README

AQL(Annotation Query Language) Subset Compiler

What Is The AQL Subset?

This project is not a complete Compiler of AQL, but it realizes the major part of it. The BNF grammar is listed below:

aql_stmt → create_stmt ; | output_stmt ;  
create_stmt → create view ID as view_stmt  
view_stmt → select_stmt | extract_stmt  
output_stmt → output view ID alias  
alias → as ID | ε  
elect_stmt → select select_list from from_list  
select_list → select_item | select_list , select_item  
select_item → ID . ID alias  
from_list → from_item | from_list , from_item  
from_item → ID ID  
extract_stmt → extract extract_spec from from_list  
extract_spec → regex_spec | pattern_spec  
regex_spec → regex REG on column name_spec  
column → ID . ID  
name_spec → as ID | return group_spec  
group_spec → single_group | group_spec and single_group  
single_group → group NUM as ID  
pattern_spec → pattern pattern_expr name_spec  
pattern_expr → pattern_pkg | pattern_expr pattern_pkg  
pattern_pkg → atom | atom { NUM , NUM } | pattern_group  
atom→ < column > | < Token > | REG  
pattern_group → ( pattern_expr )  

Lexer

Scan the original language and output the terminator to gramma parser.

Parser

Create a gramma tree and execute the program.

Tokenizer

When executing the program, we need to get the tokens of the document to be executed, it would help us to divide token from the original document.

environment

System: Ubuntu14.04LTS 64-bit
Compiler: GNU g++ version 4.8.4

examples

We have prepared some examples for you:

• America_2016_election_news

aql_master/src/# make clean
aql_master/src/# make
aql_master/src/# make test1

• Revenue

aql_master/src/# make clean
aql_master/src/# make
aql_master/src/# make test2

• PerLoc

aql_master/src/# make clean
aql_master/src/# make
aql_master/src/# make test3

• For More information, pls view:

/dataset/dataset_ReadME.txt

extras

• If you want to use your own data, you should put the ".aql" file (for instance, "your.aql") and the article (for instance, "your.text") into the "aql_master/dataset" catalog. Then:

aql_master/src/# make clean
aql_master/src/# make
aql_master/src/# ./AQL your.aql your.text

• If your have more than one articles, you can put all of the articles into a folder (for instance, "articles", and certainly, it should be under "aql_master/dataset/"), and each article should end with ".input" (for instance, "a.input", "b.input"). Then:

aql_master/src/# make clean
aql_master/src/# make
aql_master/src/# ./AQL your.aql articles

P.S.

each output will be seperated by

"==========================================================="