[转]POI : How to Create and Use User Defined Functions
本文转自:http://poi.apache.org/spreadsheet/user-defined-functions.html
How to Create and Use User Defined Functions
Description
This document describes the User Defined Functions within POI. User defined functions allow you to take code that is written in VBA and re-write in Java and use within POI. Consider the following example.
An Example
Suppose you are given a spreadsheet that can calculate the principal and interest payments for a mortgage. The user enters the principal loan amount, the interest rate and the term of the loan. The Excel spreadsheet does the rest.

When you actually look at the workbook you discover that rather than having the formula in a cell it has been written as VBA function. You review the function and determine that it could be written in Java:

If we write a small program to try to evaluate this cell, we'll fail. Consider this source code:
import java.io.File ;
import java.io.FileInputStream ;
import java.io.FileNotFoundException ;
import java.io.IOException ; import org.apache.poi.openxml4j.exceptions.InvalidFormatException ;
import org.apache.poi.ss.formula.functions.FreeRefFunction ;
import org.apache.poi.ss.formula.udf.AggregatingUDFFinder ;
import org.apache.poi.ss.formula.udf.DefaultUDFFinder ;
import org.apache.poi.ss.formula.udf.UDFFinder ;
import org.apache.poi.ss.usermodel.Cell ;
import org.apache.poi.ss.usermodel.CellValue ;
import org.apache.poi.ss.usermodel.Row ;
import org.apache.poi.ss.usermodel.Sheet ;
import org.apache.poi.ss.usermodel.Workbook ;
import org.apache.poi.ss.usermodel.WorkbookFactory ;
import org.apache.poi.ss.util.CellReference ; public class Evaluator { public static void main( String[] args ) { System.out.println( "fileName: " + args[0] ) ;
System.out.println( "cell: " + args[1] ) ; File workbookFile = new File( args[0] ) ; try {
FileInputStream fis = new FileInputStream(workbookFile);
Workbook workbook = WorkbookFactory.create(fis); FormulaEvaluator evaluator = workbook.getCreationHelper().createFormulaEvaluator(); CellReference cr = new CellReference( args[1] ) ;
String sheetName = cr.getSheetName() ;
Sheet sheet = workbook.getSheet( sheetName ) ;
int rowIdx = cr.getRow() ;
int colIdx = cr.getCol() ;
Row row = sheet.getRow( rowIdx ) ;
Cell cell = row.getCell( colIdx ) ; CellValue value = evaluator.evaluate( cell ) ; System.out.println("returns value: " + value ) ; } catch( FileNotFoundException e ) {
e.printStackTrace();
} catch( InvalidFormatException e ) {
e.printStackTrace();
} catch( IOException e ) {
e.printStackTrace();
}
}
}
If you run this code, you're likely to get the following error:
Exception in thread "main" org.apache.poi.ss.formula.eval.NotImplementedException: Error evaluating cell Sheet1!B4
at org.apache.poi.ss.formula.WorkbookEvaluator.addExceptionInfo(WorkbookEvaluator.java:321)
at org.apache.poi.ss.formula.WorkbookEvaluator.evaluateAny(WorkbookEvaluator.java:288)
at org.apache.poi.ss.formula.WorkbookEvaluator.evaluate(WorkbookEvaluator.java:221)
at org.apache.poi.hssf.usermodel.HSSFFormulaEvaluator.evaluateFormulaCellValue(HSSFFormulaEvaluator.java:320)
at org.apache.poi.hssf.usermodel.HSSFFormulaEvaluator.evaluate(HSSFFormulaEvaluator.java:182)
at poi.tests.Evaluator.main(Evaluator.java:61)
Caused by: org.apache.poi.ss.formula.eval.NotImplementedException: calculatePayment
at org.apache.poi.ss.formula.UserDefinedFunction.evaluate(UserDefinedFunction.java:59)
at org.apache.poi.ss.formula.OperationEvaluatorFactory.evaluate(OperationEvaluatorFactory.java:129)
at org.apache.poi.ss.formula.WorkbookEvaluator.evaluateFormula(WorkbookEvaluator.java:456)
at org.apache.poi.ss.formula.WorkbookEvaluator.evaluateAny(WorkbookEvaluator.java:279)
... 4 more
How would we make it so POI can use this sheet?
Defining Your Function
To 'convert' this code to Java and make it available to POI you need to implement a FreeRefFunction instance. FreeRefFunction is an interface in the org.apache.poi.ss.formula.functions package. This interface defines one method, evaluate(ValueEval[] args, OperationEvaluationContext ec), which is how you will receive the argument values from POI.
