JPH0764801A - User word retrieval system - Google Patents

Abstract

PURPOSE:To quickly retrieve a user word in a compiler without repeating the same retrieval procedures at the time of continuous retrieval of the user word retrieved just before by changing the order from the head of chain addresses so that the user word detected as the result of retrieval is placed in the head of a hash chain. CONSTITUTION:A name table head acquiring means 21 acquires the head of name entries 26 connected to a hash table 25 in a name table 24. A name entry next retrieval means 22 traces chain addresses 261 of name entries 26 to retrieve a name entry having the same hash value. When the name entry 26 having a user word 262 coinciding with a user word 102 in name information 10 is detected by the name table head acquiring means 21 and the name table next retrieval means 22, a name table reorganizing means 23 reorganizes the chain so that the detected name entry is connected to the head of the chain connected to the hash table 25.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a user word search method, and more particularly to a user word search method for searching a user word in a name table constructed on the basis of hash calculation in a compiler.

[0002]

2. Description of the Related Art Conventionally, this type of user word search method is a chain in which name entries having the same hash value are connected to the name entry corresponding to the user word once detected by searching the name table. It does not reorganize, but keeps the chain connection order fixed.

[0003]

In the above-described conventional user word search method, the name entry having the same hash value is connected to the name entry corresponding to the user word detected by searching the name table. Since the chain connection order is fixed without reorganizing the chain, when repeatedly searching for the same user word, the chain in which name entries with the same hash value are connected is searched repeatedly. However, there is a problem in that a search process for the same user word takes an unnecessary time.

[0004]

According to the user word search method of the present invention, a user word is obtained from a user word by hash calculation in a user word search method for searching a user word in a compiler. A hash table that manages by a chain address that connects user words that have the same hash value as a name entry based on the hash value,
A name table consisting of a user word and a name entry having a chain address of the user word having the same hash value as that hash value, and a name entry having the same hash value as the hash value of the user word from this name table Name table top obtaining means, a name table next retrieval means for sequentially obtaining name entries next to the name entry obtained by the name table top obtaining means by tracing a chain address, and the name table top obtaining means. A name entry reorganizing means for reorganizing the chain by positioning the name entry corresponding to the user word detected by the name table next search means as the first name entry of the chain based on the hash value. It is composed by having a search part.

[0005]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of a compiler system including an embodiment of the present invention. The compiler system shown in FIG. 1 includes a source program 1, a compiler 2, and an object program 3, and the compiler 2 includes a lexical analysis processing unit 4, a syntax analysis processing unit 5, and a semantic analysis processing unit 6.
And a code generation processing section 7 and an object program output processing section 8.

The semantic analysis processing unit 6 includes name information 10, a name search unit 20, and a detected name entry 30 detected by the name search unit 20. The name information 10 includes a user word 102 and a hash value 101 obtained by hash calculation from the user word 102. The name search unit 20 includes a name table 24, a name table head acquisition unit 21 for acquiring the head of the name entry 26 connected to the hash table 25 in the name table 24 corresponding to the hash value 101 of the name information 10, and the name entry 26. Name entry next search means 22 for searching the next name entry having the same hash value by tracing the chain address 261, and name table top acquisition means 2
1 and the name table next retrieval means 22 for name information 1
When a name entry 26 having a user word 262 that matches the user word 102 in 0 is detected, the chain is reorganized so that the detected name entry is connected to the head of the chain connected from the hash table 25. And knitting means 23. The name table 24 is a hash table 25 that manages name entries for each hash value, and a user word 2 that has the same hash value.
62 and a name entry 26 including a chain address 261 connecting the user word 262 from the hash table 25.

FIG. 2 is a diagram showing a detailed structure of the name table 24. The hash table 25 shows a chain address 3 for a name entry connected by a chain for each hash value.
It consists of a sequence of 01.

FIG. 3 is a flow chart showing the processing of the name search unit 20, and FIG. 4 is a flow chart showing the processing of the name table reorganizing means 23.

