

The correction module 230 is configured to determine the modification mode by the error type, and then correct the code file corresponding to the file name with the correct code segment according to the modification mode. In addition, the correction module 230 can be used to open the code file corresponding to the file name. The correction module 230 can also determine whether the modification mode is the insertion mode or the replacement mode according to the error type. If the error type is an undeclared type or an undefined type, the correction module 230 determines that the modification mode is the insertion mode to insert the correct code segment into the code file, wherein the position of the correct code segment is at the position of the error code segment. prior to. If the error pattern is a path non-existent type, the correction module 230 determines that the modification mode is the replacement mode, and replaces the error code segment with the correct code segment.

In summary, the present invention stores the generated message into the message record file at compile time, and establishes a set of matching rules according to the compiler used by the code file file. Accordingly, once an error occurs during compilation, the error message can be retrieved from the message log file, and the error message can be further parsed using the above matching rule to find out where the error occurred, and the correction is further automatically performed. As a result, the time for modifying the code file is greatly saved, and the convenience of debugging is also increased.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the invention, and any one of ordinary skill in the art can make some modifications and refinements without departing from the spirit and scope of the invention. The scope of the invention is defined by the scope of the appended claims.

S105~S125‧‧‧ steps of the debugging method of an embodiment of the present invention

200‧‧‧Debug system

210‧‧‧Capture module

220‧‧‧analysis module

230‧‧‧correction module

1 is a flow chart of a method of debugging in accordance with an embodiment of the present invention.

2 is a block diagram of a debug system in accordance with an embodiment of the present invention.

S105~S125. . . Steps of the debugging method according to an embodiment of the present invention

Claims (10)

A debugging method is suitable for debugging a code file. The debugging method includes: extracting an error message from a message record file; parsing the error message according to a matching rule, and obtaining an error file a name, an error type, and an error code segment, wherein the matching rule is established according to a compiler used by the code file, the error type including an undeclared type, an undefined type, and a path There is no one of the types; according to the file name and the error type, a correct code segment corresponding to the error code segment is searched from a core source; a modification mode is determined by the error pattern, wherein the modification The mode includes a permutation mode and an insertion mode, and the step of determining the modification mode by the error pattern comprises: determining that the modification mode is the permutation mode if the error type is that the path does not exist; If the error type is the undeclared type or the undefined type, determining that the modified mode is the insert mode; and according to the modified mode, repairing the correct code segment The file name of the corresponding code file. The debugging method of claim 1, wherein after the step of searching for the correct code segment corresponding to the error code segment from the core source, the method further comprises: opening a code file corresponding to the file name. The debugging method of claim 1, wherein the step of modifying the code file corresponding to the file name by the correct code segment according to the modification mode comprises: when the modification mode is the insertion mode, The correct code segment is inserted into the code file, wherein the correct code segment is located before the location of the error code segment; and when the modification mode is the replacement mode, the error code segment is replaced with the correct code segment. The debugging method of claim 3, wherein when the modification mode is the replacement mode, the method further comprises: if the error code segment is to be deleted, using an empty set as the correct code segment, In the replacement mode, the empty code set is used to replace the error code segment. The method for debugging according to the first aspect of the patent application includes: storing the error message to the message log file when an error occurs in the code file compilation. The method for debugging according to claim 1, wherein the matching rule includes a first keyword, a second keyword, a third keyword, and a fourth keyword, wherein the matching rule is used according to the matching rule. The step of the error message includes: identifying the first keyword to divide the error message into a first segment and a second segment according to the first keyword; and in the first segment, according to the second key Word extracting the file name; and in the second segment, according to the third keyword and the fourth keyword The error code segment and the error pattern are respectively taken out. A debugging system, configured to debug a code file, the debugging system comprising: a capture module, extracting an error message from a message record file; and an analysis module, according to a matching rule, Parsing the error message, and obtaining a file name, an error type, and an error code segment in which the error occurred, wherein the matching rule is established according to a compiler used by the code file, and the error type includes one One of the undeclared type, an undefined type, and a path nonexistent type, and, according to the file name and the error type, source code from a core to find a correct code segment corresponding to the error code segment And a correction module, wherein the modification mode includes a modification mode, wherein the modification mode includes a replacement mode and an insertion mode, and if the error type is that the path does not exist, the correction module determines The modification mode is the replacement mode; if the error type is the undeclared type or the undefined type, the correction module determines that the modification mode is the insertion mode; afterwards, the modification Group according to the modification mode, and code segments to correct the correction code file corresponding to the file name. The debugging system of claim 7, wherein the correction module further comprises opening a code file corresponding to the file name. The debugging system of claim 7, wherein the correction module inserts the correct code segment into the code file when the modification mode is the insertion mode, wherein the correct code segment is located at the The position of the error code segment is before; and the correction module is in the modification mode in the modification mode The error code segment is replaced with the correct code segment. The debugging system of claim 7, wherein the matching rule includes a first keyword, a second keyword, a third keyword, and a fourth keyword, and the parsing module recognizes the The first keyword in the error message, the error message is divided into a first segment and a second segment according to the first keyword; and, in the first segment, the second keyword is taken out according to the second keyword The file name; in the second segment, the error code segment and the error pattern are respectively extracted according to the third keyword and the fourth keyword.