The evaluate() method as defined above is where you will convert the ValueEval instances to the proper number types. The following code snippet shows you how to get your values:
public class CalculateMortgage implements FreeRefFunction {
@Override
public ValueEval evaluate( ValueEval[] args, OperationEvaluationContext ec ) {
if (args.length != 3) {
return ErrorEval.VALUE_INVALID;
}
double principal, rate, years, result;
try {
ValueEval v1 = OperandResolver.getSingleValue( args[0],
ec.getRowIndex(),
ec.getColumnIndex() ) ;
ValueEval v2 = OperandResolver.getSingleValue( args[1],
ec.getRowIndex(),
ec.getColumnIndex() ) ;
ValueEval v3 = OperandResolver.getSingleValue( args[2],
ec.getRowIndex(),
ec.getColumnIndex() ) ;
principal = OperandResolver.coerceValueToDouble( v1 ) ;
rate = OperandResolver.coerceValueToDouble( v2 ) ;
years = OperandResolver.coerceValueToDouble( v3 ) ;
The first thing we do is check the number of arguments being passed since there is no sense in attempting to go further if you are missing critical information.
Next we declare our variables, in our case we need variables for:
- principal - the amount of the loan
- rate - the interest rate as a decimal
- years - the length of the loan in years
- result - the result of the calculation
Next, we use the OperandResolver to convert the ValueEval instances to doubles, though not directly. First we start by getting discreet values. Using the OperandResolver.getSingleValue() method we retrieve each of the values passed in by the cell in the spreadsheet. Next, we use the OperandResolver again to convert the ValueEval instances to doubles, in this case. This class has other methods of coercion for gettings Strings, ints and booleans. Now that we've got our primitive values we can move on to calculating the value.
As shown previously, we have the VBA source. We need to add code to our class to calculate the payment. To do this you could simply add it to the method we've already created but I've chosen to add it as its own method. Add the following method:
public double calculateMortgagePayment( double p, double r, double y ) {
double i = r / 12 ;
double n = y * 12 ;
double principalAndInterest =
p * (( i * Math.pow((1 + i),n ) ) / ( Math.pow((1 + i),n) - 1)) ;
return principalAndInterest ;
}
The biggest change necessary is related to the exponents; Java doesn't have a notation for this so we had to add calls to Math.pow(). Now we need to add this call to our previous method:
result = calculateMortgagePayment( principal, rate, years ) ;
Having done that, the last things we need to do are to check to make sure we didn't get a bad result and, if not, we need to return the value. Add the following code to the class:
private void checkValue(double result) throws EvaluationException {
if (Double.isNaN(result) || Double.isInfinite(result)) {
throw new EvaluationException(ErrorEval.NUM_ERROR);
}
}
Then add a line of code to our evaluate method to call this new static method, complete our try/catch and return the value:
checkValue(result);
} catch (EvaluationException e) {
e.printStackTrace() ;
return e.getErrorEval();
}
return new NumberEval( result ) ;
So the whole class would be as follows:
import org.apache.poi.ss.formula.OperationEvaluationContext ;
import org.apache.poi.ss.formula.eval.ErrorEval ;
import org.apache.poi.ss.formula.eval.EvaluationException ;
import org.apache.poi.ss.formula.eval.NumberEval ;
import org.apache.poi.ss.formula.eval.OperandResolver ;
import org.apache.poi.ss.formula.eval.ValueEval ;
import org.apache.poi.ss.formula.functions.FreeRefFunction ; /**
* A simple function to calculate principal and interest.
*
* @author Jon Svede
*
*/
public class CalculateMortgage implements FreeRefFunction { @Override
public ValueEval evaluate( ValueEval[] args, OperationEvaluationContext ec ) {
if (args.length != 3) {
return ErrorEval.VALUE_INVALID;
} double principal, rate, years, result;
try {
ValueEval v1 = OperandResolver.getSingleValue( args[0],
ec.getRowIndex(),
ec.getColumnIndex() ) ;
ValueEval v2 = OperandResolver.getSingleValue( args[1],
ec.getRowIndex(),
ec.getColumnIndex() ) ;
ValueEval v3 = OperandResolver.getSingleValue( args[2],
ec.getRowIndex(),
ec.getColumnIndex() ) ; principal = OperandResolver.coerceValueToDouble( v1 ) ;
rate = OperandResolver.coerceValueToDouble( v2 ) ;
years = OperandResolver.coerceValueToDouble( v3 ) ; result = calculateMortgagePayment( principal, rate, years ) ; checkValue(result); } catch (EvaluationException e) {
e.printStackTrace() ;
return e.getErrorEval();
} return new NumberEval( result ) ;
} public double calculateMortgagePayment( double p, double r, double y ) {
double i = r / 12 ;
double n = y * 12 ; //M = P [ i(1 + i)n ] / [ (1 + i)n - 1]
double principalAndInterest =
p * (( i * Math.pow((1 + i),n ) ) / ( Math.pow((1 + i),n) - 1)) ; return principalAndInterest ;
} /**
* Excel does not support infinities and NaNs, rather, it gives a #NUM! error in these cases
*
* @throws EvaluationException (#NUM!) if <tt>result</tt> is <tt>NaN</> or <tt>Infinity</tt>
*/
static final void checkValue(double result) throws EvaluationException {
if (Double.isNaN(result) || Double.isInfinite(result)) {
throw new EvaluationException(ErrorEval.NUM_ERROR);
}
} }
Great! Now we need to go back to our original program that failed to evaluate our cell and add code that will allow it run our new Java code.