Next, the operation of the embodiment shown in FIG. 1 will be described with reference to FIGS. The compiler 2 inputs the source program 1, checks whether the source program 1 is syntactically correct in the lexical analysis processing unit 4 and the syntax analysis processing unit 5, and then activates the semantic analysis processing unit 6. The semantic analysis processing unit 6 activates the name search unit 20 to perform a double definition check and a non-definition check for the user words in the source program 1. The name search unit 20 uses the name information 1
0 is received, and the hash value 101 of the name information 10 is passed to the name table head acquisition means 21 (step 401). The name table head acquisition unit 21 acquires the chain address 301 to the name entry 26 corresponding to the hash value 101 from the hash table 25 and returns it to the name search unit 20. Based on the chain address 301 received from the name table head acquisition means 21, the name search unit 20 checks whether the head name entry 26 exists (step 40).
2), if the name entry 26 does not exist, name information 1
The name entry 26 corresponding to 0 does not exist (step 408) and the process ends. If the first name entry 26 exists, it is checked whether the user word 262 in the name entry 26 matches the user word 102 in the name information 10 (step 403). If they match, the name table reorganization means. 23 to reorganize the chain (step 40
7) If they do not match, the name table next search means 22 is called. The name table next search means 22 indicates the next name entry 26, and the chain address 26 of the name entry 26.
Returns 1. The name search unit 20 uses the name table next search means 2
It is checked whether the next name entry 26 exists based on the chain address 261 received from step 2 (step 40
5), if the name entry 26 does not exist, name information 1
The name entry 26 corresponding to 0 does not exist (step 408) and the process ends. If the next name entry 26 exists, it is checked whether the user word 262 in the name entry 26 matches the user word 102 in the name information 10 (step 406), and if they match, the name table reorganization means. 23 to reorganize the chain (step 40
7) If they do not match, the name table next search means 22 is further called to continue the search (step 40).
4). The name search unit 20 uses the user word 102 of the name information 10.
When the name entry 26 having the user word 262 matching with is detected, the name table reorganizing means 23 is called (step 407). The name table reorganization means 23 determines whether the detected name entry 26 is the first name entry 26 connected to the hash table 25 (step 50).
1), in the case of the name entry 26 at the head, nothing is done.
If it is not the first name entry 26, the value of the chain address 261 of the detected name entry 26 is put in the chain address 261 of the immediately preceding name entry 26 (step 502), and the chain address 261 of the detected name entry 26 is set in the hash table 25. The value of the chain address 301 to the name entry corresponding to the hash value of is entered (step 503), and the chain address 3 to the name entry corresponding to the hash value in the hash table 25 is entered.
The value of the address indicating the name entry 26 detected as 01 is entered (step 504), and the name table reorganization means 23 is terminated.

After the name table reorganization means 23 is finished, the name search unit 20 returns the detected name entry 30 to the semantic analysis processing unit 6.

After the completion of the semantic analysis processing unit 6, the compiler 2 produces a code generation processing unit 7 and an object program output processing unit 8.
The target program output processing unit 8 outputs the target program 3.

[0013]

As described above, according to the present invention, the order of the chains connecting the name entries having the same hash value is
When a name entry corresponding to a user word is searched, the detected name entry is re-edited so that it is positioned at the beginning of the chain, so it can be quickly detected when repeatedly searching the same user word. Have an effect.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a compiler system including an embodiment of the present invention.

FIG. 2 is a detailed configuration diagram of a name table in the embodiment of FIG.

FIG. 3 is a flow chart of processing of a name search unit in the embodiment of FIG.

FIG. 4 is a flow chart of processing of a name table reorganization unit in the embodiment of FIG.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 source program 2 compiler 3 target program 4 lexical analysis processing unit 5 syntactic analysis processing unit 6 semantic analysis processing unit 7 code generation processing unit 8 target program output processing unit 10 name information 20 name search unit 21 name table top acquisition unit 22 name table Next search means 23 Name table reorganization means 24 Name table 25 Hash table 26 Name entry 30 Detected name entry 101 Hash value 102 User word 261 Chain address 262 User word

Claims (1)

[Claims] 1. In a user word search method for searching a user word in a compiler, a user word having a same hash value based on a hash value obtained by hash calculation of the user word from the user word. From the name table, which consists of a hash table that manages with chain addresses that connect to each other as a name entry, and a name entry that has a user word and a chain address of a user word that has the same hash value as that hash value. The name table head acquisition means for searching the head of the name entry having the same hash value as the user's word hash value, and the name entry next to the name entry acquired by this name table head acquisition means are sequentially acquired by tracing the chain address. Name table next retrieval means, name table top acquisition means and name table A name entry reorganization means for reorganizing the chain by locating the name entry corresponding to the user word detected by the bull next retrieval means as the first name entry of the chain based on the hash value. A user word search method characterized by having.