Registering Your Function
Now we need to register our function in the Workbook, so that the Formula Evaluator can resolve the name "calculatePayment" and map it to the actual implementation (CalculateMortgage). This is done using the UDFFinder object. The UDFFinder manages FreeRefFunctions which are our analogy for the VBA code. We need to create a UDFFinder. There are a few things we need to know in order to do this:
- The name of the function in the VBA code (in our case it is calculatePayment)
- The Class name of our FreeRefFunction
UDFFinder is actually an interface, so we need to use an actual implementation of this interface. Therefore we use the org.apache.poi.ss.formula.udf.DefaultUDFFinder class. If you refer to the Javadocs you'll see that this class expects to get two arrays, one containing the alias and the other containing an instance of the class that will represent that alias. In our case our alias will be calculatePayment and our class instance will be of the CalculateMortgage type. This class needs to be available at compile and runtime. Be sure to keep these arrays well organized because you'll run into problems if these arrays are of different sizes or the alias aren't in the same relative position in their respective arrays. Add the following code:
String[] functionNames = { "calculatePayment" } ;
FreeRefFunction[] functionImpls = { new CalculateMortgage() } ;
UDFFinder udfs = new DefaultUDFFinder( functionNames, functionImpls ) ;
UDFFinder udfToolpack = new AggregatingUDFFinder( udfs ) ;
Now we have our UDFFinder instance and we've created the AggregatingUDFFinder instance. The last step is to pass this to our Workbook:
workbook.addToolPack(udfToolpack);
So now the whole class will look like this:
import java.io.File ;
import java.io.FileInputStream ;
import java.io.FileNotFoundException ;
import java.io.IOException ; import org.apache.poi.openxml4j.exceptions.InvalidFormatException ;
import org.apache.poi.ss.formula.functions.FreeRefFunction ;
import org.apache.poi.ss.formula.udf.AggregatingUDFFinder ;
import org.apache.poi.ss.formula.udf.DefaultUDFFinder ;
import org.apache.poi.ss.formula.udf.UDFFinder ;
import org.apache.poi.ss.usermodel.Cell ;
import org.apache.poi.ss.usermodel.CellValue ;
import org.apache.poi.ss.usermodel.Row ;
import org.apache.poi.ss.usermodel.Sheet ;
import org.apache.poi.ss.usermodel.Workbook ;
import org.apache.poi.ss.usermodel.WorkbookFactory ;
import org.apache.poi.ss.util.CellReference ; public class Evaluator { public static void main( String[] args ) { System.out.println( "fileName: " + args[0] ) ;
System.out.println( "cell: " + args[1] ) ; File workbookFile = new File( args[0] ) ; try {
FileInputStream fis = new FileInputStream(workbookFile);
Workbook workbook = WorkbookFactory.create(fis); String[] functionNames = { "calculatePayment" } ;
FreeRefFunction[] functionImpls = { new CalculateMortgage() } ; UDFFinder udfs = new DefaultUDFFinder( functionNames, functionImpls ) ;
UDFFinder udfToolpack = new AggregatingUDFFinder( udfs ) ; workbook.addToolPack(udfToolpack); FormulaEvaluator evaluator = workbook.getCreationHelper().createFormulaEvaluator(); CellReference cr = new CellReference( args[1] ) ;
String sheetName = cr.getSheetName() ;
Sheet sheet = workbook.getSheet( sheetName ) ;
int rowIdx = cr.getRow() ;
int colIdx = cr.getCol() ;
Row row = sheet.getRow( rowIdx ) ;
Cell cell = row.getCell( colIdx ) ; CellValue value = evaluator.evaluate( cell ) ; System.out.println("returns value: " + value ) ; } catch( FileNotFoundException e ) {
e.printStackTrace();
} catch( InvalidFormatException e ) {
e.printStackTrace();
} catch( IOException e ) {
e.printStackTrace();
}
}
}
Now that our evaluator is aware of the UDFFinder which in turn is aware of our FreeRefFunction, we're ready to re-run our example:
Evaluator mortgage-calculation.xls Sheet1!B4
which prints the following output in the console:
fileName: mortgage-calculation.xls
cell: Sheet1!B4
returns value: org.apache.poi.ss.usermodel.CellValue [790.7936267415464]
That is it! Now you can create Java code and register it, allowing your POI based appliction to run spreadsheets that previously were inaccessible.
This example can be found in the src/examples/src/org/apache/poi/ss/examples/formula folder in the source.
[转]POI : How to Create and Use User Defined Functions的更多相关文章
- Read / Write Excel file in Java using Apache POI
Read / Write Excel file in Java using Apache POI 2014-04-18 BY DINESH LEAVE A COMMENT About a year o ...
- [Hive - LanguageManual] Create/Drop/Alter -View、 Index 、 Function
Create/Drop/Alter View Create View Drop View Alter View Properties Alter View As Select Version info ...
- PostgreSQL 之 CREATE FUNCTION
官方文档 语法: CREATE [ OR REPLACE ] FUNCTION name ( [ [ argmode ] [ argname ] argtype [ { DEFAULT | = } d ...
- MySql批处理的小窍门:排行榜类数据生成
MySql批处理的小窍门:排行榜类数据生成 最近在做新版本的开发,其中涉及到排行榜的批量预生成,在此分享给大家. 关键点 名次的计算(不考虑用游标) 单榜单查询 对于排行榜这种类型的数据,当只查一个排 ...
- 超级顽固的流方式读取doc,docx乱码问题
因为工作中需要一个把doc或者docx的office文档内容,需要读取出来,并且也没展示功能.代码中第一考虑可能就是通过读取流方式,结果写了以后,各种乱码,百科的解决方案也是千奇百怪,第一点:可能是文 ...
- Phoenix综述(史上最全Phoenix中文文档)
个人主页:http://www.linbingdong.com 简书地址:http://www.jianshu.com/users/6cb45a00b49c/latest_articles 网上关于P ...
- 【repost】JavaScript Scoping and Hoisting
JavaScript Scoping and Hoisting Do you know what value will be alerted if the following is executed ...
- PHP7函数大全(4553个函数)
转载来自: http://www.infocool.net/kb/PHP/201607/168683.html a 函数 说明 abs 绝对值 acos 反余弦 acosh 反双曲余弦 addcsla ...
- 有关binlog的那点事(mysql5.7.13)
binlog作为mysql中最重要的日志之一,能实现异常恢复以及主从复制. 我们主要讨论的是主从复制中的binlog,这里将以mysql5.7.13的源码为主要依据来分析binlog. 在主从复制中, ...
随机推荐
- nginx + ffmpeg
第一种方案:ffmpeg+nginx 新的ffmpeg已经支持HLS.(本人也参与了代码供献,给自己做个广告:)) 点播: 生成hls分片: ffmpeg -i <媒体文件> -c:v l ...
- Logcat不显示Application的解决办法
Window - show view - devices - debug ----2014.12.1------ 只有在DDMS的device中显示进程名,logcat中的Application标签才 ...
- Linux下抓包工具tcpdump应用详解
TCPDUMP简介 在传统的网络分析和测试技术中,嗅探器(sniffer)是最常见,也是最重要的技术之一.sniffer工具首先是为网络管理员和网络程序员进行网络分析而设计的.对于网络管理人员来说 ...
- 基于Jenkins+Gitlab的自动化部署实战
故事背景 一个中小型企业,是典型的互联网公司,当初期的时候可能运维只能标配到2~3人,此时随着公司的发展,项目会逐渐增多.前期部署项目可能都是手动的, 俗称“人肉部署”,这简直是无比的痛苦,不能忍受的 ...
- codevs 3324 新斯诺克
3324 新斯诺克 时间限制: 1 s 空间限制: 64000 KB 题目等级 : 白银 Silver 题目描述 Description 斯诺克又称英式台球,是一种流行的台球运动.在球桌上, ...
- JavaScript-Tool:jquery.vaidate.js
ylbtech-JavaScript-Tool:jquery.vaidate.js 1.返回顶部 1. 2. 3. 2.返回顶部 3.返回顶部 4.返回顶部 5.返回顶部 0. https ...
- bzoj3489
kdtree 3维kdtree,就是三个维度轮换着切,我们把每个元素看成一个点,坐标是上次出现的位置,下次出现的位置,自己的位置,第一个<l,第二个>r,第三个[l,r],然后kdtree ...
- 2.CSS 颜色代码大全
确实使用,不用重复造轮子了!!! 摘自:http://www.cnblogs.com/circlebreak/p/6140602.html
- linux中用管道实现兄弟进程通信
1 使用fork函数创建两个子进程.在第一个子进程中发送消息到第二个子进程,第二个子进程都出来并处理. 2 在父进程中,不适用管道通信,所以什么不需要做直接关闭勒管道的两端 3 代码实现 #inclu ...
- Atom 基本使用
插件 simplified-Chinese-menu:汉化 Markdown-preview-plus:Markdown 实时预览 Atom 自带 Markdown-preview,这个插件是其增强版 ...